Robert Glaser · 2026-05-05 · notable
Robert Glaser: When Everyone Has AI and the Company Still Learns Nothing
INNOQ's Head of Data & AI argues license metrics and token counters miss the real adoption gap. Sketches a 'Loop Intelligence Hub' that turns scattered agent runs into reusable organizational knowledge.
Counting tokens isn't learning — Glaser sketches a 'Loop Intelligence Hub' to surface what AI actually changes inside an org.
What is it?
An essay on why second-stage enterprise AI rollouts plateau even when individual productivity goes up. Robert Glaser, Head of Data & AI at INNOQ, argues the failure mode is management measuring license seats and token usage instead of capturing what experiments and patterns emerge from distributed teams.
How does it work?
Glaser's three-pillar model: Agent Operations (governance, runtime), Loop Intelligence (signals from real workflows pulled into a shared hub), and Agent Capabilities (skills and templates fed back into the org). The Loop Intelligence Hub is the central claim — a layer that turns scattered agent runs into reusable organizational knowledge without sliding into employee surveillance.
Why does it matter?
317 points and 222 comments on Hacker News this week. The piece has become reference material in the ongoing debate about why enterprise AI investment isn't translating into the productivity numbers boards expected. Sits next to Drew Breunig's '10 Lessons for Agentic Coding' and Addy Osmani's 'Agent Skills' in the same week's reading list.
Who is it for?
engineering managers, CTOs, AI program owners