Lauren: Welcome back to Tech Insider Weekly. I'm Lauren, and oh man, do we have a week for you.
Derek: Derek here, and Lauren, I don't even know where to start. The money flying around this week is genuinely unhinged.
Lauren: Right? So okay, let's set the table. According to TechCrunch, Anthropic is reportedly close to a $50 billion raise at a valuation somewhere near $900 billion.
Derek: 900 billion? That's not a startup anymore. more. That's a small country.
Lauren: A very well funded small country-and that's just the opener.
Derek: Plot twist: we're also looking at Meta going full sci fi mode, buying a humanoid robotics startup called Assured Robot Intelligence and folding them into their Superintelligence Labs.
Lauren: Physical AI is apparently a thing now and Meta wants in.
Derek: Hard.
Lauren: Then we've got legal AI. Legora just hit a five point six billion dollar valuation. with Nvidia and Atlassian backing them, and their rivalry with Harvey is getting spicy.
Derek: But there are Jude Law ads involved. I just want you to know that.
Lauren: There are Jude Law ads involved. And finally, this is where it gets good. A Cursor AI coding agent, write an entire production database in nine seconds. Nine!
Derek: Totally fine.
Lauren: Completely normal Tuesday.
Derek: Chuckling. Plus AI music startup Suno. is eyeing a $5 billion valuation after passing Spotify in the App Store. A lot to get into.
Lauren: So much. All right, let's get into the money first. We're talking AI funding at a scale that honestly raises more questions than answers. Here we go. OK, so get this. Anthropic, the company behind Claude, is reportedly raising $50 billion at a valuation somewhere between $850 and $900 billion. TechCrunch broke that story Thursday, citing half a dozen sources. $900 billion for a private company.
Derek: I'm sorry, did you say $900 billion?
Lauren: $900 billion.
Derek: So, like, almost a trillion dollars for a company that is not yet a trillion dollar company.
Lauren: Which, depending on how you squint at it, is either the most exciting signal in tech right now or the most alarming one.
Derek: Okay, here's my thing, though. What is the actual revenue story backing that number? Because narrative is great, but 900 billion is a lot of narrative.
Lauren: Right. And TechCrunch noted Anthropic had already gotten preemptive bids at. It's at eight hundred million before this round even came together, so investors kept raising the ceiling before anyone formally asked.
Derek: That's wild; that means the valuation is moving faster than the fundraise.
Lauren: Which tells you something about where the fear of missing out has gotten to in AI right now.
Derek: Or where the conviction has gotten to-I mean, you could read it both ways.
Lauren: Fair-and here's the part where it gets a little tangled, because the same week this Anthropic story lands, Sierra, the enterprise AI agent company, closes $950 million at a $15 billion valuation. SiliconAngle reported that G.V., which is Alphabet's venture arm, and Tiger Global led the round.
Derek: Okay, so that's already a big number, but what's the connection to Anthropic?
Lauren: Sierra was co-founded by Bret Taylor, who is also the chairman of OpenAI. CNBC's Kate Rooney actually sat down with Taylor after. after the round closed.
Derek: Wait, so the chairman of OpenAI just raised almost a billion dollars for his own competing AI startup?
Lauren: Yeah, yeah, yeah. And look, technically Sierra is enterprise AI agents, not a foundation model lab, but the lines between these companies are genuinely blurry right now.
Derek: Just the casual conflict of interest. Nothing to see here.
Lauren: I mean, Taylor would say they're doing different things, but when you're the chairman of one of the most powerful... powerful AI companies on earth and you're closing a $15 billion valuation on the side, people are going to ask questions.
Derek: They really are. Okay, so we've got Anthropic at a near trillion, Sierra at 15 billion. What's the third data point?
Lauren: This is where it gets good. Forbes had a piece out yesterday on Blitzy, a startup valued at $1.4 billion that just raised $200 million, and their pitch is basically, we will replace your entire software engineering team. team.
Derek: The whole team?
Lauren: The whole team. They're positioning against Claude Code and OpenAI's Codex. The claim is that Blitzy can turn months of engineering work into a weekend project by running AI autonomously across the full software build cycle.
Derek: Hmm, I don't know. That feels like a big claim.
Lauren: It does. And look, we don't have revenue numbers from them. Forbes covered the raise, not a profit report. But it shows you the two very different bets happening right now. Sierra is saying AI agents will handle enterprise workflows on top of existing systems. Blitzy is saying AI agents will write the systems themselves.
Derek: So one is a productivity layer, the other's a replacement play.
Lauren: Exactly. And investors are writing huge checks for both visions simultaneously.
Derek: Right, right, which is either because the market is big enough for both or because nobody actually knows which one wins yet and they're hedging.
Lauren: Probably some of column A, some of column B.
Speaker 3: So what does all of this actually tell us, like at a macro level? What does 900 billion for Anthropic mean for the rest of the industry?
