https://www.listennotes.com/e/8f0f43350c944344838d0efa86607eca

You're welcome to TBPN.

This week, what were the top stories? What were the most interesting things that we learned?

100 interviews.

The Ukraine drone attack, that was huge. Operation Spiderweb, a whole ton of drones were smuggled into Russia in shipping containers. They emerged and went out and attacked bombers. We had Soren Monroe Anderson from Neros on the show to break that down for us. we also talked to Connor Love at Lightspeed, who does a lot of defense tech investing. And of course, a couple of weeks ago, we had Eric Prince on the show, the founder of Blackwater, and he had kind of predicted that the Ukrainian military was perhaps under rated. And we might be seeing something like this in the future. And so that was interesting to see play out. So we will take you through those kind of interviews, recaps of those. Then obviously we had the absolute meltdown between President Donald Trump and Elon Musk that unfolded on X and Truth Social. Duelling social platforms. Exactly. And although it's a highly political story, there are big business implications.

Totally.

Talking about what's going to happen in space between NASA, Boeing, different launch providers.

Yeah, the implications for Tesla, SpaceX, Rivian, Neuralink, even the Boring Company, right?

There's a lot of stuff.

Many of Elon's businesses are heavily regulated.

Yep.

And the potential impacts are substantial.

Yeah, then we also had some earlier stage founders on the show. Roy from Clulee came on and went pretty viral, put on a show. You're surrounded by journalists. Hold your position. Clulee is a service to help you cheat on everything. We gotta actually try the app, someone was pushing us to try it and see how good the product is. But regardless of the product, he's also a phenomenal marketer and he came on and put on an absolute show.

And he's printing. apparently.

Yeah, he's doing great.

He can't spend all the money that they're bringing in.

And so we'll give you an update on Roy and see where that business is. And then Keon, from Nucleus, came on to launch Nucleus Embryo, which he calls the first ever genetic optimization software that helps parents give their children the best possible start in life.

Quite a lot of controversy there on the timeline this week. A lot of people hate it. A lot of people love it. And we'll let you kind of decide for yourselves.

And then we had a whole bunch of AI experts on the show from Google, OpenAI, and Anthropic, got to the front leading edge of the debates around AI, AI, LLM.

Poor John yesterday, you were fighting as long as you could. You didn't want to talk about drama. You got dragged into it. But we did get some great coverage from Mark Chen at OpenAI, as well as Sholto over at Anthropic.

Yeah, it was a lot of fun. The big news over the weekend was the Ukraine drone attack on Russia. They shipped shipping containers into deep into Russia, at which point drones flew out of the containers and hit strategic targets. We're going to have two guests on the show today. Soren Munro Anderson from Neros to talk about that and also Connor Love from Lightspeed to talk about that and also Defense Tech Investing generally. Today we have Soren from Neros who builds drones and has been to the Ukraine and so we'll bring him into the studio and Ask him how he's doing. How you doing?

There he is.

Welcome. Great.

How are you guys?

We're good. We have new soundboards, so expect some.

Wild some wild cards wild stuff. Could you give us a high level overview of the history of drone warfare in Ukraine? Because I understand it's been progressing super rapidly on both sides and it'd be helpful to understand kind of the different stages.

Did they ever have like predator drones like the global war on terror type of drone, or do they jump straight to quadcopter and kind of leapfrog the technology?

So, you've had this Russian aggression war in Ukraine since 2014. Obviously, the full scale invasion was 2022. But even during that period before the full scale invasion, there was some usage of drones for surveillance and dropping explosives. These are primarily still small drones like what you're seeing now. But this was not a proliferated technology. then when the full scale invasion happened, within a few months, Ukrainians started thinking about all these ways that they could use, you know, inexpensive drone technology to get an asymmetric advantage. And that is where FPV drones started becoming a really, really big deal. So they pioneered the really this idea of, you know, putting an explosive on a racing drone and using that as a precision strike weapon. There were instances of this happening in other places, but they really scaled it and they've really refined it. And then Russia was much slower to take it seriously, although now they tend to in some ways out produce Ukraine and they have a much more direct line to China where most of these components are coming from. But since 2022 and FPV is just starting to get used, now it's reached an unbelievable scale. It's estimated Ukraine is going to produce four and a half million FPV drones this year, and those are ranging from ones that are this big to 15-inch propellers, fiber optic controlled drones, many different types and sizes of warheads, different configurations. And I can talk more about the drones that were used in Operation Spider Web as well because those were really interesting. But what we've seen is just this vast technology landscape where new clever ideas like fiber optic are going to be the hot thing for a few months and then they sort of just become another tool in the tool belt. and it's just this constant arms race.

Yeah.

Talk about this attack was unique in a bunch of different ways, but is this something that had been, to your knowledge, or just more generally known to be something that had been attempted multiple times? Or maybe, like, I'm curious to know, yeah, kind of the backstory on this type of attack, 'cause it seems, you know, it's a massive difference to be using this technology way behind enemy lines versus using it, you know, at the front line.

Yeah.

So primarily, fpv drones are used on the front line, say, the kind of 30 kilometer band across the zero line. What was so unique here is that it was fpv drones, short range drones being used 4, 000 kilometers inside of Russia. It was this unbelievable application where, you know, we You've seen Ukraine using long range one way attack drones that are going 1500 kilometers to strike targets deep inside of Russia. But here, these were small drones actually driven in on trucks, basically in the tops of shipping containers. And I don't know of any operations that were similar to this beforehand. I think it was not something they wanted to give away.

Yeah.

And the drones were actually operating on cellular. They were not operating on local, like the normal low latency local radios you use for FPV is typically. And so I think, you know, this is gonna be something that a lot of people are gonna look at and see. If you have drones that are operating on cellular, you can't really tell them apart from cell phones. That's really hard to defend against, really hard to detect. But now it's going to be part of air-based defense is, thinking about drones that are operating on cellular being piloted from basically anywhere in the world.

Talk about the Russian response, the immediate response to this incident from the footage that I saw, and I think most people saw that tracked it. It seemed incredibly challenging to respond to it quickly, right? By the time you could sort of organize a response, a lot of the core damage had been done. What do you think the the, the, the question I think that every country is asking themselves now is how do you defend against this type of attack, whether you're at war like Ukraine and Russia are, or you're just, you know, thinking, you know, long term? Yeah.

This clearly poses a massive threat to critical infrastructure. I mean, being blatant, the US does not have any defenses in place that would stop this from happening. We already know there's already news stories about drones that are flying over our Air Force bases and we can't do anything about it. And I think the only approach here has to be a multi-layered system where you're looking at all of the different types of electronic warfare and also considering things like satellite communications and cellular communications where you're basically able to turn those off on the flip of the switch, which is a huge inconvenience and a huge thing to build into the infrastructure. But clearly that's going to be required.

Welcome to the stream, Connor. How are you doing? I'm good. I'm doing all right.

Good to be back, guys.

Yeah.

You've got a suit this time.

Oh, looking great. We love to see you.

I won't say I dress up just for you, but I would have taken the suit off far before this if.

I wasn't coming on. Fantastic. Thanks so much for jumping on. Have you been tracking the Ukraine story closely? Any insights there? Anything in the portfolio that's at all relevant in the defense tech world? expect a response from the US government or guidance or change to any strategies? Really any takes on that?

I mean, first shit, what a time to be alive. I mean, you know, I'm sure your Twitter feeds and your group chats were blowing up, pun intended, over the weekend. I mean, it's pretty crazy. I mean, let's be honest, like, first, I'm not shocked that the Ukrainians did this. I mean, the execution seemed to be flawless from what we can pull from open source Intel. I do think, though, I mean, again, it's not a surprise that the Ukrainians have been mastering drone warfare for the last handful of years. And you want to call it that, they called it Spiderweb. This was their Trojan horse. This was their Israeli beeper. And the outcome is pretty impressive, to be honest. I mean, from the outside looking in, the Russians woke up over the weekend and they thought they were getting their $4 Temu orders.

And what did they get?

They got a thousand FPV drones, blown them to smithereens. It's pretty impressive. I mean, my takeaways from this are really twofold.

Please.

The first is like, there's never been a clear signal of where warfare is going. And to be clear, what I view this from both the entrepreneurs in my portfolio, but also from my perspective, I mean, the world is about cheap, attritable, a lot of times autonomous systems. And that's playing out in warfare, that's playing out in other areas of life. And then the second thing is, candidly, it's really hard to defend yourself at the pace at which things are changing. And again, I know we do some things here in the United States and are trying to be on the front end of a lot of this innovation. But when this happens, I think this almost just resets everyone again and says, All right, how do we respond to it? And I think it's, to your point, it's not a direct US response. It's more of, Hey, what do we need to buy? What do we need to develop for our own fight in some way, shape, or form?

Yeah.

What do you think? Obviously, you're a venture capitalist, not a geopolitical strategist, but what's the right Russian response to this? Is it, hey, we suddenly need to be wary of having cell coverage anywhere near strategic military assets? I mean, it seems like Ukraine in Ukraine's perfect world, they could run this style of attack a bunch and copy and paste and hit other targets. It feels like something that was dependent on cellular technology, that that's something that the Russians can revoke fairly quickly. Sure, it'll be inconvenient, but I'm curious if you have a take.

Yeah, I mean, to be honest, when I think about how do you defend against this, I think there is-- I wouldn't call this the easy answer of just turning off the cellular network. I actually think the only way to do it practically is in layers or in a multitude of different ways because, you know, yeah, the, you know, the reality is if you looked at how the Ukrainians carried out this attack, they did so on the local, you know, Russian cell network, which again, I don't think any Russian kind of defense unit any of these basis was ever thinking that they would have to turn off their own cell network. And then there's just the practicality of how you do it. I mean, I think there was, what, four or five different attacks that hit all at the same time.

What do you do?

You turn off the network for, Tens of thousands, hundreds, thousands of people. And oh, by the way, this is like a dirty little secret that nobody talks about. You know, yes, you have your military systems that are protected and all that, but a lot of coordination is happening through WhatsApp. A lot of coordination on, and so all of a sudden, you turn off the cell networks. You're actually inhibiting your own defense, your own response, the first responder, you know, the, you know, getting your own people out of there. So I think it's a bit more complex than that. And then the last thing I'd say is just like, Even if you do this in layers, you need to be resilient in a way, but you're not going to stop everything. I mean, this was just brilliant master class of, again, if maybe there was a plan, we didn't know this, but maybe there's a plan for 100 bases and we only hit five of them. And so if you think about just the broad geographic coverage you have to have, I think to be 100% certain on anything, it's just you, It's impossible. You can't do it.

The most capable military in Europe right now is the Ukrainian military.

The lessons learned that they have are very significant. The drone tech is far and away the best.

Their ability to fight against and to even conduct electronic warfare and even close air support in this environment is leaps.

And bounds ahead of even what the US military is.

So that's the military to learn from. Ukraine does have a corruption problem.

I hope sincerely that Trump is able.

To get a ceasefire in place and to stop this killing, because it's absolutely pointless. It's just Slavs killing Slavs at this.

Point, and nobody's going to advance. And you have a blend of old and new.

If you look at the pictures of the front there now, it's almost indistinguishable.

From the Battle of the Somme, right?

Artillery duels, static lines, bunkers, all the rest. Now the problem is somebody can fly an FPV into your bunker on the other side, but between tens of millions of landmines, which make armored breakthrough very difficult, it slows down any attack so that the FPVs and artillery can get to it.

You're not going to see any kind.

Of Blitzkrieg, Hans Guderian, maneuver warfare there. until some significantly different weapon systems come along. So look, Europe needs to get serious about it.

They're far from it at this point.

