Someone just walked out after 27 years at the company — and that same company put money on where they landed.
The day Jeff Dean handed in his resignation, Alphabet's stock slid nearly 5% intraday. And yet Alphabet ended up on the list of founding investors in the startup he just launched.
Why leave now, after 27 years?
Dean joined Google in 1999 as employee number 30 and stayed for 27 years. He built core pieces of Google Search's infrastructure, co-founded Google Brain in 2011, and later drove development of Google's custom AI chips, the TPUs. Until recently, his title was Google's Chief Scientist.
The three who left with him aren't lightweights either: Sanjay Ghemawat, a Google senior fellow who's worked alongside Dean for over 20 years; Oriol Vinyals, a VP at DeepMind; and Quoc Le, a co-founder of Google Brain. Together they founded Discovery Loop — a Palo Alto-based public benefit corporation.
At a UW commencement speech in June, Dean said he felt "the same urge I felt in 1999 to join a startup". He didn't hint at any concrete plan in that speech — in a later interview, he said the idea had only crystallized in the weeks since. What the company wants to do is simple to state: automate the research cycle of hypothesize, experiment, evaluate, repeat — at massive scale, using AI. Run thousands of experiments in parallel, and let a small team move faster and more precisely than a huge research organization. It'll start with automating ML research and engineering, then expand into hardware design, drug discovery, and clean energy.
But Alphabet put money in anyway
Here's where it gets interesting. On the day of the announcement, Alphabet's stock slipped as much as 5% intraday — the market's immediate reaction to losing its chief scientist and three top researchers at once. CEO Sundar Pichai said in a statement: "After 27 amazing years, Jeff is at a point where he wants to try something new."
Instead of lawsuits or competitive posturing, Alphabet made a different call. It became a founding investor in Discovery Loop, and committed to being its cloud compute partner for at least a year. The seed round is co-led by Radical Ventures and Khosla Ventures, with Kleiner Perkins, Lightspeed, and Doerr Capital also participating — and Alphabet's name sits right alongside theirs. Valuation hasn't been disclosed, and the round hasn't closed yet.
If you can't stop your best people from leaving, don't sever the relationship — turn it into equity and infrastructure instead. That's the move Alphabet made here.
| Typical response | Alphabet's response | |
|---|---|---|
| Relationship | Treat as a rival, cut ties | Become a founding investor |
| Infrastructure | Cut off access | Committed cloud partnership for 1+ year |
| Information flow | Control via NDAs and litigation | Secured upside through equity |
| Public message | Silence or defensiveness | CEO statement framing it as a friendly exit |
What if this isn't a one-off, but a new playbook?
This departure isn't an isolated event — it's part of a bigger reshuffle. Demis Hassabis moved from CEO of DeepMind to Chairman and Alphabet's Chief Scientist, and Koray Kavukcuoglu was promoted from CTO to SVP overseeing Gemini model development. The leadership gap Dean left behind had already been filled before this announcement went out.
Discovery Loop isn't going to compete with Google's core business anytime soon — its focus is research automation, not search or ads. The long-term roadmap is a different story, though. Hardware design is on that roadmap, and given that Dean is the person who designed Google's TPUs, it's hard to rule out overlap somewhere down the line.
What this structure really signals is this: treating the loss of top talent not as an unrecoverable loss, but as an asset you can convert into equity — that's now a real option in Big Tech's talent playbook.
5 things to check when your best person says they're leaving
- Figure out the actual competitive overlap
Are they really going to a direct competitor, or an adjacent space that doesn't overlap yet? Like Discovery Loop, if the business doesn't overlap, there's less reason to go on the defensive. - Check if an equity stake is on the table
If you can't stop the departure, see if you can negotiate a share of whatever they build next. It's the most direct way to turn a loss into an asset. - Keep the relationship alive through infrastructure
Cloud, data, tooling access — look for a connection point you can keep. Alphabet chose cloud compute. - Line up succession before you announce
Filling the leadership gap ahead of the announcement signals stability to both your org and the market. - Design the message to be friendly, not defensive
One CEO statement is the difference between "betrayal" and "something to celebrate." Factor in the effect on your employer brand.
Want to dig deeper?
Discovery Loop's official site Read the company's mission statement and open roles directly discoveryloop.com
The original TechCrunch report The first article to break the news of Dean's departure and the new company techcrunch.com
Radical Ventures' investment note The lead investor's own explanation of why they backed the deal radical.vc
GeekWire's in-depth interview Traces the arc from Dean's UW commencement speech to the decision to found the company geekwire.com
Investing.com's market reaction coverage How Alphabet's stock moved on the day of the announcement investing.com
