Weekly Recap: Apple Debuts iPhone 17, Einstein of Wall Street, YC Demo Day, Declining AI Adoption

You're watching today, there is a chart that is tearing up the Internet about AI adoption potentially going down. This is from the Census. The census says that the AI adoption rate by firm size. They do this poll every two weeks. And Apollo, the giant private Apollo Management. Yeah, Apollo, on their Apollo Academy blog, has a post. AI adoption rate trending down for large companies. The U.S. census Bureau conducts a bi weekly survey among 1.2 million firms. And one question is whether a business has used AI tools such as machine learning, natural language processing, virtual agents, or voice recognition to help produce goods or services in the past two weeks. Recent data by firm size shows that AI adoption has been declining among companies with more than 250 employees.

So is that good?

It's, I mean, I.

It's good for your job.

There. There are a bunch of different reads on this, so I think it potentially could be good. I don't know. I'll walk through kind of my, my thinking on it, but there's a bunch of interesting stuff in here, so. So the chart's going viral because everyone's been feeling that AI has been overhyped and the valuations are crazy. And so people are hunting for top signals, as we have a month ago.

Top signal on Jordan. We put out. We got to play the video.

Oh, yeah, yeah. The Hype Cycle.

It's released.

Oh, it is, it is. Yeah. Let's pull that up.

Let's pull up.

Yeah. So let's pull up the video.

The video that Tyler just posted. Did you post it from the TVPN account?

Yes, I did.

Yeah.

Of the Hype Cycle. We'll play that. But people have been hunting for, for bearish signals about AI because it feels too good to be true. Everyone's getting rich and you aren't. It has the same sort of economic trends as the, as the crypto boom, as a few other booms. And so people are hunting for data points and this is one of them. But let's first review the Gartner Hype Cycle video that we just put out.

I think Nvidia is undervalued.

It may not mean nothing to y'. All. Understand Nothing was done for me so I don't plan on stopping at all. I want this Forever mine Never mind, never mind shutting shit down in the mall It's Sunny girl she the one for me and I ain't even planning to call I won't be shit Forever mine Ever mine Ever mine Ever mine.

Yeah. So everyone's been hotly debating where we are in the Gartner Hype cycle.

Some people call it the TVPN hype cycle.

We'll give Gartner their credit for this one. So people have been hunting for potentially bearish signals, bearish data points. There was that result from Meter that showed that developers that were using AI coding tools were actually less productive. They thought they'd be 20% more productive.

And I saw something else. I don't know if this was misinformation, fake news, somewhat real or just anecdotal, but there was somebody that was saying developers are producing more. Like a lot more code, but an order of magnitude more security vulnerabilities.

Yeah, it's something like they're producing twice as much as many, 10 times as many security vulnerabilities. And I know the post you're going to talk about now, it is in the stack, but it's deep. There's a post.

I'm going to try to find it, but it's.

It's.

Somebody searched, like Vibe Coding cleanup specialist. And there's people on LinkedIn now that are finding employment around cleaning up Vibe coding.

This is their brand.

And so if you're worried about losing your job to Vibe coding, pivot to Vibe coding cleanup. It's a bull market. There's always a bull market somewhere.

Guys, can we pull up the chat on the tv if possible? Anyway, let me run through this. So from the data, the chart is going viral because it's dropping off and it looks bearish for AI and thus for tech America, for humanity broadly. Everyone is saying, please consult the Gartner hype cycle graph. There is a question on where we are in the Gartner hype cycle. I have been saying that maybe we still have a ways to go up. I can't tell if I'm in the left side or the right side of the graph at this point. Like, I've kind of already been through the trough of disillusionment. And so when was that for you? That was probably 2024. 2024 was. Was when it felt like everyone was saying AI is so insane. AI so insane.

You were going like this in the mirror, like the joker a little bit.

Because, you know, AI is so insane. But then, you know, you go back to the ChatGPT moment. That was 2023. Waymo, you know, really like rolling out to general event. That was 2023. There wasn't as much. There was a lot of hype, but there wasn't as much like actual progress. It felt like. It felt. I don't know.

So maybe now we're have 5,000 AI agents for Blank.

I don't know. Just. Just as I process the last few weeks, I feel like I'm climbing up the slope of enlightenment. I don't feel like Satya posting, red.

Posting and showing how he's getting value across.

That feels like the plateau of product.

That feels like that.

That's plateau.

Like we're plateauing and productive.

Exactly. And even the narrative of, oh, is AI progress plateauing? That's not. Is it going to crash? Yeah, that's not the trough of disillusionment. That's the plateau of productivity. So anyway, people. People love to, you know, give their takes.

Anyway, I think that. I think the. The real. An interesting study. I'm sure there'd be flaws with it.

But.

But if you went to companies and you'd say, like, how much would I have to pay you to not use AI for the next 12 months across your entire company.

Yeah.

I think you could argue that, like, in engineering. Org, specifically in terms of, like, things like fraud detection and areas that there's probably a lot of money, but it's very possible that in certain organizations. But then you have the question of, like, okay, does, like, sales call transcription count as AI? Right.

Well, I will tell you, Jordi, because 90% of American businesses told the census that they do not use AI at all, including transcription.

So they would take a lot of money to.

They would take a dollar.

Take a dollar.

They would take any amount of money because it's all net positive. But there's a lot going on in this data that we should run through. So.

And I thought. I thought. I thought people didn't trust data from the census.

Well, yeah. Ara Kharazian has a good. A good take about the size of the data set and how spend and where AI is actually having an impact. We'll run through that. But. So first off, small firms haven't declined even by this metric. So if you're a firm that has one to four employees, you're a very small firm. You're still increasing in AI adoption.

Should we switch over to these, by the way? Guys, we are testing some new microphones so that we can walk around.

The old boys wanted the mics back. They're coming back hot.

Daniel. We can't top till Dwarkeship ships his book.

He already shipped his book. I thought I got a preview. It was. Yeah. That doesn't make any sense.

I don't know.

Dwarkesh's book.

It's released.

Oh, yeah. That came out on straight press. Like we had him on the day it was like on sale, right? Oh, maybe it hasn't shipped, but I got an advanced copy, I printed it out, maybe, maybe it hasn't fully like actually delivered. But the first time we had them on the show was because of.

Yeah, it's still a pre order.

No problem.

Still a pre order on Amazon.

Oh, it's still a preorder. Okay. So okay.

It's going to be released on October 8th. Okay. So yeah, market Nvidia, please don't nuke until Dwarkash can get the book out.

Yeah. So the main headline number in this census data set is that firms with over 250 employees are declining in their self reported AI usage. It's not a huge drop. It goes from like 12%, 14% to 12%. It's a little blip but the chart does look scary. We can pull it up, it's like the third slide. So I think what's going on here is the following small firms see bottom up adoption of AI tools, like can I literally directly use ChatGPT for this? Like you run a two person company at various times, a lot of times you're just thinking okay, I gotta select an accountant or something like let me like ChatGPT.

Right.

And this kind of bottom up adoption just happens very naturally. Whereas if you're in a larger firm, you're gonna get sold on some crazy AI transformation. Mambo Jumbo needs a whole training program and is everyone up to speed at the same time? You need a bunch of consultants and they so of misunderstand what AI even is and they fall into a bunch of weird analogies like we'll hire 100 AI agents. And like that's very different from just hey, Google exists, ChatGPT exists. We expect you to use both at this company. And so this is where I thought the data was really, really, really weird. Because if you look at the data for ChatGPT adoption, we're way off of what this 10% number is for for business adoption of AI tooling. So, so this, this data says that about 10% of companies are using AI at all at work. Which is crazy because there was a data point back in March that said 52% of US adults use AI large language models like ChatGPT. So think about what that means. It means like out of 100 million Americans, there's 100 million Americans that are like yeah, I use AI tools. Like there's maybe 200 million adult Americans or something. There's 340 total Americans. But let's call it, let's call it 100 million Americans that are using AI tools. Just like ChatGPT generally.

ChatGPT Americans?