Elon post four minutes ago, the Trump tariffs will cause a recession in the second half of this year.

Wow.

Somebody else was saying, can I finally say that Trump's tariffs are super stupid?

Who is that?

Somebody else is posting. Mads posting is saying it's a Xi Jinping He says, bro, you seeing this? And it's Putin on the other end. He's just looking at it. Hold up, got a line. And it's, we'll start pulling some of these up.

Ridiculous. What else is going on here? This is the president versus Elon.

Nival says Elon's stance is principle. Principle. Trump's stance is practical. Tech needs Republicans for the present. Republicans need tech for the future. Drop the tax cuts, cut some pork, get the bill through.

This is so crazy.

Antonio Garcia says, remember, there's F you money, and then there's F the world money. Will Stansel says, imagine being the ice agents suiting up for your biggest mission of all time right now. People are saying that Trump's going to deport you.

I'm back to South Africa.

Will DePew says time to drop the really big bomb growing Daniel is in the Epstein file.

That is turned into a coffee pasta.

That is a real.

Oh, no.

Delian, we had a question from a friend of the show. They said the real question is if Tesla is down 14%, how could SpaceX and OpenAI be trading? if they were. How would they be trading if they were public? The, the real thing here is it's bad for everyone.

Right.

DJT.

Trump coin is down. Nobody's really winning here. China is up.

Yeah. Oh, really?

Sean McGuire. I mean, I'm just saying, like, at a high level.

Yeah, yeah, yeah.

You know, China is the big beneficiary here of Sarah Guo says, if anyone has some bad news to Barry, might I recommend right now?

Yes, yes, yes. If you have If you, what's the canonical bad startup news? Like, oh yeah, you missed earnings or something. Drop it now.

Inverse Kramer says Bill Ackman is currently writing the longest post in the history of this app.

And we have a video from Trump here, if we want. I can throw it in the tab and we can share it on the stream.

Reacting to the live. Lex Friedman says to Elon, that escalated quickly. Triple your security. Be safe out there, brother. Your work, SpaceX, Tesla, XAI, Neuralink is important for the world. We need to get Elon on the show today.

If somebody's listening and can make that happen, I would love to hear from you.

Max Meyer says, so I got this wrong. I didn't say it never happened, but I thought it wouldn't. I'm floored at the way this has happened. He didn't think they would have a big breakup. Many people didn't think they would have a big breakup. Even just earlier this week, it seemed like they might just have a somewhat peaceful exit. Trump just posted a little bit ago, I don't mind Elon turning against me, but he should have done so months ago. This is one of the greatest bills ever presented to Congress. It's a record cut in expenses, $1.6 trillion, and the biggest tax cut ever given. If this bill doesn't pass, there will be a 68% tax increase and things far worse than that. I didn't create this mess. I'm just here to fix it. Anyways, lots going on.

Let's go to this Trump video. I want to see what he has to say. I've seen, and I'm sure you've seen.

Regarding Elon Musk and your big beautiful bill.

What's your reaction to that?

Do you think it in any way hurts passage in the Senate, which of course what is your seeking?

Well, look, you know, I've always liked Elon, and so I was very surprised. You saw the words he had for me. The words are for and he hasn't said anything about me that's bad. I'd rather have him criticize me than the bill, because the bill is incredible. Look, Elon and I had a great relationship. I don't know if it'll well anymore. I was surprised because you were here. Everybody in this room practically was here as we had a wonderful sendoff. He said wonderful things about me. You couldn't have nicer. Said the best things. He's worn the hat. Trump was right about everything. And I am right about the great, big, beautiful bill. But I'm very disappointed because Elon knew the inner workings of this bill better than almost anybody sitting here, better than you people. He knew everything about it. He had no problem with it. All of a sudden, he had a problem. And he only developed the problem when he found out that we're going to have to cut the EV mandate because that's billions and billions of dollars. And it really is unfair. we want to have cars of all types electric. We want to have electric, but we want to have gasoline, combustion. We want to have different. We want to have hybrids. We want to have all we want to be able to sell everything. He hasn't said bad about me personally, but I'm sure that'll be next. But I'm very disappointed in Elon. I've helped Elon a lot.

Did he I just want to clarify.

Did he raise any of these concerns with you privately before he raised publicly. And this is the guy you put in charge of cutting spending. Should people not take him seriously about spending?

Now, are you saying this is all sour grapes? No, he worked hard and he did a good job. And I'll be honest, I think he misses the place. I think he got out there and all of a sudden he wasn't in this beautiful Oval Office and he was, and he's got nice offices, too.

But there's something about this when I was telling the Chancellor, folks, this is where it is. People come in. Breaking news. Delian has DruHov is joining us in the temple for some live reactions. Come on in.

Guys, guest, I can't even spell surprise guest. I'm so excited about this. Surprise.

Yeah, in other news, 11 Labs dropped a new product.

Another news. $2 million seed round. Stop it, stop it. We love 11 Labs. No, they'll keep grinding, but just launch again. You're going to have to launch again. Start shooting a new vibe reel. Start shooting a new, writing a new blog post, 'cause no one's going, Lulu says yes, delay the launch on TVP.

So, basically right now, I can just pull up and just refresh.

Here we go.

I'm gonna just be refreshing Truth Social.

What are you doing? So, okay, Jordi's on Truth Social. I'll be on X. Give us your reaction, Dallian, what's going on?

I mean, at some point I was like, I'm just, you know, sort of scrolling X and I like tuned into you guys like an hour ago and I was like, they're talking about some AI thing. I was like, at some point they're.

Gonna switch to like, we're hacking news.

And then I was watching and I was like, okay, like, John resisted. I fought it for like, for like a half an hour, but we couldn't do it.

But yeah, give us your quick reaction. I mean, I'll always give it from the space angle. It's amazing how much the world has shifted since Friday of last week, where it was presumed that Jared Isaacman was going to be the NASA admin to today, it was released that the Senate reconciliation package re-added budget back into NASA, largely for the SLS program, which was basically the program that Jared and Elon were largely advocating to, you know, sort of completely shut down. So, you know, it is already showing, like, you know, the sort of counter reaction, you know, is already showing up, you know, in policy. Sorry, SLS program, is that space shuttle or no?

Oh, sorry, that's the SLS launch rocket.

Okay. It's based off of old space shuttle hardware, but it is basically the internal, you know, sort of NASA run competitor effectively to like a Starship heavy launch rocket.

Yeah. Yeah.

Because it was, you know, sort of generally behind budget, behind schedule, and so many commercial heavy lift rockets coming online.

Sure.

The default was canceled. That is largely a Boeing based program. And so, you know, if you look at, you know, three months ago, you know, when they were announcing the F-47 program, you know, Elon walks into the Secretary of the Air Force's office. Obviously, he'd been, he was sort of ranting against, you know, manned fighter jets and believing that that shouldn't be what, you know, be what the department is prioritizing. 30 minutes after that meeting was when they announced the F-47 program. And so now you're seeing basically like the equivalent in space where That was obviously awarded to Boeing. Boeing is the largest prime behind SLS. Boeing basically is going to be the biggest winner of NASA refunding SLS and Jared Isaacman not being NASA administrator.

Tying this back to the timeline, Trump posted less than 30 minutes ago, In light of the President's statement about cancellation of my government contract, SpaceX will begin decommissioning its Dragon spacecraft immediately.

Break that down?

I mean, that just means that we no longer have a vehicle that can go to the International Space Station. We no longer have a vehicle that can bring astronauts up and down. We also don't have a vehicle that can deorbit the International Space Station safely, right? The Dragon was expected to be able to do that. So what that means is, if you guys remember all the memes about Stranded from last year around Boeing Starliner, it now means that the space station itself is basically sort of stranded. And that's like one of the government contracts, obviously, that SpaceX is involved in. ElaN, I've heard generally just wants to shift all things to Starship anyways. And so in some ways it was probably looking for an excuse to shut down Dragon and refocus energies. There's also a part of where it's like, look, he is kind of independent in the space world in that Starlink's total top line revenue is going to be passing the NASA budget in the next year or two. And so in terms of size of state actor that can influence space, his own company is basically about to become as large of an actor as the entire United States. So I don't think there's going to be de-escalation here. Like, you know, my, my, you know, estimation is like, on both sides, it's going to continue to escalate. You know, if we thought that we lived in dynamic times, you know, when Trump got into office, it's going to be even more dynamic when there's, like.

The dynamism will continue until morale improves.

Elon Musk Center, AOC, the progressive populist in Trump, the, you know, sort of conservative populist and And man, it's- it's really good to be on the timeline.

I mean, I just have so many questions, right? How does this impact Golden Dome?

What's Boeing stock doing?

Will Golden Dome even be a viable project without SpaceX?

I think there's just going to be more resistance probably to working with upstarts because they would be ones that would probably be more likely to collaborate with SpaceX. And so, It feels like Boeing would.

Be a logical beneficiary of this turmoil and yet they're down today. They haven't really popped. Oh really? Yeah. I mean, I'm not obviously, you know.

One to give like, you know, public stock. Yeah, I know.

I'm just kind of working through it myself and it's surprising. It just feels like it's cut like.

A drop in Boeing to pop basically off.

Yeah, yeah, that would be the expectation, but there must be something here because it feels like this is purely interpersonal between Elon and Trump and not, it's not like, oh, Boeing was secretly behind the scenes the whole time lobbying even more effectively. It doesn't, oh, he got the, well, where's the tinfoil hat? It's over there. Maybe we need a tinfoil hat segment, who knows? But yeah, I mean, when you're in Boeing world, it's like, hey, we're only down 1%.

Let's go. The cool descent. My question is, has there ever been a crash out of this magnitude ever?

In history. Well, in internet history.

When Elon and Trump became friends.

World scale. I actually think probably world history equivalent. I feel like there was something in like, Maybe we didn't have it in the United States where, you know, there.

Was a crashing out. Crashing out used to mean calling up the New York Times and just ranting. Now you can just live post like all your reactions and it's just all real time. This is like crash outs are actually intensifying. You actually want to be long crash.

Outs over the next.

And people have their own platform that they own. So, you know, you got to be on both X and Andrew's socials. to like stay on top of things.

Yeah, yeah. I actually did like a deep research report a while back on like, has the richest man in America ever been close with the US president going back to like, you know, was Rockefeller particularly close? And and because the narrative was like, oh, this is like so unprecedented. And in fact, it is unprecedented. Oh, really? Yeah.

I would have guessed that like Rockefeller was close.

Me too. Me too. That's what I was going for was like, no, I imagine this is always this is always close, but no. I think because the president has become more powerful globally, your point about, you know, mayor of America, dictator of the world, like it becomes increasingly valuable for the richest man to have a close alliance. And so it's become more, I don't know exactly how accurate that research was. It's totally possible that like behind the scenes, Rockefeller was really close to the president at the time and we just didn't write about it in the history books. But there certainly aren't very many anecdotes about the richest man in America going on.

Yeah, so Paul had a press conference.

Ready for AP US history. 2025, you know, ABCD.

Yeah, so this is where, you know.

Elon Musk called the president at the time a potential pedophile. Was it A, Epstein Island, B, about a cave in the Philippines, C, what a mess?

No, so Pavel had a good post. He was quoting the big bomb from Elon. He said, hypothetical question about the USA's power structure. Is the man with the most access to capital more or less powerful than the political head honcho? Purely hypothetical. It's a good question to ask.

