Feature 2.4 · “Could this be AGI?” questionnaire

Could this be AGI?

Pick a real system you know — a chatbot, a coding assistant, anything — and answer six questions honestly. You'll get a qualitative read (High / Medium / Low), not a score that pretends to certify anything.

Solves unfamiliar problems?

Handles tasks it was never directly trained on, not just variations of examples.

Transfers knowledge across domains?

Applies an idea learned in one field to a completely different one.

Plans long-term?

Breaks a distant goal into steps and sticks with it over days, not minutes.

Uses tools on its own?

Discovers and combines outside tools (search, code, APIs) without being wired up each time.

Adapts to novelty?

Adjusts when the world changes instead of repeating a failing approach.

Operates with limited supervision?

Makes steady progress without a human checking every step.

Low0 / 12 · 0 of 6 marked above “No”

AGI-like characteristics: Low

Reads like today's narrow or assistive AI: capable in places, but not broadly general. That is the normal, expected result for real systems — and this checklist agrees with the simulator's AI profile.

Weakest areas

Solves unfamiliar problems?
Transfers knowledge across domains?
Plans long-term?
Uses tools on its own?
Adapts to novelty?
Operates with limited supervision?

Scoring: Yes = 2, Partially = 1, No = 0. High ≥ 10 · Medium 5–9 · Low ≤ 4.

Qualitative resemblance only — never a finding that a system “is AGI”. A High read means “worth a closer look,” nothing more.

Cross-check in the simulator →← Back to the comparison table