Yeah, ChatGPT America. What kind of Americans? The ChatGPT Americans. And then 80 million of those are just like, no, not at work though. That makes no sense. That makes zero sense.

And so it doesn't work like that.

There's gotta be something weird going on with this data. And I think it's a matter of the definition. And so the census defines AI is it makes no sense that 10% of companies would fit this definition.

You said ARA at ramp was chiming in on this.

Yeah, yeah, he has a post in the deck.

Yeah.

So the definition.

Yeah, he says these sample sizes are extremely small for these large firm breakouts. Extremely volatile survey to survey. Which is why he's using the six survey moving average business AI adoption is back up in the most recent read. So I feel like that's important. I. Yeah, it's one of those things. I feel like payments providers like, like Ramp. Right. Would, would have better insight here. Just like obviously Ramp is a sample of companies in the United States, but it's at a scale now where they could see like, okay, what percentage of companies are paying for any type of AI products? Yep, that's better than Are you adopt. Are you using more or less AI than you were last time?

Yes, yes. The question is like how Power Law distributed, is it Right.

So the RAMP estimate of share of US businesses with paid subscriptions to AI models, platforms and tools is 43%.

43%.

The US government's estimate is 8%.

This data is weird, right? Also listen to how broad this definition of are you using AI at your business? Anything in this list, if you're using any of these buzzwords, you count as an AI user. So it should be very high based on this. Machine learning, natural language processing, virtual agents, predictive analytics, predictive analytics. Literally just like a linear regression forecast of like, oh, we did 1 million, then 2 million, then 4 million. Looks like we're growing exponentially. Next year we're predicting that we're going to do 8 million. That counts as AI, which is ridiculous.

There's a legend out there that doesn't do any of that mumbo jumbo.

Yeah, yeah, exactly.

Maybe they're just considering it. They're doing it in their head. So it's like organic intelligence.

Yeah. Seriously. I mean you can do that in a spreadsheet, machine vision, voice recognition, decision making systems, data analytics. Just if you're using data analytics at your business, not Even like Julius Powder. No, no, that would be the next step. Just if you're analyzing data, you count. And the last one is image processing. If you're processing images, that counts as AI. And so that's basically any SaaS tool, like any consumer LLM product used by employees should count. And it should be really easy to fulfill this definition. I'm thinking of like, if you're using Ramp, obviously when you scan a picture of a receipt that uses image processing, your business, technically.

You're telling me the people that, that are responsible for the Department of Motor Vehicles are getting into measuring the usage of AI?

Yeah.

In the workplace.

And we should all this is incredibly botched. It's incredibly botched. Can you say it back to me?

But it is really funny because one post like this, it's a timeline and you have people like Daniel just saying, it's over. Obviously he's joking. It's over.

He's having fun.

But I am very interested to see the results of hype.tbpn.com yes, go fill it out, please.

Hype TVPN. I was trying to dig into this more and was wondering, is this chart bearish? And I don't think so. I think that there's a big shift that's going to happen. The number of business owners who pick up the call from the census and actually say, yes, I'm using AI, that will go down, but the real number will actually go up. So if you're using Stripe, there's machine learning running under the hood to do fraud detection. So you are an AI user, but it's happening at a different layer in the stack. If you use Ramp, our sponsor, save time and money. Go to ramp.com when you scan a photo of your receipt that is processed using AI. But you aren't thinking about yourself as like an AI user. You're just using software that happens to be AI enabled. And the same thing will happen, you know, with your email. Like, it doesn't matter what email client you use. You could use Google Apps, you could use Outlook. Like there's going to be AI baked in there. You're going to be an AI user. Even when you just go to Google, you're gonna see search overviews that are LLM powered. You're gonna be using ChatGPT. Like AI is going to be 100% used. And yet everyone will say, no, I don't use AI at my company because I'm not an AI company. And there's this weird.

I use SaaS.

Yeah, I use software. Basically it's like, are you specifically a software company?

Well, and that's part of the way that people have been waking up recently being like, wait, AI is SaaS? And it's like has been.

It's the same thing with non relational databases or like in memory caching.

AI had to become SaaS in order.

To destroy it really, really did. And so I think AI is going to seep into every crack of the small business data of small business day to day operations. But the operator won't be proudly telling the census worker, I'm all in on AI for much longer. I think that trend will continue and less and less people will be saying I have adopted AI even though they will have, which is a weird dynamic, but it certainly tracks.

So I'm Bora, this is Jerry. We're co founders of Reacher and the outfits are because we help brands reach more creators and make viral content to help them reach more people.

How do you do that? Like what is the actual product?

Yeah. So we build AI agents to help brands find the right creators, personalize the.

Outreach, get them to respond, pay them.

And then also generate scripts and hooks so that they can make viral content.

We have some lightning round questions. Favorite entrepreneur. What you got? Oh no, you can say anybody. Yeah, I know, just a buddy.

I was gonna say Elon, but I think it's Jensen.

There we go.

The first Jensen. Let's start counting him up.

I'm Bruno. We're building gauss, a personal AI investment analyst for retail investors.

Oh, interesting.

B2C. If you invest in stocks, you're welco join our platform.

Yeah, I mean there's been a ton of moves in the, in the retail world. Tons of meme stocks going on. Is this going to help me find the next great meme stock or avoid getting caught up in the meme stock stuff? Is it more for the retail investor who doesn't want to get burned or the retail investor?

Can I tell that? To not make mistakes.

Exactly. That's actually our number one use case. We want retail investors to make less emotional mistakes.

Emotional.

Right now we're focusing on thesis driven investors. So not the meme stock price. So you have a clear thesis. You want to monitor the market, but you're a busy professional. You don't have time to keep track of everything that's happening. We can find opportunities, you know, you connect your portfolio to us and we can start giving tailored advice.

Favorite entrepreneur? We've been asking everyone who comes to mind throughout history, I have to say.

David Velez from Nubank. He's an inspiration for me. He challenged, you know, the most traditional industry in Brazil, big banks, and he managed to create a generational company.

Fantastic.

I really like him.

Introduce yourselves. Who are you? What are you building?

Cool.

So my name is Jason Cornelius, this is Steve. Hey, what's up, guys?

Steve with Perseus Defense. What's up?

And we're building 16 inch missiles to shoot down drones.

16 inch missiles to shoot down drones. Okay.

Okay.

Do we not have 16, do we not have 16 inch missiles in the arsenal currently? Is this something that's like, we're familiar with like the Patriot and the bullets.

How much do they cost and how do you fire them?

Yeah, it's a good question. So there are some existing solutions for like the counter UAS problem, right?

We walk through that. Well, yeah, walk through the solutions for sure.

So there's, you know, electronic warfare.

Yeah.

So GPS jamming, unreliable.

Right.

There's, there's easy ways to defeat them. You know, you can make a more advanced drone. And this is what you have, the.

Fiber optic cables, fiber optics, like already.

Defeating some of those solutions that we have.

Yeah.

And then there's like the Death Star laser beam, which is very expensive, immobile, very heavy, requires a lot of power. There's AI machine guns. So like there's a million solutions and.

We were, who's your buddy that does that? Acs, acs, gone on in the truck. Yeah.

And so there's some really good solutions out there for different use cases.

And we need more.

Yeah, we, we need more. And they all have pros and cons and they have like a layered. So some of them, like ACS is very short range, some of them, like a Patriot is very long range. And so we looked at it. Who's going to fill this gap that's in between when we don't want to spend a $10 million missile to shoot it at 10 miles away.

Yeah.

But we also don't want it to get so close to us that it's, you know, right on top.

So what's the, what's the range? It's a 16 missile.

Half a mile.

Half a mile.

Okay. Lightning round. Favorite entrepreneur, who do you look to for advice or, or as a role model?

We're looking at Dino at Saranic right.

Now and trying to match that growth trajectory that they're having a lot. Introduce yourself for the stream. Who are you? What are you building?

Yeah, I'm Pranav.

Nice to meet you.

We're building Meteor and this is Farhan, the AI native browser that's going to kill Google Chrome.

Oh, okay.

We got the browser wars going. Let's go. Yes.

