TBPN - Amazon Bets $10B on OpenAI, Ford’s Reality Check, Paramount Deal Unravels | Diet TBPN

Episode Date: December 18, 2025

Diet TBPN delivers the best of today’s TBPN episode in under 30 minutes. TBPN is a live tech talk show hosted by John Coogan and Jordi Hays, streaming weekdays 11–2 PT on X and YouTube, with each episode posted to podcast platforms right after. Described by The New York Times as “Silicon Valley’s newest obsession,” the show has recently featured Mark Zuckerberg, Sam Altman, Mark Cuban, and Satya Nadella.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://www.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

SpaceX is potentially going out, going to hoover up 30 billion of capital. They're going. They're going out at $1.5 trillion. They're talking to bankers now. They're going to hoovering. The public market's going to be tapped out, right? Well, Sam Allman would like a word with Andy Jassy. And he says, I need $10 billion.

And Andy says, sure, as long as you buy a bunch of tranium chips, that's basically the story. Closing out the story with the Ford F-150. Of course, this broke earlier this week. CEO Ford did a round of press interviews talking about the news, which is that Ford, the historic automaker is killing the F-150 lightning, their electric truck. Sales fell 72% year-over-year. That is a 72% decrease specifically in last month, which is post-EV tax credit going away. I mean, the first question that I was sort of toying with that we've been debating is, did truck buyers ever really want to go electric?

Was that ever a good idea? Because it always seemed like who's the last person that's going to buy an electric car, the truck buyer, right? So one thing that I was thinking about is I feel like the cyber truck probably got truck buyers to like traditional truck buyers to go electric. But it wasn't because it was electric. It was because it looked electric.

Yeah, it looks crazy. It looked wild. I completely agree with this. Yeah, there's this weird thing where like the F-150 silhouette is iconic, but you sort of... I forgot I had my elf ears on. Ballpark, how many ads, how many billions of dollars have been spent on ads that associate trucks with like...

Being a cool dude. Guy, dude. You know, driving through the mud. And a big part of that is the engine note. And a big part of that is the actual exhaust coming out of the back. Like, all that advertising worked on me. I grew up in a Toyota family. We only had Toyota's growing up.

And we at one point had two Priuses, right? Yeah. But as soon as I was an adult and I could afford it, I bought a Ford Raptor. It was black on black on black. It was lifted. I just wanted the truck that I was advertised to me as a kid. Yeah. Yeah, yeah.

There was an interesting data that Ford was sharing that they were framing as positive when the F-150 Lightning launched. But I think in retrospect might have actually been sort of a canary in the coal mine Totally. So the first stat was that of the people that reserved the F-150 Lightning, 50% had never owned a truck before. And then 75% of the reservation holders had never owned a Ford before. And so Ford was celebrating this.

It's like we did it. We did it. New Hero product. It's going to bring new people into the Ford ecosystem. It's going to bring new people into the truck ecosystem. We are expanding the market. And in hindsight, what it feels like is the truck buyers didn't want it. The board buyers didn't want it.

And they're the two biggest markets. And so, yes, there were a class of people that were like, oh, I would always, I've always, an electric truck. That sounds really interesting. I love the idea of a 220 volt. I'm a unique purchaser. And they're like, this is a niche product. And they go hard for the niche product.

They show up immediately. And they'll do it no matter what. And you wrote in the newsletter, the first electric truck was the Rivian. Yeah. And that had only launched a few months. Six months earlier. Yeah. So the Rivian came out in September of 2021.

That's the R1T. Ford shipped in April of 2022. That's actually very impressive to me. I was very impressed with how fast Ford was able to respond to the idea of electric trucks happening. This feels like they were like, no, we're moving in the first wave. They did successfully. A lot of that's because they built off of the F-150 platform. They were able to reuse a lot of equipment there and the supply.

chain, but ultimately they didn't ship a product that delivered at the level of the R1T. I was thinking the Rivian name. Do you like the Rivian name? I think it's fine. It sticks out to me. It was also weird when it first heard it. I'm neutral on it. It was weird when it first stuck out.

