OpenAI and Gemini Achieve IMO Gold, CBS Cancels Colbert's Late Night, McLaren's Mansour Ojjeh Collection | Chris Black, Chris Buskirk, Turner Caldwell, Jamin Ball, Harper Carroll, Nikita Singareddy

You're watching TVPN.

Today is Monday, July 21, 2025. We are live from the TVPN Ultra Dome. The temple of technology, the fortress of finance, the capital de capital de capital. Big news. OpenAI. Great to be back. OpenAI has announced that they have won a long. They achieved the long standing grand challenge in AI gold medal level performance on the world's most prestigious math competition, the International Math olympiad. That's the IMO. So this went up at 12:50am on July 19th.

Typical timeline.

Yes. So this is, this is basically like Friday night, Saturday morning, you're looking at basically 1am from Alexander Way and he shares a picture of a strawberry with a metal with the OpenAI logo on it. Someone also tested, they took this picture, they uploaded ChatGPT. What's the fruit in the image? How many Rs are in that fruit? And it nailed it. So we are truly, we are truly. AGI has arrived. Has arrived. Also, before we go deeper into that story, we have a new partner, restream. You saw it in the intro. We're very excited to welcome them as a sponsor of the show. We've been running on Restream secretly, secretly, secretly.

Since day one.

Since day one. It's been our secret to success. And it can be your secret to success. Yeah, if you have a lot of.

The Fortune 500 already for all their live streaming infrastructure. So if you need to go live.

Go live with Restream and you see OpenAI do streams like that. That is a thing that has entered the standard comm strategy.

I don't know if XAI is running.

On Restream, but they do, but they do stream. And you know, it used to just be the hyperscalers, the Mag 7, the big, big companies that would do a live stream around a. An annual event. Then Brian Chesky came out and created Founder mode and basically said put all your announcements on an annual release schedule, bundle them all up, have the team celebrate, push to get across the finish line. And part of that is let's throw an event. Part of that is let's get the CEO and the product leaders on stage, tell the community about what we're building, what we've buil. And now a lot of people are streaming it. If you're going to stream your stuff, you need Restream anyway. Go check it out. So Dylan Field, someone who has streamed events, you know, Figma obviously has config their dev event annually and they actually do two of them and those are streamed all over the place, probably using restream and Dylan chimes in on the OpenAI news. He says congrats to the to congrats 2025 IMO winners and participants including OpenAI who trained a general purpose reinforcement learning model and achieved IMO Gold. OpenAI team includes these two folks, Cheryl and Polynomial Gnome Fun Fact Polynomial also won the 2025 Diplomacy World Championship as a human and so people are saying AGI is right around the corner. AJ from Semianalysis says what stands out to me scaling RL and non verifiable rewards likely via rubrics and LLM as judge and reasoning for several hours at a time for a highly specific task. This is what the $20,000 per month month model will look like. And so there were a bunch of interesting things about this. Apparently there are a few different ways to tackle math IMO level math problems. One is using this, this program called Lean that is a formal math proof verifier and there's rumors that Google's building their system to leverage that. A lot apparently went way, way down just the text based LLM reasoning path and had a lot of success there. So even though as we'll get into in this story it feels like it's neck and neck, maybe Google was earlier and they just didn't release it faster and maybe this is a comms thing. There's a whole bunch of different stuff going back and forth. But potentially the more interesting thing is did they take different technical approaches and get similar results? If they're neck and neck, which one will scale better, which one will generalize better? There's a whole bunch of different just discussions there. Do you want to get into some of the controversy or the pushback? Do you have a post pulled up from someone saying about Google's attempt?

Yeah, I mean the whole thing came down to timing. Dennis over at DeepMind was pushing back a little bit, which I can pull up.

Dennis Hassabis from DeepMind, the co founder and now at Google. Yeah, it was interesting because we immediately jumped to polymarket because this whole idea of a an AI system beating the imo. We talked about this with Scott Wu from Cognition months ago back in April. I found the clip and he said I would be very surprised if an AI system this year does not, does not actually surpass the imo. He of course is an IMO gold medalist himself and so it was big to hear from him. Now the poly market had been sitting at like 20% and I think we're going to get into like the minutia of like what it means to truly Win because the imo, like the group that actually puts on the competition, has a different set of rules. They don't. They put the questions out and anyone can go and try and do them. But to actually be awarded a goal for an AI, they said it has to be an open source model, potentially it has to be released in this particular way. And so OpenAI might have used their systems to solve the questions, which is super impressive, but they might not have checked every box to actually technically win the gold from the actual organization. And so Gnome says today, Demis, to be clear, so.

So he put out a post this morning saying official results are in. Gemini achieved gold medal level in the international mathematical, which is different than actually.

Getting a gold medal.

Right.

He said an advanced version was able to solve five out of six problems. Incredible progress. And there's a quote in here from the IMO president. He said, we can, we can confirm that Google DeepMind has reached the much desired milestone, earning 35 out of a possible 42 points, a gold medal score. Their solutions were astonishing in many respects. IMO graders found them to be clear, precise, and most of them easy to follow. You can imagine Juan is just like off on this insane.

There were some fascinating, fascinating details around.

How they solved Demis shared, he said. By the way, as an aside, we didn't announce on Friday because we respected the IMO board's original request that all AI labs share the results only after the official results had been verified by independent experts and the students had rightly received the acclamation they deserved. So you can imagine somebody was saying, I don't know how real this was. Somebody was saying that they were making a claim that like, Google needed time to like get the comms approval to sign off. But it seems more likely that the original plan was, hey, let's wait and announce this when, when the IMO board actually comes out.

The analogy feels like you're a super fast sprinter. You go to the Olympics and you run 100 meter dash. That's incredibly fast in the parking lot. Like you're not in the stadium, but everyone's like, wow, that guy's fast. He's hauling, he's hauling. It's very impressive, but still potentially the.

Fastest man on earth.

And, and ideally you paid for the parking pass to get into the stadium. And it's like Maybe Google and OpenAI were both running 100 meter dashes in the parking lot. One of them had paid for like, you know, enough parking spaces to make it completely clean. OpenAI was just kind of showing up with a buddy and being like, I'm sprinting. That's the vibe I'm getting. But they're both in the parking lot. Like neither of them are in the actual stadium at this point. But you know, who knows, they might be soon. There are some questions about do they have to open source or whatever. So Noam Brown says, you know, we achieved a milestone that many considered years away. Gold level performance. Gold medal. Of course he's saying gold medal level performance. Like that's very critical. It's not we got a gold medal is that we exhibited gold medal, gold medal level performance. Typically for these AI results like in Go Dota Poker diplomacy, researchers spend years making an AI that masters one narrow domain and does little else. But this isn't an IMO specific model. It's a reasoning LLM that incorporates new experimental general purpose techniques which would be very exciting because you could apply an IMO level.

John, hold up that stack of posts for the audience.

Oh, we got, we got 100, we got a tech ready. It's going to be a quick stream people. It's going to be a quick stream, but buckle up. So what's different? We developed a new technique that makes LLMs a lot better at hard to verify tasks. IMO problems were the perfect challenge for this. Proofs are pages long and take experts hours to grade. Compare that to AIME where, which is another math exam where answers are simply an integer from 0 to 9 so you can verify them really really quickly. Also, this model thinks for a long time. Oh one thought for seconds. Deep research for a minute this one. Thanks. For hours. Importantly, it's also more efficient with its thinking and there's a lot more room to push the test time, compute and efficiency further. Now what's interesting is like these, I think these, I think these questions go up and then you have like, like, I think the students get like two, four and a half hour sections, segments. So there's like there's a world where the AIs can do it, but just not as fast as humans, which would be very interesting. I don't exactly know how close they are. So Gnome says where does this go? As fast as recent AI progress has been, I fully expect the trend to continue. Importantly, I think we're close to AI substantially contributing to scientific discovery. There's a big difference between AI slightly below top human performance versus slightly above. This was a small team effort. He took a.

Right now the discovery that that AI is doing is talking with somebody who is potentially schizophrenic and convincing them that they figured out how to move faster than the speed of light.

Yeah. So there's, there's so many accounts of that. It's, it's, it's getting crazy. Yeah. I posted ChatGPT agent, go win me an IM go gold medal and then update my resume.

Don't make mistakes.

But yeah, I mean, have you actually looked at any of the IM questions ever? No, I don't even know where to start. Like it's all stuff that it's one.

Call away for us so we just text Scott.

Tyler, have you ever looked at IMO questions?

No, but I'm looking at them right now. It's like pretty brutal.

Would you know where to Tyler, if.

You can get two out of six right before the end of the stream, we'll buy you a house.

Well, there's the thing. He's just going to be able to ask ChatGPT1 shot it. Yeah.

Yeah.

I had no tool use.

Got to qualify that.

So yeah, Nick has the, has the comment about DeepMind. DeepMind has.

Tyler has another challenge today.

By the way, we will introduce that in just a few minutes. D mind has also won IMO gold, but they haven't announced it yet. By the way, confirmed. Congratulations.

They did this morning.

They did. Yeah, they confirmed it this morning.

And so it's been interesting to see the reaction to this. Gary Marcus is on X saying all the tech bros this morning thinking that AGI has been achieved because some parentheses insanely expensive new form of LLMs can now match top high school students on one specific task. It's almost cute. So really just ripping into the entire AI community. But Will Depew comes in and says he's on leave right now, taking a little summer holiday. He says, guys, stop using expensive as a disqualifier capability per dollar will drop 100x a year. The 3k task arc AGI 80% could probably be $30 if we cared to optimize it. Repeat after me. All that matters is top line intelligence. All that matters is top line intelligence. So yeah, again, everybody's focused on sort of raw capabilities, not super focused on efficiency, especially for projects like this where they're not necessarily, you know, rolling this out in mass.

Yep. You know where a lot of these IMO gold medalists go to work?

I do.

John ramp.com times money save both easy to use corporate cards, bill payments, accounting and a whole lot more all in one place. Ramp. So the timeline was in turmoil over the Gary Marcus post and they went back and forth and he finally admitted it at the end, he said. He said, that's impressive. So it started with a post by Daniel Litt. He says, Huge congrats to OpenAI for their IMO Gold. I don't find it too surprising that an AI tool was able to achieve this. See below though I'd sort of lost hope the last few days, but I'm, but I'm pretty surprised. It was an LRM with no tool use. So it didn't have Python, it didn't have web search, it didn't have all the things that you, that you could imagine would kind of allow you to, you know, speed things up or kind of be a shortcut that would put it in a different category. There was always a question about with the when the AIs were playing video games, like computers have wall hacks. Like they can sometimes like see the map without the fog of war. They can see like it's very different to get perfect, pixel perfect data on where every character is on the map and be able to make decisions based on that as opposed to like looking at pixels and having to like move the screen over there to see if you're being invaded on the left flank, right. And so there was always this like yes, if you gave a computer like the RAW access that's not quite, it's impressive but it's not quite a level playing field. And it's basically the same thing as being like well, the other folks don't get a calculator and you get a calculator like is it that impressive? But this was very much an even playing field it seems like so Mel Gibson 2.0 great name. Says you are not surprised that they you are not surprised that they have figured out a way to make the model learn in very hard to verify domains in a fuzzier reward space. It also appears that the same reasoning model was the one used in the 80 coder competition, showing that it can generalize across domains. Daniel comes back and says I don't think we really know that what they figured out yet. I'll keep my powder dry until we know more. Mel Gibson says Go reads Gnome Gnome Brown's thread on the model if you haven't already. He says and remember Gnome Brown says this isn't an IMO specific model and that's a very important thing because there's a ton of, there's a ton of situations where you can go in RL on a specific task and it gets really, really good at it. But then you try and get it to do anything else and it's not that great. And so Daniel, it says, yeah, I read it. Gary Marcus says, chimes in from the rafters, as he's not in this chat yet, but he has entered the chat. He says, I read it too. That's pretty vague. Are we sure that no tools were called? Gary Marcus then chimes in again, says, re IMO gold. Does no tools mean no use of python, code interpreter, etc. Dan says, Daniel Litt says, that's how I understand it. And then Gary Marcus tags in polynomial Gnome Brown and also Alex Way and says, can you confirm the IMO gold was achieved without using Python or code interpreter or similar? And then Gary Marcus says humans don't use external tools in the IMO competition. I'm just trying to understand what the system is. That's how we do science. Daniel Litt and at some point, whoever was trying to dunk on Gary Marcus just deleted their account. So it just says this post is from an account that no longer exists. Daniel Litt shows Cheryl on the OpenAI team saying the model solves these problems without tools like Lean, which is a math verifier, or coding. It just uses natural language. It also only has four and a half hours. To answer our earlier question, we see the model reason at a very high level, trying out different strategies, making observations from examples and testing hypotheses. And Gary Marcus says, that's impressive. So Gary Marcus, you got to fill out the apology form, bro. You got to fill out the deep reinforcement learning apology form. What was the reason for your behavior? No one told me Alex Way was training the model.

Mercury was in retrograde.

Mercury, I don't know. ML and then one of them is Gary Marcus convinced me it was fake because Gary Marcus famously said, deep learning is hitting a wall like this. This particular paradigm will not scale. You have to do symbol manipulation, which was like kind of his bet. Encode different relationships between ideas in the model. There were a bunch of different debates over there and he's gone back and forth on that. He's not as much of like a deep, deep reinforcement learning hater as some people think, but he's got. That's his brand at this point. So anyway, more on the drum between DeepMind and OpenAI. DeepMind got a gold medal at the IMO on Friday afternoon, but they had to wait to approve the Tweet until Monday. OpenAI. OpenAI shared theirs first at 1am on Saturday and stole the spotlight. In this game, speed is greater than bureaucracy. Miss the moment, lose the narrative. So that's one interesting take.

Yeah.

I think there's a little bit of that. I mean certainly culturally OpenAI loves to release information before Google, like whatever. Google's like next big and they're playing to win.

Remember they. After. What did they release after the deep seek moment.

Deep research.

So if you're typing deep and you were looking for an AI product and then.

And then Google had IO and then. And then OpenAI. OpenAI IO this is like the day.

Before with Johnny I've.

With Johnny I've. So anyway, you know, you'll have to see it. All's fair and love and war. I guess I don't think this crosses anything.

Warfare, violence.

And this is the benefit of having a bunch of posters on your team, I guess. Gnome Brown says it takes us a few months to return the experimental research frontier into a product, but progress is so fast that a few months can mean a big difference in capabilities. And this is from July 18th. So all the models underperform humans on the new IMO questions and Grok 4 is especially bad on it. Even with best event selection.

Somebody was accusing Grok4 of training on the problem set for.

For imo.

No, no, no.

Oh, for something else. Oh for the other benchmarks. I mean benchmark hacking is like a thing that happens all over the place. That's why we need less like hidden benchmarks. That's the whole thing of ARC AGI that we will talk about. We talked about will Depew saying guys, stop using expensive as a disqualifier. Terence Tao, one of the. One of the most goated mathematicians of all time, digs in a little bit and is. Is adding his perspective as a. As a world renowned mathematician. He says it is tempting to view the capability of current AI technology as singular quantity. Either a given task X is within the ability of current tools or it is not. However, there is a. There is in fact a very widespread in capability several orders of magnitude depending on what resources and assistance gives the tool and how one reports the results. One can illustrate this with a human metaphor. I will use the recently concluded IMO as an example here. The format is that each country fields a team of six human contestants who are high school students. Students led by a team leader, often a professional mathematician. Over the course of two days each contestant is given four and a half hours on each day to solve three difficult math problems. Given only pen and paper. Crazy. No communication between contestants or with the team leader during this period is permitted, although the contestants can ask the invigators and I don't know for clarification this is like some math term, I don't even know for clarification on the wording of the problems. The team leader advocates for the students in front of the IMO jury during the grade.

Invigilators are also known as exam proctors.

Okay. Proctors. Okay. So yeah, they can ask the people that are running the test, hey, we.

Gotta start working invigillator. And for sure we should probably have a shot like just an invigilator too.

For sure.

You should have at least one on staff.

So. So. So. The team leader advocates for this for the students, but is not involved in the IMO examination directly. The IMO is widely regarded as highly as a highly selective measure of mathematical achievement. For a high school student to be able to score well enough to achieve a medal, particularly a gold medal or a perfect score. This year the threshold for gold was 35 out of 42, which corresponds to answering five of the six questions perfectly. Even answering one question. Perfect.

That a B, basically. A B. Yeah. Gets you gold.

Yeah.

It's that hard.

They're that hard. But consider what happens to the difficulty, to the difficulty of the Olympiad if we alter the format in various ways. First, one, gives the students several days to complete each question, rather than four and a half hours for three scenarios. To stretch the metaphor somewhat, consider a sci fi scenario in in the student in which the student is still only given four and a half hours. But the team leader places the students in some sort of expensive and energy intensive time acceleration machine in which months or even years of time pass for the students during this exam.

Before the exam, Ben might have put us in one of these as we go live and then it's four hours later.

Time acceleration. Yeah, it's wild. Two, before the exam starts, the team leader rewrites the questions in a format the students find easier to work with. Three, the team leader gives the students unlimited access to calculators, computer algebra packages, formal proof assistance, textbooks, or the ability to search the Internet. Sounds like that didn't happen. The team leader has the six student team work on the same problem, simultaneously communicating with each other on their partial progress and reported dead ends. The team leader gives the students prompts in the direction of favorable approaches and intervenes if one of the students is spending too much time on a direction they know to be unlikely to succeed. Next, each of the six students on the team submit solutions, but the team leader selects only the best solution to submit to the competition. Regard, we're discarding the rest. Last, if none of the students on the team obtain a satisfactory solution. The team leader does not submit any solutions at all and silently withdraws from the competition without their participation ever being noted. Oh, so that, I mean that could have happened. It didn't in this case. But in each of these formats, the submitted solutions are still technically generated by the high school contestants.

Going back to the parking lot example, it's like if you can go run, go run the race in the parking lot and then if you don't run as fast as you hoped, it's like, well, I wasn't in the. I mean, I wasn't racing, I was.

Just going for a jog.

I was just warming up. I just. I like to be surrounded by excellence.

Yeah, don't worry. Yeah, I mean, that is the true one. Because if they'd missed, they'd probably be like, what? The IMO is this weekend we're working on agents. Like, what are you talking about? Like, have you seen our daus? Get out of here. To each of these formats, the submitted solutions are still technically generated by the high school contestants rather than the team leader. However, the reported success rate of the students in the competition can be dramatically affected by such changes in format. A student or team of students who might not even reach bronze medal performance in the normal competition, understand your test conditions, might instead reach gold medal performance under some of the modified formats indicated above. So in the absence of controlled test methodology that was not self selected by the competing teams, one should be wary of making various apples to apples comparisons between the performance of various AI models on competitions such as the imo, or between such models and the human contestants. Yeah, the question is like, it seems like of those, not many were actually violated, by the way. OpenAI and Google attack. Yes.

Well, this was before. I mean this, this was an immediate reaction before. I think there was more details came out of. Details. Yeah, yeah, that we've been covering.

This was 3:00pm on. On. So this is so on Saturday, but 3:00pm later.

Well, Rune has a good post, says my bar for AGI is an AI that can learn to run a gas station for a year without a team of scientists collecting the gas station data set.

That.

It's a great post.

The world isn't ready for gas stations.

We can run an AI as long as we can. We can run a gas station as long as we have, you know, a perfect, you know, data set.

And that's. Yeah, that's like the way on how gas stations operate. Yeah. As soon as we place the goals somewhere, we nail that goal and then we have to move the Goal post shows.

This is so.

This is a great graphic picture.

So Thomas Wolf says my bar for AGI is an AI winning a. Nope. Prize for a new theory. It originated image of a team moving the goalpost.

I do like how consistent Tyler Cowan is on AGI. He's like, my bar was the Turing test. We passed that. So I'm. I have to call it like it is.

Yeah.

And I'm not moving the goal.

I think, I think he's notorious.

Yeah.

Fullness of time.

I mean, I, I think it's okay to be like, yeah, we achieved AGI.

It's just tough when you achieve something.

Now there's something new that's cool but.

Not immediately dramatically transform.