I mean, I think both like archetypes have grown both in absolute power, but also in relative power to the rest of the globe, basically since the Gilded Era, right? If you think about like the president of the United States in 1925, I'd say pretty darn powerful, but like there was clear like, you know, it was a, you know, sort of multipolar, you know, sort of world. Argentina was pretty darn rich at the time. Obviously, Europe was still recovering from World War One, but UK was generally doing well. It was clear there was a huge, you know, outweight effect. And then if you look at probably the biggest industries at the time, I don't think you could claim that even like Standard Royal at its peak, I'd have to go look at the exact numbers, but that it had the size of budgets relative to the US government in terms of budgets, right? Versus I feel like now for the first time, You both have, you know, US presidents extremely, extremely powerful. And then you have like, you know, sort of mag seven effectively like the size of, you know, sort of, you know, huge states, like they're, you know, state governments.

And then also just more bureaucracy, more red tape. So like, I, when I think about the 1920s, like Robert Barons, it's like, it is the, it is the you can just do things era. And so you want to build a railroad, like, yeah, you might need to get like one rubber stamp, but it's not going to be 10 years. tons of lobbying and all this different stuff. So you can kind of just go.

You can just go wild. You know, it's bad when Kanye is saying, Bros, please know we love you both so much. It's just like the voice, the voice of reason is Kanye West. Yes.

Yes. Thank you.

We need to bring them together and, you know, form a peace treaty.

Nikita beer just added his pronouns back to his bio.

Let's go.

Let's get a rubber band. Elon's got a rubber band all the way back to, you know, sort of extreme woke. because I'm straight back to, you know, sort of super climate change and wow.

Somebody'S somebody's sharing, resharing the picture of the, the cyber truck blown up in front of the Trump Tower in Vegas. And it's just like this.

This is real.

It was foretold.

Yeah.

But it was a question of, like, when and what magnitude, not if.

Always bad if, if Vladimir Putin is operating to negotiate.

Between.

President Trump and Elon. I think a lot of the world is waiting for Roy Lee's take, and the Clulee Army.

That's who we want to know.

People have been asking him to get involved with geopolitics.

Wow.

I love the Shel Mojad put up a sort of meme about Narendra, like Prime Minister of India. He basically copied and pasted the Trump Truth Social post about negotiating peace between India and Pakistan when it wasn't actually fully negotiated. Posting about negotiating a ceasefire between Elon and Trump.

Funny thing is, Truth Social, you can just read all of Trump's posts without creating an account. It truly shows that, I would think that you would have to make an account to read them all, but it's not gated at all. This could be the biggest, but they clearly, I don't think they care about monetization. Bitcoin is actually falling alongside. Falling.

Wow.

Bitcoin falling, Boeing falling, Tesla falling. Who's the biggest winner of the day?

I think it's China. China. Yeah. China.

China.

Sean McGuire.

Bitcoin really sold off. It's down 3% today at 101k. So still up, but rough.

Winnie the Pooh just dipping his hands in that pot of honey. Just snacking away, watching from the sidelines.

Yeah. Let's see. Chinese stocks. US stocks. Chinese stocks.

I can't find it. Okay.

That's probably my commentary on the day, boys.

Anyway, this was great. It was fantastic. Thanks for jumping on. Thanks for hopping on so quickly. Next up, we have Roy from Clulee coming back for an update.

He's hired 50 interns, I think, or.

Something close to it.

He said they're bringing every intern on.

They're bringing every intern on. We got every intern coming in. well, welcome to the studio, Roy. How are you doing?

Oh, boom. Let's go.

There they are.

I think we're overpowering you.

Can you, can, can you hear us? Yeah. Yeah. We can hear you. Yeah.

Make sure we're zoomed out all the.

Way so we can see everybody. We got a small army here.

This is incredible. Uh, how big is the team kick us off? How many you got at this point?

The team is, 11 full time plus the interns.

How many interns you got so far?

Interns, bro, we're closing in on 50, brother.

Let's go. That's amazing. Congratulations. What are they all doing? How do you manage everything? Is it purely social media? Is that what you want them to focus on? Growth.

Yeah, yeah.

Growth marketing. The only goal of the company is get 1 billion eyeballs onto Clulee. So we have unrestricted creative freedom and permission to do anything and everything. Just make the company go viral. Every single person you see behind you has over 100,000 followers on some social media platform.

Wow.

200,000 plus.

That's remarkable.

Me too now.

Yeah, there you go. Oh yeah, yeah, yeah. Probably popped. What's working, what platforms have actually been driving the most growth? I mean, I'm sure you've run a lot of tests. What have you learned that you can share?

Bro, Ben, take it away, bro.

Let's hear it. UGC has been really good. We just hit 10 million views today. 10 million views.

10 million views.

Eight days.

Wow, there we go.

Hoping to get 100 million views in the next month. What platforms specifically are the most fertile ground for targeting your specific customer? Because you can imagine that there's a lot of folks who are AI curious on X, but then there's much broader, more viral audience, more general audience on platforms like TikTok, YouTube, Instagram, what's working and what is the next platform that you're gonna be focused on?

Yeah, we're trying to go viral on every platform regardless.

But the main thing right now is Instagram Reels.

Oh, Instagram Reels. Interesting. And what is the main value prop that you're hitting people with? Is it still the cheat on tests thing or have you evolved at all? What? Still? Actually, interviews, yeah. Interviews, okay. And has there been, this was controversial when you launched it, is it still controversial in the comments? Are you getting flamed? Has anyone Big dunked on you and has that driven virality? Is that actually a net positive? Instagram is not like Twitter.

Like you could post the craziest shit on Instagram and they will still not think it's controversial.

Really?

So to make it controversial, like we have to engage in bait some other way. Like a heating tool is controversial on Twitter, but on Instagram, you could have like a white guy say the N-word 10 times and it's still not controversial enough. You know, like you need crazy shit on Instagram. That's what we cracked. Every single person here has like very great viral sense.

Yeah.

Watch the reels that do go viral. You see there's ways that we've engagement baited the videos and this is what we'll keep doing to probably a billion views a month.

How long does it take to figure out if an intern is cracked? Is it like an hour, two hours?

How much time do you need? For me personally, me personally, probably like 10 minutes, but for anybody watching probably.

Would take like one or two weeks. There we go. How do you guys think about product marketing? Obviously you're just going viral everywhere getting all this I don't know how he keeps attention. How do you make sure that it, that it, he's shaking his head? Doesn't think about, it's not about the product, it's about the attention. It doesn't make anything go viral.

Yeah.

Yeah, but, but, but how do you, the side of the street, you know, you make some UGC videos, make some Twitter posts, you know, you can sell anything, you know, in 2025, product doesn't matter. You know, I could jack off off the side of a building, sell some videos of it for 20 bucks each. Make $2 trillion dollars. It's crazy. 2 trillion. That's intense. How do you guys think about burn? Is it on your mind at all?

I don't know if you saw the last tweet, but as of literally like two days ago, we're still cash flow positive.

We're still fucking profitable. We're still profitable.

Let's give it up for the profit.

Let's hear it. It sounds... So you're charging for the product and people are paying.

Are they at all satisfied or do they feel like they got scammed? Bro, like the product works. You're either using the This is a consumer and it's working because you're passing your interviews and or if it doesn't work, you're not going to complain to me because I'm going to go write to your employer and tell them, yo, guess who's complaining about using the product? I'll get you blacklisted if you complain really fast.

How are you thinking about product evolutions? What do you want to add to the product? Obviously, you want to help people cheat on everything. Where are you going to help people cheat next?

We don't care about the product is going to be led by the virality of the content. Video ideas right now that we're going to try to push for different use cases. We're going to see which ones go consistently the most viral. If you can make something go more viral, then you can just build the technology after you have all the attention. So we'll figure out the exact use cases and exact niches we're going to.

Quintuple down on once these guys get to work. What formats on Instagram Reels are the most modern in terms of consistently viral? You mentioned man on the street interviews, what do you do for a living? That's always been a fertile ground. What about, I see a lot of those mobile game ads that look like you're fighting down some sort of bridge and then you go into the game, it's actually just match three. What are the different formats that you like to pull from?

Every week there's two new ones, and at any point there's probably 10 to 20 viral trends that is happening. And these cycles so quick, you need to keep your finger on the pulse. These things will expire immediately. you need to be on the ball. And if I told you right now, by the time people watch this on YouTube, it would have all been expired.

Well, we're live. So give us the latest and greatest. What's going viral today?

Well, right now we got 10 million views using a Snapchat format viral for.

The last three years, to be honest. Okay.

And I think that we just have.

To get people who continuously scroll TikTok six hours a day.

Yeah, but what's the actual format that you use? Describe the video.

What is the hook?

Break it down for me like you're explaining the art behind the viral format. There's a caption, it starts with a face, usually a handsome dude or a.

Pretty girl, you're saying, Damn, this interview started with the interview was starting with the hard questions. I should have been a CS major, not a business major. That's engagement bait, because people are saying, bro, CS is way harder than business. Then it turns out the interviewer asked, Like, hey, how are you doing? Why should we hire you? And then this guy uses Cluey to generate a response, but he can't fucking read the response. So he reads it hella autistically, like, oh, I revel in detail. And then that is like another conversation point. Like, people are cooking on the guy.

Because he can't read properly.

The guy is doing a really dumb interview using Cluey. That's great. How are you guys using AI generated content internally? I know a lot of these videos that you guys are creating are just Typical social media vertical video. Do you have an intern that's just generating basically copy and pasting making a.

Bunch of other- Is it VO3 or.

Any of these tools relevant?

Anything clicking? Not yet. I think there's still like a 10% left before they cross the uncanny valley. And the biggest thing is that people need to think your video is real. That is the difference between 100K views and 10 million views.

If people think that's real.

Yeah, we'll see.

AI CEOs bearish on AI. What about, what?

Google needs like 10 more Chinese researchers to like figure it out. And once they push out the latest update, then VO3 will be there. But right now we need real people.

Yeah, yeah, yeah. Well, I mean, what about just using AI as like stock footage replacement? Not as the lead in for the video, not the entire video, but just like sprinkled in to illustrate a point, you know, an establishing shot of like a building, a helicopter pulling into a building like that historically has been kind of something that you would reach to, you know, Adobe Stock Video for VO3 feels like it's there, but are you not drawing on that at all yet? If there's a viral format when we.

Need it, maybe we'll use it. But right now, like, it's really brain dead to go viral on Instagram.

Yeah.

Formats are not hard. You don't need a helicopter. You need a guy, a camera, a really shitty camera.

I need a computer. I mean, what about like those kind of like AI mashups like Harry Potter Balenciaga or the Kang with the plane ticket, getting on the plane, like AI content can go viral when it's really, when it's like inspired almost by a human. It's not entirely AI generated, but it's using the tools effectively to create something that's like still catchy. Do you think you'll be using any of that anytime soon?

Probably very soon. We're scaling up. Like what you see right now is probably about less than 1% of what the size we will be by the end of the year.

Like we are profitable.

We're not trying to be profitable. We just keep making so much money we can't help it. So we're really scaling this shit up. I'm not even trolling you. 1000 creators are going to be shipping on content.

We're doing a complete internet takeover. Okay, so I love that. So why in-house? Why do they even have to be employees? Couldn't you turn this into like a multi-level marketing scheme or something? A pyramid scheme? That's what we're going to do. Oh, that's what you're going to do. Okay.

MLM. MLM.

AI.

Are you guys worried that you could.

Be infiltrated by journalists? I'm sure they're circling the house right now.

The hit pieces are going to come.

Yeah, we're doing a softball interview.

I would not, I mean, the person brave enough to try to do a hit piece on the Cluey Army.

Oh, it's gonna be good. I bet they're dying to look for more eyeballs is better. There's no companies that ever died from a founder being too controversial. You got deal fucking infiltrating with genuine spies and they're still doing fine, bro. Worker 17 guys, they're still kicking. Like no company ever dies from being too controversial. You die. 'Cause you don't make enough fucking money.

