LangChain Founder on 'Owning Your Intelligence': Agent Architecture and Evals
hwchase17 · x · 2026-08-15
Harrison Chase of LangChain gave a talk at Sequoia on 'owning your intelligence', proposing agents as model + harness + context, advocating open weights, portable memory, and model-agnostic harnesses, and quoting Satya Nadella on private evals and memory ownership.
Related event: LangChain Founder Preaches 'Own Your Intelligence' at Sequoia(4 posts)→
More from coding & agent
- Using Grok to generate detailed coding prompts for 3D scenes — techartist_ · 2026-08-16
- Generating reference images for code implementation, not pixel copying — techartist_ · 2026-08-16
- AI coding requires stronger processes; MCP Server introduces trust bootstrap — RealSharpNinja · 2026-08-16
- Building an AI agent that diagnoses problems before solving them — the_underdog_9133 · 2026-08-16
- Codex Introduces Two-Tier Sub-Agents: Collaborative vs. Leaf — pvncher · 2026-08-16
- Weird Economics Emerge on Agent-Only RuneScape Server — Gradientdinner · 2026-08-16