Okay. Okay. Browser wars. Are you guys going after Comet too?

Yeah, absolutely. Everyone? Yeah.

Okay, well, what's the plan? I mean, everyone has installed Chrome at this point. Everyone runs it. It's so hard to rip it out. How do you create a solution that's 10x better?

Yeah. Well, I mean, we all spend seven or eight hours every day on our browser. We do most of our work on the browser. Most of it is redundant work. We're saying we can automate all of that away. We're saying we want to spawn AI agents that work with you as your copilot to the web, that can do your work for you so you don't have to waste your seven, eight hours.

Okay, but what does that look like as I use the product? How do you pitch it to somebody who says, Chrome works well enough for me?

Right. Chrome does work well enough for you, if you're okay wasting many hours a day.

Okay.

What we're saying is we're giving you time savings. It's sort of like people pay thousands of dollars for personal assistants.

Sure.

Right. People pay thousands of dollars for employees, like, to create automations. We're saying we want to create virtual employees that run on your browser.

Yeah.

Right.

And that's meteor. So like, let's say you want to handle your calendar.

Sure.

Right. Just automate all, like meeting creation and, you know, based on your. Your emails.

Who's your favorite entrepreneur?

It's got to be Steve Jobs.

Steve Jobs. Okay. Introduce yourself in the company.

Cool. I'm Has Hubble. I'm the founder of Pali. We're bringing every contact and every conversation from every platform into one place.

Okay.

And using AI to help you keep track.

Who's your favorite entrepreneur, who you learn in a lot from?

My favorite entrepreneur of all time actually would be like Richard Branson. And here's a fun story. So when I was. I dropped out of high, and I remember when I was like 14, 15, speaking to my mom about this, I pointed to him as someone who was successful, who dropped out.

Yeah.

Actually, did he drop out of high school, though?

Yeah. Oh, wow. And then 10 years later, I was invited to spend a week on his island with him.

No way, nothing.

And I spent actually time with him on Necker island. And so it was like a very cool full circle moment. He's been a great, great mentor over the years.

That's amazing. That's amazing. Introduce yourself. What do you do? Yeah, my name is Peter. I'm the CEO of mcpus. We help people use mcp of course build open source dev tools and infrastructure. I was going to say, what's the open source strategy?

Yeah, Open source, I mean is a.

No brainer strategy is to open source.

Yes.

License.

Everything is free. Okay. Everything's free. And then there's probably a GitHub repo that's growing and some measured by stars or something like that. 7k stars here for 7k stars. Congrats. And then at some point do you want to be more like Red Hat Linux? Like who are you looking to in terms of well run? Is it GitLab like super base Supabase? Okay. Break down. How will their business model look like yours?

Yeah.

Because they.

They have a bottom up approach.

They get developers and then those developers introduce inside the company the tool.

Sure.

And I love it.

Who is your favorite entrepreneur or someone in business that you look up to?

Yeah.

Brian Chesky, I think. Brian Chesky.

There we go.

Introduce yourself for the stream. Who are you? What do you do? What are you building?

I'm Anwal, I'm the CEO of Closera and we build AI agents for commercial real estate.

Oh, okay. What, what, what part of the commercial real estate stack, the actual transaction do you sit on? The buyer side, the seller side like.

Yeah, it's a great question.

Administration. There's so much that in commercial real estate.

Exactly. It's a huge industry. So yeah. My co founder and I are families actually grew up in the commercial real estate space. Both our families are developers.

Cool.

But you.

Let's hear it for the developer building. We like developers. It's time to build first.

Usually say that in a different context. Exactly.

Yeah. But yeah, we. We kind of did a survey of the whole market and we realized actually the most insane amount of manual, tedious, boring work that of course can be automated with AI agents is actually done by brokerages and so that's where we decided to focus.

Favorite entrepreneur. Someone you look up to.

Yeah, I think you know classic answer. Definitely Steve Jobs. Like reading that biography in middle school is like what got me to want to move to the Valley. Go to Stanford. Yc.

The Walter Isaacson. Exactly.

Book.

Introduce yourself for the stream. Who are you? What are you building? How's demo day going?

Yeah.

Hi.

Do I look at you or do I look at you?

Whatever you want. Okay.

I'm going to look at you.

Fantastic. I'm Anson.

We're building normal and we're automating. Hardware testing and compliance.

Hardware testing and compliance. Any specific verticals are we talking, like, semiconductors? Rockets, spaceships, cars?

Yeah, we're starting with, like, robotics and drones and other. Any electrical components that have, like a radio?

Yeah, yeah.

Part to it.

So that's mostly it. Who is your favorite entrepreneur? Someone you look up to. Can be anyone. Favorite founder? Favorite founder. Can I say, throughout history.

Sir.

Sir John A. MacDonald. Who's that?

First Prime Minister of Canada.

Okay. Did he start Canada?

We don't have any buttons for Canada. Yeah.

This is something American. I don't know if I can fully endorse it, but it's. It's outside the box and I appreciate it for that reason. Introduce yourself. What do you do? What are you building?

So I'm Aiden, the CTO of Effigove. We're doing the AI operating system for local governments, starting with 311 in a box.

31 1. What do you. Is that information hotline?

Yes, the informational counterpart to 911.

Okay.

So interestingly, only around 2% of cities have a 311 line. Even though every city has 91 1.

Yeah.

Yeah. And what we're doing with our first product is of creating a 24.7call center that's available to anyone from. If you're in San Francisco.

Yeah.

Or if you're in, like a city of 10,000 in the middle of Iowa. Like, you should have the same level of service.

Call three on one. It automatically routes you to your service. And then if I'm in my city, it'll give me information. What are the. What's the. Obviously, everyone knows textbook 911 call. Someone's breaking in my house. What's the textbook 311 call?

Yeah. I think.

Is this road open or when does the trash pickup come? Stuff like that.

Stuff like that. It definitely varies a lot more city to city. So we're live in Florida and we're taking calls about alligators roaming around the city.

And it can kind of route. And I'm sure you've set the system up so that you can route to 911 if you need to. Absolutely. Or Right. To animal control. Kind of sit above the rest of the stack of all the different city services.

Yeah. And that's kind of why we started with this is because it gives us a really great insight into what are the calls and services that the city's providing. So now we can go down and actually handle these workflows end to end with our agents.

Who is your favorite entrepreneur of all time? Someone you look to for advice.

Peter Thiel.

First.

First person to see pt. So, yeah. Why Is this sitting here? What do you do? Introduce yourself.

Who are you?

Yes, my name is Zane.

I'm the founder of Knox Metals and we cut metal. So we're trying to build the most efficient metal service centers in America to empower every factory in America to work with better margins and to move faster.

Okay. There's a lot of ways to do that. He's doing the meme.

He's doing the meme.

He's doing the meme.

He's re industrializing.

Who's your favorite entrepreneur? Alex Karp of all time. Alex Karp. We've had him on the show last week. Introduce yourself. Who are you? What are you building?

Yeah, so hi, my name is Vihar and we're building Orange Slice. We're basically AI search infra to help you find customers that already want your product.

Who is your favorite entrepreneur of all time?

Oh, that's a good one. I'm not, I don't actually know that many entrepreneurs in general.

Someone that you read a biography of or look up to or maybe just saw a recent interview with and thought was I could learn something from them.

He's Lisan Al gaib of Enterprise SaaS. He doesn't need a blank slate.

He doesn't need any influence. Well, honestly, I don't think, I don't draw that much inspiration just because I think the biggest motivating factor in this might be contrarian is not actually inspiration, but like the fear of mediocrity, of it not working out has been way bigger for us. That's the biggest reason. I don't have that much inspiration. But Eric from the Ramp, from Ramp. My co founder used to work at Ramp. So I mean they are absolutely killing it.

Introduce yourself. Who are you? What do you do?

My name is Arlan. I am building Nozomio, which is an API that gives more context to agents.

Okay.

So if you're a cursor user, you can pretty much just go to cursor, index any documentation or code base and give it as a context. An agent.

