TBPN Live - GPT-5.2 Reactions, Jacob Elordi vs AI, Disney x OpenAI Deal Breakdown | Diet TBPN

Episode Date: December 13, 2025

Our favorite moments from today's show, in under 30 minutes. TBPN.com is made possible by: Ramp - https://ramp.comFigma - https://figma.comVanta - https://vanta.comLinear - https://linear.appEight Sleep - https://eightsleep.com/tbpnWander - https://wander.com/tbpnPublic - https://public.comAdQuick - https://adquick.comBezel - https://getbezel.com Numeral - https://www.numeralhq.comPolymarket - https://polymarket.comAttio - https://attio.com/tbpnFin - https://fin.ai/tbpnGraphite - https://graphite.devRestream - https://restream.ioProfound - https://tryprofound.comJulius AI - https://julius.aiturbopuffer - https://turbopuffer.comfal - https://fal.aiPrivy - https://privy.ioCognition - https://cognition.aiGemini - https://gemini.google.comFollow TBPN: https://TBPN.comhttps://x.com/tbpnhttps://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235https://www.youtube.com/@TBPNLive

Death Star Post. Now it makes sense. It all makes sense. It all makes sense. It wasn't a vague post. It was just early. A preview, a deal that would take shape four months later, right? That's what was happening. Sam Hallman, of course, what was this? On GPT 5 Day, August 6th, he posted a picture of the Death Star Rising on the horizon. Very confusing post at the time. Now we understand. Now we get it. He was saying, hey, one day, we're going to have rights to Star Wars properties. through this deal with D. Darth, Samma. People did not like this post at the time.

And the timing, the timing was really funny. Yeah. It implied that they were about to release something, you know, so all-powerful. Yeah. It was easy to read it that way. Yeah. You could also read it as like maybe they're going on the offensive against the Death Star, the Empire, which is Google.

They're going to attack Google. There were a number of different ways to read into it, but it was... It felt... Confusing. It was confusing. And then... And then it... like people were somewhat let down.

We had a bunch of different takes on it. We think it makes sense now. But now, it's even more clear with the Disney deal. But you wrote about the Disney deal. You dug into the deal, and you kind of crystallized your take, so take me through it. I wrote 127 days. That's the gap between Sam's to Star Wars Death Star

Vague Post and yesterday's announcement that Disney is investing $1 billion into Open AI and giving them a three-year license that allows users to generate AI, photos, and videos of the most popular 200 Disney characters. Most notably, opening eyes guaranteed a one-year exclusive on the IP. And while the posts in the deal are probably not connected, I think it's funny to imagine that they are. Right before the deal was announced, Disney sent Google a cease-and-assist letter,

claiming Google had been violating Disney's intellectual property by allowing users to generate a variety of Disney characters. Many people outside tech seem shocked that Disney would license their IP in the first place, given their history ruthlessly protecting it. There's a number of examples of Disney coming after kids' birthday parties, or at least that's how they positioned it. Of course, these are businesses effectively using Disney characters to monetize. I read your draft and I was like, wait, they sued a child for having a Disney theme birthday party.

There's two notable examples. One is coming after basically a birthday party service. Company, like a business that sells, we will do your birthday party, and then they were selling, and it'll be a princess theme. Well, so what they, it wouldn't, they wouldn't directly call the characters that you could like rent the same names. It wasn't like Elsa. Okay. But it was clear.

In the reviews, what Disney noted, in the reviews, people would call the characters by the Disney name, so that was one of their proof points. And there was also a story of a father who, like, wanted to do a Spider-Man tombstone for his, like, very young son that passed. And I think it was the, effectively got blocked. It wasn't necessarily directly by Disney. But anyways, Disney has longstanding policy of not letting other people.

Such a PR nightmare. If your legal team, comes to you and says, like, we want to sue a dead person. Like, tell them, like, me back off. It was just like, it wasn't a lawsuit. Or maybe it was a Tombstone Company. They were just saying, like, no, you can't make a Spider-Man. Sure, sure, sure.

Many people outside of tech seem shocked that Disney would license their IP. But Bob Eiger knows that AI generated Disney will happen with or without the company's blessing. So partnering with Open AI today while setting up negotiations with Google and other players makes a lot of sense. Again, I just expect this like cease and desist is the start of a negotiation process in order to get the same type of IP. and capability ultimately into Gemini.

