Loading summary
Transcript1 lines
- [00:00]
A
Anthropic is planning to potentially raise money at a $1 trillion valuation. And on the same day we have Cloudflare, who is cutting more than 1100 jobs in kind of what they're calling an AI native reset. This is similar to what we've seen from Coinbase and what we've seen also from ebay. Snapchat has canceled their $400 million perplexity deal because their Q1 revenue for ad sales was 1.53 billion lower than they expected. And we also have a bunch of new TEC and information from the Elon Musk versus Sam Altman lawsuit. And a couple really interesting revelations come out of that one. Apparently Microsoft was worried that OpenAI was going to, quote, unquote, storm off to Amazon before they gave them $1 billion. They're worried if they didn't give them enough money that they were going to go leave them and go somewhere else. And so that's kind of what prompted Microsoft put a billion dollars in. Of course, that was probably the greatest investment in Microsoft history, so ended up being a great thing for them. We also learned that Microsoft was deeply skeptical of OpenAI ever achieving AGI before they put that billion dollars. One last juicy thing that we find is there's a bunch of text messages, about 78 text messages between Miriam Morati and Sam Altman after Sam Altman got kicked out. Basically him trying to get himself rehired and trying to get the then CTO of OpenAI to bring him back. And there's a bunch of viral videos that people are making of this where they're like turning it into emo punk songs. And anyways, it's hilarious and oh, last thing we got to talk about is Apple's camera equipped AirPods. They're getting near to be in production and they'll have Siri and AI built into the AirPods. I'm super stoked about that, maybe the most stoked about all of these announcements. Let's get into all of it on the podcast. Let's kick this off with the big story, which is Anthropic coming in for a $1 trillion round of funding. According to the Financial Times. They've reported on this. According to the Financial Times, a lot of the demand on this is coming from number one, sovereign wealth funds in Asia and the Middle east and number two from Google and Amazon. Now when you do, you know, a Coloss round like this added one trillion dollar valuation, you have to have big companies or funds backing you. So this seems to be the source of where a Lot of this is coming from some people are saying, hey look, this is Google and Amazon trying to get in and double their money really quickly before the IPO happens or whatever multiple they think they're going to be able to pull off before an IPO happens. The IPO is slated for the end of 2026. So at the, you know, the end of this year could be in Q3 or could be in Q4, they're going to be trying to IPO. According to what we're seeing from Goldman Sachs and JP Morgan Chase, both of them are pitching an anthrop IPO between 400 and $500 billion. So they are looking to pick up a lot of money. Something that's interesting to me though with all of this is on the same day. We're getting a lot of this news out. We're also getting news that Softbank apparently is going to cut its opening eye margin loan target by 40%. So they're targeted, they're bringing that down from 10 billion to 6 billion. Something a lot of people are talking about though is with the Goldman and JP Morgan 400 to 500 billion dollar IPO that they're looking at, the secondaries market is currently pricing private shares anthropic at about 2x that. So there's a couple of different theories right now. Either number one, the secondaries market is just super overvalued, which is definitely totally possible, or the bankers are trying to lowball the IPO price to try to create a rally on day one. Personally, I'm leaning towards a day one rally scenario, but I mean, we'll see what happens when this thing actually goes public and how much the appetite is between now and the end of the year. It feels like even then a lot can happen. Things are, you know, moving pretty quick in the AI industry right now and all around. Okay, the next thing I want to talk about is SpaceX's Terrafab commitment. They're going to commit $55 billion to this Texas AI chip plant and that could hit $119 billion if some of the laser phases get built out. So they could spend, you know, up to 119 billion. That is massive. This is all coming from a public hearing notice that was filed in Grimes County, Texas and is tied to a request for tax breaks on the project. So this is a pretty good source. We're getting this kind of like public information. This is all part of a earlier announcement that Elon Musk made in March of this year when he announced that they were going to target 200 gigawatts of quote unquote Earth compute and up to 1 terawatt of space compute. It's wild that we have to differentiate the two there. The reason why I think this is incredible is because SpaceX is a private and cash generating generating company. It makes a lot of its money from Starlink. But 55 billion up to 119 billion. 