Kavukcuoglu takes the DeepMind helm as Brin pushes all-in on Gemini

Share
Kavukcuoglu takes the DeepMind helm as Brin pushes all-in on Gemini

Google's AI reshuffle now has a named leader: Koray Kavukcuoglu, DeepMind's CTO, is stepping up to run the lab with a direct line to Sundar Pichai, while Reuters details how Sergey Brin has been pressing staff to commit everything to Gemini. The moves put the new org chart in writing after months of reported drift — and they land with Google's flagship model still in catch-up mode.

Koray Kavukcuoglu is becoming head of DeepMind, reporting directly to CEO Sundar Pichai, as Google reorganizes its AI effort around a single all-in push on Gemini. Kavukcuoglu, previously the lab's CTO and Google's chief AI architect, inherits a unit that hasn't unveiled a frontier model since early 2026, according to CNBC, while Anthropic's Mythos and OpenAI's GPT-5.6 have reset the pace. A Reuters exclusive fills in the maneuvering behind the reshuffle: Brin addressed hundreds of employees at an April town hall urging DeepMind to move faster, and has since urged key AI staff to go all in on Gemini. At an all-hands on August 6, staff learned some teams are moving out of DeepMind and into corporate Google — a structural shift that, alongside Demis Hassabis's move to chairman and Alphabet chief scientist, concentrates more control in Silicon Valley.

The internal picture explains the urgency. Gemini briefly overtook competitors last November with Gemini 3, but fresh updates from Anthropic and OpenAI put it back in catch-up mode; by August, Google had delayed the next flagship by two months after internal testing showed it still lagged rivals on coding. The two original technical co-leads of Gemini have quit to co-found a startup, and Google declined to answer Reuters' questions about Brin's involvement or DeepMind's restructuring.

We covered the opening act yesterday — Brin takes over Gemini as 3.5 Pro is reportedly shelved — and the follow-through is the real story: a named operator, a direct reporting line to the CEO, and a lab folded into the parent company's commercial priorities. CCS Insight analyst Ben Wood put the mandate plainly: "The goal will undoubtedly be to close the gaps with Anthropic and OpenAI in some areas." Brin's all-in directive makes the bet unambiguous — but as we argued in Brin can't fix Gemini by showing up, authority was never the missing input. A reorg only works if it stops the talent exits and ships a frontier model on a named date.

What to watch: whether the newly structured DeepMind commits to a ship date for the next Gemini flagship — and whether the co-lead departures are the last of them.

Kavukcuoglu gets a direct line to Pichai — but can a reorg close a coding gap that has already cost Google two months and two co-leads? Tell us in the comments.

Read more

Byteification turns Qwen3 and Llama 3 into byte-level models

Byteification turns Qwen3 and Llama 3 into byte-level models

The tokenizer is the last piece of the language-model pipeline that was never learned end to end — and a Nature paper out this week shows it can be swapped out for less than one percent of a pretraining budget. Researchers have retrofitted four open subword language models into true byte-level models using less than 1% of a typical pretraining budget, and the result — published 7 October in Nature — closes the performance gap that has kept byte-level models a niche curiosity for nearly a decad

OpenAI's dots go mobile and can start your Codex work

OpenAI's dots go mobile and can start your Codex work

One product update and two benchmark papers this hour, all circling the same question: how much of the work can you actually hand over, and what happens to the part you didn't? OpenAI's dots — the always-on agents that get their own cloud computer — can now be created and configured entirely from the ChatGPT app on iOS and Android, and a dot can start Codex work or pick up an existing Codex thread on your behalf. The October 9 release, billed as the first of a fresh run of updates for dots, mo

Coding agents as a hospital: Cockroach's five-month results

Coding agents as a hospital: Cockroach's five-month results

The agentic-coding debate keeps being framed as speed versus trust. Cockroach Labs just published the most detailed trust-first ledger yet — five months of numbers, including the parts that didn't work. Cockroach Labs ran its codebase like a teaching hospital for five months and says the pipeline merged well over a million lines of code with seven total reverts. In the setup the company calls MOLT Sinai, GitHub issues are patients, merging is discharge, and the humans in charge are Chiefs of M

Today in AI — October 10, 2026

Today in AI — October 10, 2026

A day where the boring machinery took center stage: proof checkers, order books, warehouses and the humble text message — plus fresh arguments about who gets to declare any of it safe. Models & Research * Mathematicians get a reliability primer for the Lean Theorem Prover. A guest post by Thomas Hales on Terence Tao's blog walks through what Lean actually guarantees — and devotes itself to the "Summer of Soundness Bugs," the string of kernel bugs found this summer that let false proofs thr