Should AI Make Decisions for Us?
The Big Question About Autonomy
When AI agents can take actions in the world, we have to decide: how much should they do on their own? Some decisions are fine to delegate. Others should always have a human in charge.
- Fine to delegate: organising your calendar, finding information, writing first drafts
- Be careful: sending important emails, spending money, making medical decisions
- Never without a human: decisions that could hurt someone
- Good rule of thumb: the bigger the consequence, the more a human should be involved
- Your job isn't to be replaced by AI — it's to be the human who decides wisely
Why a Human Stays the Boss
Imagine handing an agent your pocket money and saying "buy me the best trainers." Efficient? Maybe. But what if it spends everything on trainers you hate? Or picture a quiz-app agent that answers for you — you'd get the points but learn nothing. Now make the stakes bigger: an agent choosing who makes the team, or what medicine someone takes. Mistakes there don't just waste money — they hurt people. So the rule is simple: agents can do the legwork, but a human makes the final call on anything big. The agent gathers, compares, and suggests. You decide. Being the decider is a job that should never be handed over.
- Let agents fetch, sort, draft, and suggest — the boring legwork
- Keep the final "yes" for humans on money, feelings, health, and fairness
- A good agent shows its work so you can check before saying yes
- Practice now: when AI helps you, always ask "do I agree with this?"
Prefer slides, quizzes, and saved progress? Read this lesson in the library — free, no sign-up.