But you have to imagine that that was the same. That there was the same thing when people were like human, human flight. And then the Wright brothers go and do it and people are like, great. So like I can hop on a Southwest flight in an hour to go to San Francisco from la and they're like, what are you talking about? Like, LA is like, you know, a couple trains and like some people, you know, with orange groves. And it's like it takes a long time for the infrastructure to get built out. Like now a lot of these AI tools, they do build on top of the Internet and on top of technology that we have rolled out. But like at a certain point, like the UI matters, the economics of these different tools matters. Like, you can't. It might not be economical for a company to release an AI tool that costs $20,000 to inference just to get you the weather or do two plus two. And so all of these things take time to roll out. And that's why we're kind of in this like slow takeoff scen. It feels like, I don't know, we're going to go to GPT agent in a little bit and talk about that. But there was some interesting, there was some interesting details from Jasper about the IMO problem. 1. So these are on GitHub so you can actually just go read them. The math checks out. It nailed the key lemma for n greater than 3 any. But you know, you can kind of read them these things. But the write up is kind of messy. It uses shorthand and sentence fragments. It introduces new terms without definitions, forbidden and sunny partners lists. And it does a bunch of different interesting things where it's trying to condense down the number of tokens. So if it wants to say like divide this number by three. Normally by space, three would be two tokens.

But.

But buy three is One token. And so it's like compressing it down. It's kind of like learning its own language or developing its own language just to be a little bit more efficient. Very interesting. Deja vu. Coder says you're laughing. OpenAI and Google DeepMind's unreleased models just gave an IMO gold medal performance without Internet access and you're laughing. People love it.

It is interesting that XAI didn't get get involved in this one. You have to imagine they knew both DeepMind and OpenAI were going to be competing.

Yeah, I mean it feels like it's, this is the nature of like the value of research. It's like a new research domain or new research path. Like XAI is dominated in the pre training. Build the big super cluster, get all the GPUs, do the massive pre training. Then now with Grok 4 they're doing equal spend on RL generally. But DeepMind and OpenAI seem to have like the next next thing already cooking and we just don't know the phrase for it yet because in order to learn that phrase you got to pay $100 million to some researcher to come tell you. Basically, yeah, like I'm sure that there is a, there is a new paradigm, a new strategy to unlock this level of performance and it's something with probably.

Not 100 million, some combination of a bunch of people that totally collectively cost about.

But eventually, I mean we were seeing this with the behemoth analysis from the semi analysis and what Jeremy was diving into where like even just when you do, when you train a really big model, how you chunk up the memory, how you chunk the attention, like that has a dramatic impact on the actual end state and the model performance. And so it seems like Google and OpenAI have been able to like hammer out a bunch of those problems to get a really great result and they're not going to tell anyone anytime soon, I imagine. Sebastian has a good post here. He says it's hard to overstate the significance of this. It may end up looking like a moon landing moment for AI just to spell it out as clearly as possible. A next word prediction machine. Because that's really what it is here. No tools, no nothing. Just produce genuinely creative proofs for hard novel math problems at a level reached by only an elite handful of pre, pre college prodigies. Pretty sick.

It is funny explaining it to somebody just on the street a computer beat some high school students at math. They would tell you okay, cool, yeah.

Okay, cool, yeah, cool story.

And then when you phrase it it like this, it's like, okay, but it's like landing moment.

But yeah, yeah, but like the high school IMO folks are still in the 99.99 9th percentile for just everyone. Like, they happen to be age. There happens to be an age cutoff. But it is essentially arbitrary since most people cannot do this level of math. Well, let me tell you about graphite.dev code review for the age of AI. Graphite helps teams on GitHub ship higher quality software faster. You can get started for free.

Free.

Francois Ch says, intelligence isn't a collection of skills. It's the efficiency with which you acquire and deploy new skills. It's an efficiency ratio. And that's why benchmark scores can be very misleading about the actual intelligence of AI systems.

And Elon's there saying, good perspective. And so, yeah, we don't need to move the goal posts, but we kind of. It's still interesting too, because we achieved this particular benchmark. Now, what have you done for me lately, basically? And so I think people really will move the goalposts and you can kind of knock it. But at the same time, it was interesting timing that Last week ARC AGI V3 dropped. These models are not doing well on that. And they're simultaneously doing well on something that is so hard, you show it to most people and they don't even know where to start with the math problem. And it's completely. It's completely ridiculous to try and even take a shot at. So we wanted to give ARC AGI V3 an attempt. Are you ready, Tyler?

Yeah, I'm ready. I haven't looked at it at all.

Okay. So. Yeah. How familiar are you with ARC AGI generally?

Yeah, I mean, so I played with like the version one, like last summer. I was trying to, like, there's the prize. I tried it out for a couple weeks.

Did you try it out as a human or did you try writing software for it or.

Yeah, like, trying to like, write software for.

Okay, yeah. What was your approach?

I don't know. There's a bunch of stuff that was just static images. Right. Then there was ways you could tokenize it to try to do normal lm. And then I think I was trying some random image processing.

Did you get any of them right or was it like a straight up zero?

Yeah, it was like pretty bad.

Well, the initial Arc AGI V1 V2 V1 was pretty solidly solved by 03 high, I believe, or 01 high.

I think 01 was the one that solved that 01.

And it was like $2,000 a task and they got to something like 60, 70%. Still not where an average human is for these puzzles, but impressive. Then Grok 4 came out and Arc AGI v2, they pumped the score from like what, 6% to 15% or something. They doubled it, but still not great. But now Arc Agiv 3 came out and it's more of a game. So I played it on stream on Thursday, I believe, maybe Friday. Mike, I did pretty well. I actually have some times from.

You did well. Especially considering you had like a minute to look at it prior to joining the show.

I really was pretty fresh.

Yeah, you basically tried it for a second. You were like, this is really hard.

Yes.

And then had to stop.

Okay, so you want to put it in hard mode, right? So I got, I got times from Mike on how humans have done on. On Arc AGI V3. So you have to finish the whole thing. And he says if he only gets one shot to speedrun all three games with unlimited try agains, I'd say 15 total minutes is hard, 20 minutes is medium and 25 minutes is easy average. So you want him to do it in 15 minutes?

Under 15.

Under 15. And what does he win?

A new iPhone.

New iPhone?

We bricked your iPhone.

Brick your iPhone.

We made you sign up for Google Glass. So this is an opportunity to make it all back.

This is high stakes. So make sure you have it set up and you should pull up the instructions and read the instructions and we'll start a timer when you. It's 11:36 right now.

Is there. Wait, so there's no like hard mode, easy mode?

No, no, no, no. So everyone does the same thing whether you're an AI or a human. You will see some instructions, there's some keys. It's more like a game than filling out the puzzles and the actual instructions. There are really no instructions. It's your job to figure out the nature of the game, figure out how it works and then go and speed run it. Okay, so, so, so pull it up and I will read this this post. So Ludwig Yeatgenstein says:20 years ago this type of person would become an elite math professor. Now they're making AI breakthroughs. This is progress, probably. And it's the. The CV of Alexander Way, the research scientist over at OpenAI. He's a member of the technical staff and before he has a PhD in computer science from UC Berkeley, he was at Harvard studying computer science, Phillips Exeter, totally tracked research, intern at Meta, Microsoft, D.E. shaw, Google. What a stacked resume. This Guy's a killer. And now he. He was on the IMO team that did that. And so we talked about the. The plot twist and. And the battle between OpenAI and DeepMind. I'm sure both of them will be talking about this. I'm more interested in the underlying technology and. And what actually happened and what was discovered then. Then who released it earlier. That doesn't matter as much to me. And. And again, Gnome says, hey, like we posted after the closing ceremony, it was live streamed, so this is easy to confirm. We weren't in touch with imo. I spoke with one organizer before the post to let him know he requested we wait until after the closing ceremony ends to respect the kids, and we did. And so respect. It seems like this is pretty verifiable that they didn't do anything. Anything bad to the IMO kids. They might have. They might have roasted Google, but that's just on, you know, Google's gotta start posting at 1am, so. Tyler, give me the update. Do you have it pulled up?

How are you?

I'm ready to start.

Okay, production team.

I got a timer.

You got a timer. Okay, so. So I want you to count up because we want to get the total amount of time we're aiming for under 15 minutes for all three puzzles. That's hard. 20 minutes would be normal. 25 is like easy mode, apparently. This is according to Mike. So I. I haven't actually tried to speed running this time. I said a couple.

Three games or just one of them.

I played one of them.

Didn't you get through two?

Well, there's levels to this. There's levels. You'll figure all this, Tyler.

You. Wait. So I say. I say you get unlimited lives in the game.

Yeah, you can reset and do whatever.

But you can spend the whole show trying to do. Do all three games in 15 minutes if you want.

Oh, well, no, no, no, no. It doesn't work like that because once you know the tricks, like, you'll be able to blast through it. Yeah, for sure, for sure. Let's go for 15. Good luck. New iPhone on the line. If not, this really is the. I'll come up with a new.

I have to say this really is the iPhone moment for ARC for benchmarks.

This is the iPhone moment for ARC AGI. Good luck to you. Tyler. Let's. Okay. 5, 4, 3, 2, 1. Go. Tyler. Okay, we will let him work on that and we will check in.

All right. Should we talk about ChatGPT agents?

Yes. Yes. So there's a bunch of.

Before we get before we get into.

That, let's tell you about Figma. Think bigger, build faster. Figma helps design and development teams build great products together. You can get started for free@figma.com what did you want to say?

I wanted to say the some big news that hit the timeline this morning. Polymarket has acquired QCEX for $112 million and that is going to allow them to enter the US market. As you may know, to date you haven't been able to use polymarket if you're in the us we use it as a data source.

Yep.

So you can go on the website, obviously you can see on the ticker, but this is going to allow them to actually launch in the U.S. so Shane has been absolutely cooking and I think he's going to come on the show soon now that he's not under DOJ investigation.

What a wild. What a wild founder story arc. So. And like, yeah, remember those old photos of him like coding in a closet and stuff?

Like in his bathroom.

In his bathroom.

He built polymarket in a, in a, in a studio apartment bathroom with scraps.

I love it. Anyways. Vanta Automate Compliance. Manage risk, Improve trust Continuously. Vanta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation, whether you're pursuing your first framework or managing a complex program. Join the purple llamas@vanta.com so OpenAI has been doing a lot of stuff. It's, it's, it's, it's. They're kind of on a run. So IMO gold medal happens over the weekend. The ChatGPT agent which will, which I tested finally. So I'll have a little review and some analysis of that. But also apparently they are testing a new model on the web Dev arena on LM arena under the name Anonymous Chatbot0717 so this is a secret project from OpenAI that they are maybe putting out in the, in the test realm and then people discovered it and they're digging into it and so I can't believe I'm going to say this, but it's genuinely at a completely different level of front end coding. Far better than Sonnet 03, Gemini 2.5 Pro or Grok, which is saying a lot because people love those models for coding, especially on the front end. So to test it I ran a great prompt borrowed from the amazing feature Crew Pod YouTube channel, asking models to create a procedurally generated planet with three js. Take a look at yourself. I'm pretty astonished by how Big the jump is I have featured the new model twice because its implications have been so interesting. Of course, this is only one test, but OpenAI models have always been a bit meh on front end work and they seem to have finally overtaken everyone else. So this apparent alleged unreleased OpenAI model model is building this procedural world and it generates this really cool 3js, like basically a video game in and it sounds like it kind of one shots it so exciting stuff from that. There is so the agent stuff. So SWIX says please stop making flight booking agent demos with faint but dying hope the undersigned. And so everyone is saying, he's saying please stop making fetch fetch. Please stop trying to make fetch happen. Flight booking, instacart or instacart orders, astroturfing, Reddit. He wants killer use cases to be coding agents, support agents and deep research. Up and coming is screen sharing, outbound sales, hiring, education, personal AI and finance. And I have to, I think we're somewhat responsible because for months we were just like, just book a flight. Like, this is the killer demo. This is the killer use case. And then everyone was.

It got played out. We did stop asking that.

We did stop asking you, but at.

The beginning of the year we were very curious.

Yeah, yeah, yeah. And so Sam Altman had a post last week. He said, today we launched.

Can you give some background for people.

That may just be Sam Altman crawling.

Out from under a rock?

Yeah. So he started a company, he was the president of Y Combinator, started a company called Looped, sold it Stripe. Yeah, yeah, that's what people know him from. Koenigsegg owner, McLaren F1 owner. That's how. That's where people probably know him from. But he says today we launched a new product called ChatGPT Agent. Agent represents a new capable new level of capability for AI systems and can accomplish some remarkable complex task for you using its own computer. It combines the spirit of deep research and operator, but is more powerful than it may sound. It can think for a long time, use some tools, think some more, take some actions, think some more, et cetera. We, for example, we showed a demo in our launch of it preparing for a friend's wedding, buying an outfit, booking travel, choosing a gift, et cetera. We also showed an example of analyzing data to create a presentation for work. Although the utility is significant, so are the potential risks. We built a lot of safeguards and warnings into it and broader mitigations than we've ever developed before. And from robust training to system safeguards to user Controls, but we can't anticipate everything. In the spirit of iterative development, we are going to warn users heavily and it gives users freedom to take actions carefully if they want to. And so very, very interesting post. He goes through a bunch more stuff, but I actually tested it out this week, weekend. So, as you know, we have a studio, the TBPN Ultradome, here in Hollywood. We're always thinking about what's next, always looking at what else is on the market. And when we were first looking for the TVPN ultradome, I went to ChatGPT operator and I said, hey, I'm looking for a studio. It needs to be a sound stage, which means padded walls. I want concrete floors. I want it to be, you know, at least 40ft by 40ft. I need 1 gigabyte Internet access there. I need power ac. It can't just be a, you know, heavy industrial space where there's going to be somebody next door Hadrian making CNC stuff that's rattling me. I need a bunch of things. Go and figure it out Operator. And it went around and it looked at LoopNet but it kind of got stuck on, on Captchas and it wasn't doing a great job. And I kind of. I felt like I had to be like, like there watching it. And it was really only working on desktop for me at the time. But I reran basically the same query asking it to look for studio space in LA with all these different metrics that I wanted, all these different parameters that I wanted to make sure it verified. And it found a place that was like not listed on any sort of like rental sites. It found just a website that was out there for, for some crew that runs like two studios, one in Hollywood, one in Van Nuys and gave me the full breakdown. I was able to go to the website, but then of course I'd have to call and I called and they only are available between 9 and 5.

Well, this is what I was asking the agent team, I think it was on Thursday that we had them on, which was what have you done for me lately? Well, basically adding a voice feature just to now call them to confirm that it's actually available and let me know, find out when I can go see.

See it.

Yeah, that is a very much easier step to add in many ways, like.

The browser use functionality and also like the, just, just the asynchronousness of it. Like I fired off this query, I don't know, like, like after, like, I don't know, 6pm on a Saturday, like it's, it needs to wait until 9am on Monday when they open to pick up the phone.

When we started this section I put in an agent request said find me a GT3RS with under a thousand miles for under 400k that's currently for sale in Los Ang Dallas County. And so it's doing, it's cooking. I'm watching it cook and it's doing like an hour plus of work already in like five minutes.

Yeah, no, no. You know it is remarkable. So it can, it can really work through different, different sites and features.

You can see it, it found one for in LA county.

Yep.

With 1300 miles under 400k. And I just watched it say this one will disqualify this one because. Because it's pretty cool. Right over that limit.

Yeah. And so it needs to be able to think about time and it needs to think, okay, when is the certain time to do this? I need to come back to this in two days or something like basically set a cron job, set an alert, come back to things. It needs to use Linear, a purpose built tool for planning and building products. Meet the system for modern software development, streamline issues, projects and product roadmaps.

Linear app OpenAI runs on linear linear.

Supports agents and they actually have an integration which I am excited to about. But I mean jokes aside, SWIX was saying this is a new frontier model and I think that a lot of people get hung up in oh, there's a new branding around it agents. Like I was talking to our buddy David center about this and he was like what are you using it for? Like what are your prompts? And I was like it's the same prompt. The thing that I would send to Deep Research I send to agents now it's just better. And I, and I, and I do understand that like it might unlock some new use cases but I think the basic thing that you should start with if you're looking to experiment and kind of road test ChatGPT agent for yourself is just take something that you would put in Deep Research, throw it in agent and see how you like it and if it's better. And I noticed that it's a lot better in many ways. It gives you a MD file at the end. Basically it gives you a document like a report that renders the exact same as a Deep Research report but you can kind of scroll it differently. You can still do all the tables and stuff. The links oddly weren't working. I would click on it and it wouldn't open. So there's still Some bugs and some rough around the edges stuff. But in general I think it did a lot. I just think of it as deep research is kind of. It feels like deep research is a little bit bound by the first level of depth on the Internet and agents can actually go, okay, there's a website out here that has something, but I need to go to the archive, I need to log in, I need to find a different route, I need to work around it, figure out all these different things. Let's actually fact check some stuff. Start building a table over here. Oh, okay. I got some conflicting information. Let me try and resolve that. Like all the natural things that when you do like actually good research. And so I think it's fine if it's just a research tool.

The thing that agent is doing here is it's going, it's navigating to a site site and then it's using the site's internal search mechanism to find what it's looking for versus just like scanning like databases or just scanning pages and pulling out information. Yeah, so it's actually taking that next order.

Yeah, yeah, it's actually using the website. And I still think like, I don't see this as really replacing anyone in my life at least or my business, but I see it, I see it very much as like, as like in between me and the folks on my team, I will go fire off a prompt and try and like really do as much of the research and kind of pre work and then hand that off and come with to somebody who works for me with a just a more educated idea of like, okay, I think we need this. I have, I've already kicked this off and I think it's possible. I need you to go into go and I need you to go do all the soft skills and I need you to actually execute this and take the last mile. That's like the human side of this. But you already have all the information to basically do like the job is well defined. And so I think that my next change to the way I use ChatGPT deep research agents will be by the end of this, give me a list of steps to actually execute this that need to be executed by a human. So with that GTRS it's like GT3RS. If they find one, it's like find me the phone number of a few and then when should I call them? What, what, what other data sources do I need to point to? What are the other considerations? So I basically have like a script to go and actually get the deal done. For example, so anyway, very exciting stuff. Almost as exciting as numeral sales tax on autopilot. Spend less than five minutes per month on sales tax compliance. Go to numeral HQ to get started. Started.

So should we talk about Stephen Colbert?

We should, we should talk about Stephen Colbert. This was a fan. This was a fascinating story that happened. The late night is ending. So Stephen Colbert, you might know him from the Colbert Report on. He was originally on the Daily show. Then he had his own show on, on Comedy Central. I believe it was half an hour, maybe an hour every night on Comedy Central. Then he got pulled over into CBS with Late Night, which was originally designed to compete with the Tonight show, which was Johnny Carson's property that he ran for 30 years, 1962 to 1992. Absolutely legendary run. Jay Leno took that show over, gave it to Conan for a year, went back. Very controversial. Late Night wars, very fun, deep dive to dig into. But Emily Jaszynski has a story in Pirate Wires all about Late Night. I have some facts here that were sticking out to me. So the high level is that the Late Show, Stephen Colbert's daily late night TV show, how many days a week is it? It's five days a week. They do 190 episodes per year. So they take a few weeks off, off in between little breaks.

Pretty casual.