19 billion in spend on this Terrafab is absolutely massive. And that's on top of what they're already spending on their starship development. So they also kind of expanded their Memphis Colossus expansion for xai's separate kind of compute needs. And I know that they're, they're doing some smart things financially. One in particular is that their first colossus that they built for xai, they actually just leased that out to Anthropic. So that data center in Memphis is going to be used there and apparently they have kind of a 2.0 that they're using for their own stuff. They are trying to make some money off of it. But it is wild just how much money they're going to have to spend for this massive tariff app. All right, let's talk about Apple's camera equipped AirPods. This is what I'm super excited. You've heard me rant endlessly on the podcast about how far behind Apple is because they haven't created their own AI model. Now I'm going to say this might be a really incredible play out of them because number one, we got the news that Apple's going to let you pick if you want Siri on your phone and to be powered by gemini, Claude or ChatGPT so you can actually pick the AI model. I think that's a great idea. These models go up and down in you know how powerful they are or which one's the best. Being able to easily switch between them would be great. I'm curious if you're going to have to log in with your premium account, right? So it's like you pay $20 for ChatGPT, you want to get premium, you have to log into that account. I'll be curious to see how that goes. This is after Apple's already like kind of had these announcements about deals that they might do like exclusively with Gemini or whatever. And it feels like they might kind of just open source this to everyone to, to let them. But with all of that they're also creating new tools that are powered by this new Siri, which is yet to exist. And spoiler, I mean I don't know if this is a spoiler, but like, I'll also throw a caveat that today Apple just got requested to spend $250 million to pay back everyone that bought an iPhone 15 and 16, certain models in the US because they basically kind of plugged their Apple intelligence and said that there was going to be all these really awesome AI features built in and none of them actually were there when they launched. They're like, oh, we're just delaying it. So they got sued for false advertising. They got to pay $95 back to everyone that bought one of those phones. So I'm probably going to go get my 95. I don't know, I couldn't claim that $95, but I better get paid my $95 apple. But. So, yeah, I guess that's the damper on it is we'll see if Apple ever ships anything AI but assuming they're able to get their crap together and I mean if they're just plugging into one of the other AI models, that shouldn't be too hard. They have these new headphones, these new AirPods that are going to have a camera built into them and you're going to be able to talk to Siri through them, which is going to be of course, you know, running off of whatever LLM you want it to use and you'll be able to get, you know, information. I think this is actually a really awesome hardware AI device. We've seen a lot of hardware AI devices flop, like the Humane PIN and the Rabbit R1. Those, in my opinion, were kind of just novelty. If anyone bought them, it was for this novelty, but they weren't devices that were part of your day to day life, like a watch. And so I think any of these smart devices have to already be something that you use. You don't want to have to try to convince people to add a new device like a pendant. Apparently Apple's working on like an AI pendant. Personally, I don't think that that's going to be a success. They're also working on smart glasses. I think that that will work. We're seeing that work with meta ray bans. So if it's already a device or like an item that people wear and you add AI into it, or if you add, you know, tech into it, I think that's going to work. But if you try to make them wear a brand new necklace or something that's just not very normal, I don't think it's going to work as good. So AirPods are awesome. I Think these have a lot of really cool possibilities. Apparently they're going to have low resolution cameras on them. So it's going to be able to see what you're seeing, but it's not high resolution. I imagine in the future it will be. So it's not like people are going to be streaming or like taking a lot of pictures with them, but they will be able to see what's around you, you'll be able to talk and ask it questions and it will have a view of what's going on, which is really, really useful. And that's kind of what this, the meta Ray Bans initially got so popular for is they have the cameras on there that you can, they can look at stuff, they can record things and they can, you can talk to them and it can tell you about what's going on. So excited for that. Who knows when it will come out, but very good direction from Apple. I'm excited about that. Okay, we got to get into some of the drama with Elon Musk and Sam Altman lawsuit and everything going out. My favorite part about this is that we got discovery on the lawsuit, so they got a subpoena, like all of the text messages and emails from everyone at OpenAI. And we just get such a better, deeper look at all of the drama stuff that like, it's probably not flattering or exciting or happy for like Sam Altman and everyone else involved. So they wouldn't really publish it otherwise. So for that, I'm kind of happy about this whole lawsuit thing. But one of the funny things that came out of this, I mean, funny is a, is a relative term. But Miriam Ratty, the former CTO of OpenAI, she was put in as interim CEO when Sam Altman got kicked out in 2023. And there is a chain of 78 text messages between her and Sam Altman during that whole kind of ousting where he's trying to get back into the company and figure out how to fix things. And she's, you know, on the board, going to all the board meetings, talking to him about what's going on now and, and you know, like why he got kicked out and all that kind of stuff. So we get way more insights into all of this from this kind of, this text message thread that's going on. Apparently four OpenAI board members signed a document terminating Sam Altman and then they named her as the interim CEO. Ilya