It sounded like something that came from like a brainstorming session at a pharmaceutical company, you know, because it's like this weird. Like what does the name actually mean? I guess it means river and Indian kind of portmanteau. The name itself is a blend of. syllables from the river symbolizing adventure and connection to nature. Sure. I always looked at Rivian as something like the whole foods of cars, right?

Like the REI of cars, right? It's like people go to REI. Sure. Like the average person going into REI is not necessarily like buying gear for the most rugged adventure. Yep. Like they might be buying gear for their backyard. Or like going on a hike that weekend.

And again, like I feel like the Rivian cars again, I mean we had the CEO on but like have that range where it's like it really is just like a good daily yeah but they've built it it's super powerful it's very capable ford's plan is to pivot so they're going to be pivoting to hybrid trucks and hybrid designs but what's interesting is that it's one of those uh it's this extra long range hybrid where you have an electric power train that is charged by a gas motor and so you can get like 700 miles of range here you're Here's a question. Here's something that people don't like about Rivians.

They don't have car play. Is that a deal breaker for you? I think it's solid. The thing that I find annoying is the fact that the cars that I've owned are all defaulting back to the actual operating system. Oh, really? Wait, what do you mean they're defaulting? So I have two Mercedes. They have like the regular Mercedes operating system.

And then like Apple. Car play is layered on top. But I still find myself like turning on the car sometimes. and it's the stock system. I'm like, so I just wish there was a single operating. I wish I'd. So Apple's trying to do this because I've noticed this in my car. But the manufacturers are like, well, we sell the Android

and we don't want you to control us forever. Amazon is in talks to invest over $10 billion in OpenAI. Yeah, the valuation would be higher than $500 billion. The Amazon investment would help opening I afford some of the commitments it has made some to rent servers from cloud providers, including from AWS. Yeah, this is like, it's like there's somewhat, there's some circularity, but it's not entirely fully beating the circular allegations on this one. Open AI last month announced it would spend $38 billion renting servers from AWS over the next seven years,

making AWS one of at least five cloud providers opening I use uses to develop AI. The deal also could help Amazon find a new customer for its trainium AI chips, which compete with the Nvidia chips. This is kind of like a rebate. You know, it's like they said, hey, we're going to buy 40. and they said, here, take 10 back. And we'll take a piece. Honestly, the more notable news here is that Amazon and OpenAI have discussed commerce partnership opportunities.

That's very interesting. Open AI wants to turn ChatT into a shopping hub and has discussed earning fees for referring customers to retailers. It isn't clear whether Amazon opening I deal would involve any arrangement related to such features in Chad CPT or AI powered shopping features that Amazon is developing. So I just look at this in the same way as like the Disney deal, which is like, hey, we're going to invest, but we're going to give you access to this thing. And I would expect that, I mean, you can imagine opening I has been working on getting referrals to from, from basically getting a revenue stream from referring products out to Amazon for a really long time, right?

They've done the Etsy deal. They're doing deals with Shopify. They have not done eBay notably and they have not done Amazon notably. And I think there's been some, there's just been some general hesitance to let, again, let the Fox into the hen house, right? Because you can think about it, like the Google search experience is like, or sorry, searching for products on Amazon is extremely profitable for Amazon, right? If consumers start just going to chat CBT to find products on Amazon, that like Amazon needs to be really careful around that because, yes, they can get a referral fee.

Or they'll, they're getting a customer, but then they Amazon, or sorry, opening, opening eye wants them to pay them for that customer. and that's a customer that didn't just go look at a bunch of ads, right? And I do not like searching for products on Amazon because the experience is I'm just trying to find, I always use the example, like paper towels, right? And it's like, it's so frustrating to search on there because I just want to buy, like, I'll spend $20 to $30 on this thing.

And then it's like three pages of like $6 versions of the product that I know are going to be terrible and a bunch of ads for those things, right? And so being able to go into chat GBT and just say like, hey, I want to buy this item from a manufacturer or a brand that has been in business for more than 30 years, like pre-ecommerce. Yeah. I want a brand that has just been making this thing well for a really long time. And so I would be defaulting to the LLM and skipping Amazon entirely. Yeah, you want to fight.