What exactly is happening? I know Cursor does kind of roll ups of what's happening. There's, there's some rag infrastructure that's popular. What else is going on in, in the actual product to unlock value.

Yeah, so the two reasons I built this is because like I was so fucking pissed when I would like go to Google copy, paste all the docs and paste them right in the cursor. Yeah, the memory is not persistent, so they will fucking forget that in like two seconds and second One is that you also pollute context window, which like causes context fraud. It's actually a study done by Chroma, which is like a company.

Oh yeah, we know Jeff.

Well, yeah, saying like post 4000 tokens. I think like the performance of any LLMs goes like this.

Yep.

So that's like the reason I built this.

Who's your favorite entrepreneur of all time?

Yeah, who the fuck the most money, bro? Like Elon Musk. Yeah, yeah. He's the richest man, bro.

Yeah, yeah.

But otherwise, yeah, Brian Chesky is my goat. I love Brian Chesky.

Brian Chesky, Founder Mode. Would you mind introducing yourself and your company?

Sure. Hi everyone, I'm Alessia. Vive Flow is the all in one solution to build production ready apps in no code.

Okay. We've seen a lot of news recently that there's now a whole industry of vibe code cleanup specialists. What's your reaction? Can vibe coding ever make it to production?

No, no, not now at least. Like all the solutions out there are really broken and they can they give you this wow effect in the front end, but on the back end side there is so much to do and there is a huge need. People are like, they have hair and fire problems and they need to solve these problems. The market is huge.

Who's your favorite entrepreneur of all time?

My favorite entrepreneur is Taylor Swift.

Taylor Swift.

I'm a good one. I'm a big fan and she is amazing.

Introduce yourself for the stream.

All right.

What are you building?

Well, I'm Joshua March. I'm the co founder and CEO of Veritas Agent. We build AI agents for the consumer lending industry.

Okay.

So interfacing with borrowers to help doing sales, servicing collections for like fintechs, banks.

Credit unions, services, that kind of thing. Who's your favorite entrepreneur of all time?

So there's this guy, Felix Dennis, who.

Was a British entrepreneur and he wrote.

This book called how to get Rich, which I think is the best book on entrepreneurship.

What you actually do.

Yeah.

My name is Kevin.

I'm the CEO and co founder of the prompting company and we help product get mentioned in ChatGPT.

Okay, okay, okay. Not just another generative engine optimization tool, but yeah, yeah. How do you differentiate?

Yeah, for sure. So we plan the infrastructure layer. We're not just creating articles for you. We create shadow sites for our companies to, you know, influence all.

Who is your favorite entrepreneur of all time? In history.

In history.

I mean anyone alive or dead.

Dude, Steve Jobs.

Steve Jobs, good answer. Introduce yourself. Who are you? What are you?

Bob.

Hi, I'M Bob, CTO of Embedder. We build coding agents of firmware. Okay. Coding agents for firmware.

Firmware, yeah.

What. What device are they betting couldn't be done?

So we do microcontrollers, Linux computers. So think about any micro you can think of. STM32, ESP, TI, NXV, any, any manufacturer.

When did you learn you wanted to do agents for coding? Agents for firmware.

That's good.

Good, good question.

So I did hardware all my life. Started like, you know, started when I was small doing robots, moving to more professional stuff in high school, in college.

School.

Let's go. In college.

Yeah, in college I built humanoids and then went to a startup, the med tech stuff and then went to Tesla, worked on Robo Taxi. Yeah. So we kind of saw it. I saw this problem from like, you know, like a 10% research lab to a trillion dollar company.

Yeah. Who's your favorite entrepreneur of all time?

Oh, Elon has to be.

There we go. Introduce yourself. What do you do? Hey, my name is Scion.

We're Kalinda.

That's a great name.

Yeah, we're basically a deep research for class action law firms.

Okay, interesting.

Yeah. Let's give it up for litigation.

What's the secret sauce? Is it. Is it just the bad data in, bad data out? You got to get private data that's may behind some sort of wall. It's not scraped into the pre training weights of GPT5 and so you have some cornered resource in the data or is it something in the architecture or something at the UI level? Like where's the interesting. Well, kind of like most startups at.

Yc, a lot of it sits on the application layer.

Sure.

Most of it for class action law firms is the scale at which they do things. They can have maybe 10 million pages for litigation.

Oh, wow.

So with that kind of scale, you can't obviously just toss them.

Yeah. Be a hassle and just click upload another PDF, upload another PDF, Upload another PDF.

Yeah. And then the kind of second thing is like most of the data sort.

Of sits in silos and kind of.

Aggregating it all and making it sort of, you know, comprehensive like human wood is.

Who's your favorite entrepreneur of all time?

Sam Altman.

Sam Altman.

Just because I met him twice, but he's.

Yeah, he's the goat. Introduce yourself.

We haven't pitched yet.

Who are you? What do you do?

Raf Garcia, founder of Kernel. We're crazy fast browsers in the cloud. For your AI agent.

The browser wars, the Browser wars are.

On, but these are computer use. So it's API to go and interact with the website.

Yeah. So nowadays a lot of AI agents, you want to hook them up to the exact same tools that humans use. And the browser is one of the main tools that we use on a day to day basis. And so we're powering automations in health care, fintech, qa, Lots of different use cases that AI is really good at.

Who is your favorite entrepreneur in history? Who do you look to?

Favorite entrepreneur? How far back are we going?

We can go live or dead. Literally thousands of years if you want.

You can go.

Somebody has. Somebody has mentioned. Jesus Christ. Somebody has mentioned Eric Lyman at Ramp. So I mean, I didn't make the comparison, but it was made today.

I mean, cliche. I'll have to go Steve Jobs. I mean, I grew up reading that.

Was Jordy's answer too.

And studying Apple. Like that was my first computer as a kid.

So who are you? What do you do? Explain yourself.

My name's Adi. I'm a co founder of AgentMail.

It's the first email provider built for AI agents. The key distinction. There we go.

The key distinction is we're not AI for your email, we're email for your AI.

Okay. Oh, interesting. Like browser use for agents. I was wondering about this because I go to agent mode and oftentimes it hits a wall where it's like, okay, it should just. Now it should just email the person to actually kick off that process. And it can't do it. Yeah.

How much money are you making?

How much money are we making?

Yeah. Did you share a KPI at demo day?

Did you share what KPIs?

We did not.

But we love our customers. Okay, but you see customers plural, so I know there's more than one.

Yes.

Okay, I'm learning, I'm learning, but keep going. Okay.

Did you come in with this idea or did you pivot to it?

So we were trying to get rich quick.

Is that why you're. Is somebody vlogging out there? Is that someone else?

Exactly. We're like hustlers.

YouTube, you're getting that?

YouTube. I know what question I'm asking. So we wanted agents to go on.

The Internet, make us money, farm, free trials.

And then.

Interesting.

Exactly. We realized, okay, it's kind of hard to give agents their own inboxes, but they should have them. Sequoia is talking about them. LangChain is talking about them.

Agents need their own email inbox.

Okay, who's your favorite entrepreneur in history? Alive or Dead.

I like Warren Buffett.

Warren Buffett, Yeah. Interesting. I think he's one of those guys.

Who'S like, actually humble. Yeah, right.

Like, it's not like, okay, hey, I'm worth 100 million. I think he just genuinely is in it for the love of the game. He does. He's love of the game.

Hopefully one day, you know, know, I could be like that.

Who are you? What do you do?

Hey, I'm Odit, building Riff, which is Cursor, for music production.

Ooh, that's fun. Are you sitting. So Cursor sits on top of an open source ide. Are you sitting on top of Ableton or something?

That doesn't exist for music. So we actually built everything from scratch.

Okay.

Trained our own models.

Sure.

We have our own, like, editor. It kind of looks like GarageBand, but you can use AI tools to generate sounds and mess around.

Interesting. Do you bring your own, like, samples or do. Do you plug into a sample library? Like, where. Or is all this. Are all the samples AI gener?

Yeah, we have like a user community of samples, but we also, like, let you generate your own.

Cool.