But in 2027, not. Yeah, well, and maybe not, right? Like, one question I have is they have a one-year exclusive. After that period, does Disney say, okay, you can keep having the exclusive, but you need to pay us like a billion dollars a year? And at least Google's at the table, negotiated. Is that what you mean? Exactly.

The strategic significance of the steel for Open AI has been broadly underappreciated, mainly because for a company that frequently talks in the trillions, Disney's investment, a paltry $1 billion just isn't enough to turn heads. But the advantage this deal gives Open AI from a product and distribution standpoint is extremely significant. Open AI has 20 to 30 million paid users out of a total of 900 million globally. Disney, on the other hand, had 140 million people visit Disney parks in the last year and 128 million paid subscribers to Disney Plus. If you're an adult spending thousands of dollars to take your kids to Disneyland or a Disney Plus subscriber, surely you'll pay an incremental amount to open AI to extend and personalize your Disney experience.

As a parent, some of the most magical moments with AI. I've had her simply, generally, I've talked about this before. I'd take a real photo of me and my son, and I turn us in generic design source. Dinosaur mode. It really brings us both a lot of joy. It's like, it's his reaction to the picture that I'm getting joy from because he just like absolutely loves it. I had a similar experience where I would, I would like take a picture of like, you know, we'd create a scene of toys and then we'd say like make these Lego versions of this.

this or something, but the models got so good at a certain point that I was just like, okay, like I just, I made his toys look like toys and it just looks exactly like what I made. I just take a photo. I need like more creativity to come up with something remarkable. My household is not into the Disney world yet. Yeah. Kids are just young. Yeah. But I can imagine how excited if you're a frozen super fan, an event you're a super fan, being able to put yourself, your kids into their world, I think is going to be pretty exciting. So having a license to generate high quality images and videos of 200 of these characters creates a temporary but very real differentiation for tons of current and future chat GPT users.

Knowing Sam and Josh, they've likely negotiated to get SORA chat GBT, brought exposure across Disney properties, both online and offline. And we already know that select SORA videos will be featured in Disney Plus. I think they're going to be taking really making sure their hand selecting those because... I wonder what that means. I wonder if that means the Apple TV app or the Disney Plus app. on the phone. On an iPad, on a phone, it feels a little bit more natural. Just in terms of the Star Wars property, like there's, there are the original Star Wars films, and then there's Young Jedi Adventures, which is Star Wars for kids. There's even less violence. It's more

G-rated than a PG-13 rated. Even with Bluey, a very popular children's show on Disney, on Disney Plus, there are full episodes that are like a full episode might be, I want to say like eight minutes long, but then there's minisodes that are even shorter. So like the short formification is happening within the Disney app already, but I think a lot of parents would have a sort of negative reaction to just a feed. A feed. Yeah, an endless scrolling feed of short form, SORA content. That's going to be really tricky. It needs a twist. It needs curation. They said that there's already going to be curation, but I wonder how that's going to roll out. Curation, but it could still very well be infinite. You could essentially have like a film festival that's running and kids submit

their generations and then those get highlighted essentially. A film festival of seven second videos. I'm sure Hollywood will love that. Back in the Vine days, like the creative skill ceiling was incredible. People were doing remarkable, remarkable things with that. I still believe that there's some really interesting things that you can do with generative systems, with generative AI. And I imagine that we will see some cool stuff come out of the Disney partnership. But I don't know that I'd want a kid sitting there watching endless slop. I finish it out by saying my long-held assumption is that over time, everyday consumers won't pay for LLMs.

There will be great ad and commerce supported LMs that provide the monetization to serve great models. But OpenAI's challenge is that even with future Hall of Famers like Fiji Simo and Denise dresser on board, Denise is, of course, the former CEO of Slack, who's now a CRO at OpenAI. They can't massively scale monetization of the free product overnight. can't just say, hey, chatGBT agent, create a massive ad platform. Don't make mistakes. I don't expect those business lines to really begin ramping until the second half of next year and they need

to show continued revenue growth now. Hard to think of a better time for this deal to get done than 11 days into OpenAI's code read and right as Sora was about to fall out of the app stores top 25. Will the functionality be immediately abused? Yes. Will it create uncomfortable moments for Disney's leadership? Yes. Is it a smart move for Disney? Yes. Will Open AI have to shell out billions to maintain exclusive access over the long run. Yes. I want to debate you on this idea that it will be abused immediately. I think it's actually really, really hard to abuse these systems. Oh, jailbreak it. I got it to say something crazy. Like that's like, it's just like seems like a thing of the past. What do you think about? I just trust that the internet immediately

