Reinforcement Learning for Board Layout

Sergiy Nesterenko explains Quilter's RL approach, focusing on custom environments, high-level routing actions, and conservative reward functions grounded in geometry and physics.

Play episode from 23:10

chevron_right

Transcript

chevron_right

Transcript

Episode notes

This special AI in the AM episode features Sergiy Nesterenko of Quilter on using reinforcement learning for circuit board design, Andy Hall of Stanford on AI behavior in politics and new governance models, and Lukas Peterson and Axel Backlund of Andon Labs on their AI-run retail store in San Francisco. Nathan and Prakash also reflect on the pace of AI progress, the public reaction to existential risk, and why constructive civic action matters as AI systems grow more powerful and autonomous.

Sponsors:

Roboflow:

Roboflow's free 2026 Vision AI Trends report analyzes 200,000+ real-world projects to reveal how top companies are deploying Vision AI and turning proprietary data into an edge. Download it now at https://roboflow.com/trends

VCX:

VCX, by Fundrise, is the public ticker for private tech, giving everyday investors access to high-growth private companies in AI, space, defense tech, and more. Learn how to invest at https://getvcx.com

Tasklet:

Build your own Cognitive Revolution monitoring agent in one click.
Try it for free and use code COGREV for 50% off your first month at https://tasklet.ai

CHAPTERS:

(00:00) About the Episode

(07:57) Live stream kickoff

(09:52) Sam Altman attacks

(16:37) Quilter from SpaceX

(19:02) Why autorouters fail (Part 1)

(20:52) Sponsors: Roboflow | VCX

(23:09) Why autorouters fail (Part 2)

(28:14) Compute and odd layouts

(34:19) Simulations and safety margins (Part 1)

(39:22) Sponsor: Tasklet

(41:01) Simulations and safety margins (Part 2)