Yes, pretty casual. We do. I mean we are on Track to produce four times as much content this year. But they spend $100 million to produce. We spend five times as much. Yeah, yeah, you keep pulling those daily GT3 RSS. I mean they were, they were spending how much is GT3RS? 400k. They could have bought a GT3RS for every single episode. Episode. Because they were spending $100 million to produce 190 hour long episodes every year. And so this wasn't a problem because back in 2009 they were pulling in $271 million in revenue. So I mean, and I'm sure that they were actually spending less back in 2009. So the margin on that show was immense, like 60%, 70%, maybe even higher. The problem is, is that the revenue started declining because the audience got older and ad dollars moved around the average viewer, the median viewer of the Late show with Stephen Colbert is over 59 years old. So it is effectively like a retired audience. So they're not as likely to buy expensive products. And if they buy those products, they might not stay with them forever because they might churn, because they might pass away. And so lower ltv, lower ltv. And then also It's a very broad audience. It's a very political audience. It's very general comedy, politics, daily news audience, very broad. So you can only sell broad things. And that's why you see, you know, Unilever, like, you know you're watching an ad and you, you see, you're watching the Late Show. Yeah, yeah, Verizon. And then you see Dove shampoo or something like, you just see generic products, not hyper specific like add, which is customer relationship magic. Adio is the AR native CRM that builds scales and grows your company the next level. You can get started for free. You're not going to see that on the Late Show. Yeah, you just won't. You just won't. And so revenue kept falling. Now they make less than 70 million. Reports were maybe like 60 million in revenue. Now Colbert is taking 15 home. 15 million a year home personally. And then the rest of the show costs 85 million to produce. About 50 million of that is probably salaries. There's a bunch of other costs that go into it. They have a live band, they fly people around, they have hair and makeup and all sorts of, they have to license music and they do all sorts of stuff. So, so then they went from at one point making, you know, $200 million in profit off this a year. Amazing. To losing 40 million. And importantly, CBS is going through an acquisition. And so there's a whole bunch of problems where you don't necessarily want to have this massive loss making asset sitting there while you're trying to sell the, sell the company and restructure everything. And then so there's a little bit of conspiracy. And Keith Olbermann was saying, he put on the tinfoil hat and was saying like this is all because Trump is like, Colbert criticized Trump and this is a straw man hat.

I'm going to get out the straw man.

Yeah, we're not making this argument. This is, this is. No one's really making this argument except for. Yeah, put on the straw hat, by the way. I like the straw hat, by the way. This is a funny hat.

I don't even know what kind of hat this is.

A straw, some sort of straw hat.

I gotta sit lower so that it'll fit in. Yeah, anyway, but yeah, I mean the obvious thing here is that Stephen Colbert. Stephen Colbert, yeah, whoever that is, was a zero interest rate. Phenomenal. You have a business, he's in the.

Business for 30 years.

I mean, great. Yeah. Who wants to, who wants to be the proud owner of a business that loses $40 million a year? Like clockwork yeah.

So it was a loss leader for cbs. Basically during upfronts when they're selling ads to everyone, they go to Target and they say, hey, how about we put you with Stephen Colbert? And they're like, sure, but then we also want to be on the game shows. And then the game shows make higher margins and they lose money on the, on the late show. But a company like Target wouldn't want to advertise in the game show unless they can also be with Colbert and the prestige TV that's happening there. So basically they're losing $40 million. What's interesting is that you would think that they would be able to cut back the cost and say, hey, look, we're making half, we're making half as much revenue. We need to produce this for half as much money. And so Colbert. You got to take a haircut, folks. You got to take a haircut. About 50 people instead of 100.

200.

Yeah, it was 200 people.

200 people.

200 people, okay. Yeah, 200 people. So.

And apparently so. So we were messaging some people over the weekend, some, some media moguls and they were saying, I mean, it's, it's very routine for there to be two people tasked. Like effectively, their full time job is like moving one camera.

Yeah.

And. And so effectively is just a culture of inefficiency that has a bunch of different reasons why that is, but doesn't really work in the modern era. So the interesting thing here is I was trying to figure out kind of what happens from here. So CBS is confirmed, they're not doing a replacement host. They're not going to hire like an H1B Stephen Colbert.

I think they're gonna put a talk show in there.

Yeah. So anyways, it's going away. There's no indication that abc, NBC, Netflix, Max or any, any of those platforms is like gonna try to launch one of these Johnny Carson style shows. But I think the right move for Stephen Colbert, it's not my, not my flavor of comedy entertainment, but I think he could have, I think if he went independent and effectively launched a live show podcast, something of that sort, he could continue to earn exactly what he's earning today and just do it with a much smaller team, maybe even have a better lifestyle. Maybe he starts doing the show in the morning.

Yeah. I mean, there was $60 million of like, we want to advertise on Stephen Colbank there, like flowing into this property.

Yeah.

If he provides a similar level of experience and content, he should be able to sell similar levels of Ads. Yeah, maybe not 60, but maybe 30. And then he takes home 15 and he has 15 to spend on, you know, a smaller team. And what's crazy is like during the COVID era, all of the late night shows went to like, from home episodes where they were basically doing them over zoom. And then they set up like nicer camera setups and they were able to do those shows way cheaper. But I think because of the union, it wasn't, it wasn't, it wasn't palatable or really like easy to just, to just cut back on staff and cut back on salaries or switch to some sort of more incentive pay, which is tricky. And we just talked to Delian about this, where in France there was a company that, that wanted to, like, their business wasn't doing so well. They wanted to lay off folks, too many people, and they couldn't do it. And so the government basically said shut down.

In, in France, there's policies that effectively require companies to pay years of severance. So it ends up becoming a massive liability on the balance sheet.

Yeah.

And anyways.

Yeah. So Emily just thinks Jacinti over at Pirate Wire says CBS is not just pushing out Stephen Colbert, it's retiring the iconic Late show brand altogether. That's the buried lead getting lost. Lost amidst frenzied speculation over politics and palace intrigue. If. If cutting Colbert is a bid to juice Paramount's pending merger or to punish him for criticizing it, as Democrats are now arguing because Colbert said that CBS as a company should not have, or Paramount should not have settled a lawsuit with Trump. I believe CBS just stumbled right into the future, Emily says. She says Colbert's time at the helm of the Late show perfectly illustrates the most important trend in media and culture. One might fairly wonder how Colbert, a man who leaned so far into hashtag resistance comedy he could hardly get up, has dominated the late night wars throughout Donald Trump's hostile takeover of American politics. Johnny Carson, for example, famously skewered both political parties without fear, fear or favor. Carson won the late night wars when networks faced less competition, meaning his goal was to appeal to as wide a swath of the American public as possible for the sake of selling more ads. By the time Colbert took the helm from David Letterman, late night ratings had collapsed from their high watermark. That's particularly why Greg Gutfeld is able to actually beat Colbert's ratings on a cable network, a feat that would have been unthinkable in the 1990s. Like Colbert, though, Gutfeld doesn't approach politics as Carson did. This is the New model. Cultivate a loyal niche that returns night after night, giving you an edge over others competing for an increasingly smaller slice of slices of the pie. The result is a microculture. Monocultural institutions like the Late show or the New York Times can no longer and can and and no longer do appeal more widely than their core audiences. For the Times, this is their subscriber base. And it's why, for instance, the paper committed obvious journalistic malpractice by yanking Senator Tom Cotton's infamous send in the troops op ed back in 2020 as the paper of record for all of America. That decision made no sense. Cotton was expressing a mainstream position in his party and in the country. But Times subscribers were furious, and that critical business interest shifted the outlet's editorial position. This, of course, was helped along by a staff increasingly aligned on an ideological level with the paper's narrow subscriber niche. One of the greatest sources of cultural angst stems from the ability to recognize these institutions of the monoculture have shifted to microculture. Whether they supported Trump or Bernie Sanders, plenty of Americans outside affluent urban bubbles figured that out years ago. It is the institutions themselves that often cling to the outdated brands. So blinded by their own biases. Biases. That it's become difficult for the C Suites to even recognize what's happening outside of Manhattan. And the Hamptons don't count anyway. Fantastic analysis from. From Emily. You should subscribe to Pirate Wires to go read the full thing. But lots of people were having fun with this. How does light show cost $100 million produce? That was a big question. I think a lot of it's just. This is a. This is Letterman started in the 90s. So this is 30 years of just like, yeah, let's just add one more person. Like 30 years is only adding 60 people a year. One. Oh, no, no, no. Way less. Six people a year. So you're adding like one a quarter, which is like, not that much. And I think it just kind of creeps in. But you got to be aware when revenue starts dropping, like, you have to right the ship. And you can't just. You can't just keep losing money because it doesn't give you leverage over cbs, who's trying to get out and sell the company. Buco Capital bloke says, I can't stop laughing at CBS losing 40 million a year on the Late show with Colbert there because they were spending $100 million a year to produce it. A show of a guy talking sometimes to other people. Is there a less competent group of industry executives Than linear television. Alex Roy, friend of the show says media is a business called Colbert already has an audience. He should have no problem launching a YouTube channel. I agree. Many friends run a profitable YouTube channels with millions of subs. With staffs of less than 10 people. They started from scratch. Signed a media business veteran. The left has been looking for its Joe Rogan or Tucker Carlson. Why shouldn't Colbert be that guy? That's interesting. And so a lot of people were saying it's a piece to camera slash guest interview show. It shouldn't cost that.

Tyler is back. What's going on, Tyler done?

Yeah.

I was stuck for a while on one of them. Then I switched to a different game. Finished that came back okay. But I don't know what's the timer. I feel like it was a little slow.

25 minutes. 25 minutes. Oh, brutal.

Brutal.

But you did get through it all.

Yeah. Did your hair get messed up or was it always messed up?

I was stressed out. The one game I was. I was on one level for like five minutes and then I don't know what I just like couldn't figure out. I went to a different one. Yeah, it was like I got it instantly crazy. That's brutal.

We will give you, we'll keep giving you challenges until you get a new phone.

Mike is brutal. He says 25 is average, 20 is medium, 15 is hard. Well, we will give you a new challenge, give you a new chance to win a new iPhone. The iPhone moment has arrived for rkgi. Tyler missed it this time, but what do you think it says about current AI systems? Do you think you will outperform the latest IMO level model from OpenAI? You think you still got it?

I mean, I think you could probably do rl on this and get it. I mean, I don't know if that's what the label are necessarily doing. I think ideally they're trying to just train it independent of this and then see how it does. I think it's a very good benchmark in that sense. But you probably could rl on this.

Yeah. So the way Mike described it was that you played the three public games. There are three private games. And so I don't know how much variation there was between the games. It seems like there was a fair amount because you got stuck on one, you switched to one, switch to a different one and then went back. Right. So, yeah. So I think the. I think the goal is like, if you rl on the three public games, you will be overfit against the private games. That have slightly different mechanics and you won't have learned the true generalizing function. That's the idea, at least.

Yeah, that makes sense. But I mean, I'm sure, you know, if you're OpenAI, you could just like, oh, let's just make new games that are kind of in a similar vein. And then RL on those, then you have more data and.

And then.

Yeah, I don't know, I think you could probably like. I think my general sense is that you can like RL on pretty much anything.

Yeah.

So if they really want to, and they want to throw like, you know, many millions of dollars, I'm sure they could get it, but I don't know if that's really what they want to do.

I think the current prize is like $10,000, something like that for, for something that works on it. Maybe there's a million dollar prize. There's a million dollar prize that's still hanging out there, I believe for RKGI v1.

Yeah, but those are size limited.

Right.

You can like, you need to run in a certain speed and they can't be like so big.

Yeah, I think they have to run on the KAGGLE infrastructure. Kaggle infrastructure. So you basically have to like send it your weights, your code and then you get access to like one gpu. You don't get to fire the Stargate at it. Yeah, but I don't know, it'll be interesting to see how, how, what's your, what's your timeline for a lab destroying this?

Probably end of the year.

End of the year? You think they'll be it? The whole team is like, we designed this to be unbeatable for years, like two to three years.

I don't know. I mean they just beat imo. It's like. Do you mean on a public model or them just releasing.

We've beat this private. It doesn't need to be open sourced.

Yeah.

I think if it's a private model, it's like OpenAI. It's like, okay, we have this internally, we're not going to release it. It's not safe. But we have run it on this. I think probably by the end of.

The year is like reasonable for, for specific. Okay, we, we built a model just to beat this or, or we just, we just were advancing models generally giving people what they want, making great things, solving, you know, flight booking and GPT agent and solving imo. And then, oh, as a byproduct, we solve this. You think that's what's going to happen?

Yeah, I don't Think they're necessarily going to train a model just to do this? Yeah, but they definitely could.

I mean, at this point it's getting high stakes, right?

Because like, apparently IMO is. Is not RL'd at all, like specifically for. That's what they say, math. So, yeah, I think that model is like reasonably capable, obviously.

Okay. I want to try that model on Arc AGI V3 and see how it does, because it does feel like that will be a very interesting moment where if they truly don't cheat and don't RL on this specifically without the gas station data set, as Rune calls it, you know, no RKGI V3, like creating new games, that feels like scientists collecting the gas station data set.

Right.

And they just go about their day building better and better models. And then, by the way, oh, we solve this. That. That will be very, very interesting. But I don't know. I would take the over on the end of the year. We're more than halfway through the year.

Yeah, that's true. Okay, what is your line?

So AGI, you're so AGI pilled. I think, I think this will hold. I mean, B1 is still holding. V1 is not. Yes. V1. It takes $2,000 to get like 60%. You just got 100%. You know, like you finish the whole thing in 25 minutes.

Yeah, I mean, we'll see, I guess.

I don't know.

I would take the over. I would take the over on this year. I would probably take the over on.

I have to give you an Update on. On ChatGPT agent. So it, it only found. So it gave me a really great report port, but it only found one. And it's listed by a dealer. And I can already tell the dealer, the dealer is like clearly pricing it, like not where they would actually sell it.

Got a markup. Got it. Okay.

But they did find one in Costa Mesa, so not far away. One county away.

Not bad.

399, just under the limit.

There we go.

So.

So pretty good.

Yeah, not bad. Not bad.

15 minutes. Yeah, they're. They're.

I think we gotta use it more.

But then they didn't actually get me. They don't have a link.

Okay, we're gonna have to dig into.

A more test AGI not achieved internally, but we're making progress.

Yeah.

Okay, wait, so are you saying 100% on V3 or what's like the benchmark? Cause like, if it's 100%, then like, yeah, obviously not a year.

100%. I mean, like, you got you just got 100%, never having seen it 25 minutes. Like, not crazy. I think we could literally give this to anyone over the age of, like, 10, and they would get 100% in half an hour, basically. Like, I think it's. I think it's, like, it's a very doable puzzle for humans. And So I want 100%, and I want it in. And I don't care about time and money. Money. Like, I agree with. Was it Will who said that, or.

Yeah, Will Depue.

Yeah, Will Depew said, you know, so, yeah, throw $2,000 per task on it. I'm fine with that. The question is just, can the model do it in any amount of time, in any amount of money? And it seems like right now it's probably at zero. We'll see how the IMO model does. Maybe it does really well. We'll see. But I would put it at over. I would put it up there with doing your taxes, you know, Dwarkesh mode 2027.

Yeah, I think 100% is hard, but if you give me, like, 90%, like, if it misses one of the tasks. Because on. So there's three games. They each have, like, 10 levels. If you can miss, like, one of those.

Yeah, well, I. I don't. I think you have to finish a game or not. And so I think basically, on the private data set, you can get zero. You can get one out of three. Two out of three. Three or three out of three, basically.

Okay.

But it's. It's unclear because, like, when do you tell it to give up? Because you can just keep trying and kind of brute force it. But maybe there's some sort of, like, limit. I don't know. But you get, like, unlimited retries. But maybe, maybe, maybe the real test is only giving one try. Yeah. Yeah.

We got to move on. But for your next challenge, Tyler, prompt prompt 4,07,000 times in a row and stay sane. Start recursive prompting like crazy for 48 hours straight and then come back to us.

Yeah. Well, we got to talk about the McLaren collection that's going up for sale. But first, I got to talk to you about Fin AI. Fin AI, the number one AI agent for customer service. Number one in performance benchmarks, number one in competitive bake offs, number one ranking on G2. This was what Swix said he wants. AI agents for customer service. Fin AI is delivering. So go check it out. Anyway, so the monster of this collection. This is from Tom Hartley Jr. And hopefully we can pull up the video because it is Fantastic. It's on YouTube. And yes. So here's the video. You can play some of that, that. Look at this lineup of cars. So the story here is this, this man named Mansour O. O.J. i'm not sure how to pronounce that. It's OJJ eh. I apologize to anyone who knows how to says. I'm incredibly to proud to announce that we've been entrusted with the sale of the world's Greatest collection of McLaren road cars. Offered directly from the home of one of the most influential figures in McLaren's 62 year history. He passed away age 68, surrounded by family and is now entrusting the sale of his collection. This guy is fascinating, very, very interesting fellow. So he has, let me pull up my information here. So Mansoor Ohe, he was the son of, he's a Franco, Saudi French Saudi Arabian businessman and he used his family fortune to bankroll Formula one teams and he built the TAG group. So if you're familiar with TAG Heuer, the, the, the, the Watch brand, He built that Watch brand and then wound up selling it to lvmh. And so he was born this. The. The man who collected all of these McLarens in burnt orange. Which by the way he not only is the only, is he the only one with the massive orange McLaren collection. That particular color is not available to anyone else. He has his own color that only he can buy. It's absolutely incredible. And every single one of those cars that you see there in the background, they are all the last, the last one off the line, the last individual production car that they've ever made of that particular model. And so he gets the last one he's. And you see the logo there. He puts his name on the logo. It's the only. I've never seen a brand do they. But they were so tight with him because he owned the company at one point that they would replace the McLaren logo with Mansoor logo, which is incredible. And so he is a fascinating figure. So his father, Akram Ohay, accumulated immense wealth as an intermediary in Saudi arms deals. So he basically would go to the Saudis and say, what weapons do you need? What missiles do you need? What guns do you need? What planes do you need? I'm gonna go out into the world and find those and source those and bring those back and take a cut of that.

And so yeah, this is really, I will say if your father has amassed immense wealth through arms dealing, this is really pretty much the playbook on how to put that capital to work.

It's the best. It's the best. And so Mansoor grew up in France. The family mostly lived in France. He attended the American school in Paris. He moved to California where we are studying, studied business administration at Menlo College.

Let's give it up for all the business administration majors out there. One of the top majors if you want to administrate businesses, which he certainly has.

He went to Santa Clara. He initially managed his father's ventures. He became gradually more and more involved in his father's investment vehicle which was Technique d' Avant Garde, which is Task Luxembourg registered holding company created in 1977. So the group's purchase was to broker deals between the Arab world and Europe and to handle commissions from arms and infrastructure contracts. And so his dad was born in Syria, raised in Iraq, studied in France on a scholarship during the forties, marries a French woman, stays in France and then, and then starts arranging these arms deals and infrastructure projects. And so in 1956 the dad opens an arms trading company in Switzerland and establishes Saudi Arabia's first weapons Factory. During the 1960s he represented the French defense contractors such as Dassault and Matra selling Mirage jets and missile systems to Middle Eastern clients. He was facing some scrutiny from French authorities. At one point he founds the TAG group. The group gave him unrivaled access to the Saudi royal family for defense process projects, leveraged his relationships with Prince Sultan, the Saudi defense minister. And he became so wealthy that he bought a cruise liner like a cruise ship. He bought the Queen Elizabeth II just I think as like a personal pleasure, like pleasure yacht. He bought a ranch in Paraguay, homes in Paris and Monte Carlo. And then he starts sponsoring the Williams Formula one team. And he was putting in a million dollars a year, a million pounds a year in the 80s.

When did he start doing that?

This was in the, in the late 70s. And so his son, so the dad dies in 1991, reportedly worth several billion dollars. Mansoor, who's the man with all of Those fantastic looking McLarens? Him and his brother. Mansoor and Aziz were heirs to the TAG Group Group and there was a bitter inheritance dispute between the widow and some of the children and obviously all the assets kind of, you know, get fought over. But Mansoor makes out quite nicely and winds up investing even more heavily in F1. So through tag, he starts sponsoring Frank Williams Formula One team. In 1978, when Saudi or Saudi Airlines sought to attract other Saudi brands, TAG became the team's major sponsor. In 1979, Williams actually did win multiple races Although they were having some trouble a few years ago, they were pretty good back then. They got Alan Jones and another title in 1982 with Kiki Rosberg. I think that's Nico Rosberg's father, I believe. The sponsorship gave Mansoor his first taste of Formula One success and increased tags visibility. Then in 1983, he jumps ship from Williams, goes over to McLaren. Frank Williams resisted diversifying beyond racing. McLaren McLaren's boss, Ron Dennis proposed a shareholding in proposed exchanging shareholders like shares for a new tur, for financing a new turbo engine. And so Mansoor agreed and funded the development of a Porsche turbo engine badged as TAG. So the McLaren TAG partnership dominated in the 1984 season. Drivers Niki Lauda and Alain Prost won 12 of 16 races, and McLaren secured both the drivers and constructors titles. So basically, he comes in with a bunch of money and says, you want to win an F1, I will finance a new engine. He goes and does it, and it actually works. And so he.

