Shawn Presser is a programmer, writer, and scientist. In 2019 he started Tensorfork, the first open-source AI research community, which spawned Eleuther. In 2020 he created Books3, a dataset of ~197,000 books that became part of The Pile and was later used to train LLaMA and Claude. In 2021 he joined Groq, where he helped implement LPUs. In 2022 he was employee #2 at Keen, John Carmack's AGI lab.