Safety research is inevitably dual-use

Ryan argues all safety work affects capabilities, discusses alignment MVPs and the tradeoffs of empirical versus secret research.

Play episode from 30:08

chevron_right

Transcript

chevron_right

Transcript

Episode notes

Ryan Kidd, Co-Executive Director of MATS, shares an inside view of the AI safety field and the world’s largest AI safety research talent pipeline. PSA for AI builders: Interested in alignment, governance, or AI safety? Learn more about the MATS Summer 2026 Fellowship and submit your name to be notified when applications open: https://matsprogram.org/s26-tcr. He discusses AGI timelines, the blurred line between safety and capabilities work, and why expert disagreement remains so high. In the second half, Ryan breaks down MATS’ research archetypes, what top AI safety organizations are looking for, and how applicants can stand out with the right projects, skills, and career strategy.

Sponsors:

Tasklet:

Tasklet is an AI agent that automates your work 24/7; just describe what you want in plain English and it gets the job done. Try it for free and use code COGREV for 50% off your first month at https://tasklet.ai

Agents of Scale:

Agents of Scale is a podcast from Zapier CEO Wade Foster, featuring conversations with C-suite leaders who are leading AI transformation. Subscribe to the show wherever you get your podcasts

Shopify:

Shopify powers millions of businesses worldwide, handling 10% of U.S. e-commerce. With hundreds of templates, AI tools for product descriptions, and seamless marketing campaign creation, it's like having a design studio and marketing team in one. Start your $1/month trial today at https://shopify.com/cognitive

Claude

Claude is the AI collaborator that understands your entire workflow, from drafting and research to coding and complex problem-solving. Start tackling bigger problems with Claude and unlock Claude Pro’s full capabilities at ⁠⁠https://claude.ai/tcr⁠

CHAPTERS:

(00:00) About the Episode

(03:50) MATS mission, AGI timelines

(13:43) Evaluating current AI safety (Part 1)

(13:48) Sponsor: Tasklet

(14:59) Evaluating current AI safety (Part 2) (Part 1)

(28:11) Sponsors: Agents of Scale | Shopify

(30:58) Evaluating current AI safety (Part 2) (Part 2)

(30:59) Safety research versus capabilities

(40:01) Frontier labs, deployment, governance

(51:51) MATS tracks and governance

(01:04:11) Research archetypes and tooling