creators

Chris Olah - Anthropic Co-Founder, Interpretability

Chris Olah co-founded Anthropic and leads its interpretability work, reverse-engineering neural networks into algorithms people can actually read.

Chris Olah is a Canadian AI researcher and a co-founder of Anthropic, where he leads interpretability research: reverse-engineering neural networks into algorithms people can read. He studied maths at the University of Toronto for about a year, left, and in July 2012 took a $100,000 Thiel Fellowship. He interned at Google Brain in 2014 and 2015 and was on staff there from October 2015 until 2018, latterly as a research scientist; he co-founded the journal Distill, which launched in March 2017, and led OpenAI's interpretability team from 2018 to 2020. In May 2024 his team matched groups of neurons inside Claude to concepts such as bias and scams. In May 2026 he spoke at the Vatican at the presentation of Pope Leo XIV's encyclical on AI, arguing the industry cannot be left to supervise itself.

Specialization

mechanistic interpretability, neural networks, AI safety

Country

United States