And also, like lots of bells and whistles to make it easier. Like, for example, you can hum out a melody.

Yeah.

And we'll play it on an instrument for you.

That's very cool. Favorite entrepreneur in history, live or dead?

Dr. Dre.

Dr. Good. I knew I had to ask you. Anyway, thank you so much.

Great to have you.

Who are you? What are you building?

I'm John Ferrara. I'm building Juxta, which is a GPS alternative that can track anywhere on earth with no hardware.

I feel like I've seen this on X.

You might have seen our. We had a pretty cool launch video.

Very cool. Yeah. Okay, so how does it work? There's some company that spun out of Google that just sold that does something similar. It's like Magnetics. But how are you doing it?

Yeah, so we basically use simulated technology to basically render any environment using satellite imagery or indoor floor plans into 3D.

So you do a geoguessr?

You could put it that way. Yeah. We should have the company that way.

Yeah.

And then we simulate millions of people or objects moving throughout a space.

Cool.

Simulate their sensor measurements with like accelerometers, gyroscopes, and then train models on that and then basically deploy to software. And we can leverage the IMUs, which are like sensors that are already into every phone and most other devices today to track you.

I can see you through your wifi. Lots of people. Yeah.

If You've seen that film in like the Dark Knight there's like a scene where Batman.

Oh yeah, I know the one you're talking about. That's how you can envision it. Who's your favorite entrepreneur in history? Live or dead?

Oh my God. I would say, I would say. I know the Rift guys just said Dr. Dre, but I'll say Drake because he just came out that Amazon store and Drake's my favorite artist, so.

Oh, there we go.

What Amazon store?

I don't know. You haven't seen this. This is his actual plot to take over Amazon Music as a distribution platform. But he's working with Amazon to sell all of his merch and all of his goods through their storefront and basically facilitate all their distribution through there. So that's my cop out answer. But I just love Drake.

And we have a special guest with us opening the show. Introduce yourself for the stream.

Hi, my name is Peter Tuckman. I'm known as the Einstein of Wall Street.

Einstein of Wal street and I'm not.

That smart, I just.

Look who was the first person did you give yourself.

Aaron Burnett gave me that. And I've had a show on her show now on CNN for the last four months. Ever since the beginning of the sort of mini crash that was sort of self induced by our, our. The new leader. The new, the new leader. The new sheriff in town, but correct. She called me in. So she used to work on the floor with CNBC.

Sure.

And she nicknamed it probably 20 years ago and called me Einstein and that sort of picked up.

But your latest, your latest collaborator is not on TV but on streaming, correct?

Correct. So I had the fortunality of actually meeting ishowspeed. Right. Young 20 year old kid who I got invited. So there's a team that's doing his stream thing. There's a young guy named Adam Faze.

We know he came to our event. We were here about a month and a half ago. He came to us.

Really.

So Adam, Adam's something special. He found me about four years ago and asked me if I would let him shoot me every day. Walking out of the stock exchange for 30 days and just do what I do. And we did this whole TikTok thing called Einstein elementary went viral like me and like marching bands from Brooklyn and walking around with a 50 pound Hershey's Kiss and doing all this crazy stuff. So he called me last week, asked if I could bring speed to the floor and it didn't seem appropriate at the time and he said well how can I get you to meet him. And I said, come up with an idea. And he's an idea guy. So he called me on Thursday morning, said we're going to be at the Bull. Yeah, he's going to be at the Bowl. Would you come and welcome him to Wall Street? I said, absolutely. So they gave me a police escort. I went down there and met this young guy and talked about some of the shtick I do, you know, investing in stocks and not stuff and about, you know, the fact that, you know, that everything that his life is based on is actually a publicly traded company and that instead of, you know, buying the next iPhone, he could probably buy a couple shares of Apple and then maybe, you know, he could invest in his future besides the amazing future he's got set up. He received it totally. He talked about it a couple days down the road. So I got the message across and, and I've blown up. I've got literally 200. So I have now I came in at 260,000 followers on Thursday and now I have 419 and counting and I've got 40 million views on some of my posts. 7,000 DMs.

7,000.

I actually just did a video to answer that.

Anyway, what was the first stock you ever bought?

You know what I very rare. I did not ever buy a share of stock until recently. I went through most of my career because I'm a registered trader broker on the floor and I'm not allowed to be in a stock for a customer and for myself in a 30 day period. And so I built a trading strategy 20 years ago, trading the S&P 500 based on information that I have as a broker on the floor. And I trade all 347 names in the S and P every day. So basically I made a decision years ago, you know, and also money does a funny thing. If I'm in a stock for myself and I get an order for a customer, it's going to impact the way I trade. And I didn't want that to ever be the case. So I, I do well enough here as a commission broker that I said, you know what, I'll invest in my kids, I'll put them through college and they both graduated without any debt. And that's where my money went.

What about some of the first stocks that you were brokering?

First stocks that I broker. So we're here, we're talking about IPOs, right? And so, you know, I probably have traded more IPOs than anybody on earth. I've been in the Crowd on all of them. I mean, I remember the first IPO I think I traded was LinkedIn, right? It was priced at 30, it opened at 60, ran up to 180 same day. And that was where I got the sense of how. How amazing the enthusiasm around an ipo because it's based on confidence. Nobody knows it.

And then.

And so one of I. I've had so many fun experiences in IPOs. Jack Ma Alibaba IPO, which was one of the biggest that we've ever had here. He and I bonded, we spent. So the IPOs here take probably four or five hours. We just opened Klarna, right. And the process behind that, which we do better than anyone in the world, is the price discovery process. That what goes on, it's kind of like building a building. The way you open a stock. The way we open a stock takes time because it's a matter of the information going back and forth. And it's a public offering, so the public has to understand what's going on. And so we go back and forth with our customer. It's kind of like if you put the bricks in the mortar in the right place and you. The way a stock opens, if it's built correctly, is going to purport the way it trades, forever guarantee it's going up, but it will have structure and foundation. And, you know, you go to nasdaq, Nothing against nasdaq, but if you go to nasdaq, it's fully electronic. So it opens when the buyers and sellers meet in an electronic marketplace. Here, it's different. So then you're going to see what you see in that. It'll go up, it goes down. There's no. There are no guardrails, right. So here things are a lot different. So Jack Ma and I spent four hours together in the ipo. He and I are very short. So we bonded on the fact that we were the shortest people in the crowd. We were on the same level. Not necessarily. Not financially because he became a very wealthy guy again on that day.

Someone on our team actually was here at that ipo. Yeah.

As well.

Amazing time. And then we've had. Look, then, you know, the whole landscape has changed. When we were completely blew up, the whole IPO market and that affected.

Tell us that story. Well, remember the S1 going out, but then did they get out? Yeah.

Well, what ended up happening? Look, as I said, it's built on confidence. And when you know, everybody. It was huge money put into we work. He was lauded by now he came down to the floor many times. And it wasn't until the night before, everyone needs to know this, that on the end of the roadshow is when all the financials are given out to the public. And it turned out that it was all smoke and mirrors and he had spent all the money on prostitutes and blow and that the company had nothing, no basis behind it. And so it was aborted. The deal was aborted. And what that did was it changed amazingly enough that one stock could change the landscape of IPOs for years. It literally everyone said IPO windows closed. That's what closed that, that's what closed.

The factory on Mad Money was like, we don't want this, we don't want exactly that.

And so to think that you know something about what's going on and we didn't know anything on literally until the night before made such a huge impact on the market. It literally has taken a couple of years to grow out of that. The night before that we had had some really wonderful robust years. ReMax and a bunch of stocks they would open the huge premiums we were trading give us a quick ipos a.

Day, quick history on Klarna. Right, Because Klarna has tried to get out a couple times.