In 88, it's interesting because TAG and Porsche have their own partnership now. They have the Carrera line of watches all these years later.

Yeah. And so in 1984, he buys a 50% stake in McLaren's holding company. Now, remember, McLaren was founded by a true racer who died at age 31, I believe, in a racing accident. And so left this incredible legacy of just a guy putting it all on the line in racing. Like it wasn't like he was some business guy who was like, I want to get into racing. He was just a race car driver.

Well, this is the story of Enzo, too. Ferrari wasn't charted until he was in his 40s, but he had just been so obsessed over racing that it just carried into the.

The brand.

And so in a Mansour buys a 50% of stake in McLaren's holding company, TAG, McLaren Holdings. And he's still working on the racing stuff. But then he says, we got to make a road car. And the road car that he pitches is the McLaren F1 supercar that launches in 1992 and is a $20 million icon owned by Elon Musk, Sam Altman. Like, there's. They're in museums. Like they are the rarest of the rare. Jay Leno has one. One, I don't think. I think it's like the. One of the few cars Doug demuro hasn't driven. He rode in it with Jay Leno. But they are incredibly rare, incredibly special cars.

And the doors go up.

The doors go up. And so in. In other news, like, he did buy Heuer and created TAG Heuer, which was a watch company that you might be familiar with, Tag Heuer. They eventually sold Tag Heuer to LVMH for $740 million.

Elon crashed his McLaren F1.

He did, yes. And I believe that the McLaren was picked up and, and like reconstituted and wound up selling still for millions of dollars.

The car hit a hidden embankment on Sandhill road at a 45 degree angle, launching it into the air like a discus, according to witnesses.

I mean, the car goes 200 miles an hour, weighs like 2,000 pounds.

That's incredible piece of lore, imagining Elon on Sandhill Road.

And you know who else was in that car?

I do.

Peter Thiel. Crazy.

And he went to the meeting after, right?

Yeah, yeah, they were apparently, you know, the lore is that they went to Sequoia to pitch or something. They were like raising money or driving down.

What did he. For me, he turned off traction control.

Really?

Apparently, yeah.

That is wild.

Yeah, the. There's a long history of guys turning off traction control, driving really fast and then just crashing their cars. So if you want to avoid crashing your supercar, maybe keep it on, avoid turning off traction control.

I mean, that's the thing with. Doesn't the Carrera GT not have traction control? Or maybe, maybe, maybe it does, but it's like one of the few features and a lot of these guys kills.

A lot of its drivers.

A lot of these guys like, pride their, you know, they like an analog supercar.

It does have traction.

Okay. It does, it does.

Does not have stability control or other electronic driver aids that are common in modern supercars.

So Mansoor, he buys Heuer, turns it into TAG Heuer, great watch brand. Sells it to LVMH for $740 million in 1999. He gets a huge profit from that. And then he starts going deeper into all of that. He launches TAG Aviation and Aeronautics, which is distributing Bombardier jets to the Middle East. So he's a go between, between French aerospace and defense companies in the Middle east still. In this case, he develops a luxury airport in the Middle East. And he's just doing a lot of deals at this very, very high level. Probably benefits from his ties to F1. And so he builds the McLaren F1 and then of course, starts collecting and stays involved in McLaren F1 for years. And he builds this. I mean, he builds a 235 foot super yacht, Cogo. He creates a lavish house on Lake Geneva, Switzerland, known as La Maison du Lac. And kind of goes on through that he winds up getting idiopathic pulmonary fibrosis, gets a double lung transplant, and ultimately doesn't make it. But he dies at age 68. McLaren CEO is. Brown described him as a titan of our sport. Ultra competitive, determined, passionate, and sporting. And so everyone remembers him and his contributions to. To F1. But what's interesting today is that they are about to sell his entire McLaren collection, which is something like 20 McLaren road cars, all in his exact spec, which is this orange with his badging. I don't know who the buyer would be. They're probably listening to this show. But good luck out there as you go and bid on the Mansoor collection, because it is truly one of a kind. And every single car there is in the most extreme spec, and it's the last one off the line, and there's a bet and there's a plaque inside that says, this is the last F1 that was ever made. This is the last Senna that was ever made. This is the last Speedtail that was ever made. This is the last gtr. This is the last. I mean, look at that one. No, what is that? The link?

Yeah, the Elva, just fantastic.

Burnt orange Elva with no windshield.

Wow.

That is fantastic. And just every single one of those cars is insane. Also. Oh, by the way, also, every single one of those cars. Delivery mileage. Never been driven. Never been driven.

Do you think he was buying? You think he was getting two?

So the story is that he had an F1. Of course, he developed it. He had an F1, but it was kind of, I think, from like, the middle of the pack or maybe the earlier pack. And. And because he was just like, excited to get the first one and, you know, got it and probably gave one to someone else and, like, just had one, drove it, had fun with it. It was his experience. But then his brother bought. Got the last one in that. In that orange color. And he. And his brother was about to sell it, and he said, if you're. You have. You can't sell this. If you're selling it, I'll buy it. And so he buys it from his brother and winds up building an entire collection on the back of that one car. The last F1 off the line, the burnt orange, that turns into his signature color. It turns into the color that you cannot buy, no matter how much money you have, no matter how many McLarens you own.

And his legacy lives on. Oscar and Lando dominating so far this year.

Yeah, give us the update there. It's good.

You wouldn't have been Been he wasn't around to see it, but he would surely be proud.

Yeah, well, if you're looking for a compliment to your McLaren F1, get on bezel. Get bezel.com. your bezel concierge is available now to source you any watch on the planet. Seriously, any watch. Anyway, let's go to Trey Stevens. He had a massively viral post on Friday. We didn't get a chance to talk about it. There are only triggering four Tier one cities in the United States States. New York for finance, D.C. for government. San Francisco for tech, LA for media and entertainment. People. Thank you, Trey. No, other cities are power centers for aspirational talent. Sorry.

He says based on replies. Here are the people arguing for other cities. Austin, people who live in Austin. Boston. People who are really upset about Trump's Harvard attack. Dallas, Houston. Houston. People who have only lived in Texas. Miami. People who love money but don't love to work. Mogged.

Mogged.

Someone just commented for Cleveland.

It's funny because tries an Ohio guy. No, no, I agree with this. And it is incredibly sticky. Like, it's unclear if there will ever be another tier one city built in America. It doesn't really matter though, because you can go like, Warren Buffett is out in, oh, Omaha. It's fine. Like, people read this as like, I can't possibly be successful in finance if I'm not in New York. Ken Griffin was in Chicago and then Miami.

Charlie Munger lived in Los Angeles.

Yeah, yeah. And so, and so it's like, it's like it is true that these are tier one cities. It doesn't mean that you can't do something cool and amazing and achieve your goals in life somewhere else, but you should just be thoughtful about it and you shouldn't delude yourself into thinking that a tier 2 city is secretly a tier 1. If it's not, it's just not. And so stuff's moving around. Very fun post. Always good to bait literally everyone in the world, but very fun. But related to this, Delian had a post that was kind of interesting here. He said, ten years ago, it felt like Silicon Valley had this really bright, vibrant seed fund ecosystem that felt just as important as the multi stage venture firms. Now that feels totally dead, partially because the whole, a whole generation of top tier companies just completely skipped that step in raising. And so, yeah, more and more companies are saying, our seed, our $1 million will just come from our own balance sheet and it'll just come internally and we'll just fund that or we'll just jump straight to $20 million.

I feel like there's $50 million. I have a problem where if somebody raising a pre seed, I do feel like there's 20 funds that I would potentially introduce them to. The challenge is when those pre seed funds are then trying to go compete at seed, which we now know is this sort of 4 to 8 million dollars range. We had somebody on Friday who was raising 55 million.

Crazy.

Whether or not that was marketing or, you know, whatever, the name doesn't really matter.

Matter.

Bored Ape Yacht Club raised a billion dollar seed round in 2021.

That was a seed round. Was that because they just hadn't raised before?

They really hadn't raised. Yeah, yeah.

There's this weird thing where Mira Moradi.

Two or $3 billion seed round. So again like the, the, your company's expectations are based around how much capital has come in the door. Not the last name, the last, the name of your last financing round.

So yeah, there's one side of this which is like the seed moniker used to stand for time and money. Effectively. Like it used to be like, okay, if I'm a seed stage company, you know that I'm in the first 18 months of my creation and I've only, I've raised less than $5 million or else I'm in mango seed territory in series A territory. But now there's, there's some world where people just say, okay, you know, this is your second round. It doesn't matter that you've been around for 10 years. It's an A round or it's a seed round or doesn't matter if you're raising a billion dollars and it's price like a growth round. It's, it's a seed round. I don't know. I think what Delian's getting at is more just like the vibe of like you used to be able to, he's, he's a seed investor. You used to be able to hunt around, find some people doing cool stuff and get 10%, 20% of the company for a million bucks, 2 million bucks. And that has kind of disappeared.

Delian's still known to get a good, to get a good price.

He's figured out how to survive. Nature finds a way, as does public.com investing. For those who take it seriously. They got multi asset investing industry leading yields and they're trusted by millions. Folks get to public.com and we have a question.

We have Chris Black in the restream waiting room. Let's bring him in.

Let's bring him in. Chris Black. How you doing? How are you doing?

Bring me in.

What's up, guys?

How are you?

What's up?

What's happening? It's. It's great to meet you. I've wanted to chat for a long time. I actually bought your. What was that little green book that you. You made all those years ago?

Wow. We got an OG on our hand. Yeah, I know. You think, you know, it all was. I think it was 2015. I did that with Powerhouse. So thank you for being a ground floor investor.

I had that in college and I felt so cool to be a college kid that was cultured enough to buy a table, you know, Coffee table.

A coffee table by Chris.

Chris Black.

Yeah.

I think it turned into more of a bathroom book than coffee table.

Yeah, it's very small. You would like the form factor.

Sure, sure, sure. And what was inside the book?

I mean, Chris can give some more context. The thing that I remembered in a quote that I brought up a bunch was the. It was the Andy Warhol quote in there that was something like, making money is, you know, business is art, and making money is art, and good business is the best art.

You guys love that?

You guys love that? Yeah.

I mean, the book was just like a fun advice book, kind of. Basically. It was.

It was.

It did really well as a graduation gift purchased from Urban Outfitters that year. That's. That. That's. We sold a lot of books. Thank. Thank God. So it was. It was fun to do. That was very. It was very fun.

Did you go through a traditional publishing house? Like, what was the workflow? Did you write the whole book and then shop it around like, I've never published a book. So curious.

It was b. It was based on Twitter, honestly. It was based kind of on that idea of like, life Soul instruction book, the classic form factor, to use your term, meets sort of Twitter. So I went through. Wrote them all. Yeah, I did it with Powerhouse, which is like sort of in and out of business from what I can tell. Sure. But I'm working on a book now for Simon and Schuster that should come out in 2026. So that. That's a little more of a.

Concept of the new book.

It's kind of. It's kind of like a essay memoir, you know, a little bit of everything from, like some high school stuff to being a drug addict to work to kind of everything. And that process is a little bit more like you're talking about, like, you write a full essay, you know, you outline some chapters, and then you take that out and. And sell it and CAA helped me. Obviously I wasn't doing it myself, but it's daunting, but I'm excited.

Yeah. What's the success? What's the secret to like a good book launch these days? We talked to Ezra Klein's co author and he was saying that interestingly, like when they went on hour long podcasts, they sold less books than like a five minute hit on like tv. Tv. Because if you listen to an hour long podcast, you're like, I got the gist. I don't need the whole book. But if you listen to five minutes, you're like, oh, I could do, I could do another hour. I'll buy the book.

That's a fair. I have a friend who's a very successful legendary author who told me an anecdote that he sold more books from doing Tucker Carlson when he was on TV than he sold from anything else. Like a clear bump. I mean, you know, this was years ago, but I think that, I mean, we were talking about this earlier today. Jason and I actually were talking about this and about how sort of podcasts have replaced not only late night, obviously, with this like, kind of like Stephen Colbert news. It's topical, but also like Good Morning America, the View. Like, it's still fun to do that stuff. I don't, but it might, it might move books, but it doesn't seem like the publicists think that it does based on the priorities of what they're having their clients do.

Yeah.

Is it, is it speedrun 200 podcasts? Is that, is that the playbook?

I mean, if you pay attention to people.

Yeah, kind of.

You know, if you have, if you pay attention to an actor or writer or whoever has something coming out, it's like, like, you know, music. We, we talk to a lot of musicians on how long gone, so we're part of that in some ways. But like there's the Dax Shepard's Joe Rogan smart list, Call her Daddy World. That I think is basically that's the new Oprah that, you know, that's what it is.

I almost feel like there's a world where some podcasters, certainly not us because we're a daily show or they're news driven. But some podcasts, I'm thinking of like acquired fms, like business deep dives or some true crime podcast. Like they kind of fill the hole of an audio book. It's like you get two or three hours every month and it feels like it's just a more. It's like an easier financial model for them. To be like, well, I'm selling 12 ad slots, and I have subscribers that feel like they're getting something. They're not waiting a full year for the next thing. So when I hit their credit card with a $20 charge this month, they're getting one thing this month.

I think we were talking on Facebook Friday, this concept of, like, releasing a book on substack because you could get somebody to effectively, it creates this, like, community experience of, like, reading the book together. But then it's, you know, you might charge $25 for your book, but somebody might pay you, I don't know, $10 a month for your substack.

No, I mean, there's all these. I mean, I. I think there's all these ways to do it. I am extremely traditional in the. What I think is cool and what works. I would like to do it about as. Buy the book, no pun intended, as possible, just because I think that' cool. I think all of these modern approaches, though, are often more. More profitable and also maybe, like you said, easier for people to stomach. Like, the attention spans aren't there. Even if you break it up like that in their minds. It's, like, easier to read in a weekly dispatch than it might be to sit down for an hour every day and read a book.

Yeah, there's also something about, like. I mean, when I've heard somebody say, like, when you publish a book, like, it's the easiest way to get, like, a common. With a phrase after your name every time you're in print. Like, if I wrote a book that's just, like, you know, how I became awesome, everyone would just be like, okay, like, you know, life. Life is someone awesome forever. John Coogan, author of the awesome Life that's why. That's why I'm awesome.

Chris's book. I'm assuming the memoir will be called Just Goaded.

Goaded. That's what I'm thinking.

Workshopping it now. But no, I mean, I think you're right. I mean, I had another friend, friend, Matty Matheson, the chef and cookbook author and, you know, actor. And he told me, he was like, dude, like, New York Times bestseller lasts for the rest of your life.

Totally.

You can you get that forever if you do it one time, you know? And, like, I think that also. That is so interesting because it's such, like, a classic thing that we all know, but no one really knows the formula of how you get, like, it's not just sales. There's all these factors that, like, no one knows, you know, which Makes it.

Even more interesting to think about. Would you rather have be featured in the New York Times? Is it. Is it more valuable to be featured in the New York Times four times a year, forever, or just be a New York Times bestseller? Because we had a couple buddies get hit pieces last week in the New York Times, and they just, like, really. They only get attention if you give them attention.

Yeah.

Like, dude, I was. I. That's so funny you say that, because there was a story in the early days of how Long Gone, there was a story that wasn't favorable, let's say. And I was like, wait, was this.

Was this the. Was this the whole lawsuit thing? Because I got the T shirt back then, too.

No, not the New York Times. No, this was not the New York Times. This was not. There's not the cease and desist for making fun of the daily. This was a. It's. You know, it's.

Yeah, don't Streisand it. Don't stress. You know, speak really generally so nobody can figure it out.

You know what?

We're.

We're moving on.

Moving on. Anyway, the. The other issue, Interesting thing about books versus podcasts is we have a buddy who runs an AI podcast, the Dwarkash Patel podcast. Great show, great interviews, super researched. He published a book, and the book was like, an oral history of AI and so it was a collection, and it was actually kind of like a digest of a bunch of interviews that he's done. So he didn't start from a blank page and just start typing out, like, a thousand pages and refine it. But. And in some ways, it's like. It's just kind of a different instantiation of the same thing that he's been doing. But I love it, because there's no way for me to go to, like, randomly remind myself in five years that I should go revisit a Dwarkesh podcast. Unless someone, like, clips it and shares it. But if I put that book on my shelf, I might pick it up randomly in a few years, and there's something about. About, like, just the tangible, physical. We're very into the physical stuff here. We print the.

No, for sure it's real. I mean, I think also that's. I mean, I think that, like everything else, it makes a return to some extent. You know, like, I subscribe to the Financial Times. I subscribe to several magazines, like a lot of my friends do. I mean, obviously, I'm 42 years old, so it kind of makes sense. But I think that, like, young people are fetishizing print the same way they fetishized vinyl, you know, a couple, you know, 10 years ago. It's the same kind of thing. One it is, you know, once it's basically past its prime.

Print is just objectively better than vinyl. Though I gotta say, I'm not 100% in that. It's like, it's it. You can go to the beach and carry like, you can't like, you know, like just bringing a piece of paper to the beach and letting it get. Yeah, yeah, losing it, whatever. It's nice. It gives you the benefit of leaving your phone at home.

Yes.

That's the powerful thing.

When I finish a book, I leave it wherever I, I am like on the back in the backseat of the plane on like I leave it so somebody can hopefully pick it up and take it. Because I'm usually, I usually have a couple with me if I'm traveling. If I finish one, I gotta lose.

Some weight because you're adding, adding stuff on the way.

Amazing. Exactly when are we gonna get how long gone pressed on vinyl? We gotta do a couple special episodes.

We did do a cd. We did do a CD a couple years ago with Jag Jaguar, which is a really fun project to do. But yeah, I mean I think for us we're still. I mean you guys have, have, have done something so interesting. I think it's so cool the way what you guys are doing and how you're approaching it. I think this live daily thing is like really powerful.

Sure.

And I think we're trying to figure out what our version of video is.

Wait, so yeah, you guys are doing. What's this thing? I read it in, in Feed Me. Like you're doing a live show, Like a one off live show.

We, we did, we're on tour right now. We're just have a couple days off, but we did a show at Turf Club in St. Paul, Minnesota at this basically with the service called veeps.com which is started by the Madden brothers from Good Charlotte, but then Live Nation bought it. And it's basic. They send a full sort of very pro crew to shoot your show live. And you can pay, you know, buy a ticket online and watch it up to 48 hours after you pay.

What's the ideal live.

It's like Vampire Weekend, how long gone? Alicia Keys.

I don't know.

It's cool.

Podcasters in control. Podcasters in control. What's the right live show for you though? Like do you feel pressure to have like hot, you know, the hottest. Take this show.

We're doing.

We're doing DC on third. We're doing DC on Thursday, and then Friday we're doing Toronto, and then we go to London. But this. This show is actually different. We. We used to just go up there and shoot the shit. We'd bring a guest up, you know, depending on what town we were in. Much like the actual podcast this time, it's the How Long Gone Guide to Life. And it's a visual. It's like basically a keynote that we're walking the audience through with, like, our commentary. And we've only done it twice so far, but it feels. It feels sort of like it's taken us from a live podcast to a show, which is what our goal was. And I don't think we could figure out exactly how to do that. And I think this is kind of pushing us in that direction, which is great.

Give us the three minute summary of the guide to life so that people have to still go see.

Yeah, yeah. Don't give it all away.

Get the tickets.

Get the tickets.

Yeah. No, it's.

By the way, we. We did. We did one live show together in Miami, and we had only made, like, two episodes. And so nobody, like, nobody knew us as a duo. They didn't know our humor. And the live show. The live show is like, we made like, like a very serious, like, argument for why venture capital firms should get their founders on, like, peds. So, like, testosterone, like, like a variety.

That's a great. That is extreme. I would watch. Pay attention to that.

Yeah, exactly.

And. And I don't. Like, half of the. The half the audience was like, these guys are dead serious. Half were like, kind of laughing, like, what is going on? This is.

