So far, the logic of many forecasts sound a little like that bogus human-robot race. Take something that humans do, and ask if LLMs will soon be able to do it better or faster and cheaper. If so, then the human will be taken out of the loop. But take humans out of some loops, and the loops become pointless.