AGI Strategies

person

Colin Raffel

UofT; Hugging Face; T5 author

Associate professor at the University of Toronto and Vector Institute; previously a faculty researcher at Hugging Face. Lead author of T5 (2019), one of the foundational text-to-text pretraining frameworks.

current Associate Professor, University of Toronto; Faculty Member, Vector Institute; Faculty Researcher, Hugging Face

Strategy positions

Open sourceendorses

Release weights widely; transparency beats closed safety

Argues open-source ML research, datasets, weights, training code, is the principle that lets science around foundation models actually accumulate; closed releases break that mechanism.

“We propose a unified framework that converts every text-based language problem into a text-to-text format. The simple objective lets us study transfer learning systematically.”
§ paperExploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer· arXiv / Google Research· 2019-10· direct quote

Closest strategy neighbours

by jaccard overlap

Other people whose strategy tags overlap with Colin Raffel's. Overlap is on tag identity, not stance; opposites can show up if they reference the same tags.

  • Ada Rose Cannon

    shared 1 · J=1.00

    W3C web standards advocate; AR/VR engineer

  • Ali Farhadi

    shared 1 · J=1.00

    Allen Institute for AI CEO

  • Ali Ghodsi

    Ali Ghodsi

    shared 1 · J=1.00

    Databricks co-founder and CEO

  • Anjney Midha

    shared 1 · J=1.00

    Andreessen Horowitz general partner; AI investor

  • Arthur Mensch

    Arthur Mensch

    shared 1 · J=1.00

    CEO of Mistral AI; French frontier-model founder

  • Ce Zhang

    shared 1 · J=1.00

    ETH Zürich → University of Chicago; ML systems

Record last updated 2026-04-25.