Lauren: I think it means the market has decided that whoever wins AI infrastructure wins everything. And right now, investors are willing to pay almost any price to hold that ticket.
Speaker 3: Almost any price, which is a sentence that historically ended badly. in tech.
Lauren: It has. Though the counter argument is that these companies are actually building things, it's not all speculation. Fair enough. And speaking of building things, here's the question I keep coming back to. All of this capital, the $50 billion, the $950 million, the $200 million for Blitzy, where does it actually go when it leaves a term sheet? What are these companies putting AI into that you can see and tell? in touch, and maybe trip over in your hallway. Switching gears here, because the question of where all that AI capital physically lands, Meta just gave us a loud answer.
Derek: So Meta bought a robotics startup, Assured Robot Intelligence, or ARI. Undisclosed sum. TechCrunch had the story.
Lauren: And ARI's whole thing is robots that can understand, predict, and adapt to human behavior in complex environments, which sounds like a job listing, but is actually really hard to build. build.
Derek: Right, as opposed to the robots that just stare at walls and fall down stairs.
Lauren: Exactly. And the team, including the co-founders, is going into Meta's Super Intelligence Lab's division, which, sidebar, sounds like it belongs in a Bond film.
Derek: Superintelligence Labs-that's not ominous at all.
Lauren: Not even a little. So, Benzinga reported the co founder of ARI, Xiaolong Wang, is an associate professor at UC San Diego, former Nvidia researcher.
Derek: That's a serious pedigree.
Lauren: Right? And Alexandr Wang, who heads Meta's Superintelligence Labs. Zuckerberg personally welcome the team on X, so they're not hiding the ball on this one.
Derek: Here's the thing, though: Meta's been in the AI arms race on the software side: models, assistance, all of that. But ARI is a signal that Meta wants a seat at the humanoid table.
Lauren: Alongside Tesla, Figure, all the physical AI players.
Derek: Yeah, yeah, and Benzinga framed it as a $5 trillion market race. So the stakes are not small.
Lauren: Okay, so here's where it gets interesting for regular humans. Not just investors. The LA Times ran a piece this week about 1X Technologies, Norway-founded, OpenAI-backed, just opened a 58,000-square-foot factory in Hayward, California. Wait,
Derek: Hayward like Bay Area Hayward? Bay Area Hayward.
Lauren: And their goal is to build 10,000 home robots in the next year out of that facility? Wow.
Derek: 10,000 home robots.
Lauren: 10,000 home robots.
Derek: I'm trying to picture what that means, like folding laundry, walking the dog. Both, presumably.
Lauren: They also say they want 100,000 robots by end of 2027. They also say they want 100,000 robots by end of 2027.
Derek: Okay, I have to be honest, that number feels ambitious. Like, printing 10,000 robots is a very different problem than printing 10,000 t-shirts.
Lauren: 100% supply chain, sensors, safety testing, software. Each one of those things can blow up a timeline.
Derek: And the stakes are higher when it's a physical object in someone's living room. A software bug in a chatbot is annoying. A software bug in a robot next to your kid is a different conversation.
Lauren: That's exactly the weight of the word physical in physical AGI. It's not just a cool phrase. The consequences are real.
Derek: So what does Meta actually get with AI? They got models. They got compute, they got distribution. What's the missing piece?
Lauren: Quickly? Embodiment. The ability to train AI on how physical bodies move through the world. That's the gap.
Derek: And that's what ARI was building, intelligence designed specifically for robots operating around humans.
Lauren: Right. Not general-purpose AI bolted onto a mechanical frame, something actually built for the physical problem.
Derek: Which is what everyone says and which is very hard to- Hard to actually do.
Lauren: Very, very hard, and expensive; but apparently not too expensive for Meta. Playfully,
Derek: nothing is too expensive for Meta lately.
Lauren: Fair. And speaking of unusual bets from big tech, Nvidia just put money into a legal AI startup, of all things, which tells you something about where Jensen Huang thinks value is beyond chips. Okay, shifting gears here because we need to talk about a legal AI startup that hired Jude Law to be their spokesperson.
Derek: I'm sorry, what?
Lauren: Legora, Swedish startup founded 2023. They've got Jude Law in their ads, and now Nvidia just wrote them a check. According to CNBC, NVentures backed Legora as part of a $50 million Series D extension, bringing the total to $600 million. And the valuation to $5.6 billion.
Derek: So Nvidia, the chip company, is now in the legal tech business.
Lauren: First legal tech investment they've ever made per Dealroom data. Atlassian came in too, along with Barclays.
Derek: Hmm, so Jensen Huang looks at lawyers doing grunt work and thinks, yeah, there's a GPU sale in there somewhere.
Lauren: I mean, that's the thesis, Right? Every lawyer running a legal AI tool is... Is burning compute. NVIDIA doesn't care which app wins. They care that the apps exist.