You want to fight to be the aggregator. You want to, like, I guarantee that although Amazon shows up on Google search results, like, like, you know, they want people to open the app and search in the app and be the main starting point for their commerce, their entire commerce journey. We've seen this with Shopify as well. Shopify obviously would love for the commerce journey to not start on Facebook or meta properties. Instead start in the shop app. They're working towards that.

The same thing is true of Amazon. And every aggregator is acutely aware of aggregation theory and acutely aware that they should not let someone come around and an aggregate on top of them. Apparently projecting 60 billion of advertising revenue. Yeah. Just growing way faster than the core retail business. They've sort of like... The core retail business is probably growing at the rate of overall e-commerce penetration,

whereas this is just like extremely high margin, fast growing, and they want to protect that. And probably bigger than what they could make off of a referral fee on top deeper in the stack, if they're deeper down. Well, the other side of the Amazon OpenAI deal is that the deal could also help Amazon find a new customer for its Traneum AI server chips, which compete with NVIDIA AI chips that OpenAI primarily uses today. As part of the deal being discussed, OpenAI plans to use Traneum chips, two of the people

said the cloud deal Amazon announced with OpenAI last month only made mention of service powered by NVIDIA. So the interesting thing here is what will they be doing with those Traneum chips? Will they have a specific model that runs on Trinium? Will they set up some sort of abstraction layer that they can run any of their models on any Hardware or any ASIC basically like like will you see or will it be like okay, we still have GPT 4-0 workloads let's Recompile 4-0 for for for Traneum and let it just chill there and Traneum is our is our pool for

for four-o or you know what? Traneum is going to be our workhorse for image gen or video gen and let's do our image gen optimized for that particular stack. The Wall Street Journal highlighted real-time video as an interesting place where Traneum could potentially outperform. They weren't making the case, at least to the journal, that Traynium is what you want if you're going to do the biggest and most massive training run. That was sort of the narrative that the TPU was pitching with the latest anthropic like runs, but they did highlight you know real-time video video generation and so I'm what I'm interested in is that is is trait does Trainium get abstracted to a point where

it's sort of like model agnostic or is open AI like the the chat GPT the app has a whole host of models because these models are now mixtures of models and there's model routers and there's different products, video, audio, image, you know, deep research. Is one of those going to be on Traynium or will Trinium be a like a liquid pool of compute that cuts across the entire stack? Do you have any, you know, instinct on this? Yeah, I mean, I think the abstraction thing is pretty hard, right? Because you always hear about TPUs and how the TPU team and like the Gemini team are so closely integrated, right? Like every, all the model architectures, like, inner-length with the GPU,

so I think it's hard to actually abstract all the way up. But it's interesting. I mean, Anthropic has been like multi-platform platform for a while now. So I'm curious how they think about this stuff. Something that's interesting, if I search on Gemini for a product on Amazon, find me the best blank on Amazon, it takes me, it says top recommendations on Amazon, and then I click the link, and it takes me to a Google search for that, product that is a sponsored result on Amazon.

Then I click- Who's paying? So Amazon is paying, Amazon is paying Google to appear in search results. AdWords. OK. And AdWords. And then Gemini is routing basically to AdWords to get the click through there. So there's no direct integration at all. Google has so many odd advantages.

It's crazy. Like the fact that the Google bot just sees so much more of the internet feels extremely important. And yet, I just don't know if it will be enough to win in consumer in some meaningful way. Does it mean 50% of the value of consumer? Does it mean that they can win, come from behind, defeat OpenAI, chat GPT? Feels so important. And yet, it also feels extremely hard to actually pass that message through. The Amazon investment would help OpenAI afford commitments, including from AWS. That's very funny. Um, well, liquidity is shown the gang standing around. All right, Jeff, you're up next to invest in Open AI.