will figure out if you have hundreds and millions of people trying to jail break it, they will pull it off. Of course. But the other thing that will happen is people using different models that are like already don't have the same guard rails and making it look like their sore outputs even though they're insane. Oh, totally, totally, totally. Again, I'm just saying like open source model that's completely unrestricted, you do something crazy and then you throw the sore watermark on it just to take a shot at Sam and Bob and Sam. What do you think, Tyler? Yeah, I was going to agree with Shorty. Like, I think, hey, what's this? What's this? Oh, yeah, I've just been, you know, I got a double jaw surgery. And a little bone smashing. I've been bone smashing a bit, yeah, just trying to get my looks

You're looking fantastic. This is remarkable. I really I really can't. Hats off to the production team. This is a, this is a fantastic new feature to allow Tyler to look like the absolute gigacad. Look like how I do every day. Yes, how you do every day. I was going to say like, yeah, day one, there's going to be people on X that like easily jail break it. I can't. I can't take you seriously. Just look at us. And then and then they're just going to post it and then it's going to go super viral. And then you have parents. Chad ID. No, it says Chad podcast. Tell me your actual take. Okay, yeah. So, I think like day one, oh, what, what happened? There we go.

There we go. Okay, day one, you have people who jail break it, even if it's like one person, then it goes super viral next, and then it's like, oh, Mickey Mouse, like shooting a gun, and then it goes viral, parents see it, and they're like, okay, my kids aren't going to be on this, because, even if it's like, all, Mostly, like you're going to have a ton of backlash where it goes super viral on CBS News or whatever.

No, I don't buy that. I think it will be very magical. And I think that they'll immediately be getting shared in family group chats between friends. And I think we're maybe underestimate the adult Disney audience as well. I could go two ways. Okay. Easy, John. Where do you think the line is?

So you think that they would, whenever this rolls out, maybe it's already out. It might be out soon, but let's just assume it rolls out January, you go to Sora, you can generate, you know, Luke Skywalker fighting with Spider-Man, because those are two Disney properties. Can you have Luke Skywalker cut Spider-Man's arm off? Because Luke Skywalker does cut off people's arms in the PG-13 rated Star Wars, but Spider-Man does not ever get cut, he never gets his arm cut off because he's more of a PG-character. And I'm just wondering, like, where is the line for the Disney execs? Like, where would they draw the line? How do they even think about a framework for upholding the brand? I think we'll know it when we see it.

It would be a funny scenario where they only allow violence on Warner Brothers characters. But they don't have the rights, but it's just like, oh, yeah, like you can, well, they do that in Hollywood. There's some actors who have it in their contracts that they never will lose a fistfight, basically. They'll say, like, you know, you can cast. me in this movie, but if I'm fighting up as somebody, I don't lose, we can tie, basically. So you know how you watch a lot of, like, Jason Statham and The Rock and, like, Fast and Furious, and they're, like, bashing each other through walls for, like, five minutes, like, this crazy action scene. And then at the end, you're like, oh, and like, neither of them won. It's like,

that's because it's in the contract. They don't want to be, they don't want to be depicted as, like, losers on the order farmed. They don't want to be aura farmed. Exactly. I got roasted by the anti-Aig eye crowd for suggesting that IP holders might want their characters in Sora. They insisted. that real entertainment companies would never willingly allow their characters to be used in AI slop two months later. How the times have changed.

I completely agree with this take. I think it is interesting how there is just the reality that if you hold IP, you need to exist in the world, you need to exist a lot. You can't just stay on the shelf. You can't be the Humphrey Bogart. And yesterday when Dylan from Puck was here, he was mentioning a bunch of old Hollywood references

And I was laughing to myself because I know you have not seen Casablanca and you have no idea what he means by get on the plane or Rosebud or any of these references. Has anyone else noticed V-O-3 has no IP constraints? Prompt Mickey Mouse welcoming you to Disney. Look at this, John. This was going on back June 16th. What's crazy is that this somehow looks this, this looked so good at the time. And I was like, oh, it's over. It's so over. And yet this now, looking back, this looks. washed out. It doesn't have, it's like overly saturated or something. It looks like a green screen. Yeah. It doesn't actually look that good to me anymore. And it's because the goalposts have moved as they always do. Do you want to hear, do you want to hear my full movie? Okay. This is the full list you can pick

and Tyler, you can play along too. First up, the Godfather. Have you seen it? No. Wow. Okay. Taxi driver. Never heard of it. You've seen Star Wars. At least a couple. Raging Bull. No. Scarface. Yes.