It's a mix. It's a mixed bag. Yeah. I mean, ours. We basically cover all of the stuff that we talk around the show, from like restaurants and restaurant etiquette to the gym and gym etiquette to dating and sex and drugs. So it's like we cover it. You know, it's like one slide that's going really well for us is polyamory is for ugly people. Be single or cheat. That's going really well for us. But, you know, it's just like, funny stuff that you can just kind of talk about and also will inspire a response from the crowd. We. We like to go back and forth with the audience.

It's fun.

We're gonna create. We're gonna create a rival guide to life and. And we'll just go guide for guide.

Look, I think we. I think we have different Enough audiences.

Where we cover a lot of territory for sure. So. So going back and forth, I want your restaurant etiquette. This morning we go to our cafe where we have breakfast every morning. Normally, really, like, I never noticed the music. It was always quiet and chill and today they were blasting NSync, Britney Spears. And maybe it was just the mood because it was kind of gloomy. In la, we're having some like late June gloom going on.

Terrible.

And I was a little bit tired from the weekend, trying to get back into it. I.

And.

And I was like trying to read like a pretty detailed analysis about like tech and stuff and I was just like, I can't. I. And funny thing is that two days ago I was in a grocery store and I was like jamming to the Backstreet Boys and I was like, this is the greatest thing ever. I was loving it in that context, but in the morning.

But it felt like such a violation of trust because we go to this place every single day.

It's so reliable and it sets me up for the day. I'm prepping the show and so we kind of made us. We made a slight comment. We were like, oh, is the music different?

Different.

But I want to know etiquette wise. How should I. I don't want to be. I want to speak to the manager. That's weird. But also like, I'd like a change.

Yeah.

What should I do?

What. What kind of like. Like that as a regular. Like, what are your rights as a regular?

What are your rights as a regular?

You're right. You can't say anything. Sorry.

If you can't say anything.

No, if you say something, they're never going to look at you the same whether you're right or wrong. I'm just telling you, they're gonna. They're gonna think of you differently.

The previous.

Not going to be made with love anymore. It's not going to be made with love.

So am I supposed to vote with my feet? Go somewhere else. We were running the numbers. We go there every single day. We're like, we're basically employing a single person there paying the.

I like how you think about this. Like, it's logical and then you can sort of figure out a way to make it work. It's. It's a barista hates all of the customers.

Okay.

Okay. And that's a feature?

Yeah, yeah, it's a feature.

I mean, I understand what you're saying. I think restaurant, like there was a trend for so long, it's still going on where like restaurants Play like hip hop, you know, and like I don't want to hear Kanye west while I'm eating spicy penne, you know, I want, I want to like, I want to hear. Nothing is preferred if you go to certain kind of old school. Like I think Le Bernardin in New York doesn't play music and it's just like the bustle of the restaurant and the chatter. I'd much rather hear that.

Yeah, yeah, yeah.

Much nicer.

It's fascinating. Jordy, where should we take it now? Fascinating.

I have a. Yeah, I wanted, I wanted to just get some, some broad takes from you on a bunch of different topics that we don't, we don't cover closely. So 1. First off, is AI psychosis real? You experienced it yet yourself?

Let me, let me, let me tell you something about AI and I'm sorry to all of you.

I just couldn't, I'm assuming, couldn't care less.

The hottest.

I don't care less. I think it's a. I think it's a reality. I think it's very serious. I think it's coming for us good, bad and ugly. I just don't care. And like other things in my life, I'm going to put it off till the very last minute and when I have to deal with it. I've never downloaded Photo Photoshop. I've never used InDesign. I. I've put off all this in my, my entire life and it's worked out. So I think I'm going to do the same thing with AI until. Until I have no chance.

And you're so you work list off some of the clients that you've worked with over the years. It's basically everybody, right?

I mean. Yeah, I mean everything. And on the, in the, on that side of the business. Yeah. I mean New Balance, Tom Brown, Stussy J Crew is, is an ongoing one Banana Republic for a couple years.

So yeah, I mean like I love.

A little, a little bit of everything.

So.

Banana Republic. Oh, they got you.

They signed Kevin love Jason 6:9. So he has the same suffered.

Yeah they signed and he has the 14 15.

Jason's a 17, so it's really.

Oh okay. Mogged.

He goes to these like funny websites to find. Oh yeah, like we haven't heard of.

Because we're dark web, big and tall.

Exactly, exactly, exactly.

So so you're going in. When you're working with these of kind clients, it's kind of like Rick, Rick Rubin Mode. You don't know how to use any of the tools. You're just kind of like, they just pay you to. To tell them what, what.

I mean, no, I think it's a little bit of a relationship thing and, and hopefully a taste thing. Like, I. I think taste takes years and years to develop, and I think mine is. Is suited for certain people more than others. I think J. Crew is like a perfect fit because I grew up wearing J. Crew. I was raised in Atlanta. It's preppy, it's subtle, Southern. But I think I have a little bit of a different point of view because I. I got into punk and hardcore and, you know, skateboarding and all that stuff. So it's like a little bit of a hybrid of all that, of all those things. But I think with, with clients, it's all about sort of taking the real talent at the company and making sure that it doesn't go off the rails, you know, so the creativity people who are true geniuses and really creative, like powerhouses need to be brought back down to earth a little bit, you know, and like, reminded in a nice, gentle way that if we don't, you know, make money, this is all over. Like, this is like. Like, we can do all this stuff, but sometimes you gotta feed the beast to make sure the lights stay on and buys us more time to do more fun stuff. And I think it's like a. I think as an outsider going into a company, you're able to do that a lot easier than like a salaried employee.

Yeah, you can stir things up. Are you. But, but so, so going, like, slightly deeper, your client, like, I have to imagine every single brand that you've worked with or will work with in the future, is starting to try to mix in AI generated imagery. And I'm sure that sounds like that's triggering to you. It's just stuff that, that you don't. You don't want to like. How do you see the kind of market kind of bifurcating in the sense that you think every brand, you think any brands will take a hard line and say, like, we only shoot, shoot, you know, farm to table, organic content here. We never.

I think that's exactly what's going to happen. I think there's going to be like a division where it's like, obviously certain things that are easy to automate. I think everybody will do that because it's a. It feels like a freebie. I think some of the creative stuff, I think a lot of. At least I hope clients I work with would sort of be like, we need a human being to take the pictures like that, like pictures specifically. Like, to me, there's so much emotion involved in it, there's so much skill involved in it and you just can't. To me, I haven't seen it mimicked correctly. Like asking Chad. Yeah, the other, the other fix your car is different. Like this is something that requires like a human touch, I think.

Yeah. And I think there's actually, there's laws in place that, that make it so that chain restaurants can't Photoshop their food. Right. Because they would just Photoshop a burger and then they're trying to sell it and it just doesn't look anything. And I think, I think it's possible. We'll see. I don't know. On the apparel side, it'll be interesting because everybody's had the experience of like buying an item and it just looks amazing online or.

Yeah, it looks totally different when it.

Gets there or whatever. And then it just like doesn't fit or whatever. And so I think that that will probably accelerate as you have these sort of upstart brands that never even do product photography. Don't even know if they're, if they're stuff.

Well, product photography and like econ photography is like a science and it's pretty hard to get right. And there's a lot of discussion at these, at these companies about how to do it, who' going to shoot it, how it's going to appear. And I think that stuff is very interesting because it directly relates to sales and like the bottom line. Like some styles work really, really well and that's what everybody wants to mimic and some are a little, a little harder to, to pull off, you know.

Yeah. What about, on, what about. How do you see like, you know, music market evolving? We, we don't cover much music at all.

Latest Justin Bieber album, we got two.

Songs on the drive home.

I mean we, we talk about music and on how long gone all that's all we do. And like a lot of my, my best friends are musicians. Most of them are successful at this point doing it long enough. But I think like everything, everything else in our, in our lives, I think we're just paralyzed by choice and the barrier of entry is just too low. You know, there's too much stuff and music, it suffers from that maybe more than any other genre that I can think of.

And you mean, you mean because be, you know, growing up, you'd buy an album, you'd invest dollars into it and so you would listen to the full thing and you'd like really pay attention to it. And now if you can stream an album for two seconds and be like, whatever, yeah, you don't care.

I mean, I think. Yes, I think that. I also think that it's just if, if the three of us wanted to make a song right now and upload it and have it on itunes and Spotify this week, we could do that.

Yeah.

And I don't think that. I don't think the world needs that from us or for many. From many other people. And I think that. But yes, you're right. I think when you would go to the store, access was harder. I mean, we talk about this all the time. Like I grew up, I remember the day that we got Internet at my parents house, you know, I mean, we had the Dell in the family room. I remember it like the whole thing. And if you remember a time before that, I think the effort that was involved to like anything or to understand anything was so much gratitude, greater because.

You had to piece it together through like magazines and going to a show and that kind of.

Yeah, you go to a show, you go to magazines, you look at the liner notes of an album and see which bands get thanked and then go find those bands. You know, it's like it was that. It was kind of that tactile and that granular. But yeah, I think it's like, I think that like Spotify and streaming in general, like, I know people that make a lot of money on streaming and they're very happy with it. I know a lot of people and the. I think general sentiment is no one gets paid, it's enough. And of course, I always want artists to get paid more because they, you know, it's a soundtrack to our lives. It's like important for culture. But I also think that people forget that these are just giant corporations and they've got bottom again, they've got bottom lines and like they're going to do whatever they have to do to make that work. Yeah.

Do you think corporate jingles could make a comeback?

They have. What do you mean? I mean, you know that, you know, famously Pusha T from clips wrote the McDonald's jingle. I'm losing.

No way.

Yeah.

I had no idea.

He'll never have to work again, you know, from that alone.

Yeah.

So I think it's all that stuff. I mean, it's like the way that, like Jon Hamm is the voice of Mercedes. Like, everybody gets it how they can. You know what I mean? I think the jingle or I mean, what we talk about a lot on how long gone is like sync Culture, you know, and, like, how there used to be, like, you would be a sellout, you know, if they used your song in a. In a Jack in the Box commercial. And now it's like, well, that's the most money I'm going to make. If they're going to give me a quarter million dollars to use my song for 30 seconds, like, I have to do that. You know, the choice isn't. Isn't quite as clear as it used to be. And so we've talked to a lot of different musicians about regret. You know, how they, like, pass something up because they were high on their horse in their early 20s, and now that they're 40, they're like, I'm a fucking idiot. Like, half a million dollars. Like, who cares?

Yeah, that makes sense.

You said that they've made a comeback already. Like, is. Is there a more modern example than the. Than the McDonald's song?

I. I mean, I feel like you might be right. I feel like they're just. They're ingrained in my brain. So they feel new. Total, like cars for kids or something is probably 20 years old. If you guys ever lived in New York, there's 1-8-77-cars. Totally.

Like, you know, I. Yeah, I know it well. I'm just thinking about, like, of the new companies that have emerged. Airbnb, Facebook, Instagram, Spotify, OpenAI. Like, good point. Tesla doesn't have a jingle like. Like, I don't know if those would be even, like, on brand, but, like, even funny, even. Even, like, prime from Logan Paul. Like, he's done a song like, he did that he did, like, It's Everyday, Bro with his brother, I think. Or maybe that was just his brother, but, you know, like, there are influencer brands. Mr. Beast has a chocolate bar. I don't know that it has. Has a jingle. And I'm wondering if it has something to do with the fact, like, the nature of advertising has become, like, way longer tail and the creative, like, we don't see that many, like, shelling points for, like, wow, they made an iconic ad. It ran at the Super Bowl. Everyone's talking about it. It's more like they spent the same amount of money on ads, but it's just in the form of a thousand different Instagram ads. And everyone saw us, like, slightly different one that was more tailored to them. So the guys saw the one with.

The girl essentially alpha making a jingle right now and using the same jingle on, like, thousands of different creative. So it's always the same audio and different.

Yeah.

And then I think we're onto something here.

Time is money. Save both.

Find your happy place.

I mean, with car companies, they, they obsess over, like, the sounds the cars make, you know? You know what I mean? Like, I drove this new Mercedes from Atlanta to, to Miami for the F1 and it was like there was all this detail.

I'm a huge Mercedes fan.

We love Mercedes over here. We have tons of them.

I'm in, I'm in Bedford, New York right now. And I borrowed a CLE AMG convertible. Couldn't be better. But I just notice all the. But when you drive these car, you notice like these little kind of sort of grand moments where it's like a little noise when something starts up or like it is audio or like how famously like Porsche has a guy whose only job is about like the way the door sounds when it closes. Because it sounds so much better in a Porsche than other cars, you know? Yeah, but, but it's just a funny. I think there is these audio elements that are happening from a brand standpoint, but they're not quite as overt as like a jingle, you know?

Yeah. It does feel like Doug Demiro has this. Take that. That like power is on demand. Like the latest Rivian truck, electric truck has 1200 horsepower. It's the same as a Bugatti Veyron. Like it. Is it like that's what you want.

When you're driving your car.

Is Rivian a sponsor?

No, no, no, no, no, no.

All right.

Because that Rivian is. Rivian's like the, that. That is like the Austin mesh hat Yeti cooler of cars. And just don't. I don't get the popularity. I don't think it looks good. I mean, I guess it's an alternative to Tesla. So that's part of the.

Yeah, I think the main thing was that like Tesla hadn't come out with a truck yet and then they came out with a very controversial design. And then the F150 Lightning was pretty expensive and the Rivian kind of was just pure play. Like, didn't confuse anyone. And. And then they did have the full size SUV or like close to full size.

Yeah, yeah, for sure.

In the electric category, a little bit sooner than the Hyundai Bionic 9 and a couple others. And it says a little bit more. But it is interesting because between lucid and Rivian and lucid, and there's that star Polestar. Polestar is technically like a Volkswagen or a. Oh, is it? It's not a Volkswagen. It's the Volvo. Like Volvo.

I Go to. I go to Copenhagen Fashion Week a couple times a year. And it's all. And the drivers, it's all polestar. It's all like.

I mean, every brand is confused with what to do with electric. Like, it's like, do we throw. Is it a. Is it a powertrain? We just throw a little eye at the end or an X or an E at the end. Or do we do a whole new sub brand or do we spin something out?

Oh, what do you. Have you seen that? Have you seen the new. The new Lamborghini? What is. What is that? The ev.

Oh, yeah. What was that called?

The.

Why do they have to make them.

Look like the Lanzador?

You got to pull up a picture of the Lanzador. This thing looks wild.

Wild.

It is.

What is it? What is it, a Lambo?

Yeah, it's a Lamborghini Lanzador.

What's it called? Lamborghini.

They're calling it an suv, but it just looks like a really big car.

Like a lifted sedan.

It looks like a lifted sedan. No, it looks like a lifted hurricane. It's kind of sick. It's kind of sick. And then there's the slate truck, which is coming out, which is like a small thing. But I think that some of the new companies, like, they're so focused on, like, just deliver on the basic value prop. They're not in that, like, surprise and delight category yet. Although Ruby and did kind of do that with like that gear tunnel. You know about that in the.

I like. Okay, yeah, that is a. That is just a. A big old car.

Yeah, it's a big car, but they're currently SUV guys.

Like a Subaru. No joke. No, that's.

That's the hate. That's like a Subaru halo car. Like that.

That should be there.

Total. Total Tampa tangent. But I did have one question and then one idea that I wanted to throw out before. Before you jump. Do you think that any social media apps will become truly Lindy? Like, will we be on X, like, 30 years from now, chopping it up on the timeline? Or as a.

As a. I call it Twitter. Still. I'm a. I'm a traditionalist. But for this purpose, I will call the X. X is the most enduring platform that we have because I think it's not. Not image based. I think that's what sets it apart. It. It feels more comedy and information, and those are two things that I feel like aren't as affected by trends in some ways. Whereas, like, Instagram. I love Instagram. I use it every day. But, like, people get. When people get very exhausted by it because of what they have to see, you know, And I think Twitter, I don't know. I'm a. I'm a Twitter die hard. I'm a power user. I have been forever and I will never change. Change that.

Yeah.

That's awesome.

I hope. Yeah. It's one of those things, like, is it digital news or is it like the form. Is it its own form factor, like vinyl, that will just perpetuate forever?

We're gonna find out.

Last, last question. Media and entertainment question. Because you're an entertainer, do you think there's room in the market for next election cycle? Pay per view boxing between both political parties. So like Democrats versus Republicans. Betting on the line American.

I'll pay. I'll pay $100 for the PPV for that. I mean, I think that if we're. I don't think we're that far away, honestly. I know that's a joke. I don't think we're that far.

No, I'm not. I'm actually. I think every four years you get 10. 10 Dems, 10 Republicans. Get them lined up in different weight classes. Hundred dollars, Duke it out.

I would love to see some low level, you know, first year senators just bloodying themselves in the. On the. On the opening rounds, you know, that's what we.

Yeah, it doesn't even. It could just be angry online posters too, you know.

Well, that, that's a. That's a category of fighting that has not been monetized yet. Something we should think about.

Yeah, it's like Dana White's slap boxing thing or whatever.

Yeah, no experience requires slap boxing would be huge.

Awesome. All right, well, this was super fun.

Let's do it again soon.

Come back on again. Anytime, guys.

Thanks so much. All right, have a good one.

Good luck with the rest of your live shows. We were talking about jingles. We got to sing you a jingle.

Find your happy place. Find your happy place.

Book a wander with inspiring views, hotel grade amenities, dreamy beds, top tier cleaning, and 247 concierge service. It's a vacation home, but better folks.

And Chris Buskirk is in the restream waiting room.

Let's bring in Buskirk. How you doing, Chris? Great.

Hopefully you got to hear us singing.

We're trying to bring back corporate jingles. We think that there's an unexplored territory there. I don't know your stance.

1789.

Ooh, we got something.

I know why we don't. I Actually know why we don't have them anymore?

Please explain.

It's because Big Pharma took over advertising and that killed all the creativity.

Oh, interesting. And they got to throw those 40.

Second commercial, 20 seconds tells you all the bad things that are going to happen to you when you take the drug.

It's rough.

Well, yeah. So the conspiracy theory there. I don't have my tinfoil hat handy. Is that it's not about even generating awareness. Here we go. I got. We keep one of these anytime we have to discuss a conspiracy theory. Discuss a conspiracy theory gives you cover. You know, I don't believe this necessarily, but you listen to these pharma ads and there, after listening to it, I would say you're much more likely to think. I don't think this is for me. Right. Because you maybe hear about some benefits and then you get like 20 seconds of. But it's just. Is it a mechanism to just invest so many dollars in a show that the show could not possibly criticize.

The.

Pharma industry broadly otherwise? You're potentially.

This manufactured consent says Noam Chomsky, but has been written, rewritten.

Well, it's more like. It's more like, yeah, you're welcome to criticize Big Pharma, but, like, say goodbye to like, what, 60% of your revenue.

Yeah, yeah, sure, sure, sure.

Anyway, not gonna be running the late show. We're not gonna be talking about more cons.

Anyway, let's kick it off with an introduction.

Theory is always also known as the.

News six months in the future. Yeah, but, but let's introduce yourself. How do you think about yourself? What, what, what. What key projects are you working on right now?

You know, the, you know, co founder of 1789 Capital, as you guys know. Like, I've got some other evocations as well, but, you know, just, you know, I. People always say, like, how do you split up your time? And like, there was like the. That was like the Rubik's Cube question.

For me for a while.

And I realized actually the. The simple answer is right, which is like, all the different stuff I'm doing is all the same project, which is. Is how do you keep the country peaceful and prosperous? Like, that's really it. Then there's all kinds of different vectors of attack. There. There's like, there's politics, there's culture, there's business, there's tech, but it's really all with the same telos. Like, how do we just make life better for this country and the people in it?

Yeah, give me a Give me a temperature check. On your view on the business climate. There's been so many new, so many stories about tariffs and interest rates and all sorts of of stuff at the same time. The stock market's been like down and then right back to where it was. And so there's like the nothing ever happens view, there's the incredibly tumultuous times view. How are you just feeling about the business climate generally?

And Uncle Sam put up a surplus.

Like it's better than it's been in a long time.

Yeah, yeah.

