LANDMARK · 5 MIN · USE
Creative exploration
After this landmark, you can use a model to widen your options in the divergent phase of creative work while keeping the selection and the voice your own.
Some uses of AI live where being wrong doesn’t cost anything: brainstorming names, angles, what-ifs, twenty openings you’ll throw nineteen of away. Here the model’s weakness — that it produces the average, plausible option — barely matters, because you only need raw material to react against. It is genuinely good at quantity and variation; it is weak at taste, at knowing which of the twenty is actually good. That judgment is the part you keep. The risk isn’t a wrong answer, it’s a bland one: lean on the model too hard and its generic average quietly becomes your output. Use it to diverge, then do the converging — the selecting and shaping — yourself.
This is the canonical concept. It stays the same across learner lenses so personalization never changes the underlying facts.
◌ EVERYDAY LENS
What this looks like for you
Stuck naming something or planning a trip? Ask for fifteen options, not the best one. The value is in having things to react to — “not that, but the third one’s close” — which is faster than staring at a blank page.
MAKE A DECISION
You need a distinctive name for a new project and ask a model for ideas. It returns a list of competent, safe-sounding options. What’s the best use of it?
CARRY THISOn your next creative task, ask a model for ten options, then write an eleventh yourself that reacts against all ten.