Context isn't the agent bottleneck — enterprise knowledge is an untrusted mess, argues Google researcher
_AustinCalvert_ · x · 2026-11-01
Google's kaseyklimes pushes back on the idea that "context" is the real bottleneck for enterprise agents: meetings live in Granola, async comms in Slack, docs in Notion, agent logs in shared repos — all with ample MCPs and CLIs. The problem is that this knowledge is stored, retrieved, and acted on non-deterministically: unstructured, stale, and overwhelming, so neither agents nor humans can trust it, and results fall far short of agentic-system visions. Austin Calvert agrees, noting the industry still awaits a new work-OS UX layer.
More from coding & agent
- FrankenSim dev shares a prompt trick: swap physics terms to boost simulation realism — doodlestein · 2026-11-01
- claude -p burns ~1.8× more subscription quota than Claude Code TUI, Magpie switches — dotey · 2026-11-01
- LangChain's Harrison Chase: new context engineering challenges as agents go proactive — hwchase17 · 2026-11-01
- guizang-pages-skill: turn Markdown into polished single-file HTML pages via AI agents — op7418 · 2026-11-01
- One Pipecat pipeline, six voices: classifier-fronted handover between AI characters, one-shotted by Opus — alexcovo_eth · 2026-11-01
- Build a 5-bot Grok content crew: drafts, video, replies and calendar on autopilot — nima_owji · 2026-11-01