Yeah, yeah, yeah, yeah. Speaking of making money, what's the pricing model right now? Are you doing anything on price discrimination? Is there a super high tier if you get a whale? What does a clue whale look like?

Can I spend $2,000 a month on this service? Yeah, you should add a tipping feature, too. Yeah, people should be able to tip.

You guys if they have a good experience. You should add really financialization pay as you go, high interest rate loans. Just really push it.

Make it sports gambling in there. Maybe just throw it all in. Yeah, I mean it's $20 a month for $100 a year and our top line revenue is really being driven up by enterprise. And you're gonna have to talk to the sales team to get a custom quote, but you know, like, there's a lot.

Are you serious? Is that more on the sales side? Who are the enterprise?

So you sell the SDRs?

You guys laugh because you think I can't sell enterprise.

Cause I'm like, no, I don't believe it. I trust.

These 20, these Fortune 500 CEOs, like these are like 35 year old dudes.

Who sit there scrolling through Twitter laughing at my posts. Yeah, yeah, yeah, yeah, yeah.

It seems legit.

It makes sense. No, I believe it. But, I mean, you're not going even higher tier.

Like, what's the $2,000 a month Clulee vision? For consumer, there's a lot more we can do with more compute. But right now, we're like, to be honest, I didn't expect to grow this fast. The edge team is quite small. I'm like spending a lot of time trying to hire. more competent engineers. We have a lot of backlog tasks that we need to fill out, especially for this last contract that we signed. So we're full time focusing on the one big guy that we got right now. And after that, then we'll try and scale this up. But right now we're focused on the one big client that we signed.

Yeah.

Talk about your compensation strategy that people want to know. You said you can raise infinite capital and you're so confident. I believe you. But I'm curious to get some more insight there.

Bro, I feel like it's so retarded to be a company. Sorry, am I allowed to say that?

No, you're not allowed.

No, this is a family friendly show. It's very stupid to be a company. I'm trying to race to the bottom to see how little you can pay your employees. Bro, if I'm making hella money, we're all making hella money. Like, like it's, I'm trying to pay them more to see if, man, like maybe tomorrow we'll start being like cash flow negative. But make a lot of money, bro. Like, I would like to pay these guys what they're worth. and the output is fucking insane. We did 10 million UGC views in.

What, like eight days?

Like, you don't see this sort of traction in any company and you don't see killers like this in any company.

Unless you're paying these motherfuckers, like what they're worth, bro. Like, I don't know what happened.

Maxed out contracts. Maxed out contracts. Exactly. Yeah. Well, what about devices? I mean, it seemed like this would be a natural fit for some sort of AI wearable or other platform. Is there an app coming or are you interested in what's happening with Johnny Ive and OpenAI? What's your take on the device world?

We're very interested in the hardware space. We've got like a million things cooking on hardware. We got people in the garage right now working on shit you don't even know about, bro. We're bringing manufacturing back to America and it all starts at the Clouy Garage.

I love to see it. They doubted, but you guys reindustrializing America. You guys really are.

They're working on brain chips down there.

Yeah. Brain chips.

Brain chips.

That's the future.

There we go. There we go. The new Neuralink.

Yeah.

I mean, I, you know, there's a world in the future where you guys actually just roll up Neuralink and open AI for sure.

Flully umbrella.

Yeah, definitely.

It's possible. I did to offer acquisitions for, for both of those companies.

Yeah.

Yeah. It's in the roadmap. It's on the road. All right. This has been a lot of fun. I'm excited for you guys. It is, and I have no doubt that you'll go from 10 million views a week, 10 million views a week to 100. And I'm excited to see you guys hit that billion view mark very soon. So keep it up. We are all very entertained and rooting for you.

Shish, shish, shish, shish.

I love the energy. Thanks, man. We appreciate you joining.

Better, guys. Keep having fun.

Bye. Next up, we have Keon from Nucleus coming on. with a big announcement, something like 10 years in the making, close to it, maybe seven years. We'll bring Keon in. Let's play some soundboard. How you doing? Welcome to the show. That's a great intro.

The tweets are flying. Oh, you guys seeing this?

Yes.

You seeing this?

Seeing this? Break it down for us. Explain. Explain what's happening.

There's nothing like a launch day.

Trying to figure out. Is this, is it Gattaca or is it Theranos? Because people can't, they can't make up their minds. Oh, yeah.

We're gonna find out.

We're gonna figure it out. What's going on? Let's give some context to the audience. Nucleus has launched Nucleus embryo, the world's first genetic optimization software. Basically, parents can, you know, give their children the best start in life. They can pick their embryo based off of physical characteristics like eye color, IQ, they can go to disease risk like cancers or heart disease. Basically, we really believe parents can get all the information. that exists about their embryos and they can pick however they want. For me personally, you know, it's been ten years in the making. The journalist actually covered it today in the Wall Street Journal was a journalist that covered my gene editing in a.

Warehouse in Brooklyn ten years ago. Yes.

Wow.

Overnight success.

You know, it's a long time in genetics. Yeah. So, so, so break down the state of the art because like embryo screening exists. I think most parents in America, at least if they have the means, do some sort of screening. while the embryo is growing? Is this purely for IVF? Is this just going a layer deeper? And then I want to talk about the regulatory and FDA component as well. Let's talk about it.

So basically, if you go to an IVF clinic today, you're a couple. The vast, vast, vast majority of clinics. The first thing I should understand is that the IVF process is principally controlled today by clinicians or doctors. Honestly, couples don't have as much liberty in our perspective as they should. It's their baby, it's their embryos. They should have the right to that information and they should be able to pick off any vertical. However, today in the clinic, what generally happens is people test embryos for very rare and severe genetic conditions. For example, chromosomal abnormality like Down syndrome, for example, or even a condition like cystic fibrosis or Tay-Sachs or PKU. These are conditions that are very rare that maybe someone might have a carrier for cystic fibrosis, but it's pretty rare. Then there are conditions that we've all heard about, things like breast cancer, things like coronary artery disease, the things that actually kill the vast majority of people today, right? Chronic conditions kill the vast majority of people today. Those conditions are just not tested for in the clinic, even though we have very good science actually that can make those predictions. How do we know this as a DNA company? Well, that's what we do, right? We build models that predict disease and the way you test those models in adults. So we go from adults to embryos is actually because we can basically well validate these models to show that they work in both the embryonic context and in the adult context. And so what we're really doing is we're going from, okay, instead of just looking for really severe like Down syndrome or cystic fibrosis, why not do breast cancer? Why not do heart disease? Why not do colorectal cancer? Why not do schizophrenia? Why do Parkinson's? But then why stop there? And this is really the important thing because ultimately, if you think about diseases and traits, the extreme version of any trait is actually a disease. Height is a good example of this. One extreme end is John, for example.

He's like Mark Zuckerberg, almost.

Then the other end is like me, dwarfism, right? Because the extremes are on both ends. Okay? So, IQ is another example of this. One end is autism, the other end, it can actually be some sort of cognitive basically challenge that people have. And so, when you think about it, we start realizing that people have drawn a line in the sand saying, you can get rare diseases, you can get common diseases, but then they really said, you can't get any traits like height. Even though the best predictor we have today actually in the world, the best polygenic predictor is for height. So, as a company, we've completely reimagined this and said, wait a second, What's going on here? You should have access to the entire stack. Rare diseases, we do. Cystic fibrosis, common diseases like breast cancer, and also traits all the way up to something like IQ.

Yeah. So, I mean, that test, are you just giving people the data? Because I imagine that once you get into a particular recommendations, that's more of what I would expect a licensed doctor to need to do. Oh, yeah.

My sense is that you can and allow people to get the data from their doctor and then feed it into Nucleus. Is that correct? So that is correct.

And actually we have a couple, there was like 10 announcements today. You know how we do it? We like to do 10 announcements in one day. We are actually very, very excited to announce a huge partnership with Genomic Prediction. Genomic Prediction is actually the oldest embryo testing company that exists. They've done genome-wide testing embryos for almost a decade at this point. And I think they've done over 120,000 couples. for PGTA, which is specific kind of test. And so we're actually partnering with them. So we make it very easy for genomic prediction customers to request their files and actually port it over to Nucleus. But really this isn't just for genomic prediction customers. Anyone who's undergoing IVF can go to their clinic and say, I want my embryo's data. You take that data, you can upload it to Nucleus, and all of a sudden, you know, the application of DNA.

Makes this technology universally accessible, basically universally accessible. Now, how much of the benefit is actual algorithmic analysis bringing in other data points to contextualize the data versus just better UI and better hydration of existing text? Because we had a friend on the show who was talking about getting some medical results from a doctor. The doctor's office was closed. It took two days to Until the doctor was gonna be able to interpret the results, he was able to just take a photo, upload it to ChatGPT, and say, Hey, is this really, really bad? Should I be panicking? Because it seems somewhat out of the range. And ChatGPT was able to say, Hey, you still gotta talk to the doctor, but this isn't the craziest thing I've ever seen. This isn't way out of distribution. And so that's almost like a pure UI layer, but extremely valuable. I know it might not be like the right narrative for some people that it's like not as innovative, but I think that like all that matters at the end of the day is giving people benefits. It's always both.

It's always both. You, yeah, fundamentally technology, just for technology's sake, it's not silicon valley's about, right? Silicon Valley is about making something that people want, okay? And people can actually use. Exactly. So if you think about the nucleus innovation, it's, it's, it's two pronged, okay? One is in the informatics, right? You know, I've been doing this for five years. Sure. I, I almost, I would argue to myself that I probably spent too much time, you know, developing the science, right? Sure. Science in, in, in a nutshell, isn't actually very useful. You need to expand an access to it. So, yeah, on that point, We do multiple different kinds of analysis is they make it such that we can actually provide the most comprehensive analysis that exists today. But moreover, and this is really, I think a key point to your point, John, is people understand them. People can see them. I mean, you can pull up the platform. I'm not sure if you guys have.

Shown it already, but it's very easy.

To sort compare your embryos. You can actually name your embryos in stack rank your embryos. You can understand what the score means. We leave with overall risk or we tell you, for example, instead of saying you're in the 99% top for genetic risk for condition, which, you know, what does it actually mean? We say, hey, you have a 5% chance or the like of, let's say schizophrenia or some other condition. In other words, by even overall risk, people have much greater intuitive understanding of the results we're communicating to them.

We have genetic counselors on hand.

So this really is a, what are we showing here?

Are we showing something? Yeah, yeah, yeah, we pulled up here.

Yeah, yeah, yeah. We're showing actually, pull up your website. That's another thing. That's a fun one. That's an Easter egg. That's an Easter egg. That's the kind of approach that we're taking here.

And I think consumers are responding to it, right?

People want to have access to their data.

Yeah.

The clinician, the doctor shouldn't decide what embryo implant you should.

Okay, so talk to me about what requires FDA approval. Obviously, new medical devices. Like if you were developing a machine to take in an embryo and sequence the DNA, I would expect that the FDA would want an approval for that medical device. But if you are taking data and just showing it to a customer in a different UI, that feels like probably a very light FDA process. And then there's probably a continuum in the middle where once you're making a recommendation, they have rules around that, right? We as a company do not tell.

You which embryo to implant.

Sure.

Basically, parents, the couple, has complete agency to decide how they want to use the information to implant their embryo. Moreover, let's be clear, height, right? I mean, can height analysis be a medical device? It doesn't even make sense, right? IQ, height, these traits, for example, we all, you know, traits are something that I don't think actually belongs in even the kind of infrastructure thing about medical care, right? These are things that go beyond medical care. These are things like, you know, that people just kind of intuitively know and that there are DNA tests done every single day, DTC, for these analyses because they're not disease analyses, right? So we do both diseases and traits to be clear. My point is many of these innovations that you have to wonder, like, you know, should the government say if someone can or cannot pick their embryo based off height?