Amazon, $10 billion investment in Open AI in the form of AWS credits. You got Sachs Nadella. I couldn't find where... 2% for Amazon, maybe less if it's at above a $500 billion valuation. So Andy Jassy's getting 1.5% of Open AI. Satch is sitting there with over 20. He's pretty happy. Pretty happy. Looking at the screen. Jared Kushner is pulling out of the Paramount bid hours after his father-in-law took aim at the Ellison clan apparently. The latest news in the Paramount bid for Warner Brothers, the story that just keeps on giving is in back-to-back Salvos Tuesday, the president and his former or his family distanced themselves from Paramount's hostile bid for Warner Brothers discovery. I think we know what's going up. on there. It's about Foghorn Leghorn. It's about Tweety Bird. It's a rebuke to owner David Ellison's attempt to leverage relationships with the White House to close the $108 billion takeover effort. President Trump Tuesday afternoon said he had been treated far worse by the Ellison-owned CBS since

the family closed a deal for CBS parent paramount. Which is so interesting because I've seen a bunch of people have been riled up about Barry Weiss running CBS. The reason that you maybe would say that she's doing an effective job as a manager of that asset is because people are talking about CBS content in a way that I have not seen ever. Do you ever remember, like maybe a couple times here you'd see something? And she's, she's clipping CBS content. It's like she's doing stories.

It feels like it's working. No, no shade to the people that were writing CBS before, but like what content was on that? Yeah, we just don't know what they were doing before. It's like it didn't exist. And now it exists. And you can like it or you hate it, depending on your political persuasion. But you can't deny it. And I always looked at this as like, it's a thing.

The Ellison's were like, hey, we can get a truth engine. Yeah, I mean, there's definitely like the brand is still great. Like CBS feels like a solid news source. So I agree with that. But the distribution was so far behind that people weren't talking about what was going on there. And I would say that the reason, the reason. The reason, one way to think about the value of CBS is what would it cost and how long would it take to recreate a brand like CBS? Probably cost, it would take you decades.

I don't think you can buy it. I actually don't think I think you could be Sam Altman and Marshall a $50 billion fundraise around. Yeah, you can't just snap your fingers and get it. And it would still take 50 years to get there. If you get $50 billion, what do you have to do? You have to go buy the legacy IP because there's only, you can't just. You can't just, you can't snap your fingers and create a brand overnight. Like, it just takes time.

So Warner Brothers sent a letter to shareholders this morning basically saying that they're riding, they want to, the board of directors still wants to go with Netflix. They believe it's superior in a number of different ways. One thing that stood out to me is that Paramount has consistently, they said Paramount has consistently led WBD shareholders that its proposed transaction has a full, backstop from the Ellison family. It does not and never has. Paramount's most recent proposal includes a $40 billion equity commitment for which there is no Ellison family commitment of any kind. Instead, they propose that you rely on an unknown and opaque revocable trust for the certainty

of the crucial deal funding, despite having been told repeatedly by WBD how important a full and unconditional financing commitment from the Ellison family was. And despite their own ample resources, as well as multiple assurances from Paramount Skydance during our strategic review process that such a commitment was forthcoming, the Allison family has chosen not to backstop the Paramount Skydance offer. And a revocable trust is no replacement for a secured commitment by a controlling stockholder. The assets and liability of the trust are not to publicly disclose and are subject to change. So they're basically like have this entity being like, yeah, we're guaranteeing it, but it's not actually them saying, like, it's.

you know, they could move assets out of that trust. Yeah, yeah, got it. So strength, the offer, not as strong, potentially as Netflix. You know Netflix is good for it. It's a huge company. They've already signed a deal with a massive termination clause, and I believe they've raised debt for this. Like, they're ready to rock.

So burden hand is worth not too in the bush. Yeah, the other thing is Paramount has not offered to reimburse the breakup, the termination fee. It's a $2.8 billion fee. Yeah. There's also financing costs that they're going to have to take that Warner Brothers would have to take on if they don't, you know, complete the debt exchange. So yeah, yeah, at the end of the day, what are the, what are the Ellisons do at this point, right? They've been, they've been doing deals, right? They've got CBS now. They've got the UFC. They're trying to build this streaming platform. Again, going back to some of the conversations that we've had, like this entire, the entire strategy to date has been predicated on getting this Warner Brothers. asset. Yeah, yeah, and it seems like it might not happen. But game's not over. Announce the one trillion dollar backlog. Every out-of-home agent I've ever talked to has offered 50% reductions in price

when doing a large-scale campaign. Most of the inventory is actually pretty cheap if you don't focus on the most premium assets. Where haven't you seen a friend.com billboard? The 101. You haven't seen it, you know, in the iconic places. He hasn't done the Times Square buyout. He's in the subway, right? Like when we saw him, you always make fun of this one. There's one that's like up against a wall. I saw one just at a random bus in my hometown.

