Priya: A hedge fund that lost billions last quarter just bet four hundred million dollars on a single AI chip startup.
Theo: This is Theo.
Priya: And this is Priya.
Theo: And this is your money briefing for Monday, August 10th. We've got three big stories today about where the money is moving.
Theo: First: that struggling hedge fund we mentioned is making a huge, risky bet on the future of AI hardware.
Priya: Next: the AI that helps you code is about to get a lot more… hands-off.
Theo: And finally: a top historian argues that Silicon Valley leaders are terrible readers—and that it's a threat to democracy.
Theo: Alright, let's start with Situational Awareness. That's the AI-focused hedge fund that had a very public, very bad second quarter.
Priya: 'Bad quarter' is putting it mildly. They took a bath on some speculative AI software bets. We're talking billions in losses.
Theo: Right, but they aren't running for the hills. TechCrunch reports they're now leading a four-hundred-million-dollar funding round for a startup called Source Foundry.
Priya: Four hundred million. On one startup.
Theo: And Source Foundry isn't another app—they make specialized AI chips. The actual hardware that all this software runs on.
Priya: So this is a 'picks and shovels' play. Instead of betting on one gold miner, you bet on the company selling tools to all of them. It's a classic de-risking strategy.
Theo: Exactly. It feels like a smart pivot. They're still in the AI game, but they're moving from the frothy, high-risk apps down to the essential hardware underneath.
Priya: Is it a pivot, or is it a Hail Mary, Theo? Hardware is notoriously difficult. It's capital-intensive, development cycles are long... This is no easy win.
Theo: Maybe not, but it's where the real value could be. So even as AI hype seems to be cooling off in some areas, the big money isn't leaving. It's just getting smarter—and moving deeper into the tech stack.
Priya: And the AI boom is creating this huge gravitational pull on the global semiconductor supply chain. We're talking about real, physical things with geopolitical risk, not just code in the cloud.
Theo: A very expensive, very strategic supply chain.
Priya: Alright, let's stick with AI, but move from the hardware to who—or what—is doing the work.
Theo: You're talking about Anthropic's Claude Code. It's an amazing tool for developers that suggests code and helps fix bugs... but it's about to change in a big way.
Priya: It is. And TechCrunch is reporting that Anthropic is turning 'auto mode' on by default.
Theo: Wait. So right now, it's a co-pilot—it suggests things, I approve them. What does 'auto mode' do?
Priya: With auto mode, you give it a high-level command and it just goes—writing entire functions, fixing bugs, building features, all with minimal human oversight. That step-by-step approval is now something you have to opt into.
Theo: Whoa. So you tell it 'build a user profile page' and just... watch it go?
Priya: In theory. You're still supposed to supervise and review it. But this fundamentally changes the workflow from collaboration to delegation.
Theo: I get the appeal. It could free up developers to think about the big picture—architecture, product—instead of just boilerplate code. That's a huge productivity boost.
Priya: Or it's a huge risk. What about subtle security flaws a human might have caught in a line-by-line review? And how do junior developers learn to code if the AI is doing it all for them?
Theo: That is the big question. And it means if you work in tech, your job is shifting. You're moving from creator to reviewer, from coder to AI director. Prompting and auditing are the new essential skills.
Priya: And if you're hiring tech talent, you need to find people who can handle these tools safely. It's a whole new world of liability.
Theo: The co-pilot is becoming the auto-pilot.
Theo: Alright, for our last story, let's zoom out from the 'how' of tech to the 'why.' Harvard historian and New Yorker writer Jill Lepore has some serious criticism of the philosophy driving Silicon Valley.
Priya: Serious is right. Speaking on TechCrunch's Equity podcast, she argued that tech leaders are bad readers, and their misinterpretation of science fiction is actively undermining democracy.
Theo: Oof. So what's her argument? That they're just reading the wrong books?
Priya: It's not that, it's how they're reading them. She says figures like Elon Musk read classic sci-fi from authors like Asimov or Heinlein, and treat them less like cautionary tales and more like instruction manuals.
Theo: So they read a story about a futuristic society and think 'Great business plan!' instead of, 'Wait, what is this story trying to warn us about?'
Priya: Precisely. They completely miss the irony and the critique. Lepore worries about what she calls 'government by machines'—this belief that messy human problems and democratic processes can just be replaced by optimized, engineered solutions.
Theo: Because that whole mindset dismisses debate, compromise, and the will of the people as just... noise in the data.
Priya: It frames them as bugs to be fixed, not features of a free society.
Theo: So... what does this mean for the rest of us? We can't exactly assign them remedial reading.
Priya: No, but it means that when you hear a tech billionaire propose some grand solution for society, be skeptical. Ask what human process is being 'optimized' away. Whose voices are being silenced for the sake of 'efficiency'?
Theo: Read the source material and see if they completely missed the point.
Priya: [laughs] Exactly. And maybe pay attention to the parts with all the messy human feelings. They're usually pretty important.
Theo: Alright, let's do a quick recap. A struggling hedge fund is making a four-hundred-million-dollar bet that the real money in AI is in the hardware.
Priya: Your AI coding assistant is going from co-pilot to auto-pilot by default, changing the job from writing code to reviewing it.
Theo: And a top historian is warning that when tech titans treat sci-fi like a to-do list, it's a huge problem for democracy.
Priya: And before we go... a little tech history, since we're on the subject. On this day in 1945, Percy Spencer filed the patent for the commercial microwave oven.
Theo: No way! That came from his work on radar technology during World War II, right? He noticed a candy bar melted in his pocket.
Priya: Yep. A technology developed for war that ended up mostly being used to reheat leftovers. A slightly less world-altering path than the one Jill Lepore's worried about.
Theo: And that's our show for today. We'll be back tomorrow with more.
Priya: Until then, maybe go read a book. The whole thing.
This show is made with AI: the hosts’ voices are synthetic and the scripts are AI-assisted. Every story links to its original source.