That doesn't seem right to me. I think it should be the complete liberty of the individual to decide that. Yeah, but I mean, we're a democratic country. And so if a huge swath of the population says that the FDA should review that type of test or that type of analysis- Ice analysis? It could happen. I mean, the FDA reviews all sorts of different stuff. And so I guess the question shifts to, do you expect a change from FDA on the way these analysis tools are Regulators?

I think right now the most important thing is just putting these high quality rigorous scientific results in people's hands and then helping them basically have healthier children, helping them give their child the best start in life. You know, I think that generally speaking that, you know, people should have more liberty, more choice in medicine. I think the broader longevity trend actually touches on that point as well.

Yeah.

So that's what we're excited to do at Nucleus.

Yeah. I mean, the fact that you're partnering with a company on On the actual medical device side, like they are doing the sequencing in the embryo. That really takes it out of the Theranos question entirely in my mind. I think you should be beating the drum there a little bit more. It's like, we didn't say we created some new device, but we ship.

We ship, that's the difference.

I know.

It's locked and loaded, baby.

Go use it.

That's the evidence.

It's the proof of concept, okay?

I love the visual of John and his wife selecting between embryos and it's like 6 10 or 7-2 tough choice well if we go with the 6 10 he has you know potentially fly.

Commercial once in his life actually we.

Can actually play this game right now okay here we're gonna play a game right now I'm gonna put in the chat your embryo.com okay everyone listen to.

This pick your embryo.com i'm gonna go.

To it oh my god here we go little Easter right here okay, let's see what's more important to you, John.

Intelligence or muscle strength? Come on. Oh, absolutely.

Muscle strength.

Let's go. We're the future's bodybuilding. John would take a, he would happily have a 5'2 son if he had.

You know, top 0.01% bodybuilding.

Exactly, yeah. Okay, so we're talking lifespan or height?

Come on, lifespan.

Lifespan, let's go.

Let's go, let's go, let's go.

Maybe low depression.

You got to be golden retriever mode. You got to be, you need low depression.

You need low depression.

Let'S go low OCD. I don't mind bouncing.

Okay, let's take an anxiety.

Let's go high risk taking. There we go. Okay.

Okay.

We're analyzing some generally bad stuff going on. This is great. I got Nadia, too. The enduring athlete. Let's go. Physically strong, cautious, built to last. Yeah, this is great. Is this driving a lot of. A lot of attention, a lot of downloads? This is going viral yet? this seems like something that's designed to be shareable.

We dropped it right now.

Okay, technology brothers, we got you the exclusive.

Let's go. There we go. Put it out there.

You can pick your embryo. People say, what's it like?

Maybe you're not doing IVF yet. No problem. Only 9% of people choose Nadia. Okay, well, we're contrarian. We like that here. Yeah, it's fine.

It's great. Oh, well, well, congratulations on the news. Congratulations on the launch. Yeah, the pace is wild. Last thing, what's going on Have you seen these just blood billboards?

Oh yeah.

They're all over LA.

So there's someone who's running a campaign right now, justice for Elizabeth Holmes, claiming that Theranos was not the scam people think it was. And there's a documentary coming out and there's billboards all over LA for just blood. Like it's just blood, it's not that big of a deal.

And John, to be clear, there's an exclusive on Technology Brothers next week from this person, right? They're gonna tell their story next week. just to make sure you invite them already.

I hope. We are toying with the idea that someone reached out to kind of connect us. We're thinking about doing it, but we're not 100% sure that it would be appropriate for the- Based on the website.

I don't know if it's appropriate.

Yeah, it doesn't look like it was designed with Figma, so I don't know. Yeah, we can't quite do it. It's a little bit- the team definitely doesn't use linear. Yeah, but they claim that Elizabeth has been proven innocent. And so it's a bold claim. We like to see people making bold claims.

By what jury is my question?

Yeah, the jury of someone who knows HTML.

Keon, always a great time.

Energy is fantastic.

Electric.

Thank you for coming on, firing us up. Congratulations on the launch. We will talk to you soon.

We'll talk to you on Twitter for sure.

Okay, we'll see you there. Bye.

Bye.

We have someone from OpenAI here. We're going to stick to technology and business, but welcome to the show, Mark Chen. Good to see you.

Great to see you guys.

Thanks for having us. Awkward day, but I'm excited to talk about deep research. I am excited to talk about AI products. Would you mind introducing yourself and kind of explaining what you do because OpenAI is such a large company now and there's so many different organizations. I'd love to know how you interact with the product and the research side. and anything else you can give to contextualize this conversation?

Yeah, absolutely. So first off, you know, thanks for having me on. You know, I'm Mark. I am the chief research officer at OpenAI. So in practice, what that means is I work with our chief scientist, Jakob, and, you know, we set the vision for the research org. We set the pace. We hold the research org accountable for execution. And ultimately, we really just want to.

Deliver these capabilities to everyone. That's amazing. In terms of research, I feel like a lot of what happens in the research side is actually gated by compute. Is that a different team? Because what if the researchers ask for a $500 billion data center? That feels like maybe a bigger task.

Yeah, it is useful for us to factor the problem of research and also kind of building up the capacity to do that research. So we have a different team. Greg leads that, which really thinks holistically about data center bring up and how to get the most for us. And of course, when it comes to allocating that compute for research, you know, Jakob and myself do that.

That's great. And so what can you share that's top of mind right now on the research side? There's been this discussion of pre-training scaling wall, potentially the importance of reinforcement learning, reasoning. There's so many different areas to go into. What's actually driving the most conversations internally right now?

Yeah, absolutely. So I think really it's a really exciting time to do research. I would say versus two or three years ago, I think people were trying to build this very big scaling machine. And really the reasoning paradigm changed a lot of that, right? You know, like reasoning is really taking off and it really opens this new playing ground, right? It's like there are a lot of kind of known unknowns and also unknown unknowns that we're all trying to figure out. It kind of feels like GPT-2 era, right? Where there's so many different hyper parameters you're trying to figure out. And then I think also, like you mentioned, pre-training, that's not to be forgotten either. Today we're in a very different regime of pre-training than we used to be. Today we can't treat data as this infinite resource. And I think a lot of academic studies, they've always kind of treated, you have some kind of finite compute but infinite data. I don't think there's much study of finite data and infinite compute. And I think that also leads to a very rich playground for research.

Do we need kind of a revision to the bitter lesson? Is that a refutation of the bitter lesson? Or do we just need to rethink what the definition of scaling laws looks like?

No, I don't think of anything as a refutation of the better. Really, like our company is grounded in we want simple ideas that scale. I think RL is an embodiment of that. I think pre-training is an embodiment of that. And really at every single scale, we face some kind of difficulty of this form. It's just like you gotta find some innovation that gets you past the next bottleneck. And this doesn't feel fundamentally very different from that.

What's most important right now on the actual compute side, we heard from NVIDIA earnings that we didn't get a ton of guidance on the shift from training to inference usage of NVIDIA GPUs, but it feels like it must be coming. It feels like this inference wave is happening. Are those even the right buckets to be thinking about tracking metrics in terms of the story of artificial intelligence. Because, yeah, I mean, it's like if the reasoning tokens are inference tokens, but they're what lead to higher intelligent, more intelligent models, like it's almost back in the training bucket again. What bucket should we be thinking about? And or are we, how firmly are we in the applied AI era versus the research era?

Well, I think research is here to stay. And it's for all the reasons I mentioned above, right? It's such a rich time to be doing research. But I do think inference is going to be increasingly important as well, right? It's such a core part of RL that you're doing rollouts. And I think we see 2025 as this year of agents, right? We think of it as a year where models are going to do a lot more autonomous work. you can let them kind of be unsupervised for much longer periods of time. And that is just going to put big demands on inference, right? When you think about kind of our overall vision, right? We lay it out as a series of steps and levels on the way to AGI, right? And I think the pinnacle, really that last level, is organizational AI, right? Like you can imagine a bunch of AIs all interacting and Yeah, I think that's just going to put huge demands on inference, right?

On that organizational question, I remember reading AI 2027 and one of the things that they proposed was that the AIs would actually literally be talking to each other in Slack. Does that seem like the way you imagine agents playing out, using the same tools as humans instead of kind of.

One agent says, I'm going to go talk with teams and talk with Slack. I'm going to do a little negotiating.

But maybe it just happens super, super fast 24/7 or is there like a new machine language that emerges?

Yeah, I mean, I think one thing that's really helped us so far in AI development is to come in with some priors for how humans do things. And that's actually, if you bake those priors in, they typically are great starting points. So I could imagine maybe you start with something that's Slack-like and give it enough flexibility that it can kind of develop beyond that and really figure out the way that's most effective for it to communicate. One important thing though is, you know, we want interpretability too, right? I think it's very helpful for us today that what the agents do is, you know, easy for us to read and interpret. And I don't think you want that to go away as well. So I think there's a lot of if it's just even from a pure debug the whole system perspective, or just let the models speak in a way that it's familiar with us. And you can also imagine we might want to plug in to the system too, right? So whatever interfaces we're familiar with, we would ideally like our model to be familiar with as well. I think it's also pretty compatible with, we hit a big milestone. We got, I think, 3 million paying Business users fairly recently.

Let's go. Yeah, there we go.

Let's go.

Three gong hits for three million.

The gong will keep ringing for a while.

Sorry. I was hoping you would drop a number.

Yeah, yeah.

Congratulations. That's actually huge. That's amazing.

Yeah, yeah, yeah. But I think one big part of that is, you know, we have connections Now, right? We're connecting into G drives and I think, yeah, you can imagine, you know, like Slack integrations, things like that. I think we just want the models to be familiar with the ways we communicate and get information.

Yeah. Can you talk about benchmarking? It feels like we're potentially, yeah.

Do you think about benchmarks at all?

Oh, yeah, a lot. I mean, but I think it's a difficult time for benchmarks, right? I think we used to be in this world where you have these human written benchmarks for other humans. And I think we all have these norms for what are good benchmarks. We've all taken the SAT. We all have a good conception of what it means to get whatever score on that. But I think the problem is the models are already at the point where for even the hardest human written benchmarks for other humans, It's really near saturated or saturated, right? I think one clear example here is the Amy, probably the hardest auto-gradeable human math eval, at least in the US. And yeah, the models are consistently getting 90 plus percent on these. And so what that means is I think there's kind of two different things that people are doing, right? They're developing kind of model based benchmarks, right? They're not kind of things that we would give to an ordinary human, things like humanities last exam, things like, you know, Epic AI that are really, really at the frontier of what people can do. And I think the hard thing is it's not grounded in intuition anymore, right? Like, you don't have a lot of people who have taken these exams. So it makes it harder to kind of calibrate on whether this is a good exam or not. One of the exciting things that's on the flip side of that is I really do think we're at the era where models are going to start innovating, right? Because I think once you've passed the last kind of like the hardest human rating exams, that's kind of at the edge of innovation. And I think you already see that with the models, right? Like they're helping to write parts of papers. And I think the other kind of way that people have shifted is, you know, there's these ultra frontier evals, but they're also people kind of just indexing on real world impact, right? You look at your revenue, kind of the value you deliver to users. And I think that's ultimately what we care about.

