Your personal, on-call AI advisors — and how to use them.
Most people use AI the way they'd use a search engine — type a question, take what comes back. Often useful. Rarely exceptional.
That's not a flaw in the technology. It's what happens without direction.
Every AI response is shaped by the context it's given at the start. Without it, the model draws from everything it knows at once: broad, balanced, general. A "persona" changes that entirely. You give the AI a specific identity — an area of expertise, a set of priorities, a way of thinking — and it responds from that position rather than from the middle of everything it knows.
Ask an AI about a legal dispute and you get information. Give it the persona of a seasoned litigation attorney — one who knows what courts look for, where arguments fail, and where most people underestimate their exposure — and you get analysis. The first response tells you things. The second one thinks with you.
These principles apply across all major AI platforms. The personas here are written for Claude, my preferred tool — but the approach works wherever you are.
Nine personas follow, lean and full versions both. Instructions included.
Copy. Paste. Ask.
See the Difference
The same question. Two completely different replies.