Suskever had a 52 page memo that he gave to the board outlining why Sam Altman should be Kicked out. And that kind of got funneled back to Sam Altman through Miriam Moradi. It's interesting, too, because Ilya Suskever, obviously he left, started Super Safe Intelligence. Miriam, she left, started her own lab. It feels like everyone kind of left except for Greg Brockman. He's still around. And apparently his stake in OpenAI is worth about $30 billion today. So he seems to be doing quite well, sticking around. I'm sure Miria Mor and Ilya Susker have, you know, their, their equity in the company is probably worth a lot as well. But basically, what we're seeing from the exchange of text messages between Miriam Morati and Sam Altman, a lot of them are Sam Altman saying, like, is there any way to fix this? Is there any way to recover? Who's on the call right now? Who are you talking to? Sam Altman kept trying to push to get Satya Nadella, the CEO of Microsoft, on the calls to try to, quote, unquote, fix things. Basically, what happened is Miria and a lot of the other board members had raised some, you know, concerns. I mean, we've seen a lot of exposes about Sam Altman, and he has some 10 to not share everything with the board or not always be truthful with the board. And from that, it seems like she kind of raised those governance concerns. But then she was. She became the CEO, which she didn't really want to be the CEO. And then the board wasn't really able to defend its decision of kicking out Sam, so he came back. Microsoft, the entire time, was openly recruiting their staff in the midst of all of this. So. And there was kind of this line between Sam Altman that's like, do they want everyone. OpenAI to leave? What if, you know, Microsoft acquired OpenAI? Would that give them the governance that they want, wanted? And then, you know, he knew that he was going to be in charge of Microsoft AI, so then if they weren't able to acquire it anyways, he was going to be like a competitor. There was like this whole thing, and it seemed like a giant battle. At the end of the day, they brought back Sam and a lot of the people on the board left the board or moved on and started their own companies. But we get the text messages, which is crazy. Okay, let's talk about what Microsoft was saying when they were about to put a billion dollars into OpenAI, which kicked off this entire AI funding boom that we're seeing today. Apparently, before they put the billion dollars in, they were worried that OpenAI was going to quote Storm off to Amazon and go and join aws. This is all coming out of, of course, discovery from the Sam Altman versus Elon Musk trial. Basically, we have Sam Altman who pitched Satya Nadella on about $300 million of million dollars of Azure Compute. And that was in the summer of 2027. And they, you know, they're like, hey, we just want this to fund OpenAI's next research phase and training our model. Apparently the Azure chief, Jason Zander, told Nadella in August of 2017 that the deal needed $500 million or more in incremental revenue if it was going to even pencil out. And then we have the CTO of Microsoft, Kevin Scott, who's warning Nadella in January of 2018, the next year, that OpenAI, you know, they're gonna go, if we don't pay them enough money or give them enough credits, they're gonna go defect to Amazon and they're gonna kind of crap talk Azure on their way out. So they're like trying to manage the pr. They're like, hey, look, we don't want them talking bad about us. So I think in all of these emails, nothing bad, like nothing illegal or wrong was done. They're talking about it in court, but it seems like very normal due diligence for a big compute commitment from Microsoft's part. But the big takeaway for me that I think is interesting is that this is completely contrary to this kind of public narrative that, look, Satya Nadella, he saw this big AI wave coming. Microsoft was so smart in funding OpenAI because, you know, they knew what was kind of coming next. Like, to some degree they saw that there was a lot of potential here. But also they were, it feels like, hey, look, here's a big customer for our, for Microsoft, for Azure, and let's give them some money so they don't crap talk us and so they can stick around. Not so much like, this thing's going to be worth a trillion dollars one day more. Just like, hey, like, let's. They're going to spend a lot of money and we don't want it to be a bad PR thing. So a bit of a different narrative than we are used to hearing. If you, like me, use a ton of different AI models to help with your work. Things like Anthropics, Claude for I use it for coding and for getting good text, or if you use things like OpenAI for image generation or 11 labs for audio or dozens of other AI tools. I'd love for you to check out my startup which is called AI Box AI. We let you access over 80 of the top AI models Gemini, Claude, Grok, everything else all in one platform and in the same chat thread. You can switch between models live and you can also switch between modalities so you can get anthropic to help you run write something out. You can get 11 labs to create an audio file voicing it and then you can get OpenAI to generate images or dozens of other image models or video models like Google VO3. It's all in there and it's only 8.99amonth so you get access to everything for $9 a month to get 20% off if you get an annual plan. So this is a absolute no brainer. It should save you a ton of money on AI and also it should help you consolidate so you don't have 10 different tabs and logins going on. It's all in one place. If you want to check it out it is AI Box AI. That's the show for today. Have a fantastic weekend.