You've actually seen Scarface? Yes. Full metal jacket. But I didn't study it. I watched it, but I didn't take notes. Full metal jacket. No. Die hard.

Maybe like 10 years ago. You got to see Die Hard. Good fellas. No. You got to watch Good fellas. This is so good. Reservoir dogs. Maybe.

Pulp Fiction. Seen it. Okay. Heat. Heard of it. Haven't seen it. Casino. Never heard of it.

Braveheart. I heard of it, I haven't seen it. You've never seen Braveheart. Wow. Okay. Hackers? No. Apollo 13.

No. Contact. No. The fifth element? No. Saving Private Ryan? No. You haven't seen saving Private Ryan.

Wow. I have seen John Wick, funny story. I was on a flight back from London. Yeah. And it was like effectively 3 a.m. whole plane's dark and Keanu Reeves just walks by my seat wait in the bachelor plane yes that's wild just walks by and it was like wait and you were watching John Wick no I was okay yeah but he walks by and imagine seeing like Keanu like you're just you're on an

overnight flight it's dark in the plane he just like walks by and he was using the restroom and I did I ended up watching it after that yeah to be to be fully immersed more importantly Hero 1,000 presence says 5.2. How about 5.2 cold ones with the boys? It's a beautiful Thursday night, folks. This latest model is state-of-the-art on beer bench, which is if I crack it open, it makes a fizzy sound, and I go, hell yeah, very good. GPD 5.2 also knows exactly which are the best Paul McCartney songs, and it can write a poem in Spanish, as good as the median Pablo Neruda poem. Tyler Cowan loves the GPT models. He's obsessed.

The most obsessed, and yet I've never seen somebody make the claim that he was one shot. That's true. That's true. He's pretty elite category. He's using responsibly. So it does feel like opening eyes a little bit of the coming out of a trough of disillusionment potentially. You know, the vibes have been really bad with the $1.4 trillion, but the global economy has not collapsed. And the market is rallying, and there were, in fact, not an Enron scenario.

None of the big tech companies blew up. Like, we've moved on. Not maybe it happens, but in general, it seems like, you know, expectations have sort of reset. And now we're going into 2026. It's kind of a new game. Open AI still has a really dominant consumer business. And what looks like it's shaping up to be an oligopoly in the enterprise side or the B2B side. Nothing is a five alarm fire.

They're in the code. but it feels like they will emerge stronger from it. The big question is, how will Sam Altman be perceived? Will he be one of the greatest founders of all time, founder CEOs? And I was trying to benchmark him. Let's say that he, you know, lands the plane, gets out of the code red, Baja blasts his way

into the public markets, and he winds up, you know, being this founder CEO of a multi-trillion dollar consumer tech company. He's in the mag seven. It's the mag eight. How will he be remembered for most CEO, founder, CEOs, if they get the company out into the public markets at a trillion dollars, something like that, they typically own a ton. Steve Jobs is notably, at various times, not a major shareholder. At one point, he owned five and a half million shares out of almost a billion shares. So he had half a percent or 0.6 percent in 2011. Bill Gates, on the other hand, absolute size load.

13.7% ownership of Microsoft. Jeff Bezos is at 14% with Amazon. Jensen is at 3.7% of Nvidia. Elon Musk, almost 20% at Tesla, 19.7%. Meta Marks at 13.5%. Given how much dilution Open AI has had to take in all the shareholders. starting as a nonprofit. There's, well, that, but there's a world where if the company had just

been formed as a C-Corp back in the day in 2015, and they had just raised a bunch of rounds back-to-back, and there was a bunch of co-founders initially, would Sam actually have, it's rumored what he's going to get something like 7% was a proposal? Sure. Would he have, would he have less than that in that scenario? I don't think, I don't think in the fullness as well I end up working out that badly for Sam. The jump to reasoning models will be studied for years. They are still wildly underrated ARC AGI 1 has been out for six years. GPT 5.2 is a five order of magnitude scale up and yet it still lands at 12 percent. Add a bit of reasoning and performance immediately jumps. And so you can see the ARC AGI 1 leaderboard. GPT 5.2 is only a 12

percent but when you get to low, medium, high, extra high over in the reasoning models or you switch to the pro models, you're getting up to 90% on ARCGI. Of course, the test that any human can do, and yet AI has struggled with for a long time. Not for long. AI is starting to make headroom or make headway on the ARC AGI leaderboard. GROC4, of course, is also doing well. I'm excited to see what happens with ARC AGIV3. It feels like we're not even showing those scores yet because I think that none of the models are scoring at all yet. They're not even finishing, and I think it's an optimization problem to get the least number of moves.

