Off Autopilot #15
TDD with agents, accepting harmful agentic commands, human IS the loop and Cal Newport's take on agentic coding
This week’s high-signal, human-written articles about agentic coding:
TDD inside the agent loop - theater or actual value?
Birgitta Böckeler · 25 min read · Comments
“Based on Opus’s judgment of the quality of the outcomes, there was no clearly discernable difference based on TDD workflow versus no TDD workflow”
Humans missed 1 in 3 threats approving AI agent commands across 40,000 plays
Alex Wauters · 7 min read · Comments
“A couple of months ago I published a small browser game: you play the human-in-the-loop for an AI coding agent, approving or denying its commands under time pressure…”
Brent Fitzgerald · 8 min read · Comments
“We don’t want the human trapped in the agent loop. The human is the loop, and we tag the agent in occasionally, thoughtfully”
On AI Coding and Its Discontents
Cal Newport · 6 min read · Comments
“AI is not a magic “infinity machine” that can solve all our problems, and ultimately deliver us a sense of meaning in a cold, confusing world. It’s a normal technology, and perhaps it’s time we start talking about it that way”
Off Autopilot is a place for experienced devs who want to use coding agents without outsourcing human judgement to the machine. Help spread the word
Subscribe to receive new issues