Most Teams Deploy AI Like Hiring Someone and Never Explaining the Company
diegobuildsai · x · 2026-11-19
Developer Diego points out that most teams deploy AI the way you'd never hire a person: no context, no memory, no knowledge of last week's decisions — then wonder why it underperforms.
The takeaway: context and memory are prerequisites for AI to be useful in real workflows, and their absence is the common root cause of disappointing results.
More from coding & agent
- Developer's self-running QA agent finds 28 bugs in PlotJuggler and counting — facontidavide · 2026-11-19
- 226K custom agents deployed to Slack as Cognition and others build on it — silasalberti · 2026-11-19
- Cognition's Poke experiment: pitch your idea by text, win up to $10,000 in Devin credits — marvinvonhagen · 2026-11-19
- Google Antigravity ships huge week: parallel multi-agent runs, SSH dev, PR cards — clmt · 2026-11-19
- GPT-6.2 randomly abandons tasks mid-step, users resort to self-set timers to keep it going — bofors14 · 2026-11-19
- OpenCode fix cuts GPT-6.2 uncached input per step from ~14,000 tokens to 3 — victor_explore · 2026-11-19