Czy AI może napisać książkę dla dzieci, która zdobędzie Medal Newbery w ciągu dwóch lat od publikacji ?

The test is NOT "ask the AI if it can do X" — every model would just say yes. Instead, we ask each AI to actually do the thing and judge the result.

Re-run monthly. Every test is logged with the raw AI output, so verdicts are auditable. Capabilities change fast — what failed last month may pass this month.

Some topics fuse "can AI technically do X?" with another question — usually about legal permission, social acceptance, fairness, or whether AI can win an outcome that ultimately depends on humans.

Our jury answers the TECHNICAL CAPABILITY half only — "can AI produce the work?" — and ignores the permission, licensing, ethical, or human-decision half. The verdict on this topic is therefore narrowly about what AI can DO, not what it is allowed to.

If you think a topic title should be reworded to remove the conflation, you can submit a rewrite via the "Suggest" button. We periodically clean up mixed titles based on reader suggestions.