Okay. So I don't know the history about Kleiner. All I know is that I watched how it opened here. So it did take a little while. It did open at 52, I think. Right. And so it's not trade, you know what, it's hard to know. So what's important about it is where the bids are, where the offers are and how the foundation looks. So where the bodies are buried within the stock when it opens. Right. There was enough stock at 50 to bid on balance with what it's called. So stocks pair off that are priced accordingly. So four point something million shares open. That means that they were market orders. They had no limits to them and that's. That was they paired themselves off. It was a 52 bid for 300,000 on balance meant that there was a stabilizing bid, whether it was by the Goldman Sachs the company or the banker or not. But that's what made it open at 52. If there had been more to buy at the market, it would have opened higher or more to sell, it would have opened lower. It opened at 52. It did trade up to 5720 on a small bit. 70,000 shares traded up and then it came in. It's trading at a couple of dollars lower, which is reasonable. It does not have the power and the impact that figma and, and Bullish did Bullish.

So, so, so, but, but all three of these are relatively like. Is there a specific strategy that these IPOs have had in terms of like pricing at a, at a, at a. At what feels like very reasonable. Like.

So you know what they're, they, they, they're pricing it based on the reaction to, to what the, what the investment community is to give.

Yeah, it is.

So some of the other ones got priced at 30. There was an oversupply over demand at 30. They upped it to 30 to 35. That's how it will go. So the night, the last night when they go out to the roadshow to give out allocations, they will get a sense of where that valuation should be and then that's where they will give it up. They left some talk the table as did those other IPOs. The other ones opened at 90 and went to 117. That was, that was Bullish. And Figma also traded a huge premium. My gut is that the sectors that they were in, there's a lot of pressure on this sector right now. And, and so those, those sectors were more icloud and, and AI based. Right. And so I think that's why there was a little more pop to it.

Well, we know you have to get out of here. Thank you so much for stopping. We'd love to have you anytime. I'm always talk for another.

Let's go.

Thank you.

I love it.

We need to dive into the Apple and the iPhone announcement briefly. Obviously this has been talked about. It happened a couple days ago, but the post that stuck out to me, there were a few about the design, but the big one was that the iPhone 17 Pro now has more graphics power than an M2 MacBook Air. I have. This is a MacBook Air. I don't know if this might be the M3 or M4 version.

Wouldn't have clocked you as an Air guy.

I just went for it because I was like, I'm traveling more. I'm less in workstation mode. And so let me give this one a try. And let me see, I have the M4, so I think I have more power here than the iPhone 17 Pro. But even just two years ago, it's crazy to think how quickly the chips are shifting from workstation, which is huge laptop to phone. And so I wanted to talk about what the downstream implications of this are.

And did you, did you see the camera functionality? Apparently if you take a, you can take a photo holding your phone vertically and it will give you the Lens horizontal.

Yeah. Because again, it's. I think it's 4. I think it's a square sensor that's maybe 48 megapixels.

How are you feeling about the orange? Do you think this is Tim Cook signaling that he's a BTC maxi?

Oh, I thought it was him talking, giving a little nod to Gary Tan Y combinator.

That's one or the other. Clearly. Had to be one or the other.

Yeah, it is.

The color, the colors I like. I like orange. I think it's a under underrated color in general. But still the color options were interesting. Like no matte. You would have thought they would have done a matte black.

They didn't do matte black this year. They just. Wow. So they have a silver and like a blue. I think something like that. I like orange. I think it's a cool, bold color. I think it's one of the ways to stand out. It's McLaren Orange. You know, McLaren Orange looks like F1. Maybe it's a nod to that. Who knows? But they're, they're having fun with it. Hopefully next year. The color I'm pulling for, ramp yellow. Can you imagine the ramp yellow iPhone? It just might happen. Maybe we'll make it happen. But on the tech side, you know, people were pretty. Like there were a couple takes early that were pretty like bearish. Like, oh, like nothing changed. Nothing changed. And Ben Thompson summed it up pretty well.

Tim Cook, you had a funny banger about this. I mean Tim, Tim Cook posted. He did post.

Yeah.

The morning of. And he was like days like this make Apple, you know, like, like basically like hyped it up as though it was going to be this dramatic announcement. And I think people were pretty, yeah.

Pretty let down at this point. If it's not a change in the overall form factor, like you're going to folding phone, it's not going to be as easy just to talk about completely. But at the same time it has a bunch of improvements that I think will have downstream effects that might be interesting and might actually have kind of these viral moments. And I'll get into my thesis here. So has a vapor chamber in it that allows it to cool. So if you're in the sun or you're running some heavy duty Apple, you're not going to get. Your phone's not going to be as hot anymore. 50% more RAM. Obviously that's important for AI. And then they have a new chip that's designed with the LLM transformer architecture in mind. So the Apple a Series of chips. They've always been somewhat AI optimized, but more for just broader machine learning tasks. Now they're doing the same thing that OpenAI is doing with Nvidia and Broadcom on chip design. Same thing that Google DeepMind TPU team is doing over at Google and Anthropics, working with Amazon and Trainium.

Here's My post introducing iPhone 4000, a newer, better iPhone that's lighter, faster, newer and better.

Where did this land? How many likes did this wind up getting?

6.

6,000?

6,000.

And the funny story here, the little inside baseball. Jordy posted this exact post like five minutes earlier.

Well, so. So I originally had this post as a draft.

Yeah.

And then I altered it.

Yep.

And I posted it and it completely. I had a different picture and I did iPhone 472 and it was just completely flopping. It got like one like on like 300 impressions. Like, people hated it, the timeline hated it. I just went back to this version, posted it, and then it completely ripped. So never give up. Never give up. I think you have a good bit.

But the funny thing here, I mean, obviously it is funny that Apple always comes out and is like, it's our most powerful iPhone ever. Because it's like, well, if it was the less. If you guys messed up and accidentally created a less powerful iPhone, you would just keep selling the current one. You're never gonna come out and be like, this year we're selling one, the latest iPhone, but it's less powerful because just sell the last one. Then just be like, no new iPhone because we couldn't make it anymore.

Yeah, we actually made it. We made it perfect.

We made it perfect. So we're just selling.

How are you feeling about the air? I think that's gonna sell incredibly well.

I don't know. I'll need to feel it and try it. I've never had any iPhone that's not just like the biggest and most of the biggest screen, the maxed out one. I've always gone for the most power, the biggest one. I've actually considered carrying an iPad mini just as a phone because I'm so big that my pockets are big enough to carry an iPad mini. Yeah, no, no. I was feeling it and I was like, you can get it with cell service, so I think you can get a phone number on there and so you can just have like, pull out your giant iPad mini. But I never did that. I've always just stuck to like the pro Max, the big one with the big screen. But it's easy I don't know. The interesting thing about the iPhone air was this. The whole computer is in the camera bump now, and basically everything that drops down is just battery and screen, battery and screen all the way down. And so people are wondering, like, you know, they could clearly take out the battery in the screen, maybe make like a pin. Like, they're getting so good at miniaturizing, like, what it means to be a computer that they can put that in all sorts of different things.

I would be. Would have been really excited to have an iPhone that doesn't. That will sit flat on a surface without a case on it. I haven't used cases in many years. I just like the. The feeling of the phone.

It is crazy that they've.

So funny that it's just perpetually like, you know, just give me, like a flat surface. Like, I'm happy for the phone to be.

Yeah.

Thick as necessary to just give me a flat surface.

And it looked like they were going to do that with the iPhone pro because it has this wider notch. So at least if you did. If you did the notch evenly all.

The way across, it might be nice if it's kind of like.

Yeah, that's fine, I guess. I don't know.

But it's all.

It's also so funny, the idea of, like, you have this incredible camera.

Yeah.

And then you're just, like, putting it on. Like, really want you to buy a case. Like, it's like a lot of the product design seems driven by.

I don't think that's the real. I don't think that's the real reason.

I think it's a. I think it's.

I think it's. I think it's genuinely. Like, people don't like heavy phones, and so you have to balance. Like, you need the optics of a camera that's this certain thickness, and then you need the phone to be thin so that people actually carry it around and aren't like, this thing's heavy tinfoil hat. Yes, it's case. I just don't think they.

I want.

How much money do you think they make from cases?

I mean, they're adding. What percentage of people do you think when they go to that. Buy an iPhone, buy an Apple case.

