On June 19, John Jumper posted a short note on social media. After nearly nine years at Google DeepMind, he was leaving for Anthropic.

Jumper won the 2024 Nobel Prize in Chemistry for creating AlphaFold, the AI system that solved protein structure prediction. He was a vice president at DeepMind and one of the leads on Google’s AI coding team. He is not a mid-level researcher looking for a raise. He is one of the most accomplished AI scientists alive.

Three days earlier, Noam Shazeer left Google for OpenAI. Shazeer co-authored “Attention Is All You Need,” the 2017 paper that introduced the Transformer architecture. Every large language model running today, from GPT to Claude to Gemini, descends from that paper. Shazeer built the thing that made all of this possible. And he walked out of the company where he built it.

Two departures. Three days. The co-inventor of the Transformer and a Nobel laureate, both gone from the same company in the same week.

This is not a talent story. It is a platform story.

When industry coverage frames this as a “talent war,” it misses the point that matters for your business. The question is not who has the best researchers. The question is what those researchers are telling you about where the real work is happening.

Gil Luria, an analyst at DA Davidson, put it directly: OpenAI and Anthropic hold an advantage over large companies like Google because they can promise less bureaucracy and a more focused effort on building the next generation of AI systems. When a Nobel Prize winner and the co-author of the foundational paper in modern AI both reach the same conclusion in the same week, that is not coincidence. It is signal.

Google is the company that invented the Transformer. Google is the company that built AlphaFold. Google is the company that trained some of the most capable AI systems ever created. And the people who did that work chose to leave.

They did not leave because Google lacks resources. They left because resources are not the bottleneck. Focus is.

What this means for platform bets

If you are a business leader evaluating AI platforms right now, this week’s departures are more useful than any product announcement or benchmark score.

Benchmarks tell you what a model can do in a controlled test. Talent movement tells you where the next capabilities will emerge. When the people who built the architecture and won the prizes choose to work at Anthropic and OpenAI, they are placing bets on where the most important work will happen over the next two to three years.

Google DeepMind employees have raised concerns internally that the company does not have a clear solution for businesses seeking AI coding tools — the primary growth driver for both Anthropic and OpenAI right now. That is not a minor product gap. Coding tools are where enterprise adoption is accelerating fastest, and the people building them are walking out the door.

This does not mean Google’s products will stop working tomorrow. It means the rate of improvement will shift. The company that loses its best builders does not collapse overnight. It falls behind gradually, then suddenly. And by the time the gap shows up in product comparisons, it is already structural.

The pattern is not new. The speed is.

Andrej Karpathy, a founding member of OpenAI and former senior director at Tesla, joined Anthropic in May. Jumper followed in June. The concentration of top-tier talent at a small number of companies is accelerating, not slowing down.

For the past two years, AI capability has been spreading outward. Open-source models improved. Smaller companies caught up. The gap between frontier and everyone else narrowed. That trend is now reversing at the research level. The best people are consolidating, and they are consolidating at the companies building the tools your team will use next year.

If your AI strategy depends on a specific platform, you need to watch where the builders go. Not where the marketing budget goes. Not where the conference keynotes happen. Where the people who actually create the next generation of capability choose to spend their time.

The business decision underneath

None of this requires you to switch platforms today. But it does require you to stop treating platform selection as a one-time decision.

The AI market is moving faster than annual procurement cycles can track. The company leading today may not be leading in 12 months. The company losing its best researchers today will feel that loss in its products within 18 months.

If you are building workflows, integrations, and team habits around a specific AI platform, you need to know whether that platform is gaining or losing the people who make it better. Two of the most important AI researchers in the world just answered that question for Google.

What matters now is not the departure itself. It is whether your organization is watching the signals that predict which tools will be best in a year — or still choosing based on which vendor had the best slide deck last quarter. Track where the builders go, not where the products are today. That is the bet that pays.