Dan Wasserman

A year of agentic AI inside a research agency. The three things still in the way were not technical: trust, accountability, and relevance. Stop checking the work; start checking the thinking.

Key Takeaways:

1. Set the standard before you hand the work over: a standard written afterward is just an opinion.

2. Don't drop your review, raise it: stop checking the work and start checking the thinking.

3. The hardest blocker to AI success isn't trust or accountability, it's relevance.