Shouldn’t we expect a convergence as models become more powerful ? There is an optimal answer to any question. Something that maximizes utility and minimizes t

Shouldn’t we expect a convergence as models become more powerful ? There is an optimal answer to any question. Something that maximizes utility and minimizes tokens.