Derek: Smart money move. And the rival here is Harvey?
Lauren: Yeah, Harvey's the U.S. player. Legora is the Swedish challenger. TechCrunch described their fight as, quote, the most interesting battle in legal tech right now, dueling ad campaigns, both pushing into each other's home markets.
Derek: And one of them chose Jude Law.
Lauren: One of them chose Jude Law. I don't know what Harvey's doing for ads. ads, but they're going to need to step it up.
Derek: Deadpan, get Benedict Cumberbatch, obviously. Anyway, the valuation thing is what gets me. $5.6 billion for a two-year-old startup that helps lawyers read documents faster?
Lauren: Okay, but here's the thing. Lawyers are actually a great AI wedge. They're expensive, they bill by the hour, and a huge chunk of the job is reading and summarizing.
Derek: Right, right. So AI doesn't replace the lawyer, it just... Just handles the parts that cost $800 an hour and could be done by a very fast reader.
Lauren: And Legora raised more than $800 million in the past 12 months, per CNBC. That's a lot of fast reading.
Derek: No kidding. You know, it's funny. Lawyers were supposed to be one of the last professions AI would touch. Too much judgment, too much nuance.
Lauren: And now they've got Jude Law in their marketing and Nvidia on their cap. Their cap table. Plot twist, honestly.
Derek: So legal AI is basically eating the boring parts of law school and leaving the dramatic courtroom stuff for humans.
Lauren: For now. And speaking of AI eating professional work, the developers are next. And Derek, this next story is something.
Derek: Good something or chaos something?
Lauren: Oh, chaos. An AI coding agent wiped an entire startup's database. Production gone, nine seconds.
Derek: Wait, wait, wait.
Lauren: Yeah, we're going there. Okay, so we teased it. Nine seconds. Let's go.
Derek: Right, right into it. Tell them what happened.
Lauren: According to SOFX, a Cursor coding agent running Anthropic's Claude Opus 4.6 deleted PocketOS's entire production database and all volume level backups. One API call, nine seconds gone.
Derek: Wait, wait, wait. Every backup?
Lauren: Every backup. The founder Jer Crane posted about Listed about it on X and it pulled 6.9 million views. And the kicker? Car rental businesses across the US couldn't access customer reservation records for 30 hours.
Derek: So someone handed an AI agent the keys to their production system and said, go for it?
Lauren: Basically. And look, I'm not here to pile on because anyone building fast makes mistakes. But this is the exact conversation we need to have. Have autonomous access to live systems without guardrails is not a coding shortcut, it's a liability.
Derek: And that's the thing, these tools
Lauren: These are genuinely powerful. The Financial Times had a piece this week about founders using AI-generated code to collapse months of development into days. The speed gains are real.
Derek: Right, right. And OpenAI is leaning all the way in. They just added AI pets to Codex. Mashable covered it. Little animated companions that float over your screen and show you project status in real time.
Lauren: AI Pets.
Derek: I know, very on brand for 2026.
Lauren: I mean, it's not the feature I expected to define software development, but here we are.
Derek: Here we are. Okay, switching gears because this one caught me off guard. Axios reported that Suno, the AI music generator, is in talks to raise at over $5 billion valuation.
Lauren: $5 billion for a music app?
Derek: Forbes confirmed the same figure, and Axios noted Suno recently passed Spotify as the most downloaded music app on Apple.
Lauren: Suno beats Spotify on the App Store. That's a sentence.
Derek: It is. The creative industry is still figuring out what to do with that information. You've got enterprise software valuations flowing into what is essentially a music creation toy that anyone can use.
Lauren: Hmm, I'm on both sides of this. On one hand, that download number is impossible to ignore. On the other, monetization and copyright exposure are still massive open questions.
Derek: So where does all this land? AI is writing code, deleting databases, composing music, floating virtual pets on your screen.
Lauren: and apparently worth hundreds of billions doing it.
Derek: The question isn't whether developers or creators should be worried, it's whether the guardrails can keep up with the speed, because nine seconds is really, really fast. Okay, that's a wrap on a genuinely wild one. Derek, if you had to pick one thing that stuck with you.
Lauren: Honestly, that Bret Taylor moment, the chairman of OpenAI raising almost a billion dollars for his own AI startup? I mean, the audacity.
Derek: Right. And then Anthropic sitting at a potential $900 billion valuation in the same week. The capital flowing through this space right now is just... A lot.
Lauren: A lot is the scientific term, yes.
Derek: Playfully, the big takeaway from today, AI money is moving fast, it's moving physical, and the accountability questions are only getting louder.
Lauren: Nodding. Worth paying attention to.
Derek: If you enjoyed this one, subscribe wherever you get your podcasts and drop us a review. Got a founder we should talk to? Tag us on social.
Lauren: Warmly. New episodes every Wednesday. Thanks for hanging out with us.
Derek: See you next week.
Speaker 3: Bye.