A truly great >ai< could make assumptions that you can't simply verify with a single quick internet search. It could be a claim that is simply distilled from- a

A truly great >ai< could make assumptions that you can't simply verify with a single quick internet search. It could be a claim that is simply distilled from- and supported by 5 separate 18th century manuscripts that don't contain the answers themselves. With an actual advanced system most of the answers would be like this, thus making explicit labels meaningless.