Megan Skiendel: I want to start with Demis Hassabis's statement. Not the stock drop, not the Anthropic announcement — the statement.
David Sterling: The public tribute.
Megan Skiendel: He and Jumper shared the 2024 Nobel Prize in Chemistry. For AlphaFold. Same podium. And now Hassabis is writing warm send-offs on LinkedIn or wherever — look, that's not how you talk about someone who simply got a better offer. That's how you talk about someone you failed to keep.
David Sterling: Really. You're reading the tone of a press statement as evidence.
Megan Skiendel: I'm reading it alongside the reassignment. Jumper was moved off frontier scientific research onto Google's AI coding tools — in his final months at Google DeepMind. After the Nobel. Honestly, after two hundred million protein structures predicted by the system he co-built. And GOOG dropped six percent when it came out he was joining Anthropic, same week Noam Shazeer left for OpenAI. That's not a poaching story. That's an exile story. The statement from Hassabis just closes the loop.
David Sterling: So the 6% — you're saying the market is pricing the reassignment decision, not the departure itself.
Megan Skiendel: Exactly that. What kind of organization hands a Nobel laureate a coding-tools ticket and then acts surprised when he walks to Anthropic.
David Sterling: Well — pump the brakes on that for a second. Think of it like a star architect leaving a massive construction firm for a boutique studio. The firm doesn't collapse. Everyone just starts asking whether the big firm still attracts the best blueprints. That's the actual story here. Not that Alphabet is broken.
Megan Skiendel: Sure. But the blueprints walked out.
David Sterling: Right, except — wait, actually the reassignment detail. That's sourced to trade outlets. Google DeepMind hasn't confirmed it. Anthropic hasn't confirmed it. Jumper's own departure statement was warm. Warm toward Hassabis specifically. That's — I mean, that's tension with the 'pushed out' read you're building.
Megan Skiendel: Warm statements are written by comms teams, not Nobel laureates.
David Sterling: Frankly, maybe. But here's the point — GOOG dropping 6% on one personnel exit at a two-trillion-dollar company is a sentiment story. Not a fundamentals story. You cannot do that math on fundamentals alone. And — look — Alphabet has invested in Anthropic. The talent is literally flowing to a company Google partially owns.
Megan Skiendel: Honestly, that detail is almost funnier than the stock drop. They're financing their own competition for the same people.
David Sterling: But that's — wait, that's actually where I think the hot take lands correctly. Not the dysfunction read. The destination.
Megan Skiendel: The destination being Anthropic specifically.
David Sterling: Right. Because — look, when Dario Amodei's founding team walked out of OpenAI, everyone said those researchers would scatter. Academic labs, other companies, whatever. Instead they built Claude. The whole Anthropic product line. That exodus is the precedent here, and the precedent says: top-tier people who leave do not come back, and they do not go dormant.
Megan Skiendel: And now Anthropic has a Nobel laureate. The person who co-built AlphaFold — two hundred million protein structures, same Nobel podium as Hassabis — that person is now defining what Anthropic's frontier looks like. Not Google's.
David Sterling: That's the repricing. The market isn't paying for Jumper's past work — AlphaFold already happened. It's paying for who hires the next Jumper. That's optionality. And Shazeer co-led Gemini before he walked to OpenAI. Two VPs, forty-eight hours. That's not noise.
Megan Skiendel: It's contagion. And honestly — this is the part that matters — Anthropic isn't academia. It's a direct, well-capitalized competitor. That's what makes this cycle different from normal researcher churn.
David Sterling: Yeah. The hot take gets the mechanism wrong but the pattern right. I'll give it that.
David Sterling: I'll wait for a third. Two data points — Jumper to Anthropic, Shazeer to OpenAI, forty-eight hours apart — that's a pattern candidate, not a confirmed trend. I mean, two VPs is two VPs. Show me the third VP out of Google DeepMind in the next six months and now we're talking structural.
Megan Skiendel: The third data point might not be a VP. It might be the researcher two levels down who watched what happened to Jumper after the Nobel and quietly started taking calls.
Megan Skiendel: But look — you want the clean sentence? The 2024 Nobel Prize in Chemistry winner is at Anthropic. You can debate push versus pull all day. The Nobel is not at Google DeepMind anymore. That's the sentence the market read.
