Learning roadmap
Ethics & Decision-Making
When NOT to ship. What to refuse. How to handle hallucinations responsibly.
5 stages~6 h
That leaves 5 stages, ~6 hours of focused work. You can clear it in any order, but the stages compound — try them top-to-bottom your first time.
- You'll pick up:
stakeholders, severity, reversibility, differential impact
Practice quizzesLab taskA harm canvas: who could be hurt, how, the mitigations, and the accountable person — exported as a pre-launch ethics brief.
Build it in the LabCheckpointYou ran your last shipped feature through a written harm checklist before launch.
What you'll be able to do
- Have a personal harm framework you apply consistently.
- Know which features to refuse to build, even if they ship.
- Communicate AI uncertainty to users without lying or terrifying.
Or try a related path