Comment by gcanyon

99.99+% of all problems humans face do not require particularly original solutions. Determining whether LLMs can solve truly original (or at least obscure) problems is interesting, and a problem worth solving, but ignores the vast majority of the (near-term at least) impact they will have.