I would. I would estimate that it's less than 1% of their overall revenue. Yeah, totally.

But you're talking to the goat of profit maximization. Tim Cook.

There's just so many other ways. I feel like you could sell 1% more iPhones if you actually choose this odd choice because people are like, well yeah, it's the lightest iPhone. It's lighter than the Samsung one, it's thinner, it's lighter, it looks great. And then there might be something about the notch kind of holding on your finger and so you can actually hold the phone easier with that as opposed to if it was just like fully smooth anyway.

Yeah, I mean when you start thinking of, of when you start like the incremental sales from design driven decisions like this, I do think it'd be significant.

But anyway, huge keynote filmed on an iPhone. Obviously a lot less graphics this time around. My take on the iPhone Pro was we're getting to the point where you could run an LLM on device and even if it's not Apple intelligence through that API, just the ability to actually distill something that's like GPT4 level and have it be free. Not even an API call that you need. Not even cheap on a per token basis but actually just have something that you can basically drag and drop.

Cost of the energy.

Cost of the energy. And no one thinks about it because they charge on the edge because they charge their phone every night. Anyway, so I was talking to Tyler about this earlier. Try and put into context how powerful an LLM running on device could be with this next iPhone because currently people have done it. I feel like people have distilled models to the point where they can run locally on phones but they're not fast or they're not smart. But it feels like we might be at a point where you get to some sort of touring, touring, test passing. Plus not waiting for Chad is asking.

If it's Nico Rosberg.

It is, it is Nico Rosberg.

It is the former F1 Formula One world champion and now a capital allocator.

Now a capital allocator. So yeah Tyler, what is your take as this gets into people's hands? How solid do you think the actual interactions with LLMs could be on device?

Yeah, so I think, I think it's 12 gigabytes now up from 8. So you can probably run like a very solid 7 billion parameter not super quantized model. We've had those for a long time that are very solid. Obviously they're going to pass Turing tests, whatever. Sure you're probably not going to be using it for doing your hard physics homework, but most things that you want from like a non device model, like helping look at your emails or like doing stuff that like you would think of like like a really good Siri product would do.

Yep.

I think it's like totally capable of Doing that. I mean, it's like. I don't think that Apple intelligence was bad because the hardware was lacking. It was like, like, I mean you can. It'll be a little bit slower probably to run on cloud maybe.

Yeah, I mean my experience with Apple intelligence was it was just instantiated in very odd ways. They didn't really have like a killer app there. And it was like a bunch of different things that kind of piped in and you could highlight text and then just say rewrite this all sort of very incremental gauge botch summaries. Yeah, the summaries are botched, but sometimes that might be the best feature. Actually.

Yeah, it's a feature, not a bug.

We just saw one in our chat where I posted some fake news and Apple summarizes it as like Gabe says, shouldn't be.

Shouldn't Tyler be back in school? Gabe, you must have missed the episode.

But he's taking a gap semester.

He is riding with us.

Yeah.

And so I think it's. I think there's like, there's something magical about getting the ability to inference a GPT4 level model on the edge for free. You can quibble about the level of intelligence that is, but it is actually too cheap to meter because you won't even have to set up an API key and so you can vend in a GPT4 level model and then go viral and not have a bill show up and not need to think about monetization, not need to think about anything that's a drag on your. Just pure virality. You can have a free app that just grows and grows and grows and figure out the monetization later. And I feel like this sets us up for maybe something interesting to happen in consumer. We haven't really had that moment of, you remember the Lensa moment, the magic avatars. You upload your picture and it turns you into Superman or something. And then the Studio Ghibli moment. We haven't had that many viral apps that are purely based on LLM interactions, like text interactions. There's been a couple. Like there was one called Generative Dungeon. Do you remember this? It was like a dungeon crawler text thing. So it would say you walk up to the dragon's lair, what do you want to do? And you could just type like I brandish my sword. And then it would say like the dragon.

That was like GPT2.

Yeah, that was super low level. Right. So imagine that plus but you can do it on device at a reasonable level and you can bring in some Other UI elements from the iPhone, bring in the camera, bring in gps, bring in all the other tools that are available just in the iOS, Swift developer kits, the SDKs. I feel like there will be a ton of people that are just building cool, interesting things on the iPhone that could potentially like go viral and then become real businesses.

Yeah, I don't think this will be that like meaningful for like Apple Intelligence because it's not like, that's like a product issue, it's not like a hardware issue. But yeah, I definitely think for like third party developers this will be like pretty big. Yeah, you can do like, you know, if you make some random game, you can now have like avatars or whatever.

Yeah.

Like if you make the SDK really easy where like it's very easy to inference an LLM. You don't have to like figure out how to like download and then do inference and all that stuff. If they like abstract that away. Yeah, I think it could be interesting.

And I just feel like we've seen so many. There's been so many companies. We talked to some folks at Demo day who are doing this where they were like, we have this cool AI agent that like scrapes your emails and does this. And it all is all predicated on you installing a Mac app, which is a really, really hard thing to do. Whereas, like.

Yeah. How many Mac apps do you think you've installed this year?

Oh, it's.

That's super rare. It's super rare. I got something to put the decks together. I literally have Chrome. That's the one I installed. I'm looking at my home row and it's just Chrome.

Well, and this is why everybody's obsessed with the browser.

Truly, truly. Because if you can get people over there, then you're there. But I have installed apps and I've installed little tiny apps. We were talking about the workout tracking app, Strong that. I saw my friend at the gym, he was using an app to track his workouts. So what app is that? I went and downloaded it. You went and downloaded it. It was very easy for that app to like loosely go viral. And it's a free app, but it doesn't have any like AI. It doesn't really need any AI. But you could imagine that the viral coefficient of someone downloads an app, they have a magical experience with some twist on top of an LLM. Something storytelling, something astrology. Who knows what it'll be? I can't predict it. It's as hard to predict as, as Instagram or Uber. But it's unlocked by Having this ability for free on the iPhone and then people are sharing screenshots of, oh, I had this hilarious interaction or this was incredible or I love this app. They create some viral loop, they do some Nikita beer type stuff and then everyone's downloading the app and then it gets actually really, really big. And I don't think it's gonna be a competition. I don't think it's gonna compete with ChatGPT. I think it's gonna be a completely different thing. And it's very hard to predict, but it feels like, like we might be seeing it. Yeah. And it's going to require some like really crack level of creativity. Like what we saw with Harry Potter, Balenciaga where it's like the technology existed and then someone came up with the great idea to actually go like instantiate it. So if you're going to be mocking that app up, you're going to do it in figma.com. think bigger, build faster. Figma helps design and development teams build great products together. You can get started for free. The Wall Street Journal was also covering the Apple's iPhone air. They had a whole interview, they had really good coverage of the iPhone event. There was a few. There was one thing in the reporting from the Wall Street Journal. I don't have this particular article actually, but they did an article with the Wall Street Journal and Apple revealed that I believe they said the majority of iPhones, 51% or more that are shipping to America are going to be made in India. And I'm not sure if it was this cycle, but it seemed like they were taking the tariff stuff really seriously, moving the actual manufacturing capacity around. Obviously Foxconn, their manufacturing partner, is a Taiwanese company and so they're worried about geopolitical risk as well. And so just another, another data point around Tim Cook. He is not loudly shouting like America, I'm all in re industrialize, but he is making moves.

What was the number he threw out at the dinner last week, wasn't it?

I don't know if he had a number because who was it? It was Zuck had a number. And did Apple have a number two that was 600 billion or something? Apple has thrown out a number at one point. You're completely right. My point is that we've seen a.

Number of thank yous data points.

Oh yeah, the number of thank yous is a lot. But we've seen he's invested in some semi, some rare earth developments in the United States and starting to think about manufacturing internationally. And you know, the pro Bullish bet on Tim Cook is that, like, if he was able to set up this in China and he's the reason he can set it up somewhere else, because he's the guy who did it. And so even though he hasn't been really loud about his moves, because obviously if he says, like, oh, yeah, I'm not going to be manufacturing in China anymore, then his important business partners over there are probably going to be like, well, yeah, we're not going to take your business seriously anymore. There are a ton of different geopolitical considerations.

