“They know what hallucinations are... if they ever just take output from LLMs and plop it to me or a client without that review, that's a strike”

“You cannot inspect quality into a product ... stop thinking about the output is bad, ... govern and manage that, test it along the way”

“I’d go through every fix a person made to an agent’s work… propose a rule… add each approved rule as a test using the original input and the accepted output.”

“You give team A asterisk 6.1 and team B asterisk 6, then you see ... compare team A's output to team B's output, were there more errors or uncaught bugs”

“Agents can be better at structured extraction than humans… if you have a harness where the agent is not only extracting the information but correcting itself, reviewing its outputs, you kind of get to unbalance that in the favor of perfect accuracy”