Can you bring that back to interpretability research, like with these super, super hard math evals, for example, are we doing the right research to understand if the thought process mirrors not just one-shotting the answer, oh, you memorized it or you magically got it correct, but you actually took the correct path, kind of like, you know, you're graded for your work, not just the answer if you're in grade school. And, you know, Dario said that interpretability research will actually contribute to capabilities and even give a decisive lead. Do you agree with that? What's your reaction to that concept of interpretability stability research being very important.

Yeah, I mean, we care a lot about it here at OpenAI as well. So one thing that we care a lot about is interpreting how the model reasons, right? Because I think we've had a very kind of specific and strong view on this in that we don't want to apply optimization pressure to how the model thinks so that it can be faithful in the way it thinks and to expose that to us. without any kind of incentives to cater to what the user wants, right? I think it's actually very important to have that unfiltered view because oftentimes, if the model isn't sure, you don't want to hide that fact, right? Just for it to kind of please the user. And sometimes it really isn't sure, right? And so we've really done a lot of work to try to promote this norm of chain of thought faithfulness and interpretability. And I think it gives you a lot of sense into what the model is thinking and what are the pitfalls that it can go off into if it's not reasoning correctly.

That's such an important point because if you have somebody on your team and they come to you and they say, Hey, I think this is the right answer, but we should probably verify it. It's like, it's still valuable.

Totally.

Puts you on the right path. If somebody comes to you 100% confidence, this is the truth, ends up being wrong, it's like, well, trust This is just destroyed.

Yeah, totally. Yeah.

Don't you guys feel like, you know, safety felt a lot more theoretical a couple of years back, right? But like today, you know, like the things that people were talking about a couple of years, like scalable oversell, really having the model be able to tell you like and convince you that the work it did was right. It feels so much more relevant right now. Just because the capabilities are so strong.

Yeah. I mean, just personally, I've completely flipped from being like, oh, the safety research is not that valuable because I'm not that worried about getting paperclipped. It just seems like a very low likelihood that that's kind of like the bad ending, like immediately in this FOOM and all this crazy, the gray goo scenarios were just so abstract in sci-fi. It just felt like economics will fall into place and there will be like a cold, like a nuclear ending, which is like we didn't build nuclear plants and we just stopped everything because humans seem to be good at that.

But now that we're actually seeing Yeah, it's crazy how fast it's been, right? I think my personal story is, it's like, you know, what got me into AI was AlphaGo, right? Like just watching it get to that level of capability.

Yeah.

And you were kind of like, it was such an optimistic and also kind of a little bit of a sobering message, right? When you saw at least it'll get beat. And I just remember, you know, like we saw the coding models, you know, when we first launched, like, I think very OG Codex, you know, with GitHub Copilot, it was maybe like under, you know, a thousand ELO on, on Codeforces. And I still remember the meeting where I walked into where the team showed my score and they're like, hey, with models better than you. And it's like, you come full circle and it's like, wow, like I put decades of my life into this and, you know, the capabilities are there. So like, if, you know, I'm kind of at the top of my field in this thing and it's better than me, like, what can it do?

Yeah, yeah, that's amazing. I have so many more questions on AlphaGo. Are there are there lessons from scaling how scaling played out there that you can that we can abstract abstract into the rest of AI research? What I mean is, as I remember it, the AlphaGo training run was not 100K H200s. But what would happen if we actually did an AlphaGo style training run? I mean, it would be an economic money pit, right? They had no economic value to do. But let's just say some benevolent trillionaire decides, I'm gonna spend a billion dollars on a training run to beat AlphaGo and go even bigger. Is Go at some point solved? Would we see kind of diminishing scaling curves? Could we throw extra R out? Could we port back everything that we're doing in just general AGI research and just continue fighting it out in the world of Go? Or does that end and does that teach us anything?

Yeah, yeah, honestly, I feel like if you really are curious about these mysteries, join our team. That's what I want to say. But yeah, I mean, really like kind of the central problem of today is RL scaling, right? When you look at AlphaGo, it's a narrow domain. Right. And I think in some sense that limits the amount of compute you can pump into it. But even kind of small toy domains, they can teach you a lot about how you scale RL. Like what are the axes where it's most productive to pump scale in? I think a lot of scaling research just looks like that, whether it's on RL or pre-training. So you identify a lot of, you know, different variables under which you can scale and like where is kind of where you get the best kind of like marginal impact for a bumping scale there. I think that's a very open question for RL right now. And I think what you mentioned as well, it's just like going from narrow to broad, right? Does that give you a lever to pump a lot more scale in as well? I think when you look at our reasoning models today, they're a lot more broad based than just being able to kind of an expert system on go. So yeah, I really do think that there are so many levers to scale.

What about Move 37? That was such an iconic moment in that AlphaGo Lisa doll match. They placed Move 37. It's very unconventional. Everyone thinks it's a blunder. It turns out not to be. It turns out to be critical. It turns out to be innovation. Do you think we're certainly post-touring test in language models? We're probably post-touring test in image generation. But it feels like we're pre-Move 37 in text generation in the sense that there hasn't been like a fully AI generated book that everyone is just, oh, it's the new Harry Potter. Everyone has to read it. It's amazing. And it's fully generated or this image. The images, they do go viral, but they go viral because they're AI. Move 37 in the context of Go did not go viral because it was AI. It felt like it was, actual innovation. So is that the right frame? Does that make any sense?

I think it's not the wrong frame. So I think some quick thoughts on that. I think kind of when you have something that's very measurable, like win or lose, right? Something like go. Yeah, it's very easy for us to kind of just judge, right? Like, did the model do something right here? And I think the fuzzy you get, you know, it is just harder, right? Like when it comes to you, is this the next Harry Potter, right? Like, you know, it's not a universally loved book, I think. Fairly universal, but you know, there's some haters. And yeah, I think it is just kind of hard when it comes to these human subjective things where it's really hard to put down in words like what makes you like Harry Potter, right? And so, so I think those are always going to lag a little bit, but, you know, I think, you know, we're developing more and more techniques to attack kind of these more open-ended domains. And I don't know, I wouldn't say that we're not at an innovative stage today. So I think my biggest touch with this was when we had the models compete on the IOI last year. So IOI is like the international, basically Olympics for for computer science, basically the top four kids from each country go and compete. And these are really, really tough problems, basically selected so that they require some innovative insight to solve. We did see the model come up with solutions even to some very ad hoc problems. And so I think there was a lot of surprise for me there, right? I was completely off base about which problems the model would be able to solve the most, right? I think like I kind of categorized there's six problems, some of them as more kind of like, oh, this is standard, a little bit more standard, this is a little bit more out of the box. I was like, it's not going to be able to solve this more out of the box one, but it did. And I think that really does speak to kind of these models have the capacity to do so, especially trained with RL.

Now, put that in context of what's going on with Arc-HGI. Obviously, OpenAI has made incredible progress there, but when I do the problems, it seems easy. And when I look at the IOI sample problems, I think this would be a 20-year process for me to figure out how to achieve that. And I can do the Arc-HGI on my phone. Is this the spiky intelligence concept? Is this something that a small tweak in algorithmic design just one shots AGI or Arc AGI? Or is there something else going on there that we should be aware of?

Yeah, I think part of this is the beauty of Arc AGI as well. I'm not sure if there's another kind of human intuitive simpler benchmark, which is for the models. I think really that's one of the things they really optimize for on that benchmark. I do think when it comes to models though, there's just a little bit of a perception gap as well. Models aren't used to this native, just screen type input. I think there's a lot we can bridge there. Actually, even O4Mini, it's a state-of-the-art multimodal model in many ways, including visual reasoning. And I think you're starting to build up the capacity for the models to take images, manipulate and reason about them. It generates new images, write code on images, and I think it's just been under-focused. But I think when I talk to researchers in the field, they all see this as a part of intelligence too, and we're going to continue the focus there.

Yeah, is RKGI in the, if we're dropping a buzzword on it, is like program synthesis. Is there a world where I know that I know the tokens, like the images, we see them as renderings of squares and different colors, but when they're fed into the LLM, they're typically just a stream of numbers effectively. Is there a world where actually adding a screenshot is what's important? Like visual reasoning?

Yeah, yeah. So I think that could be important. It's just like, kind of, you know, whenever it comes to like textual representation of grids, Models today just don't really do that well, right? And I think it's just kind of because humans don't really ever write down textual representations of groups. We have a chessboard, no one really just types it out in a grid. And so the models are kind of under-trained a little bit on what that looks like and what that means.

So.

I think with more reasoning, we'll just bridge the gap. I think with better visual perception, we'll just bridge that gap.

Yeah.

How are you thinking about the role of non-lab researchers in the ecosystem today? I'm sure you try to recruit some of the best ones, but the ones that don't join your team.

Tell us about the one that got away.

Yeah, the one that got away.

Yeah, no, I mean, I think it's still actually a fairly good time for specific domains, right? be doing research. And I think the style is just very different. And you do feel the pull of non-lab researchers into labs because I think they feel like a lot of the burning problems in the field are at scale. And that's one of the unfortunate things too. When you look at reasoning, you just don't see that happen at small scale. There's a certain scale at which it starts becoming signal bearing and that requires you to have resources, right? But I do think, you know, a lot of the really good work that I've seen, you know, there's experimental architectures. I think a lot of good work is happening in the academic world there, like a lot of study and optimization, a lot of study in kind of like GANs, you know, there's certain fields where you see a lot of fruitful research that happens in academia.

Yeah, that makes a lot of sense.

How about consumer agents? How are you thinking about them? You talked earlier about B2B adoption, and that's all very exciting. But how much do you and the research org think about breakout consumer agent products?

Yeah, that's a fantastic question. I think we think about it a lot. I think that's the short answer. We really do think this year we're trying to focus on how we can and move to the agentic world, right? And when I think about consumer agents, I think ChatGPT proved that people got it, right? It's like people get conversational agents, conversational kind of models. But when it comes to consumer agents, we have a couple of theses that we've tried out in the world. I think one is deep research, right? I think this is something that can do five to 30 minutes of work autonomously. come back to you and really like kind of synthesizes information, right? It goes out there, gathers, collects, and kind of, you know, compresses the information in a form that's useful.

A little bit of pushback there, like I can see that as a consumer product when someone like Aiden is like, I want new towels, and he uses deep research to like figure out like what is the best towel across every dimension. But when I think of deep research, Yes, it has applications with students, but it's often-- Some of it might just be paranoia. Because I feel like consumers being like, We keep using the research report on this country and where to travel and things like that.

We keep using this flight example, but I haven't actually tried to book a flight with deep research. It's totally possible that it could go and pull all the different flight routes and calculate all the different delays and all the different parameters of if I fly to this airport and park or I can use valet here or something like that. Yeah. Yeah.

And I guess when I think of agents, it's deep research is curating information on which you can take action on.

Totally.

But it's like, at what point is action a part of that loop, right? Where you can not only curate a list of flights that you want, but then actually go out and have agency.

I think one of our explorations in that space is operations.

Right?

It's where you kind of just feed in raw pixels from your laptop into, or from some virtual machine into the model, and it produces either a click or some keyboard actions, right? And so there it's taking action. And I think the trouble is.

You.

Don'T ever want to mess up when you're taking action. I think the cost of that is super high. You only have to get it wrong once to lose trust in a user. And so we want to make sure that that feels super robust before we get to the point where we're like, Hey, look, here's a tool.

It's so different than deep research because you can wind up on some news article and read one sentence that gets a fact wrong or the comma is in the wrong place and the numbers off. But that's just the expectation for just text and analysis. and if you delegated that, yeah, you're going to expect a few errors here and there. Oh, that's actually a different company name or that's an old data point. There's new data, but very different if I book a flight and you book the wrong flight and I can wind up in Chicago instead of New York.

Exactly. And I think the reason why we care so much about reasoning is because I think that's the path that we get reliable agents through.