It's such an iconic line from the dinner. He hooks it. I've always enjoyed dinner and interacting.

Yeah, if you're getting dinner this weekend with one of your boys, just tell.

How to say something that, like, no one can, you know, cancel you for. It's like, I've always enjoyed dinner and interacting. Two things I've enjoyed.

He's a. Dinner and dinner.

He's not saying, I've always enjoyed dinner with you.

Yep.

And interacting with you. He's just saying, I've always enjoyed dinner.

Incredible. Putting on a clay. I feel like for watching some of these, like, I was watching the Tucker Sam Altman interview, and there's a certain level of polish that some of these. Some of these CEOs have when they get to a certain level of media training and public speaking experience, where I think of them more as, like, word rotators than word cells or shape rotators. Like, they can so quickly, like, shift a sentence to be like that. Like, so perfectly balanced, where you're like, wow. Every word in that sentence was calculated so that it rides this perfect line. Like, Sam was going back and forth with Tucker, and Tucker's. I didn't accuse you. I didn't accuse you. And if you run back the transcript, you're like, yeah, technically he didn't. It was not an accusation. But then they're rotating the sentences around to try and pin each other. It's like watching a wrestling match. It's crazy.

Tim Cook said they would be investing $600 billion.

600 billion, yeah. Okay. So Apple says the majority of iPhones shipped to the US Are now assembled in India, meaning they face no tariff. Yet Indian suppliers aren't yet capable of delivering as many of the pro models that US Consumers demand. So most of those still come from China. So this is the majority of iPhones in total because Apple makes a ton of those, but they can't make the leading edge. And then the other thing that everyone always talks about with this, like, assembled in India, is Not the same as fully manufactured in India. Obviously. The advantage to Chinese manufacturing prowess is that you walk across the street and there's a CNC shop with 25,000 CNC machines that can just immediately make you any part and make a million of those parts if you need. Need. So obviously it's going to be a long road for them to set up manufacturing anywhere other than other than China. But it does seem I would do.

Understand Tim's actual strategy here because I don't think he's saying, I'm going to invest. He's not going to be in a huge hurry to invest that 600 billion. Right. He may, you know, he may not actually believe in. It's at all the right decision for Apple. And he's just feeling like he needs to placate the admin and then in 2028 comes around and just can pull back or pull out.

There's also so much to 600 billion of CapEx over what, five years or something like that. That includes HQ. That includes data centers for icloud that need to be local. If you're streaming the F1 movie on Apple TV, you don't want that stream coming from a data center across the world necessarily. It needs to be cached locally. So there are so many different pieces of the business. Building more Apple stores, that counts as capex. If they build them and they're so 600 billion doesn't necessarily mean, yes, we're building the one plant. This is a rainforest of complexity that actually delivers the iPhone. It's not just like, like one iPhone factory, please. It's like a whole network of things. Battery life appears to be a challenge for the air. Apple always says, like all day battery life, but people are like, allegedly. What kind of all day? Are we talking? Are we talking scrolling for 18 hours?

Doom Scroll?

Are we talking proper scroll? Proper doom scroll. Apple, somebody was an accessory that will attach to the air and obviously that's no longer as thin. The company added what it calls a plateau on the back of the air and pro models, adding thickness at the top of the device to fit bigger and hotter chips.

Not a bump.

Plateau to plateau. Yeah. Oh, they're rebranding.

Maybe they're signaling that we're. Maybe this is gartner hype cycle code.

I mean, they have plateaued in productivity for sure. The hype cycle. They've been through it. They're like, first hype cycle. First hype cycle. I've been here, I'm plateauing. I'm getting productive. OpenAI, the nonprofit is Going to control its for profit arm and own equity valued at $100 billion. I believe this will make it the most highly valued nonprofit or like the most profitable nonprofit in world history. The best funded nonprofit. Right. And it is remarkable to think about what that will be like. People have kind of written off. The nonprofit is like it's going away. Like it is not going away. Like it will continue to be incredibly back. It's. It's incredibly back. Funded forever essentially to do a ton of interesting things. I think we're going to see interesting things about come out of that organization. But anyway the news. This is from the Journal and we'll, and we'll break this down, but Nick has it summarized here. OpenAI LLC will be converted into a Delaware public benefit corporation. OpenAI nonprofit retains control of the new Public Benefit Corporation. The nonprofit will hold $100 billion plus equity in the Public Benefit Corporation. The PBC structure enables raising large amounts of capital for the mission, which obviously they're already doing. They're still aligned with ensuring AGI benefits all of humanity. I think that's good also just aligned with like your shareholders. Like if you're just a normal C Corp. I know they're a public benefit corporation, but if you're just normal C Corp like shareholders are humans. Let's make sure it benefits the shareholders. That will by definition benefit humanity. But specifically they're saying all of humanity. Even if you're not a shareholder, AGI should still benefit you, which is a noble cause. OpenAI and Microsoft signed a non binding MoU for the next phase of the partnership. A definitive agreement is still being negotiated. OpenAI says it's engaging.

Remember as it stands, Microsoft gets 20% of OpenAI's revenue.

And for the big man, something big.

Man Satya, something like that. Which it felt like a stretch to me that, that that would be sustainable given the looming costs that OpenAI needs to incur. Obviously they're ramping reven revenue really quickly. Yeah, but you have Broadcom to pay, you have Oracle to pay, you have all these.

But it is revenue, you know, it's not the same as being like crazy. Yeah, yeah, it's an unfixed cost. Exactly.

But still, but still it's a lot of, you know, these companies are under margin pressure already and wasn't the original.

Deal that it would, it would be 20% up until they paid $100 billion or something like that. There was something where like it eventually ran its course. I believe it had an ending and I think that's what's justifying a lot of the underwriting because the lesson from Google, the lesson from Facebook, the lesson from, you know, Microsoft and the rest of the hyperscalers was that you needed your DCF models to be 20 or 30 years into the future.

Right. Yeah.

And so if you just, if you, if you didn't take into account the third decade of growth you were undervaluing.

Originally reported 20% through 2030.

Okay.

Which still is.

I don't think a lot of investors are scared by thinking about value in 2035. I think they're fine with that now I think they're saying like, yeah, I have a 10 year fund and realistically, all the best venture investments, SpaceX is still hanging out in the portfolio. 20 years later, they got continuation funds. I think that a lot of investors are actually fine to say, yeah, this company's going to be like, you know, maybe a financial mess for 10 years, but if I'm super confident it's going to be a money printing machine in 2020 money, I'll do the deal all day. Yeah, I think that's the wrap.

It's still insane to think that the deal ever got done in the first place.

The Microsoft deal.

Yeah. If you were, if, if, if you talk to a founder and they were like, yeah, I was running a fundraise and like, you know, ended up taking this deal, I got the valuation I wanted but ultimately it gave up 20 of revenue off the top too. So not 20 of like profits but like a 20% tax on, on gross revenue.

Yeah.

Just for the, just till 2030. Yeah. And the founder's like, well, it doesn't matter that much because I'm in this for like 20 years.

Very few do not get the pass on that.

Yeah.

Very few businesses, very few businesses in the world, especially like highly competitive, you know, categories can sustain a 20 tax off the top.

Yep.

And still really produce any profits or be functioning at all. Obviously I'll push you profits or what.

If you're, what if your business is mobile, Mobile games in the App Store, you've been effectively paying Apple a 20, 30% tax off the top for.

There's a bunch of examples like Nvidia could do this too.

Right.

And they sort of are through.

I don't know. I mean, I agree with you. Like it is, it is a crazy, crazy deal. Unprecedented in a million ways. But everything about the entire OpenAI story is unprecedented. Unprecedented to start as nonprofit, unprecedented to have. Who are the co founders? Elon Musk, Peter Teal. Like you have like seven different co founders who have bunch of co founders have gone off and start direct competitors. That doesn't happen very often. Usually people are like, I got bags in the category, I'm going to go work on something else. Yeah, there's a bunch of weird stuff.