It's like, there's just like random places, but there's so many. Some of the alpha and out of home in LA is there's so much traffic. Yes. That you're kind of moving slowly by some areas. And you'll just see random stuff. And so yeah, I was kind of fighting on you,

fighting you on this. Like, it was this truly one of the the greatest campaigns of the year and hearing his extra context. It's incredible. He might have unlocked some entirely new strategy of just like the go big, massive billboard campaign. I wouldn't be surprised of next year is the year of the copy paste of the strategy for, you know, a company that has a million dollars to spend on a big campaign. Let's do an interesting billboard campaign. Maybe they have a million dollars in revenue too. Ideally, yes. Ideally, Ideally, yes. Ideally, yes. I mean, he clearly, like, it was, you know, he's, he's like risk on exploring, testing new things, like learning. But just the core, the core arb of, like, a big billboard campaign pang off, I think you got to credit him. You got to check in with Avi Schiffman.

Did you see his other post? He said, SF is over. Yes. Still a beautiful place to live. Hype around LMs has subsided. It's not an interesting place to be anymore. Why go to a hackathon? It's not like GPTD just came out. There's nothing too interesting to discuss at a party anymore. All the big companies are too mature now. Most of what is new is just YC Slop startups. If you're still in pre-seat exploring stage, it's mostly too late. The directions have been positioned in. It's just a performative scene left. There are always a cycle to these things, and this is fine. I've enjoyed 2022 to 2025. I hereby declare New York the new bastion of what matters in the near future. could not disagree more with every single, pretty much every single word in here.

I think Avi has shown brilliance in some ways, even though many don't. But this was, I put this up as one of the worst takes of the year. It's just like, it's literally like saying like, it's like saying in the early days of the internet or in the early days of the iPhone, like, hey, like, yeah, it's over. Just don't build anything. Also, you should get in, if you're bored with the hackathon, you're bored with the YC Demo Day, get into shark diving. Go dive in the bay. Put on the 7 mil wetsuit. Swim out to Alcatraz, take on a shark head to head, and emerge victorious.

I think that will really give you the sort of the glory. You'll be excited again. You will have survived a shark attack. That will energize you in a way that GPT 5.2 might not be energizing you. Totally. Fighting head on with a great white shark in the San Francisco Bay. That's something you can only do in the Bay Area. Or who's making friend.com for sharks, right?

Like a wearable pendant that a shark could use to, you know, better navigate. Maybe they're lonely out in the high seas, right? It's cold. It's dark. Yes. Maybe, you know, in between hunts, right? They're just kind of hanging out, right? Yeah, just having a digital companion.

Why reserve digital companions for just for humans, right? Like all life, all life matter. Think bigger. The other thing I was thinking, why has no one made like a telemedicine for anabolic steroids for your pets? Somebody has, right? Isn't that a real thing? I want to see a golden retriever as a mass monster.

I think that's just a Rottweiler. I mean, you can make your like cattle really jacked. Yeah, that's like what SARMs are. So you could just, couldn't you just give it to your dogs? You know way too much about performance and things, right? Anyway Yeah, just saying the word Sarms is like just

Just say that you've been Deep in bodybuilding form, Styler Okay, okay, we got a Christmas Present from Friends of the show Sah Hill Bloom Let's open it up It's on Christmas Santa Belt it up

Okay, so this is From Saw Hill Bloom himself Look at this, look at this new brand alert I love this So he said I got sick of putting things on my skin that I'd never put on my body. So I spent 18 months creating the perfect solution.