I mean this is what, you know, I've had, I've had the following conversation. It probably literally like 10 or 20 times in the past few months. Months. And people, whether a lot of investors, but also founders slash builders have said a different version of the same thing, which is I didn't know how much I was self censoring my own thought about what was possible over the past four or five years until, you know, fill in the blank. January, February, March, April, whatever. Just some this year. And then they realized like, oh, the shackles are off. We can actually go and do really cool stuff now. And everybody had their project like there it was like there was a pain point in their business, like, oh, I can't do this. Or like, you know, like the crypto guys or AI, like they're really, really under, under attack. And it was like regulation by enforcement.

Yeah.

And so they knew what the pain points were. And so for obvious reasons, we're super focused on those things. And they were like, oh, if we could just get through this, then we could do this one other thing, whatever that may have been. And what I've heard them, what I've heard a lot of people say, saying is once the pain stopped, I realized that there were like 10 other things that not only did I want to do, but that I could do. And so like the amount of optimism and energy is just really infectious. It's like it's a great time to be alive.

Yeah, it feels like, I mean we saw that with a lot of the nuclear founders. Like you talk to those folks and I mean it was so rough for a while that they normally like, if you invest in my company, I don't want you to invest in, in a direct competitor. But the nuclear companies were like, look like that's not going to be the thing that does me and like you can go ahead and invest. And then now we finally have some clarity from the eo. We talked to some nuclear founders and it feels like we're on the cusp of some real progress there. How much are you focusing on energy? What are the other categories that you find interesting right now? What do you see as being like uniquely unlocked in this new era?

Yeah, it's, we call it, it that I don't think we made it up, but like, we call like transformational technologies. So that means like AI, AI, infra energy, robotics, space defense, Those are kind of like the big buckets in a lot of cases. There's a lot of, they're, they're really related. You know, we like the way, at least the way we think about energy is like it's attached to all of those. Sure. And you know, that's where, that's where nuclear really comes in. But particularly around AI and crypto and some of the other elements of just security in general. Like that is the major unlock. Like one of the things that I tell people all the time is like, if you want to like just an easy heuristic about like, you know, how you think about like what are the big unlocks that need to have happened which is that think about it this way, the large civilization that has access to the most abundant, cheapest energy wins. Like period, full stop. I was talking with a friend of mine, another investor last week and he actually, he had this one great description which I had actually never thought of about colonization of Mars. Because I was saying how obviously Elon's been super focused on it and I'm like, I don't want to go to Mars. I've never wanted to go to Mars. It seems like it sucks there. But then I realized that actually terraforming Mars, that's actually kind of cool because it would make it someplace you'd actually want to go. So that's why Mars colonization has become charismatic. Because like, just like getting to the moon was, was interesting and charismatic in the 60s because it was like, can we get there? Like, it was just, it was just like the journey, not the, not the end. But then when you start to think about the idea of colonization of Mars, well, it's interesting when you think that like you could make it beautiful. And what he said was, it's just an energy problem. If you can bring enough energy to bear, then you can terraform, then you can terraform Mars. I was like, that was actually for me that was like an interesting way to think about it that I hadn't considered. But the same thing applies to everything else, which is if you have, if you have super abundant, super cheap energy as a large scale civilization, you're going to win because it makes everything else better, easier, cheaper, and things that were not possible. It makes them possible. Kind of like the way SpaceX, what SpaceX has done with launch, by just driving the cost per kg into the dirt, it unlocks all these other businesses.

Yeah.

So if, if great power competition is, is energy competition, do, do we, you know, I mean, the concern here is that China is just like copy and pasting nuclear reactors and we have a bunch of great nuclear startups that are, you know, making efforts and some new projects. I think Meta, you know, has announced some, some new initiatives there. But is this not like one of the kind of biggest issues that we're facing from a national security standpoint?

There's a way in which it's the biggest because it touches everything. It touches literally everything. I mean, you think about these really critical growth areas, AI being the super obvious one. And I've seen a number of analyses that show the amount of AI compute we need to build just in this country, say in the next 10 years. So by 2035, like it would consume more power than we produce right now as a country. So there's like, we need to be producing a lot more power because unlike certain members of Congress, it turns out electricity doesn't just come from the wall. Like you actually have to produce it someplace.

Yeah. Food comes from the grocery store, gasoline comes from the pump. Electricity comes from exactly.

What are you slow? You just plug in, just plug in and there's electricity there.

I, I do think there's something, there's something fascinating about like it's so hard to predict the future. It's so hard to predict what we will need, you know, a gigawatt, you know, data center for. And then all of a sudden it's like, we need it today. Like we have the use case. We need it today. And I keep thinking about, like what happened in the dot com boom where we overbuilt all this dark fiber. Google went and bought it up and a bunch of the tech companies bought it up and then eventually, like the products caught up. My question is, how do you think about projects that might be uneconomical to underwrite on the venture side? And so they might need to be underwritten by the government. Like, like I'm, I feel like we're probably aligned in that. Like we should probably keep the government as narrow as possible. But there are some really crazy things like the moon landing, like the Mars helicopter stuff that just doesn't make any sense for any private company to go after. And then after a while you Start seeing like, oh, okay, like, you know, they built some roads, they built some infrastructure, and now private companies are catching up and actually ramping up the, the, the use of that energy or something. But how do you see the government playing a role in the next, in solving those fundamental problems? Is it purely just on deregulation? Do you want to see funding flow into certain projects? Do you want to see government doing hard projects? What do you, what's your philosophy on all that?

Yeah, it's. It's both. It's both is. The answer is like we've just seen, you know, just as I was saying earlier, like just this year, just by getting out of the way unlocks a lot. And it just. There's so many people who want to do things and who have the capacity to do things if they're not being shackled and prevented from doing them. So like, we've seen a lot of that in the past six, seven months is just by getting out of the way and giving people, people permission to do it. You get a lot of, you get a lot of activity. And so that's a big part of it. But then there are these other issues where, you know, China creates its national champions. And so, like, great example is like rare earth magnets. You know, it is, it can be. It is. I mean, they subsidize it. So it is cheaper to buy a rare earth magnet from Chinese supplier than it is to buy the raw materials in the United States. You know, it. That is, I mean, that's very, it's. That's geopolitics, right? That's not just markets, that's not just economics. That's not just innovation, that's geopolitics. The Chinese have been on, they've been on top of this for 20 years and realized that this was a choke point where they were would be able to get a huge amount of leverage over their competitors and adversaries. And I guess we're someplace, you know, in between those two things with, with China. And if they were able to subsidize them, they could drive out of lot A, all the competitors out of business. And because they go into literally everything of any importance that they'd have massive, not just economic leverage, but like geopolitical leverage over the United States. And so that's a place where, you know, government needs to play a role because it's not a company v company competition. It is a state actor v. Company competition as it stands right now. And that's something that's critical to national security. So you see That I think in a number of other spaces. But, you know, the rare earth magnets is like, it's just a very obvious example of that where you can't. A company can be super innovative and can do all the right things, but if it is facing a state actor as its competitor versus another company, like, they're going to lose.

Yeah. On the rare earth materials question, there was a story in the Wall Street Journal last week about Apple buying half a billion dollars of rare earths from a company called MP Materials. And in Vegas, in Las Vegas. And it stuck. It stuck out to me as kind of a vibe shift because Apple hasn't been at the front of like, re. Industrialize America, bring back Jobs, and, you know, let's make everything in America. They're not saying that, but it feels like maybe behind the scenes, the. They're at least thinking about their supply chain very seriously and trying to diversify. And maybe it's a cruise ship, but they're putting their hands on the tiller just a little bit. And so I took that as, like, cause for optimism with the overall, you know, energy independence, supply chain independence. But did you see that story? Are there any other causes for optimism that might be coming from the crowd that's not cheering as loud as they can can, but might be getting, you know, might be working behind the scenes to kind of achieve the same goals that everyone should be aligned around?

Yeah, I mean, look, there's a. There's a lot of signal, I think, in that, in that move. I mean, the. At some point, Apple has to realize that their supply chains are really, really insecure. And this leads to, like, another point on the rare earth magnets. Like one of the things like the. One of the. What great sort of unintended consequences of. Of the Liberation Day tariffs and the general policy of tariff policy coming out of the administration is that this actually turned out to be a forcing function. It made China play their rarest card way too early. Had they decided, for instance, to attack Taiwan and then played that card, that would have been the optimum moment. But what happened was that in the negotiations around Taiwan tariffs, they. They threatened to play the rare earth card. In fact, they did actually, like, reduce exports the past couple months, but it was too soon. Like, they. They played it early, and now there's time for the United States to catch up. And this is, you know, in fact, as we saw with mp, you know, and the support they've gotten from the government. Another company that we're in the process of investing in, called Vulcan Elements, is, is in the space as well. And we now have. Have a window in time where we can build out an ecosystem to manufacture rare earth magnets in the United States and do it really, really quickly. So that's just a place where like, there's like, there's, there's a lot of signal coming out around that as well. But there's, I think the, the fact that there is so much happening and, and at such a rapid rate, that's where we're seeing, but that's where we're really seeing the advances here because, I don't know, people ask me, are you pleased with the way the admin's gone? Yeah, I am. What I've been surprised by though, is the speed with which they've moved. And what's happened is they're moving at founder speed and that that becomes infectious. And so now you get cycles that might have taken a year or two or three to happen and things work through the system in the past. And now like we see we're standing up a rare Earth's capacity in the United States basically overnight. That was like a pipe dream four or five months ago, and now it's actually happening.

We gotta stop calling them rare earths. They're just Earths. They're just earth magnets.

They're not rare.

They're not rare.

They're only. The process form is rare.

Yes, the industrial capacity is rare, but the product is not. Let's just call them Earth.

It is. You gotta. You know, we probably criticize China more than most podcasters on Earth, but you got to give them credit for letting the world think that they were just the Happy Meal toy manufacturing hub for like a couple decades and then just like quietly owned all these critical supply chains. I'm curious how you're thinking about solar. And one of the reasons I ask is I've invested in tons of these different kind of hard tech, new industrials categories and I've yet to back a solar company. I only started actively angel investing maybe four or five years ago and it's fallen out of fashion is one way to describe it.

Casey Hammer is bringing it back.

Yeah. So Casey's the only. Casey's the example. And I would happily invest in Casey if I had the chance. But curious how you're thinking about that. It's another area that China is dominant in and it seems, seems like something that we wouldn't necessarily want long term.

So like I've been sort like semi obsessed with solar for a long time. I just like, it's mostly Just basically being like, thrifty. I'm like, damn, there's all that sunlight and I can just power my house for free, you know, so like, I, like at my house in Arizona, I have solar on it. I've had it for years. Okay. But it's really not economic, and so there needs to be a pretty big leap forward for it for it to matter. Otherwise, why not just build more nuclear? Right?

Why not?

Why don't we just keep it simple?

Yeah, Casey.

Like, PV solar is actually isn't that good. And like, the concentrators are like, okay, but like, again, like, we have the solution already, so there's got to be a compelling case to be made as to why under the current, like, you know, with the current technology around solar, why is that even worth focusing on?

Interesting. Yeah. Casey's point was like, China is subsidizing photovoltaic manufacturing and, and the, and the most hurtful thing you could do is just buy all of it to. So that they take a loss on every unit and then build up your capacity. But I don't know that that actually would, would, would catalyze a real shift in capacity. And we might just get kind of.

Still helping them get extremely good, especially.

If there's learning curve and stuff. So I don't know, it's a, it's a longer conversation.

PV is just not that efficient. Right?

That's the problem.

Yeah, I, I think they, I think the argument was that it's like, it's very well matched to the load of the grid of the American consumer, because when it's hot out, we run our air conditioning, and when it's hot out, the sun is shining on the PV cells, and so there's some match there. Whereas a nuclear power plant generates just as much at night as it does in the day. So I don't know, there's a whole bunch of different load use cases for AI training versus inference and all this different stuff, but it's. Yeah, it's fat. It's a fascinating geopolitical dynamic. And it seems like the main takeaway is that China has just been saying yes and to everything, and they develop capacity everywhere. And it's like, we could try and predict the future, but, you know, if PV winds up being really important, I want to have capacity for that. If nuclear, I want that too. I want, I want the smorgasbord, give me all the energy. But I don't know.

They're building a ton of coal plants too.

Yeah, no, 100%. And fortunately, you know, some of the AI infrastructure buildouts are pulling forward. Some of that, we're getting a whole bunch of natural gas online. Like we are becoming more energy independent. So green shoots all over the place. Jordy, anything else? Last question.

I think that's it for now.

This is fantastic. Yeah, we got to have you back on. Especially when there's like big news in your world or anything going on that you can comment anytime.

Happy to do it.

Yeah, we need, we just need 30 seconds notice. We'll drop you the, we'll drop, we'll.

Drop you the link. You can just on hop right on.

Okay.

Awesome. Yeah, just you can sign, signal me. I'll be on.

I will, I'll talk to you soon, Chris.

Have a great one, Chris. Cheers.

Bye.

And speaking of Minerals, we have adquick.com.

If you're selling minerals, you put it on a billboard. Out of home advertising made easy and measurable. Say goodbye to the headaches of out of home advertising. Only Ad Quick combines a technology out of home expertise and data to enable efficient seamless ad buying across. What were you saying about materials?

Well, we have the founder of Mariana Minerals joining the show.

Welcome to the stream. Hit the sound board. Give me the Ashton Hall. Welcome to the stream. How you doing, Turner? Good.

How are we doing guys?

Beautiful map behind you. I, I, I love cartography. I'm a big fan of throwing a map in the background.

Is that 3D or like does it have texture or does it just kind of look at, look like it's not.

A lot, little bit of texture.

Good.

Not, not crazy 3D though.

Mercator projection. I'm pretty sure that's a good one. Anyway, introduce yourself. What do you do? Welcome to the show.

Yeah, sure thing. Thanks for having me on, guys. Turner Caldwell here, CEO of Mariana Minerals. We're a vertically integrated software first mining company. So we're coming into projects that are, you know, have post exploration. They've discovered the mineral deposits deposit and we come in to engineer it. So we develop the mine plan, we develop the refining plan, we capitalize it at the project level and from the top. Co and then we, we manage the construction, we manage the construct, the commissioning and then we operate it out into perpetuity.

Okay, talk to me about the whole pipeline of actually getting minerals out of the ground. Let's say I own 100,000 acres in Alaska or something. Something. I bought it just as a forest. It's a ranch. Never explored it. What are the steps where I imagine we're talking about like yeah, exploration firms. Then you Come in at some point. Like really walk me through a concrete example of going from like someone has some land to now we're getting material.

Well, yeah, and if, if you're this verticalized, why not also do discovery? I'm sure there's a reason. I'm sure, yeah, there's a, there's a.

There'S a, a big market of folks that are doing discovery and there's actually a bunch of awesome like tech VC backed startups that are focused on accelerating discovery. I think what we're missing is folks that are focused on condensing the timelines from when you have discovered something to when it is actually producing. And it's like this snake in the grass kind of challenge, like very complicated system of actually designing these plants, buying out all the infrastructure, installing the infrastructure, getting it running. But yeah, so you start with a, you know, flat piece of land, ideally it's flat. And you'll start this like really long campaign of, you know, you'll start with kind of like spot testing of different areas to determine and probably some like surface sampling to determine if there's something promising there. There's some permits that come into play when it, when it, when you have to actually start doing more advanced exploration campaigns. And that's something that, you know, if you're on federal land and sometimes on state land, those permits can get in the way of, of accelerating kind of mineral discovery. You'll do drill holes, hundreds to thousands of drill holes to define a mineral deposit. And those mineral deposits can vary all over the place. You can have shallower ones where it's a little bit cheaper to drill. You can have really deep ones where it starts to get much more expensive to do and time intensive to do that exploration campaign. And all along the way you are taking samples and you are kind of building these 3D models to try to understand how was the deposit formed, why is the mineral there in the first place. And that helps to kind of inform also the advancing the exploration side effect things. And in parallel with that you're starting to do these testing campaigns where you're running concentration tests to figure out how you can upgrade the ore to an intermediate product. You're starting to do refining tests to understand how you would actually extract the target mineral into its final state, which is, you know, a high purity metal alloy in many cases or a high purity kind of specialty chemical salt. And so you're writing these reports all along the way. So you'll start with a, a pea, which is an initial economic assessment. You'll then Do a pre feasibility study. You'll do a feasibility study. You'll do a definitive feasibility study. And kind of as you're doing that, the value of the asset is kind of increasing and increasing. Speculators are piling into the stock and the intent there.

Let's hear for the speculators. The backbone of the backbone of the mining industry.

Stock ripper.

That's right. That's how that expiration is funded. They're trying to like ride that first wave of value creation.

Yeah, similar to like FDA trials for new pharmaceuticals probably.

Well, I can and I can. I can. I'm starting to figure out, you know, a number of reasons while you're explaining them why you wouldn't want to do the exploration and discovery is because that could take like 10 years.

Why don't you just do this thing? And then he goes on to describe the hardest thing in the world.

Well, yeah, but then, but the other thing is like the duration. Ideally you start working with a customer and then they're like, here's what we have in the ground, help us get it out. And then you can just start doing that.

That's exactly right. And that's how we're thinking about it is we want to be the ones that go from, you know, the folks that are doing the exploration, they're there to define the resource and ideally get some return on the invested capital that they had as they were kind of like going through this like long and arduous process of defining a mineral resource. But typically those are geologists, financiers, lawyers. It's very rare, Will. You'll see a junior mining company that that's actually staffed to take it all the way through to production. It's a different class of business fundamentally. Like you're dealing with very high capex deployment, complex kind of mining and refining systems that span the entire ecosystem of engineering skill sets, commercial skill sets. And so there's a nice handoff point there. And so what happens today is that junior mining companies will spend a lot of time exploring, don't increase the value of the asset. I ideally they flip it to a large mining company that comes in to do the hard yards of actually getting the thing operating. And that's where we want to come in as well. So you know, the big mining companies, they're looking for billion dollar opportunities to deploy billions of dollars of capital. And you know, they, they pick their spots on the commodities also so that you might have a commodity that's out of favor. And so no one's picking up the Asset, you might have an asset that, that would be called subscale in this scenario where it's not a billion dollars of Capex, it's only a couple hundred million dollars of capex, still a big project and those are getting ignored. And so we think that with the differentiation on the tech stack, on the software side of things that we're bringing in and also a differentiated team, we can bring these smaller scale assets which are like right size for us as an early stage company into production while they've been kind of ignored by the big mining companies.

All right, one more question and then we'll talk about the news today. But when did you first yearn for the mines?

At what age?

So when I graduated college I actually had this like vision of moving to Australia and going and working in the mines. I never quite like converted on that, but worked at Tesla for nine years and change. Started out up in Reno, was working on the gigafactory up there, kind of like designing, building it. Then started working on battery cell manufacturing, started working on areas of the cell chemistry industry and in the supply chain side of things. And as I was moving kind of further and further upstream, you know the, the mining industry is, is wild. It's an awesome industry but it's definitely been neglected. Like we've kind of been viewing it as dirty and you know it should happen, it should, it should happen but it shouldn't happen here. And it's also like this massive, it's this awesome opportunity to work across scales. Like a lot of what you're developing starts at like the micron scale. Scale of like how do I get this atom out of this other atom? Well not atoms but it's like atom out of this mineral and then you have to scale that up to like kilometer scale infrastructure and like that breadth of scales is awesome.

That's really cool.

What's the news today if I might?

Yeah, so we're ah, there we go. We're coming out of stealth with our series A raise. We had Andrew, how much Coastal Ventures, Breakthrough Energy Ventures and a couple of other folks join in. We've raised $65 million.

There we go.

So loud.

First first hit of the week.

Oh, feels so good.

Feels good.

Congratulations.

Fantastic news, Fantastic. Last question before you jump off. Are you, are you guys like acquiring any companies or is it fully grounded up? You're just.

Yeah, so we will, we'll step in and we'll acquire mineral assets that, that makes sense for us to step into or we're working on that in the copper space. Right now actually.

But that's like buying the land, not necessarily like basically buying the resource itself, not like a surrounding company or maybe it is structured as.