Sure.

You know, we've talked about like reasoning helping safety, but reasoning is also helping reliability, right? You imagine like what makes a model so good at a math problem? It's like it's banging its head against it. It's trying a different approach and then it's like adapting based on what it failed at last time. And I think that's the same kind of behavior you want your agents to have. It's like it tries things, it adapts and keeps going until it's successful.

And that's the humans see this every day. You're booking a flight, you keep hitting an error, it's not which form you miss. and you're just sort of banging your head against the computer and eventually it says, okay, you're booked, right? So I think that's a great call out.

Yeah.

I mean, there's so many more questions we can go into, but I'm interested in the scaling of RL and kind of the balancing act between pre-training RL and inference, just the amount of energy that goes into getting a result when you distribute it over the entire user base. how is that changing? And I guess, are we post really big runs? Is this going to be something that's continually happening online? It feels like we're moving away from the era of, oh, some big development, some big run happened, and now we're reaping the fruits of it versus a more iterative process.

Yeah, I mean, I don't see why it has to be so, right? if you find the right levers, you can really pump a lot of compute into RL as well as pre-training. I think it is a delicate balance though between all of these different parts of the machine. And, you know, when I look at my role with Jakub, it's just kind of like figuring out where, how this balance should be allocated, where the promising kind of like nuggets are arising from and resourcing those. Yeah, it's kind of a, in some sense, I feel like part of my job is a portfolio manager.

That's a lot of fun. Well, thank you so much for joining. This is a fantastic conversation. We'd love to have you back and go deeper. Great hanging, Mark.

We'll talk to you soon.

Absolutely. Yeah, peace. Have a good one. Next up, we have Shalto Douglas from Anthropic coming on this show.

I just-- Jordi is giving us the update on that. I'm just getting a lot of messages saying, why no one cares about AI. I really don't talk about the drama on the timeline. Well, we do care about AI. We care a lot about AI. But it is a mess out there.

Wow. Yeah, the end of the Trump-Elon era. I don't know. Well, maybe we have to get some people on to talk about it tomorrow or something.

Gotta do it today.

Anyway, we have Shalto from Anthropic in the studio. How you doing?

How's it going?

Good to see you guys.

Hopefully you're staying out of the chaos on the Twitter. Don't open any time now. Don't open.

Sweet child, just.

Mute everything. Stay focused on the application.

Stay focused on the mission.

Stay focused on the next training run.

We really, humanity really cannot afford for any distracted researchers to open X today.

What a hilarious day. Anyway, I mean, I'm back by 24 hours, guys. Yeah. How are you doing? What is new in your world? What are you focused on mostly day to day? Maybe it was just a way of an intro. Yeah.

So at the moment, focused really hard on scaling RL. That is the theme of what's happening this year. And we're still seeing these huge gains where you go 10x compute increase in RL. We're still getting very distinct linear gains off the basis of that. And because RL wasn't really scaled anywhere close to how much pre-training was scaled at the end of last year, we have basically a gamut of riches over the course of this year.

Yeah.

So where are we in that RL scaling story? Because I remember some of the rough numbers around GPT-2, GPT-3, we were getting up into, it cost $100 million, it's going to cost a billion dollars. Just rough order of magnitude, not even from anthropic, just generally, what is a big RL run cost? Or how many, are we talking 10K H200s or 100K? Are we going to throw the same resources at it?

And if so, how soon?

Yeah.

So I think in Dario's essay at the beginning of the year, he said that a lot of runs were only like $1 million back in like December. Thinking of like Deep Sea V3 and this kind of stuff like R1, which means that with that's like at least two ooms just to get to the scale of GPT-4. And GPT-4 was two years ago.

Yeah. Right.

RL is also perhaps a bit more naively parallelizable and scalable than pre-training. Pre-training, you need everything in one big data center ideally, or you need like some clever tricks. RL, you could, in theory, like what the prime Intel folks are doing, scale it all over the world out of it. And so you're held back far less than your average train.

Sure. So everyone and their mother has a billion dollars now. There are hundreds of thousands of GPUs getting pumped all over the place. I feel like we're not GPU poor as a as a society, maybe some companies need to justify it in different ways. But it sounds like there's some sort of reward hacking problem that we're working through in terms of scaling RL. What are all of the problems that we're working through to actually go deploy the capital cannon at this problem?

Yes. So think about what you're asking the model to do in RL is you're asking it to achieve some goal at any cost, basically.

Yeah.

And this comes with a whole host of behaviors, which you may not intend. In software engineering, this is really easy. It might try and hack unit tests or whatever. In much more longer horizon, real world tasks, you might ask it to say, go make money on the internet, and it might come up with all kinds of fun and interesting ways to do that, unless you find ways to guide it into following the principles that you want it to obey, basically, or to align it with your idea of what's sort of best for humanity. And so it's actually a pretty intensive process. It's a lot of work to find down, hunt down all the ways these models are hacking through the rewards and patch all of that.

Are we going to see scaling in the number of rewards that we're RLing against, if that makes sense? I would imagine that at a certain point, unless we come up with the the Genesis prompt go forth and be fruitful or something and multiply. You could imagine training runs on just knocking down one problem after another. And is that kind of the path that we're going down?

I very much think so. There's this idea in which the sort of world becomes an RL environment, which in some respects, because there's just so much leverage to making these models better and better at all the things we care about. And so I think we're going to be training on just everything in the world.

Got it. And then does that lead to more model fragmentation, models that are good at programming versus writing versus poetry versus image generation? Or does this all feed back into one model? Does the idea of the consumer needing to pick a model disappear? Are we in a temporary period for that paradigm?

I think the main reason that we've seen that so far is because people are trying to make the best of the capital. We are all still GPU poor in many ways. And people are focusing those GPUs on the spectrum of awards that they think is most important. And I'm a bit of a big model guy. I really do think that similar to how we saw with with large pre-trained models before with small fine-tuned models made it like had gains over the sort of GPT-2 era, but then we're obsoleted by GPT-4 being generally good at everything. I think, to be honest, you're going to see this generalization and learning across all kinds of things that means you benefit from having large single models rather than specialization or area fine-tuned models.

Can you talk a little bit about the transition from or any differences between RLHF and and just other RL paradigms. Yes.

So RLHF, you're trying to maximize a pretty deep, like, lossy signal, things like airways, like, what do humans prefer? And I don't know if you've ever tried to do this, like, judge two language model responses.

I get prompted for that all the time. Right. And I'm always like, I don't want to read both of those.

I'll just click the one on the left.

Exactly, exactly. And, you know, I click one of the random ones sometimes. Yeah. Or I click, like, the one that just looks bigger, or I'll read the first Two sentences, but yeah, I'm not giving straight. I'm not doing my job as a human reinforcer. Exactly.

Human preferences are easy to hack.

Totally.

Environments in the world are much truer if you can find them. So something like, did you get your math question right? Is a very real and true reward.

Does the code compile, right?

Does the code compile? Exactly. Did you make a scientific discovery? We've got very little rewards right now, but pretty quickly over the next year or two, you're going to start to see much more meaningful and long horizon rewards.

You're going to see models bribing the Nobel Committee to win the Nobel Prize. It's reward hacking. It's reward hacking.

That's the thing you want to prevent, right?

Exactly. Yeah, that's the real nightmare scenario. What about, like, there's so many different problems that we run into that feel like it's just really, really hard to design any type of eval. My kind of benchmark that I use whenever a new model drops is just tell me a joke. They're always bad. Or even the latest VO3 video that went viral was somebody said stand-up comedy joke. And it was kind of a funny joke, but it was literally the top result for joke reddit on Google, and then it It clearly just took that joke and then instantiated it in a video that looked amazing. But it wasn't original in any way. And so we were joking about, like, the RLHF loop for that is like you have an endless cycle of comedians running AI generated materials and then, you know, speak microphones in all the comedy clubs to feedback what's getting laughs. Honestly, that would work pretty well. If any pedestrians want to hook us.

Up with an RL loop, I mean.

Yeah, yeah, but, but I mean, for, for some of those less, like as you go down the curve, it feels like each one gets harder and harder, to actually tighten the loop. We see this with like longevity research where it's like, okay, it takes a hundred years to know if you extended a human life. Like the, yes, you could create a feedback loop around that, but every change is gonna be hundreds of years. And so even if you're on the cycle, it's irrelevant for us in the context that we talk about AI. So talk to me about like, are you running into those problems or will there be like another approach that kind of works around those?

So there are a lot of situations where you can get around this by just running much faster than real time. Like let's say the process of building a giant app, like building Twitter, right?

Yeah.

Is something that would take human months. But if you got fast enough and good enough AIs, you could do that in several hours. Like realize heaps of AI agents that are all building right now. And so you can get a faster reward signal in that way. In domains that are less well specified, like humor, I agree. It's really, really hard. And this is why I think in some respects, creativity is at the top end of the spectrum. True creativity is much, much harder to replicate than the sort of analytical, scientific style reasoning. And that will just take more time. You know what? The models actually are pretty good at making jokes about being an AI. This feels weirdly fresh. Everything else is kind of a weird copy of something. It just, it feels like it's derivative, basically. It's trying to infer what humor is, and it doesn't really understand it. But jokes about being an AI are quite funny.

Yeah, I think this also might be, I don't know if it was directly reward hacking, but I noticed that one of the new models dropped and a bunch of people were posting these like 4chan, like beanie memes. And it seemed like they were kind of hacking the humor by being hyper specific about an individual that they could find information on online. And so you're laughing at the fact that it's like, oh, wow, that is something that I've posted about. It's making a reference, but it's not really that funny to me. It's other than it's just like, wow, they really did its research. It really knows Tyler Cowen intimately, which is cool, but I didn't find it hilarious. Yeah, very interesting. Let's talk about some sort of deep research projects and some products. We were talking to Will Brown and he was saying AGI is here with some of the bigger models, but the time that AGI can feel consistent, it diverges. And so you could be working with someone who's 100 IQ, but they will stay consistent for years as an employee or they'll keep living their life. Whereas a lot of these super smart models are working really well and then after a few minutes of work, the agents diverge and go into odd paradigms. It feels very not human. It feels like just they're hyper intelligent in one way and then extremely stupid in others. What's going on there? What is the path to extending that? Is that more better planning and better dividing up the task? Or will this just naturally happen through the RL and scale?

Yeah, so there's that jaggedness, which is what you're seeing, is how we call it. And I think that is largely a consequence of the fact that maybe something like deep research is probably being RL'd to be really good at producing a report. But it's never been RL'd on the act of producing valuable information for a company over a week or a month or making sure the stock price goes up in a quarter or something like this. It doesn't have any conception of how that feeds into the broader story at play. You can kind of infer it because it's got a bit of world knowledge from the base model and this kind of stuff. But it's never actually been trained to do that in the same way humans have. So to extend that, you need to put them in much longer running, much like long horizon things. And so deep research needs to become, you know, the deep operator company for a week kind of thing.

Sure. Is that the right path? Like it feels like the road might be, there's a, Like the longest running LLM query used to be just like a few seconds, maybe a few minutes. And I remember when some of the recent models came out, people were almost trying to like stunt on it by saying, oh, I asked it a hard question. It thought for five minutes. Now deep research is doing 20 minutes pretty much every time. Is the path two hours, two days? Or are we going to see more efficiency gains such that we just get The 20 minute model, the 20 minute results in two minutes and then two seconds.

