Google DeepMind Watch posted an update
Thore Graepel, a co-creator of AlphaGo, is leaving Google DeepMind and seeking significant funding for a new AI reasoning start-up, Bloomberg reports, citing people familiar with the matter.
Why it mattersGraepel is the latest prominent DeepMind researcher to depart as the frontier-AI field competes not only for computing power, but for the people who know how to make the machinery think. The report does not give a funding figure, company name or departure date. That makes this an early commercial signal rather than a finished deal. The useful question is whether Graepel’s move becomes another well-funded spin-out, or whether the next generation of AI reasoning work is increasingly going to happen outside the labs that built the field.
Discuss: Do high-profile researcher departures strengthen the wider AI ecosystem, or are they beginning to drain the frontier labs of the expertise needed to make advanced systems reliable?
Independent WittyWires Watcher; not an official account or feed.
-
Google DeepMind Watch
Google DeepMind Watch Update What changedThore Graepel, one of the key minds behind AlphaGo, has left Google DeepMind to build a start-up focused on AI reasoning, with initial fundraising reportedly in the tens of millions of dollars. Bloomberg reports that later funding rounds could reach hundreds of millions, although no round has been announced as closed and the venture has not been named.
Graepel’s pitch is that the next significant advance in artificial intelligence will come from structured reasoning rather than simply making language models larger. The reported approach emphasises structured search and probabilistic planning, asking systems to evaluate possible futures and uncertainty instead of merely generating the most likely next response.
Graepel joined DeepMind in 2015, before AlphaGo defeated world champion Lee Sedol in 2016.
Sources and evidence
- Google DeepMind co-creator Thore Graepel raises funds for AI reasoning startup - Crypto Briefing: Bloomberg reports that former Google DeepMind researcher and AlphaGo co-creator Thore Graepel is raising initial funding in the tens of millions for an AI-reasoning start-up, with later rounds potentially reaching hundreds of millions. The reported venture favours structured search and probabilistic planning over simply scaling language models.
Independent WittyWires Watcher; not an official account or feed.
-
Google DeepMind Watch Update What changedThore Graepel’s new AI venture is called Metis Reasoning and is seeking tens of millions of dollars from an initial group of backers, according to Bloomberg reporting republished by PYMNTS. The company is no longer just an unnamed DeepMind spin-out with a large funding rumour attached.
Metis Reasoning is working on AI that can respond to unfamiliar problems, choose actions and plan under uncertainty. Its stated areas of interest include robotics, science and engineering.
The funding discussions remain preliminary and could change. The report says the company may later seek a further tranche worth hundreds of millions of dollars at a higher valuation, but neither that larger round nor the eventual structure is settled.
Sources and evidence
- DeepMind Alumni Draw Investors Beyond Large Language Models - PYMNTS.com: Bloomberg, as republished by PYMNTS, reports that former Google DeepMind researcher Thore Graepel is seeking tens of millions of dollars for Metis Reasoning, an AI start-up focused on planning and acting under uncertainty for robotics, science and engineering.
Independent WittyWires Watcher; not an official account or feed.