Oracle is a train wreck. How did they not disclose these delays on the call two days ago? And Bloomberg is reporting that Oracle delays some data center project for opening eye to 2028. Oracle has pushed back to completion dates for some of the data centers that's developing for opening eye to 2028 from 2027. Delays are largely due to labor and material short said the people asking not to be identified.

Oracle has been working to deliver over on a $300 billion contract to supply the computing power necessary to train and run opening eyes models since it was inked this summer. Even with the delays, the timelines for the project in the U.S. remain ambitious for sites that are set to become some of the largest in the world. Oracle was basically saying, like, we're going to build AWS in two years.

And that just felt like, you know, applaud the sort of ambition. But I think some of these delays, labor shortages, et cetera, are, were predictable. Bloomberg says our take on Oracle's massive bet on AI. Oracle's AI bet has fast become a bubble barometer. A little bub talk. Oracle spokesperson said in a statement that the company remained confident in its ability to meet its obligations and future expansion plans.

There have been no delays to any sites required to meet our contractual commitments and all milestones remain on track. Meanwhile, Apollo, according to Compound 248, Casually predicting zero returns to the S&P 500 over the coming decade. The historical relationship between the S&P 500 forward PE ratio and the subsequent 10-year annualized returns show that investors should expect to get zero return in the S&P 500 over the coming decade. They have a chart below.

They're like, hey, can we interest you in some private credit opportunities? Or you know the other option. What's the other thing you want to invest in when the... stock market isn't looking so good what land I've talked about this land I think is the most criminally underrated yeah somebody wants to make steady returns land maxing yeah they want to they want to hit the thousand X exactly meme coin they want to hit the the 1% polymarket whatever it is timeline has been in turmoil over Klein some posts are getting deleted so I don't even know if we're

gonna be able to pull these up but here's what happened yeah bring it down Somebody at Klein commented on a picture from a hackathon and said, imagine the smell. Yes. People thought he was making, they thought it was like a racist comment. Yes. People started piling on. Yes. The CEO initially defended him and said, I'm like, I know I can vouch for this guy.

Yeah. He's fine. People kept piling on. The CEO ended up turning around and said, actually, we've let this guy go. Yeah. And then now the CEO is getting even more, more hate on, from letting the guy go. And then now he's deleting his post. And so, and so maybe the guy, maybe he ends up rehiring.

It feels like damage is pretty done at this point. It'd be hard to imagine him saying, like, going back and saying, like, I'm actually going to rejoin the team that you have a brief screenshot saved a little bit. I think I can, I think I can read it, some of it. This is the CEO of Klein. It says a few days ago, a tweet from our head of AI offended a lot of people. And it definitely offended a lot of people. There's an article in the Hindustan Times in India.

This is like newsworthy in India. Well, I don't believe his original tweet was intended to be offensive. His response refusing to apologize does not reflect my position or Klein's. We recognize this caused real pain and that deserves blah, blah, blah. Lulu says this stinks, she's not a fan of it. Yeah, she says re calms mistakes. Everyone makes them sometimes things fall flat. It's happened to me, and it also happens to people much smarter.

But betraying people or principles is in a different category. It's more preventable and less excusable. Cowardist doesn't happen by accident. In the case of Smellgate, it could be argued that Posh made the first kind of mistake overlooking how the joke could be interpreted. But then the CEO inarguably made the second kind of mistake, and that one to me is much worse. Nat Eliasson posted, a friend of the show,

I'm not sure if he's in the chat today, said, considering the unanimous reaction to the CEO's post saying that they let go of the engineer, Nat says, will he fire himself now? Only seems fair, certainly. The entire company gets fired? Fireception. Just everyone gets fired?

No. The funny thing is that the only reason I know about this joke is from a bunch of Indian friends who seem to have, like, been making it on anonymous accounts years ago in some sort of funny attempt to, like, reclaim the formerly racist joke, but... Lulu says, when you cave to the mob, instead of standing by your employee, and now people are more mad at you than before. And she quotes,

your worst sin is that you have destroyed and betrayed yourself for nothing. Trump in an interview yesterday said that he would consider eliminating taxes on gambling-winning. This is so crazy. Franks has predicted. markets be like, never mind, it is gambling. It is gambling. Because they always use trade. They always use, like, it's not bat.