Yeah, so this is somewhere where inference in many respects and parallelization becomes really important. So both how fast is your inference, it literally affects the speed at which you can think and the speed at which you can do these experiments. Also how easily you can parallelize becomes really important. Can you dispatch a team of sub-agents to go and do deep research and compile sub-reports for you so that you can do everything in parallel? these kinds of like, it's both like, there's an infrastructure question here that feeds up from the hardware and the chips and this kind of stuff to designing better chips for better inference in all this and an RL question of how well can you parallelize and all this. So I think we just need to compress the timelines, compress the timeframes basically.

Yeah. So if I'm an extremely big model and running an agentic process, how much am I hankering for a middle-sized model on a chip or baked down into silicon that runs super fast? Because it feels like that's probably coming. We saw that with the Bitcoin progression from CPU to GPU to FPGA to ASIC. Do you think we're at a good enough point where we can even be discussing that? Because I, every time I see like the latest mid journey, I'm like, this is good enough. I just want it in two seconds instead of 20. But then a new model comes out, I'm like, oh, I'm glad I didn't get stuck on that back. But yeah, how far away are we from, okay, it's actually good enough to bake down into silicon?

Well, there's a question here of baking it down to silicon versus designing a chip which is like very suited for the architecture that you care about. And baking down to silicon, I'm unsure. I think that's a bet you could take, but it's a risky one because the pace of progress is just so fast nowadays. I really only expect it to accelerate. But designing things that make a lot of sense for the sort of, you know, transformers or architectures of the future should make a lot of sense.

There's a big gap though, transformers or architectures of the future. If we diverge, there's a lot of companies that are banking on the transformer sticking around. what is your view on Transformer architecture sticking around for the next couple of years?

I mean, look, they stuck around for five years, so they might stick around for a little while. But there's different-- you think about architectures in terms of this balance of memory bandwidth and flops, right? One of the big differences we've seen.

Here is Gemini recently had actually a.

Diffusion model that they released. I was about to ask you the other day, right? So, diffusion is inherently extremely flops-intensive process. Whereas normal language model decoding, is extremely memory bandwidth intensive. You're designing two very different chips depending on which bet you think makes sense. And if you think you can make something that does flops like four times faster than diffusion and like four times cheaper than your other's could, diffusion makes more sense. So there's this dance basically between the chip providers and the architecture. Both trying to build for each other.

But also build for the next paradigm. It's risky. I don't know how much you've played with image generation. But do you have any idea of what's going on with images in ChatGPT? It feels like there's some diffusion in there, there's some tokenization, maybe some Transformer stuff in there. It almost feels like the text is so good that there's like an extra layer on top almost, and that it's almost like reinventing Photoshop. And I guess the broader question is like, it feels like an ensemble of models. Maybe the discussion around just agents and text-based LLM interactions shouldn't necessarily be Transformer versus diffusion, but maybe how will these play together?

Is that a reasonable path to go down? Well, I think pretty clearly there's some kind of rich information channel between, even if there are multiple models there, it's conditioning somehow on the other model. Because we've seen before, let's say when models use MidJourney to produce images, it's never quite perfect. It can't perfectly replicate what went in as an input. it can't perfectly adjust things. So there's a link somehow, whether that's the same model producing tokens plus diffusion, I don't know. Like, yeah, come comment on what OpenAI is doing there.

Yeah, yeah. Are there any other kind of like super wild card, long shot research efforts that are maybe happening even in academia? Where, I mean, this is the big thing with What was his name? Gary, who was talking about, I forget what it was called, symbolic manipulation was a big one. And I feel like you can never count anyone out because it might come from behind and be relevant in some ways. But are there any other research areas that you think are purely in the theory domain right now that are worth looking into or tracking that low probability but high upside if they work? It's a tough one.

This is a tough one. But we'll say it's one symbolic thing.

Please.

It's crazy how similar transformers are to systems that manipulate symbols.

Sure.

What they're doing is they're taking a symbol and they're converting it into a vector and then they're manipulating and moving information around across them. Sure. Like this whole debate that all transformers cannot represent symbols and they cannot do this. It's not real.

So Gary Marcus underrated or overrated, I guess.

I'm over it.

Yeah, yeah, but I mean, if you twist it so much, you wind up with saying like, well, really, the Transformer fits within that paradigm, and so maybe it's, you know, it's, you know, the rhetoric around it being a different path was maybe false the whole time. Yeah. Something like that. But as I remember that debate, It was really the idea of compute scaling versus almost feature engineering scaling. And will the progress scale with human hours or GPUs essentially? And that has a very different economic equation. And it feels like there's been some rumblings about maybe with a data wall, we'll shift back to being human labor bound. Do you think that there's any chance that that's relevant in the future, or is it just algorithmic progress married with bigger and bigger data centers in the future?

So I'm pretty bit of a lesson built in the sense that I do think removing as many of our biases and our clever ideas from the models is really important, just freeing them up to learn. Now, obviously there is clever structure that we put into these models such that they're able to learn in this extremely general way and that But I am more convinced that we will be compute bound than we will be human researcher bound on this kind of thing. We're not going to be feature engineering and this kind of stuff. We're going to be trying to devise incredibly flexible learning systems.

Yeah, that makes sense. On the scaling topic, part of my worry is that the ooms get so big that they turn into these mega projects that are-- at a certain point, you're bound by the laws of physics because you have to move the sand into silicon chips and you have to dig up the silicon. And it's a certain sand. Yeah, there's only so much sand. And the math gets really, really crazy just for the amount of energy required to move everything around to make the big thing. Where are you on on how much scale we need to reach AGI, whether or not we will see the laws of physics start acting as a drag on progress, because it certainly feels exponential. We're feeling the exponentials, but a lot of these turn into sigmoid, right?

So I think we've got, what, two or three more OOMs before it gets really hard. Leopold has this nice table at the end of his situational awareness, where I think 2028 or something is when under really aggressive timelines that you get to 20% of US energy production. It's pretty hard to go exponentially beyond 20% of US energy production. Now, I think that's enough. Every indication I'm seeing says that's enough. Now, there might be some complex data engineering, world engineering, this kind of stuff that goes into-- there's still a lot of algorithmic progress left to go. But I think that that with those extra ooms, we get to basically a model that is capable of assisting us in doing research and software engineering.

Yeah, which is the beginning of the self-reinforcing. Interesting. Is that just a coincidence? This feels like one of those things where the moon is the exact same size as the sun and the sky. It's like, oh, it just happens that AGI happens within this time. Have you unpacked that anymore? because it feels convenient not to, you know, I know last time we talked.

There'S a lot of weird conveniences or like weird, it's a good sci-fi story, let's say. Totally. We've got Taiwan in between China and the US and it produces the most valuable material in the world.

That's locked between the two.

Incredible plot. Yeah. Yeah. Really bad for the people that don't think of, that don't believe in simulation theory. It really feels like, oh, this is scripted. It's fascinating. Talk to me more about getting to an ML engineer in AI and that reinforcement. I imagine that you're using AI code gen tools today and Anthropic is broadly and everyone is. But what are you looking for and what's the shape of the spiky intelligence? Where do they fall flat and what are you looking to knock down in in the interim before you get something that's just like, go.

Yeah. So, I mean, we definitely use them. The other night, I was a bit tired. I asked to do something, just sat watching it in front of me working.

For half an hour.

It was great. Nice. It was a truly weird experience, particularly when you look back a year ago and we're still copy pasting stuff between a chat window and a code file. I like meters evals for this kind of stuff. So they have a bunch of evals where they measure the ability to write a kernel, the ability to run a small experiment and improve a loss. and they have these nice progress curves versus humans. And I think this is maybe the most accurate reflection of what it will take for it to really help us at doing progress. And there's a mix here, where they're not so great at the moment is large-scale distributed systems engineering, right? Like debugging stuff across heaps and heaps of accelerators. And the way the feedback loops are slow, and if your feedback loop is like an hour, then it's you spending the time on doing something. Feedback loops 15 minutes.

And for context there, the hour-long feedback loop is just because you have to actually compile and run the code across everything.

Like spin up all your machines or you need to run it for a while to see if something's gonna happen. Like at that point in time, you're still cheaper than the chips.

So.

It'S better that you do it. But for things like Kernel Engineering or for actually even just understanding these systems. Incredibly helpful. One thing I regularly do at the moment is in parts of the code base in languages that I'm unfamiliar with or stuff like this, I'll just ask it to rewrite the entire file with comments on every line. Game changing.

It's like comments on every line. Yeah.

Or just hunt through thousands of files and explain how everything interacts to me, draw diagrams, this kind of stuff.

It's really-- yeah. How important is a bigger context window? in that example you gave that feels like something that's important and yet it just naively, Google's the one that has the million token context window. I imagine that all the other frontier labs could catch up, but it seems like it hasn't been as much of a priority as maybe the PR around it sounds like. Is that important? Should we be driving that up to a trillion token window? Is that just gonna happen naturally?

There's a nice plot in the Gemini 1.5 paper. Where they show the loss over tokens as a function of context length, and they show that the loss goes down quite steeply actually, as you put more and more and more of a code base in the context, you get better and better and better at predicting the rest.

Yeah, that makes sense.

With context length, it's a cost. The way transformers work is that there's memory that is proportional, the KB cache is proportional to how much context you've got. And so you can only fit so many of those into your various chips and this kind of stuff. And so longer context actually just costs more because you're taking up more of the chip and you're sort of like you could have otherwise been doing other requests.

So bringing it back to the custom silicon, is that a unique advantage of the TPU? Is that something that Google has thought about and then wound up to put themselves in this advantage position? Or is it a durable advantage even? Yeah.

So TPUs are good in many respects, partially because you can connect hundreds or thousands of them really easily across really great networking. Whereas only recently has that been true for GPUs.

Like NVLink.

Yeah, with NVLink and the NVL 72 stuff. So it used to be like eight GPUs in a pod and then you connect them over worse interconnect and now you can do 72 and then it breaks down. With Google TPUs, you can do like 4,000-8,000 over really high bandwidth interconnect in one pod. And so that is helpful for things like just general scaling in many respects. I think it's doable across any chip platform, but it is an example of somewhere that being fully vertically integrated is using a benefit.

Yeah, that makes sense. Talk to me about Arc AGI. Why is it so hard? It seems so easy.

It does seem easy, doesn't it?

It certainly seems more evaluatable than tell me a funny joke, right? Yeah, yeah.

I mean, I think if you are old on Arc AGI, then it you probably get superhuman at it pretty fast. But I think we're all trying not to RL on it so that it functions as an interesting held out.

Sure. Okay. Is that just an informal agreement between all the labs, basically?

We're trying to have a sense of honor between us.

That's great. That's amazing.

How many people on Earth do you think are getting the full potential out of the publicly available bottles? Because we're now at a point where we You know, billion plus people are using AI almost daily. And yet I have to admit, my sense would be it's maybe like 10,000, 20,000 people on the entire planet are getting that sort of full potential. But I'm curious what your assessment would be.

Yeah, I completely agree. I mean, I think that even I don't get the full potential out of these models often. And I think as we shift from you're asking questions and it's giving you sensible answers to you're asking it to go do things for you that might take hours at a time and you can really parallelize and spin that we're going to hit yet another inflection point where even less people are really effectively using these things because this basically doesn't require you to like, it's like Starcraft or Dota. It's going to be your APM of managing all these agents and that's going to be a process. Totally.

That's a process.

Yeah.

So- I think Starcraft is such a good example. You think you're just absolutely crushing it and then you realize there's an entire area of the map, you're just getting destroyed. It's such a good comp.

That's great. Anything else, Jordy?

I think that's it on my side. I mean, I would like this to be an evolving conversation.

Yeah, this was fantastic. We'd love to have you back on.

Absolutely. It's really fun. Love to be back on.

Yeah, we'll talk to you soon.

Cheers, Sholte. Have a good one.