UC Berkeley's SkyDiscover-Synthesize: AI-generated just-in-time databases and inference engines
CShorten30 · x · 2026-10-05
Episode #147 of the Weaviate podcast features Shu Liu, Shubham Agarwal, Mert Cemri, and Alexander Krentsel from UC Berkeley's Sky Computing Lab, discussing their SkyDiscover-Synthesize research.
The core idea is "just-in-time systems": instead of fixed software, AI generates complex software components on the fly — task queues, databases, and even inference engines like vLLM.
The episode also covers how Claude Code and its cohort of AI coding tools have changed how we build software and perceive what software is, and where the limit of AI-generated software lies.
Related event: UC Berkeley's SkyDiscover: AI Generates Specialized Software On Demand(4 posts)→
More from coding & agent
- Same model, very different performance: Dylan Patel on why the harness matters — MatthewBerman · 2026-10-08
- EreNodes after a year and 44 versions: prompts as clickable pills in ComfyUI — Erehr · 2026-10-08
- localmcplib: an open-source shared component library for local MCP servers — mostly_deterministic · 2026-10-08
- e2e: open-source agent turns natural-language goals into E2E tests for web and mobile — abhishek__AI · 2026-10-08
- TERMy24k: billed as the first local neuro-symbolic AI terminal assistant — gioscarab · 2026-10-08
- Managed Agents on Interactions API: one call spins up Antigravity agents in a remote sandbox — clmt · 2026-10-08