NAS Made One-of-One Models Practical

Joseph Nelson explains Roboflow’s weight-sharing neural architecture search, how it efficiently explores thousands of subnetworks, and how users can generate custom Pareto-optimal models.

Play episode from 01:01:40

chevron_right

Transcript

chevron_right

Transcript

Episode notes

Joseph Nelson, CEO of Roboflow, breaks down the current state of computer vision and why it still lags behind language models in real-world understanding, latency, and deployment. He explains how Roboflow distills frontier vision capabilities into efficient, task-specific models using techniques like Neural Architecture Search and RF-DETR. The conversation covers Chinese leadership in vision, Meta and NVIDIA’s roles in the ecosystem, coding agents, and emerging S-curves from world models to wearables. Nelson also explores aesthetic judgment in AI, real-world applications from agriculture to sports, and why outcome-focused regulation matters.

Sponsors:

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

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

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

(04:23) State of computer vision

(12:29) Is vision solved

(19:41) Frontier models and failures (Part 1)

(19:46) Sponsors: Tasklet | VCX

(22:39) Frontier models and failures (Part 2)

(32:16) From cloud to edge (Part 1)

(32:21) Sponsor: Claude

(34:33) From cloud to edge (Part 2)

(43:25) Data needs and scaling

(50:52) Open source vision race