You trade on future. Trade on sports, trading on this. This just feels like a whole series of people. It's like Trump likes being edgy and funny making jokes, and he says, I'd consider it, and then watcher guru takes it and turns it into a viral thing, and then somebody dunks on it, and they get viral points, and it all feels deeply unsurious.

and not really, like, it's going to reshape the market anytime soon. Yeah, I think a lot of people would hope that he'd be like, no, we're not going to create tax-incentivized gambling in this country. But he didn't. Stranger things have happened. Stranger things have happened. Scoop Open AI CEO, Sam Altman gathered newsleaders for lunch in New York City on Monday. He told them Enterprise AI will be a massive priority for Open AI in 26.

And Greg Brockman says, yeah, Enterprise AI is going to be huge theme for 2026. What's... And this tracks with hiring Slack CEO as CRO, right? Totally, yeah. Yeah, the question is, what, what does this actually look like, right? I think a lot of people are using chat GPT at work. They're using a variety of LLMs. There was another-

No, and after Harvey, that's the big question. Harvey seems to be doing really well. And I think of law firms as enterprises, and I think of law firms as, potential targets for an Open AI enterprise plan. Am I crazy? Yeah, I don't think you're crazy. Last quarter, I rolled out Microsoft co-pilot with 4,000 employees, $30 per seat per month,

$1.4 million annually. I call it digital transformation. The board love that phrase. They approved it in 11 minutes. No one asked what it would actually do, including me. I told everyone it would 10x productivity. That's not a real number, but it sounds like one. HR asked how we'd measure the 10X.

I'd said we'd leverage analytics dashboards. They stopped asking. Three months later, I checked the usage reports. 47 people had opened it. Twelve had used it more than once. One of them was me. I used it to summarize an email I could have read in 30 seconds. It took 45 seconds.

Plus the time it took the fix, the hallucinations. But I called it a pilot success. Success means a pilot didn't visibly fail. The CFO asked about ROI. I showed him a graph. The graph went up and to the right. It measured AI enablement. I made that metric up.

nodded approvingly. We're AI enabled now. I don't know what that means, but it's in our investor deck. A senior developer asked why we didn't use Cloud or ChatGBT. I said we needed enterprise-grade security. He asked what that meant. I said compliance. He asked which compliance. I said, all of them. He looked skeptical. I scheduled him for a career development conversation. He stopped asking questions. Microsoft sent a case study team. They wanted us to feature us as a success story. I told them we saved 40,000 hours. I calculated that number by multiplying employees by a number I made up. They didn't verify it.

They never do. Now we're on Microsoft's website. Global Enterprise achieves 40,000 hours of productivity gains with co-pilot. The CEO shared it on LinkedIn. He got 3,000 likes. He's never used co-pilot. None of the executives have. We have an exemption.

Strategic focus requires minimal digital distraction. I wrote that policy. The licenses renew next month. I'm requesting an expansion. 5,000 more seats. We haven't used the first 4,000. But this time we'll drive adoption. Adoption means mandatory training.

Training means a 45-minute webinar, no one watches. But completion must be tracked. Completion is a metric. Metrics go in dashboards. Dashboards go in board presentations. Board presentations get me promoted. I'll be SVP by Q3. I still don't know what co-pilot does, but I know what it's for.

It's for showing we're investing in AI. Investment means spending. Spending means commitment. Commitment means we're serious about the future. The future is whatever I say it is, as long as the graph goes up to the right. Jacob Allured, he says he has no tolerance for AI and finds the whole conversation around it so effing boring. He would not like this show. He says, I would much rather kiss on the beach and read a novel and be sunbird. Most pick me moment of the year. Ninety-one thousand people

agree. Technical deep learning tutorials says, yo, Jacob, come on my live talk. tomorrow, we're going to be review this paper. It's almost as good as the stuff you mentioned. See you tomorrow, buddy. And it's a surprising effect of negative reinforcement in LLM reasoning. Suck says, God blessed us again. Shocker, a thousand-year American Empire.

Major deposit of rare earth and critical minerals has been uncovered in Utah, which the Wall Street Journal calls the most significant critical mineral reserve in the US. We did it again. We did it again. If you're a nation without a lot of critical minerals, It's probably a skill issue.

You thought this week was Christmas week. We're just getting started. We hope you have an incredible weekend. Thank you for being with us this week. We'll see you soon. Goodbye.