The perfect solution, Wild Roman. I can smell it. Everyone says the TBP and Ultramm smells bad. Now, it smells great. This actually smells fantastic. So, Wild Roman is 100% natural skin care for men made with grass-fed tallow, cold-pressed oils, and wild botanicals.

You can order today at wildroman.com. I just want to give him a shout out. And then we were going to... So this is good stuff. You've been using this? I've been using this for about, like, two. weeks. Wow. It was working so far.

Two weeks on Wild Roman and you look like that. Yeah. I mean, I think it's been, you know, it's helped with like beard growth and just just skin clarity. You look fantastic. Yeah, you're glowing. Yeah, you're really glowing. Yeah, yeah. Truly never stop using this product.

Yeah. Because I do not want you to go back. Never churn. I think you're a customer for life. What do you think it takes to win in this category? Saw Hills, obviously, a, an influencer, an author, he has a master. He has a massive newsletter. He has 1.1 million followers on X and has an audience.

But something we keep coming back to is like an audience might not be enough to truly win in a category. I think he's got to go hard on Target. This feels like a good brand to introduce like tallow to the target audience. Right? This feels again like going for the set, a bunch of products out the gates. this screams end cap to me. I was talking to a friend, and they have a brand that does over 100 million a year only in Target. Yeah.

They don't sell anywhere else. Yeah. And so it's just such a massive channel. And so I think Sahel can probably leverage his brand to just go really hard into Target early. But I'm sure he can at least get some initial traction D to C. The main thing that people miss with, like, personality-led, kind of like influencer brands like this, is that no matter how big your audience is, you can be Kim Kardashian. And in order to build a truly big business, you get this initial boost from your audience. But the nature of like any audience is that the longer that you just advertise against it, you can saturate it.

So, like, Kim K can post like five times in the first week, but then eventually you have to go find. find new, net new people that aren't necessarily getting exposure. Dave Portnoy, the founder of Barstool, says, Breaking, I am proud to announce in our continuing 20-plus year evolution. We are now partnering with Netflix for exclusive video podcasts. And the way he frames this in the video is remarkable. So let's play it. Emergency press conference time.

If you haven't heard the news, I'm proud to announce that Barstool is partnered with Netflix for three of our top podcast. Exclusive video only on Netflix starting next year. I'm talking you want to watch a video. Part of my take, Netflix. You want to watch a video of Spitting Chick-Chiclix? Netflix. I actually spit there.

That's just my brain. You want to watch a video of Ryan Rusillo show. Where? Netflix. Netflix. Six, seven. Video. Audio.

Stay the same video. Where? Netflix, Netflix, Netflix. Netflix. We're proud to partner in one of the best and breed companies. That's what we do at Barstle. Evolve, rotate, evolve. Video next year, PMT, chicklets,

Ryan Rissolo, Netflix, Netflix, Netflix. Founders, technology, founders. Next time you think, oh, I need to film this crazy cinematic video. I need a film this crazy cinema. I need a studio shot of me sitting down on a couch, looking all put together. Dave is sitting there with a bunch of windows behind it that are reflect,

actively spinning. One shot of this video. And it's way more engaging than him just being, you know, trying to be all professional. No, this is good. But I mean, to be fair, like, in order to do that, 20 years, 20 years of experience. Like, most people cannot just one shot that on day one of their career. The other big get for, I guess, the modern tech companies is the Oscars are moving to YouTube, which is a bomb-held. Okay, so explain the Oscars.

Okay. So it's like, you know how we did the award shows for, you know, random, obscure achievements of the year? Yes, absolute hitter of the year. It's like that, but for movies, of course, the Oscar's art. The Academy of Motion Picture Arts and Sciences, and they said they recently deal with YouTube for exclusive rights to show this, to, to the show, starting in 2029. It really feels like forever, but I'm sure it'll be upon us in no time, but probably the right. time but um does feel particularly it hits particularly hard because it's uh it's like the whole show is about the theater it's about the movie industry yeah and the movie industry is saying like yep

like youtube beat us it's over we're so back but also it's over thank you folks see you tomorrow goodbye