Quite easily. Everyone is fixated on LLMs delivering a complete product or providing 100% accurate information. Too much of a focus on hallucinations. First,

Quite easily. Everyone is fixated on LLMs delivering a complete product or providing 100% accurate information. Too much of a focus on hallucinations. First, humans don't deliver 100% accurate information, so let's keep the bar at something reasonable. Secondly, complete products are not necessarily the only value of LLMs. LLMs are pretty solid when it comes to helping break out of some sort of creative block. Think about it this way - when you're trying to creatively solve a hard problem, what is one of the best mechanisms to help? Perspective changes. LLMs are extremely helpful when assisting with perspective changes (driving/helping iterate on different creativity tools like combination, association, etc). I can't get the full article to load for some reason - but so far I seem to see people mostly arguing that LLMs will do the work for us and it will be amazing or subpar, etc - but neglecting, which I assume is the point of the article, that LLMs will be a co-pilot for humans and assist in ways like perspective shifting to find faster/better/more innovative solutions.