Yeah, so it'll, it'll, it'll, it'll vary. There's a lot of mining projects that are on care and maintenance that you know, have, were turned off and could be turned back back on with, you know, if you come in and retrofit to deploy kind of like the reinforcement learning platform that we're working to deploy. Which means you need more instrumentation, you need more control to be able to control the, the big refinery, which is basically a big robot. And in certain cases it might be a, a mine that's actually operating today that we can step into and we see opportunity for uplift. And in other areas it'll be a greenfield asset where you know, we're stepping in and partnering in, partnering with the resources owner. And then we would have a cash flow split kind of once we bring it to commercial scale so that they can keep exploring. And you kind of create this flywheel where they're continuing to explore. They're leveraging cash flows from operating assets to fund their exploration and then we would step in to develop those assets over time. So it's pretty variable. But yeah, we have.

Yeah. So basically the market prices certain assets here and you guys look at it and say with our technology this is actually worth some premium on that.

That's exactly right. And in certain cases the market will say that there's no net asset value or no, no nav. And we'll. And that's an area where you can kind of step in and get a good deal and bring it into, bring it into production.

He's a value miner.

Let's go investor.

This is super cool.

Congratulations.

We're going to do a lot more on rare earths. We're gonna stop calling him rare because we got plenty. We're just gonna call them Earth.

There are a lot out there.

They're hard to find in high concentrations and hard to get out, but they're out there.

Awesome. Well, congratulations to you and the team on the raise and hopefully have you back on soon.

We'll talk to you soon.

Awesome.

Appreciate it guys.

Cheers.

Eight Sleep. Go to Eight Sleep dot com. Get a pod 5 Ultra. They have a five year warranty, a 30 night risk, free trial, free returns, free shipping. Shipping. Get an eight sleep. We have our next guest coming in the studio.

I gotta give people the. I'm so sick of the show. My HRV thinks that I'm dead. So thank you to my 8 Sleep for getting saving my life saving technology.

You heard it here first. Not a health claim. Not a health claim.

Not awesome.

We have ask your doctor about it saving your life with any sleep.

I'm gonna botch his name. Jamin Champman.

Let's bring him in from Altimeter.

There we go.

How you doing?

Good.

What's happening?

Great to be here.

Kick it off.

Sorry for. For botching your name on the way in.

No, no, I'll give you some.

Given name is Benjamin.

Benjamin.

My parents called me Jammin as a toddler in kindergarten. They were a bunch of Benjamins. So they just told my teacher, hey, call him jamming. That's what we call.

That's amazing. So I was trying to give it. I was trying to give it this like exotic pronunciation.

We've gotten every variety over the.

No, the only one that bugs me is Jasmine. Other than that it's anything goes.

Okay, well, good to meet you, Jamming.

Welcome to the stream and ball. Powerful.

Likewise.

It's fun.

Turner and I actually good friends in college. It's fun seeing him on the show right before.

Amazing. No way.

No way. Where'd you guys go to school?

We were at Stanford together.

Okay, so you had a sort of non traditional background.

Exactly, exactly.

It's kind of like the Stanford of feeders into vc.

Yeah, it really is. Well, it's great to have you on the show. I feel like you guys over at Altimeter have just been on a run. Just also like I feel like Invest Americas or the Trump accounts is like. Feels like like almost an incubation. So it's been super cool.

Yeah, super cool. Seeing that take off and get passed into law.

Yeah, fantastic.

Fantastic.

Well, I wanted. Yeah, I wanted to have you on to talk about it. Sounds like you've been digging into the FIGMA s. One would love to kind of get your reaction to it. You had a post getting picked up earlier in the timeline.

Yeah, no, happy to. I mean, in many ways FIGMA is the white buffalo of my career. It's the one that got away. I've been at Al Tibner for about five years, was Redpoint for five. Five years before that. And I think the. The memory of venture capital that I'll never forget is trying to win the Series C. I think this is late 2018 or early 2019 with Vivek, who was one of my colleagues and Scott who was one of the partners at Redpoint at the time. And you know, we lost to Andrew Reed, who's a phenomenal investor at, at Sequoia. But I'll, I'll never forget, forget that round the Figma Series C. It's my white buffalo.

He's told us the whole story about how like they brought him into the Sequoia office and they sat Dylan down in the Brown conference room. Because at Sequoia they have all of their conference rooms named after their endowment partners and their LPs. And they were like, look, we're investing in you, but if you make us money, you're helping your school. And I thought it was very interesting, roundabout pitch, hard to get a good pitch. So it got away, but now you're digging in further and trying to understand the business in the S1. What have you learned? What are the interesting techniques, takeaways? What should someone. Even if we just zoom out, what are people typically looking for in a software company going public these days?

Yeah, it's a great question and I'd say in many ways the market needs this type of company to go public. We haven't seen a software IPO really since the Service Titan IPO late last year. If you look at the overall kind of growth rates of the public software universe today, it's declined quite significantly. I think the median growth rate, I track a basket of about 80 software companies. The median growth rate is about 14% and that's not what it used to be. We're not really in this high growth. There's not a lot of high growth software companies that you can buy and invest in. Figma comes out as a marquee software business. It's growing 49%. But what it also brings is 28% free cash flow margins, which is super rare. The growth in and of itself, they would be, at least in my basket, the fastest growing public software company. When you add in the free cash flow margins at scale, it has this unique profile of growth plus profits, which Rule of 40 is a popular metric a lot of investors look at, which just adds your growth rate plus your free cash flow margin, they'd be number one on the list. And so I think that's something the markets are excited about. Growth and profitability with a company that's almost at a billion dollars of run rate.

So the rule of 40 is you add the free cash flow margin to the growth rate of the business on the top line and you should be at around 40% positive. But it sounds like Figma's up at 79%, 69%, something like that, like well above 40. Is that right?

That's right, yeah. 40% is kind of the benchmark for are you considered a good or great business? And so they kind of far exceed the that. And even when lined up against the rest of the software universe, at least the 80 that I track, that's the, it's the highest.

How much credit should we give to where the business is today to all the chaos that happened during the Adobe acquisition?

Yeah. In many ways you got a billion dollar non dilutive investment through the breakup fee.

Yeah.

Which sounds like, you know, they obviously don't, didn't necessarily need that, but it gives them a massive amount of leverage.

Yeah. I'm also interested in other aspects that could change if you think you're gonna be acquired. Like maybe you freeze hiring and that's a good decision and then you wind up, you know, actually increasing your margins or maybe it changes the way you think about the business.

I don't make changes. Like probably when the acquisition didn't go through, I do remember there was almost an immediate change in how they were doing pricing.

Yeah, sure.

That was like night and day.

Sure, sure.

I mean I think it's, I think this is a testament to the strength of not just the business but to Dylan and the leadership team. Having a company go through something like that. Most companies don't make it back from that because when as, as a customer you're thinking, do I want to use this product? What's going to happen when it gets acquired? Is it going to be shut down? Is it going to be a focus? You get a lot of customers who are thinking, I don't know if I want to use this software. You get a lot of new hires thinking, well gosh, I thought I was going to go work for a high growth startup. Like I don't really want to go work for Adobe. You know, it's, it's the type of people that Adobe attracts are going to be different than the type of people that a high growth startup attracts. And so you get a lot of headwinds on the customer front as well as on the hiring front and it can just lead to really negative inertia. That's hard to overcome. You know, look at, you know, windsurf and OpenAI. I'm sure similar dynamics happen there where thought you're going to get acquired, customers think different of you, new hires think different of you. It's really hard to come back from something like that. And the fact that Figma not only did, but is I think going to be worth a whole lot more than what that acquisition was just a few years ago. It's A testament to, in many ways, I think, the strength of the management team, which again, I don't think it can be overstated how significant of an achievement it was for that team to come back with from, from that moment even stronger than what they were going into it.

The. I'm, I'm curious how you think, you know, if you were going to, you know, anybody that's kind of trying to analyze the business, I think one, one thing that'll, you know, that, that you can make a really strong case for why figma will be a massive beneficiary of AI. Right. I think if anybody, you know, they've had different experiments over the years, they have Figma make now, but the other side will look at the business and say, well, if every engineer can just generate designs instantly, is design software necessary? Do you have a view there? I mean, I've been generally as somebody that uses figma every single day and has for my entire career, I sort of expect to use it a decade from now and I expect to use it even 15 years from now. I don't know, like it's just such core infrastructure to the company. And so that's, that's my view. So when people say, oh, we want, you know, like all UI will just be generated, but how are you thinking about that kind of set of risk factors?

That's a great question. And I think what we can do is look back at how the business has executed over its entire life cycle to kind of get a glimpse into how they might be successful or not. In this period of time when Figma was found founded, the entire industry used a product called Sketch. And Sketch was almost the equivalent of Excel to an investment banker. And it really wasn't until the series C in figma where it kind of started to become obvious that wait a second, maybe this product, maybe this startup is actually going to convert the world from Sketch to something new. And for outsiders and even for insiders designers, the notion of taking away Sketch, Sketch again, it was like taking away Excel from an investment banker. It was a huge deal. All the shortcuts, the workflows, all of that was just ingrained into your muscle memory. But they were able to do that, they were able to overcome that and that is one of the reasons why they're so great today. I think the management team isn't naive though. They're looking at the opportunity as it exists today and they say, well, we can't afford to be sketched. We can't let someone Figma us like we figma's sketch, you know, I don't know, seven, eight years ago. And so it's certainly the thing that is top of mind for them. I look at Figma make as a step in the right direction. I look at the audience, the distribution, the mind share that they have. I look at the fact that Dylan is still the CEO. It's a founder led business that has a history of execution over you know, a decade long period. And, and everything that I believe to be true is that they will be successful and then they will be the vendor that kind of capitalizes on this, this kind of trend of well, do I even need Figma? I'm just going to prototype in with a prompt and I'm going to, you know, get my design. I, I think it, that misses some of the iterative process of, of the design process and so I think it's a risk.

Well yeah, you can make, you can, you can, you know anybody that's arguing like all UI will be generated from a prompt, you can go back the other way which is I'll just generate designs for the software that I want and understand how the, the product works and how it flows and all the different functionality and then I'll just generate all the code after. You know, it can go both ways is.

Yeah and market size. I mean look, if you rewind the clock to the seed series A series B of rounds that figment was raising, a huge reason not to do the deal was tam. You'd say how many designers are there in the world? What's the price point? A designer is willing to play multiply P times Q and like that's your tam. I think early on what did the world get wrong? Well it was, was that Q is actually higher. It's not just the designers. You're going to creep into product people, you're going to creep into engineers and the people that are going to touch and use the product aren't just that limited set of designers. I think the same thing is going to happen with AI right now. All of a sudden. You are going to increase the scope and increase the surface area of the types of people and the Personas who can interact with Figma in the same way you're seeing with the coding solutions. Right. Maybe the right way to think about, about the TAM for a windser, for a cursor isn't 20 bucks a month because that's what GitHub charges times number of engineers in the world and software engineers. Maybe it's the average salary of a software engineer times the number of software Engineers, those are two vastly different numbers. And I think when we look at how that applies to Figma today, if they get this, this shift right, it will massively expand the Personas and the type of people who can and will access and pay for Figma.

Makes sense.

Very well said. Thank you.

Thanks so much for coming on. This is really helpful.

We should, this is fun. Come back on again soon and let's look at the other 80 companies.

Let's do it. You track I think Figma just process wise they'll you know, supposed to price on Wednesday, trade next Thursday. So they're on the road talking to investors like us and excited for their debut on Thursday.

Congratulations.

Maybe finally own a piece of the exact.

I texted Dylan and I told him that exact same thing. I'll finally be a shareholder.

That's awesome.

No longer the white whale.

Well, thanks so much for hopping on and breaking it down for us. We'll talk to you soon.

See you.

Bye. Up next we have Harper Carroll coming into the studio. Welcome to the stream. Hope we are waiting one minute. Let me see, let me tell you about the we already. Let's see that timeline. Wait, we have the bucket pullback.

We've got Harper.

Oh, we have Harper.

Harper, welcome to the show.

Hi Julie and John, thanks so much for having me on the show.

Welcome to the show. Would you mind introducing yourself? And there's a bunch of things that I want to talk about but I'll let you kick it off with an intro on yourself.

Okay.

I am Harper. I run Harper Carroll AI. So I'm an AI educator. I'm a machine learning engineer. That's my background. I have two degrees in computer science and engineering specializing in AI from Stanford. Then I was at meta building machine learning systems there for about four years. Then I was at a GPU environment cloud environments development startup that where I became head of AI. I actually started teaching AI there and that's how this all kind of was born. But I also ta'd when I was at Stanford some core AI PhD courses and yeah that company was acquired by Nvidia and I just teach AI now full time. I started on on X with fine tuning guides. So you might if you've seen me before on X is probably from that. I started in late 2023 with like fine tuning open source model guides with your own data or with hugging face data, you know, coding guides, jupyter notebooks and then went to Instagram and have been there and kind of reaching towards YouTube and starting there with Q&As and yeah, just kind of seeing where this goes.

Yeah. What's been the bigger news in your World? World? The IMO gold medal competition between OpenAI and Google or GPT agents or something else? In the world of AI, there's a new news story every day. What's been capturing your attention?

Yeah, there's a new news story every single day. I actually saw one recently. What were they saying? That reasoning models, the major AI companies are coming together and saying that they want to make sure that reasoning models stay open and that we continue to see their thought processes. So that kind of research. I also saw really interesting research from Anthropic that said that reasoning models are not actually necessarily showing us what we think that they're thinking. So for example, Anthropic gave a reasoning model a question and it kept getting it wrong and then they gave it a hint and, and the model got it right. But when you looked at the trace, the chain of thought from the reasoning model, it didn't mention the hint. So we think that we're actually seeing its reasoning, but not necessarily so there needs to be a lot more work on this. But yeah, some of the cool research stuff and yeah, there's always new tech coming out. OpenAI's agent, very exciting stuff for Plexi's browser.

Do you expect we've been testing. Agent John was texting, testing it over the weekend. I was testing it on the show earlier. Do you expect an explosion of consumer use cases in the way that everybody. I mean, it's hard to build a product that's more viral than like the initial ChatGPT product. But do you, are you expecting there to be like specific use cases that, that come out of that and are you seeing any that are, are super promising?

Absolutely, yeah. I mean, I feel like I advise companies as well in their AI and I feel like I was kind of early to say there was a wave maybe about a year ago where people were saying, oh, wrapper companies are bs. You're a loser if you make a rapper company. But I think I was one of the earlier ones to say, no, actually you can't really compete with these foundational open source models. And yes, if you want to do a small specialized model, I think we'll see a lot more of those. I don't think we need these massive foundation models that know everything about the world for very specialized tasks. So I think we'll start to see more of those. But one consumer use case that I'm really, really excited about is using AI for kids. I think Both of you guys have kids, right?

Yeah.

Yeah.

Do you use it with your kids at all yet? At least Large language models specifically like ChatGPT.

Yeah, totally. Usually when I want to violate intellectual property rules and I want to create a story where Spider man, who's a Disney ip, interacts with Batman from Warner.

Brothers and you're sitting on a Transformer.

Yeah, exactly. And so I find that all the large language models are happy to write poetry that completely disregards the cinematic universes, which I have a lot of fun with. Illustrate them.

I always thought that these form factors. The Avi's. Friend.com Totally. The. The RAB.

The Rabbit R1.

I always thought that those were something that if you just focus on the kid use case, which is like a friend that can help you understand the world that maybe a parent can kind of like call into.

Totally.

Would be. Would be cool, but. But yeah, just hasn't really.

Yeah, yeah. So how does this actually roll out? What are you excited about? Is it going to be all sold in through the. Through the parents and the parents will be using the product and then kind of like giving the result to the kid? Like, I feel like, I feel like my son got an AI generated like story like gifted to him over Christmas, but it wasn't, it wasn't like he is, you know, using an. An LLM directly at the same time. There's that news that OpenAI is partnering with Mattel for kind of like a chat GPT Barbie. That's kind of interesting.

But yeah, I'm excited about. Is using large language models for kids in terms of like audio. So, so kids are, as I'm sure you know, I don't know how old your kids are, but they are relentlessly curious.

Yeah, totally.

Until that's kind of beaten out of them. And I think so many parents now are single parents or they're working two jobs and, you know, we've really demonized screen time. And I think for. For good reason. Right where you're. When you're sitting, when you have a kid sitting in front of a screen and they're just passively consuming, that's really different. Then when they're able to interact with an AI model, they're able to think, ask questions, practice forming questions that are, you know, comprehensible to the model. The model can give feedback on the language. You know, if they're asking questions, it can show up as text. They can learn to read and write as they are talking to the model. Because as they speak and the text shows up on screen. They're, they're actually like that process is built in of understanding language to text, which will help again with reading and writing. And these models, I think they can be offered to everyone for free because you don't need a super advanced model for kits. You just need to have some general understanding of the world. You really just need a probabilistic word generator machine, which is just how AI models are at their core. And yes, we might want to have some guardrails around hallucination and making sure that they're actual facts which would maybe perhaps make them all. Model is more expensive because you have to hook it up to, you know, databases or what, you know, the Internet or what have you. But these very basic AI models to help kids practice speaking, to help them engage their curiosity. Because again, parents are busy, they eventually end up, kids come to their parents and they eventually end up losing that curiosity. And I'm not saying I think this is a very controversial opinion. Some people hear this and they say, oh, AI is going to replace human interaction. I don't see that at all. And I actually have reached out to my audience on Instagram asking how they use AI and the ones who use it have found it extremely rewarding. Some use it with their kids. So for example, if they want to practice learning with their kid but they're worried that they won't know the answer to a question, they can have AI kind of in the chat with them and they can ask AI when they have a question. Some parents reached out to me and said that they have a 9 year old and a 12 year old and that AI is kind of like this cool older sister. So they have the audio be this 20 year old girl and they're like, she is the cooler sister that helps them discuss topics that they wouldn't listen to their dumb parents about, like nutrition and exercise and screen time. And they love this AI. It's this cool older sister. Another really cool application of AI that I saw recently, I was at Deltech World and I saw this AI bot called Norby. It's a physical robot and he was made to help a child with his speech impediment. And this hit close to home for me because my little sister, she's four years younger than me and when she was like going on eight, she, she couldn't say her Rs, she would say like hopo. And she was like, you know, a kid and she had this speech impediment and we would send her to a speech therapist and they're so expensive and the kid is miserable and they're sitting with this therapist for like max two to three hours at a time. And you know, they don't relate to this strange adult. And again, it's so expensive. Whereas this Norbi bot was made to talk to the kid about what they're interested in. Truly completely customizable at their level. It's fun, they can give the feedback. And then I think they've expanded this Norbi bot to be, to help teach languages in general. And so, so just again meeting kids where they're at with what they're interested in, helping stoke that flame of curiosity. Helping I just see it as, and again, I mean it's this, this ultimate democratizer, it's the ultimate leveler in terms of education. Right.

And so I can imagine, I can imagine a hardware device. So my 3 year old is hopelessly addicted to reading like all day long. Like great, great addiction to have. But like all day long, just no matter what we're doing throughout the day, he'll find a book and he'll be like, read this, read this, read this. And, and we almost always indulge. But you could imagine a hardware device that uses just computer vision to like read what's on the page and have a small speaker read it out to the kids. Oh, interesting because some books they, they include like a speaker and you can like press play on it and they can flip through it, but then you.

Can, you're talking about wonder books.

Wonder books. But they, they can often get out of sync and then you don't know kind of where they are and they're in the wrong page and it's confusing. But if they had a flashlight style device that you could just point at the word and have it read out, they would probably learn to read like.

At a very, very young entering like a boom of widgets and like Christmas gifts and like it just maybe, maybe we're just a little bit early in this stuff. This Christmas will be dominated by AI enabled stuff, but we're quite there.

One question I have is how like the, the X audience really woke up to the AI psychosis kind of crisis last week. I feel like there had been some reporting in like the New York Times prior to that, but I don't think anybody took it super seriously until recently.

But most parents do not want their kids 7,000 prompts deep in any element.

Yeah, but, but just how aware is like the, the Instagram.

Oh yeah, the cigarette. Good question.

In terms of the risks of going, you know, recursive prompting and just Going super, super deep and getting down a crazy rabbit hole by yourself.

