Pushmeet Kohli

Pushmeet Kohli is an Indian British computer scientist and Vice President of research at Google DeepMind. At Deepmind, he heads the "Science and Strategic Initiatives Unit". He was noted by Time magazine as being one of the 100 most influential people in AI according to the Time 100 AI list.

Pushmeet Kohli
Born	Dehradun, Uttrakhand, India
Nationality	Indian-British
Alma mater	Oxford Brookes University (PhD) National Institute of Technology, Warangal (BTech)
Scientific career
Fields	Machine learning Artificial intelligence Computer vision[1]
Institutions	Google DeepMind
Thesis	Minimizing dynamic and higher order energy functions using graph cuts (2007)
Doctoral advisor	Philip Torr[2]
Website	research.google/people/105667

He has led and supervised a number of projects including AlphaFold, a system for predicting the 3D structures of proteins; AlphaEvolve, a general-purpose evolutionary coding agent; SynthID, a system for watermarking and detecting AI-generated content; and Co-Scientist, an agent for generating and testing new scientific hypotheses.

Education

Kohli received a BTech in Computer Science and Engineering at the National Institute of Technology, Warangal.^[3] He went on to study at Oxford Brookes University, where he earned a PhD in computer vision under Philip Torr in 2007.^[3][4][2] Kohli was also a postdoctoral associate at the Psychometric Centre, University of Cambridge.^[5]

Career

Before joining Google DeepMind, Pushmeet Kohli was partner scientist and director of research at Microsoft Research. His research investigates applications of machine learning and artificial intelligence.^[1][6]

Pushmeet has made research contributions in the fields of computational biology, program synthesis, superoptimization, discrete optimization, and psychometrics.^[7]

Notable Research Projects

• AlphaFold - breakthrough AI system for protein structure prediction^[8]
• AlphaEvolve - agent for code super optimization.^[9]
• AlphaTensor - Reinforcement learning agent for discovering new algorithms for matrix multiplication^[10]
• SynthID - system for watermarking AI generated images.^[11]
• AlphaGenome and AlphaMissense - AI models for predicting the effect of mutations in the genome ^[12][13]
• AlphaCode - Competition-level code generation with AI^[14]
• FunSearch - Discovering algorithms using LLMs to search over program space.^[15]
• Neural Program Synthesis^[16]
• Probabilistic Programming^[17]
• Community based Crowdsourcing of Data for Training AI Models^[18]
• Behavioral analysis and personality prediction using on online networks^[19][20]
• Human Pose Estimation using the Kinect^[21]

• Learnt Magnetic confinement control for Fusion ^[22]
• Learnt Density Functional for solving the fractional electron problem ^[23]

Awards and honours

Pushmeet's research in computer vision and machine learning has been recognized by a number of scientific awards and prizes. Some notable ones include:

• Koenderink Prize (Test of Time award) by the European Conference of Computer Vision^[24]
• British Machine Vision Association and Society for Pattern Recognition (BMVA) Sullivan Prize for the best PhD thesis.^[25]
• IEEE Mixed Augmented Reality (ISMAR) Impact Paper award^[26]
• Lasting Impact Award by the ACM Symposium on User Interface Software and Technology^[27]
• Best paper award at the International World Wide Web Conference 2014^[28]
• Best paper award in the European Conference on Computer Vision (ECCV) 2010^[29]
• Best paper award in the Conference on Uncertainty in Artificial Intelligence (UAI)^[30]