4 sources
A useful person to study if you care about alignment proposals that try to make superhuman systems legible enough for humans to supervise in practice.
Researcher Profile
Editor reviewedPushmeet Kohli
Robotics, vision, structured prediction
Research leader at Google DeepMind
A strong person to follow if you want to understand how frontier AI gets pushed into science, security, and trustworthy deployment rather than staying inside benchmark culture.
Organizations
Google DeepMindGuru Gobind Singh Indraprastha UniversityGoogle
Labs
About This Page
This profile is meant to help you get oriented quickly: why this researcher matters, what to read first, and where to explore next.
Last reviewed
March 18, 2026
Official And External Links
Known For
The ideas, systems, and research directions that make this person worth knowing.
01
Applying frontier AI to science and public-interest problems
02
Robustness, verification, and trustworthy deployment
03
Translating research progress into high-impact systems
04
Robotics, vision, structured prediction
05
Pushmeet Kohli (homepage)
06
DeepMind
Start Here
Canonical papers, project pages, or repositories that anchor this profile.
Signature Works
Additional papers, projects, or repositories that help flesh out the profile.
Supporting Sources
Additional links that help verify and flesh out this profile.
Related Researchers
People worth exploring next because they share topics, labs, or source material with this profile.
3 sources
A strong researcher to follow if you care about reasoning-heavy language models, especially the line connecting chain-of-thought style methods, evaluation frameworks, and more agentic prompting patterns.
Start HereStanford CRFM people page
3 sources
One of the clearest people to follow for the sequence from retrieval-augmented language models to compute-optimal scaling and then into Gemini.
3 sources
A useful page for the DeepMind work that connected large-language-model scaling to the multimodal Gemini push, with a clearer safety-and-evaluation flavor than many purely scaling-focused pages.
4 sources
One of the better people to track for the sequence from Gopher to retrieval-augmented language models and then into Gemini, especially if you care about how DeepMind actually iterated on the frontier-model recipe.
4 sources
One of the most useful people to study if you care about what deployed models get wrong under pressure, especially around extraction, adversarial behavior, and practical security failures.