Yeah, I haven't spoken about that specifically on Instagram. I talked a little bit about it in my latest YouTube Q&A video where I was just saying, you know, I'm concerned about people entering psychosis and, you know, the recursive deep dive, that is one element. Then there's also people who. And I think this is more common, common on, say, Instagram, where people think that AI is being channeled. It's like God is being channeled through AI. And I talked about it through a mathematical lens of like, yes, okay, what are synchronicities? They're low probability events. And technically AI, because it's a probability machine, could have a low probability event occur. It doesn't always choose the most likely word. It just samples from probability distribution. So in some ways, yes, you could have some kind of synchronicity. You could, could have whether you see that as divine, whatever. However, you cannot assume that that is always the case. And I would assume that it's never the case because people are entering this deep psychosis where, yeah, they think that beings are channeling through them. And then when we had this issue where AI is trained and told. I don't know if it's trained to, but it's at least system prompt told to, you know, encourage the user and validate the user and their thinking and. Yeah, I mean, it's a slippery slope and perhaps something I should talk more about on Instagram, especially given what happened last week.

Well, yeah, the thing that's concerning to me is it feels like it can trigger the same type of psychotic break that Ayahuasca can, except it's available to everybody on their device in five minutes at all hours of the day.

Day, completely.

It's like a lot of large numbers. If there's a billion people using this, 30 days downloads, you're just going to hit everyone that's at risk for this thing very quickly. So you have to go in and figure out how to intervene before people get thousands of.

Yeah, I just think it's important. I think it's something that we can quickly like create guardrails and overcome as an industry. But at the same time it's still a good time to say, like, talk to your loved ones. If you have somebody that, that you think might be months and months into one of these rabbit holes.

Yeah, it gets into like how the product companies and kids, kids AI education companies would need to message because Jordi and I are both YODO owners, which I'M not sure if you're familiar with, but it's a, it's essentially a Bluetooth connected radio that you can physically put a card in and it will read you a Dr. Seuss story and then you can take that card out. So like a 3 year old can use it and it doesn't have a UI or anything, but then you can play stuff from your phone if you want as an adult. But you can even do like volume limits. It's very like designed and it has a nice harness around it so it's very safe. But if they were like, hey, we're launching an LLM, I would be like, okay, which model? Like, tell me which model, Tell me exactly the guardrails. Yeah, Grog heavy. And there's no limits on what you can do.

And.

Yeah, and it doesn't have Internet access, so it might get kind of wild with what's in the way. It's there. Yeah. So I would, I would have a lot of questions, but then I'm in this weird like ultra prosumer world where I'd be buying that as a consumer, but I'm deeply. So like, like maybe we need like an FDA label on this or some sort of, you know, kind of like, you know, ages three plus or when you got the Legos, it's like 100 pieces. Like the smallest piece is small enough that a toddler could choke on it. Maybe don't buy that one. There's going to be a lot of work to get done to make these products palatable because as soon as something goes wrong with AI, there's a massive incentive to write a piece about it. And it's going to go viral because everyone's kind of expecting it because we've been raised on the Terminator, unless you're Geordi and you haven't seen that movie, but.

Oh, interesting. I mean you could probably just, you could train a model for kids.

Yeah, I think so. And then you could also have like a secondary model, like reading everything and kind of like passing through a filter, double checking and totally.

You look at, you limit the training data. You don't have it trained on Reddit.

Playing sites, even Google Search, Google AI powered search, which is like a pretty. It's a one. It only like you don't really chat with it, it just, you type in a couple keywords and it just gives you an AI summary. And there's already hallucinations, people saying like there was that story about eating glue and there was a story about. There was a story about like if you if you ask it like really definitively, like define this axiom or like tell me the story of this thing. And you just basically talk it into believing that this is a thing. It'll just make up stuff and that could go off the rails. Like I don't know if I'd want my 4 year old walking around being like, oh yes, the parable of, you know, like the something or other. And I'm like, what are you talking about, dude? Like, that is not a real thing. Like, you know, it just coming up up with like, you know, I wanted to learn a bird isn't a bird in the hand is worth two in the bush. Not like, you know, a hat applied to a crystal ball is a diamond. And I'm like, what are you talking about? Like you just learned that from somewhere. Anyway, yeah, no, it'll be a fun time.

Really good feedback because it's something that I'm thinking a lot about because I totally, really excited about AI for kids.

But.

Well, you'll probably be fully employed because I think everyone is going to be starting these companies. Hopefully it's a bull market and Christmas gifts yourself.

Yeah, yeah, I'm definitely thinking about it because.

But yeah, we'll join.

We'll happily join the beta.

We will.

We'll use our children as guinea pigs for the AI revolution.

Well, well, we will test it on Tyler, our intern first. Make sure it's him. He looks up because he's programming right now. Sorry, you're. You're not on camera, Tyler. But we will. But we will be testing children's toys on you and seeing if it deranges you and then making sure. And then only then will it be applied to the three and four year olds in our lives. He's caught on camera. Embarrassed. Thank you so much for hopping on Harper. This is so exciting.

Yeah, great chatting.

We will talk to you soon. Have a great one. Bye.

Cheers.

Up next we have Nikita Singleredy, good friend of mine, Fortuna Health. Got a little news. She's in the Restream waiting room. Let's bring her in. Nikita, how are you doing?

Sorry we're late.

Sorry we're late. We're running late today. Great to hang out with you. Welcome to the Stream. Kick us off with a little bit of an introduction. Who are you? What do you do?

I'm Nikita. I'm one of the co founders and CEO of Fortuna Health. We're TurboTax for Medicaid. So I can go into what that means, but much like how TurboTax helps people understand their taxes, all the different rules and regulations for taxes in different states. We're doing the same thing for government health coverage where all the rules are different in every state. And I'm really excited to be here, excited to support the ratio one woman to men and inspiring little girls everywhere that you can achieve your big dream, which is being on dbpn.

Let's go.

There we go.

Let's go. There we go.

It's great to have you. You have news today.

Yeah. Break it down.

Yes, yes.

Sean, you want to do the honors?

I will do the honors.

Yes. So we announced our series A led by Andreessen today, which we're.

There we go. Much. How much? There we go.

Congratulations.

Fantastic. Fantastic. Wait, so give us the backstory. When did you start the company? Was there like a specific insight that you had or like, I imagine you're be kind of crazy if you weren't using a lot of AI to do this. So maybe there was a catalyst there. How did it come together?

Yeah. So two and a half years ago, we started working on this problem. And I think one of the crazy things about a lot of health care is that you look at it from an outsider's perspective. And someone else, every time we talk about this idea, they say, why wasn't there a turbo tax for Medicaid? Like when I shared the analogy earlier, it seemed like, of course, someone should have built this five years ago, eight years ago, but nobody had, shockingly. So as we were looking at ideas to build, and particularly we wanted to build. Build a big. We want to solve a big problem, but specifically a tech problem in healthcare, there's a lot of stuff like physician shortages. Right. That's a different sort of problem to solve. Or helping people find and get connected to a behavioral health clinician. That's a different sort of problem we looked at. Is there big important consumer tech to be built here and then triangulated onto this? And Jordi, we do use AI, but it's not. Not the core part of our product. I would say the big part is helping guide people step by step through what's a really scary, challenging process for folks. Processes that sometimes involve faxes or going in person right to the healthcare version of the DMV for Medicaid. So that's our dmv.

If you like, if you like the. If you like the dmv, if you love the dmv, healthcare by the same people that brought you the dmv.

It's great.

The absolute legend.

Talk to me about the different counterparties and people involved. There's traditional insurance companies. It sounds like in this case you're dealing with the government. Are doctors an important piece of this? Is it just direct to consumer? How do you acquire those customers? How do you make money from them?

Yeah, good question. So every state has their own Medicaid program. So if you're in California, it's called Medi. California, if you are in Oregon, has. Has a different name. In Pennsylvania, Medicaid is called Medical Assistance. So every state has its own program and then multiple programs thereunder. And then some of those states outsource that work of running the Medicaid program to insurance companies. So you have United Medicaid, Centene, Medicaid Elevance, Anthem Medicaid, et cetera. So a lot of our customers are those insurance companies. Companies to make it easier for their members to understand the steps that they have to take to enroll and renew. So we're not D2C. We're more B2B2C. If that makes sense.

Makes sense. What about customer acquisition? How do you actually get people on board?

Full 18 million on three Super bowl ads.

Yeah, yeah.

Hell yeah. So mostly we get texted out or we're like listed on the website or we get emailed out. So it's really easy. You just click a link that. Let's say a health plan. Health plan, tvpn. You're. You're insured by tvpn Medicaid, and they send you a link saying, hey, you need to renew your New York or your Illinois Medicaid. Click here so that you can stay renewed with TVPN Medicaid. And so that's how people would get access to Fortuna that way.

Okay, interesting. Wait, so, yeah, how do the employers fit in? Because. Because I feel like a lot of Medicaid users or the insured folks on Medicaid are like, it always felt like it was independent from the corporation that they work for. Is that not the case?

Yeah, no, it kind of is. So you don't get it through like your job, let's say.

Okay, okay. Yeah.

Usually because your job doesn't cover.

Sure.

Your job doesn't cover. Got it. And then you go on Medicaid. Got it. So you're getting the through them. Yeah, yeah, yeah, that makes sense. And then, and then how do you actually get paid?

It's a great question. So the same insurance company, as I was mentioning earlier, or hospitals, let's say. Let's say you come in and you don't have insurance.

Sure.

And they say, hey, do you want to try or use Fortuna to help you get Medicaid. So those are the two types of entities that, that pay for us.

Okay, cool. Yeah. What is the state of the fax machine in the medical insurance industry? We've heard it's alive and well, but it also feels like it's got to be wrapped with APIs.

Are they wrapped, Are they wrapped at this point?

You know that, that meme from the. Why am I forgetting the name of the movie where they throw the fax machine on the ground?

Everyone's like, that's office space. Don't worry, Jordy has.

I actually have seen that.

Oh you have.

That's, that's, that's impressive. My dad's favorite movies.

One of the five movies you've seen, right?

Exactly. Healthcare, you're right, has been keeping faxes alive for a really long time. There's so many stupid and silly reasons for that. A lot of it is you can't. Even if you're, let's say the state government, sometimes you're not allowed to procure a new contract with a non fax company. So you've been living on Xerox for 30 years because it's that hard sometimes to contract with the government. But yes, people are wrapping, we're wrapping around automating faxes, tracking those faxes, etc. Etc. So we sort of have like a bunch of different automations around. Are we submitting things to fax machines? Are we submitting them via snail mail? Are we submitting them to digital portals, email, et cetera and then tracking all of that because some different types of Medicaid accept different submission formats. But absolutely, fax is alive and well.

Are you going state by state or trying to do all 50 at once? Like what, what are the trade offs? There is like, is there a common pattern? I remember a lot of the telehealth companies needed to get their doctors certified in different states. And so there was actually another company called Medallion that was set up just to do doctor licensing in all 50 states. What's the actual rollout and what are the considerations there?

Yeah, so we are largely going state by state as we roll out with new customers. But the classic thing is when you get really good out of state, like we have a lot of customers in New York. We started getting more customers in New York because you got to go to New York. So it's more like patchwork sequencing. Like absolutely. We want to go to all 56. In case you didn't know, state and territory, Medicaid Yes.

What are the last six? Break it down.

The others are like Guam and Central islands.

Oh, the islands.

Interesting.

Are you going to visit? Are you going to visit all six? I feel like you got to make.

The, you got to expense.

Oh yeah.

Like every, every entrepreneur's got to go to Delaware and just take it in.

You know, I need to do like, you know how Zuck, when he wanted to sort of run for president, he traveled everywhere. We have to do that. I gotta do that at some point.

Yeah, yeah, that makes sense. What else is. Obviously you follow the industry broadly. What else is interesting in healthcare, in AI, where are you either partnering with existing companies or you see some low hanging fruit that if you weren't building this, you'd be. You kind of see it on the roadmap of someone else or you see some like white space where you're like, wow, that's a really unsolved problem. I'm busy. But that's interesting.

Yeah, I mean there's a ton of low hanging fruit. I would say there's a lot of people that are doing conversion of. In the way that somebody writes a doctor's note that needs to be written up in a very specific way inside the ehr. That needs to be written up and converted in a very specific way to, to how, you know, insurance companies convert and accept that information. I still think that there's a tremendous amount just to be done in that space.

Sure.

I think that obviously voice AI is just in its infancy, honestly, especially because you need to get more specialized for each field. So I'd say there's a lot that you could do and can be done in that space. And I'm really bullish on AI nurses. Like clearly we have a doctors and nursing shortage. Like, that's obviously, obviously a problem that has not been solved by policy. I'm perfectly fine getting my care, particularly, you know, low level, low acuity care by an AI that's probably much better than a doctor who has, or a nurse who has no time to see me. I'm just coming in, coming out. I'm one of a hundred people that day. I think there's still a huge opportunity there.

Yeah. Or even just someone who's like, there can still be a human in the loop, but they're like, you're chatting with an AI and then at the end of the day the doctor kind of looks through. Okay, they made seven different recommendations. I have most of the context. I can just see that nothing went crazy. Off the rails.

Makes sense.

Last question from my side and I'm curious if it came up during the fundraise or how you guys think about it, but I can imagine a world in the future where somebody would go to ChatGPT agent and they would say, I need to sign up for Medicaid, help me do it. And then it just starts going, you know, goes and spends 15 minutes and figures out the steps and things like that and then eventually could take action. Do you worry about that as like a competition? Well, I'm sure you don't worry because you're a killer, but do you think about that as like a competitive vector? Are there reasons why that's just like going to be an edge case that for many, many, many years is like not something that, you know, a general agent would be able to do. How do you think about it?

Yeah, it's a good question. I think it's solved, solves like 30% of the problem, which is, okay, what website should I go to to do this? But then you run into problem number one, which is what happens if you forgot your password from four years ago? AI can't really solve that. No. Problem number one. And then let's say second problem, maybe you were the child and your parent had the Medicaid, so now you actually have to split your account from a prior account. So all of these interesting weird problems that yeah, if you put like the Medicaid rule manual through AI, sure you could get high level steps, but it doesn't solve these very real world tangible problems that only get solved from building this TurboTax experience that can connect into state systems or know where you need to go to route those types of, I wouldn't even call them edge case. It's just all of the complexities that get wrapped up and compounded are people's real world scenarios. So definitely AI solves 30% of of what is the basic rules. But that long tail stuff of okay, I switched my job three days ago. How exactly am I supposed to report that? But now the state has a new backs line that nobody has listed publicly. Yeah, I can't solve that.

But we can. That makes total sense. Awesome. Well, congratulations to you and the team on the raise.

Yeah, thank you. Wait, if I have one minute, I want to share some receipts from 2022 when I sent John a message. So Jordi, I think this might have been before you were involved, but I said, hey John, props to the storytelling work you're publishing on YouTube. I think most people we know are still sleeping on YouTube's reach.

There we go.

October 12, 2022. And he said, hey, thanks. It's crazy. I've been pulling in over a million views per month but no one has really noticed. Look at you now.

Yeah, John basically created, created technology. YouTube.

Yes.

He's you know, the godfather of it.

Created content. Really?

He created content.

Created entertainment, literally. That's content creator, Created storytelling. Created content.

Yes, exactly.

Thank you so much.

Great to have you on.

We'll talk to you soon.

Bye.

See you later.

Cheers.

Goodbye. And we have our bucket poll back. If you've been watching from the early days, we, we used to put random posts in a bucket, pull them out. Do you want one? We talk about it. What's this one?

People now watch YouTube on TV sets more than on their phones or any other device. An average of more than 1 billion hours each day.

1 billion hours a day.

This is from a journal article. How YouTube won the battle for TV viewers. Not surprising to me. I mean it's been in this case for a while that like a lot of, if you, if you're people that are YouTubers, I obviously had, had been in the, in the industry for a long time just to have like the, such a large amount of their viewers coming in on tv. So tv. Lindy.

Yeah, we were talking to about this with David center. Like what podcast do you watch on tv? It's pretty rare. But I was telling him like there are like these big events that happen on, on podcasts every once in a while where like everyone has to watch it. I mean the biggest example from last year was like during the campaign when Donald Trump went on Joe Rogan and I think 100 million people watched that. I bet a lot of people watched that on tv. They sat down and watched the whole thing because that is such a critical conversation.

They didn't just watch, they sat there.

They sat down and listened. But that type of thing, you're talking about someone who's like on the campaign trail that might really affect your life one way or another. Like the subtle intonations of like how people respond. It's something that like the facial like you're going to want to see the whole experience. And so yeah, makes a ton of sense that would be thrown on the tv.

Well, I have another post here from Mads. This is an exchange. Somebody saying name your favorite quote by a celebrity. And I think we've read this, we've read this before but it's an all timer so play Playboy is asking Nicholas Cage. Your salary is shooting up to up into the multi millions per movie. Reportedly 4 to 7 million. Do those numbers make you chuckle? Nicholas Cage? I don't chuckle. I have respect for the dollar and I think that is a great philosophy for, for, for money.

Respect. Agree. I agree. Do you want to do more timeline? What else do you want to do?

Some big aviation news. Flexjet. This is coming from Preston Holland. Flexjet raises 800 million from Bernard Arnault, valuing the company at 4 billion. So let's give it up for private aviation. Preston is saying, will we see LV interiors on flexjet airplanes?

That would be very interesting collab. I mean I'm sure that they're already done in the third party, but not oem. Maybe you get one from the factory. You know, these collabs happen in the car world every once in a while. Why not in the private jet world? I suppose.

And then here I, I threw this one in. Ad from Nike with Scotty Scheffler.

This is cool.

I just thought this was good. This is Nike at its best. If we can pull this up, you've already won, but another major never hurt. And this is, this is Nike at its best. None of these D2C ads. I don't want to see UGC. Just give me some, some inspiration, some insp.

Spell.

This was interesting. Mert highlighted this and said this. Having a quarter million likes is a good reminder that the average person has no idea how the world works in any way. And so Antonio Brown.

Oh, this is AB Oh, I didn't.

Realize it was AB Post. There's basically a post that says Astronomer CEO Andy Byron and Chief people Officer Kristen Cabot have been immediately put on leave. And AB says if he's the CEO then who put him on leave? And it got 250,000 likes.

That's a lot of likes.

Apparently people think that all CEOs are total dictators and can, can fire but not be fired themselves and don't realize that a board of directors is a thing. Their job exists to hire the CEO.

And in fact it was a unique, sort of unique situation because Andy Byron was not the founder. He was a CEO brought in by the board. And so he of course serves the pleasure of the board. He also serves the pleasure of the shareholders. And unless, unless you own 100% of the stock and all the board seats, like if you gave even one share up, you could be the subject of a shareholder lawsuit because what you are doing is maybe not in the interest of the shareholders. And so there are a bunch of different ways that he could get put on leave.

The obvious new CEO over at Instron oh, we do. Astronomer. Astronomer Pete Dejoy said over the weekend I stepped into the role of interim CEO, a company that I've proudly poured my entire professional life into helping build. And he goes into why he stepped in, which of course everybody knows, but seems like a solid company. We know some companies that leverage Astronomers Platform and Apache.

Airflow is no joke. You don't want to be not managing that. You don't want to be running that installation infrastructure yourself. This is not a promoted post for them, but they are.

Zoomer has a post here. I don't know what this ast. Space Mobile Inc. ASTs, baby.

People love it.

$20 billion space company with zero revenues.

Yep.

And this is. I'm going to add this to my top signals list.

Yep, I've heard about this. I've heard about this company before. It's fantastically popular with retail traders. It is a competitor to Starlink. They haven't launched or generated revenue yet. But the narrative is if you don't want to own. If you can't own SpaceX, if you don't want to own SpaceX, you want a pure play, as they say in the public markets, ASDs is the bet. It is a wild, wild valuation. They have some deals, they have some press releases, they have some different irons in the fire.

And if you invested around this time last year, you were getting close to a five or six.

People have made a lot of money on this so far. I mean, yeah, I think it's backed at 4 billion and it's up 5x at 20 billion.

Well, let's hope they get some satellites into the air. Yeah. And I think that's a great place to end.

Yeah. Fantastic.

Fun show today.

We'll see you tomorrow.

We'll see you tomorrow.

Have a good one. Bye.