That would be the benevolent Engelbart version of AI, wouldn’t it?
Instead of dumb capitalism, clickbait and silly content marketing driving human activity, to have an AI that saves us from wasting our time figuring out the answer to questions that have been answered long before. And then, instead, points us to those questions that have not yet been asked, much less answered. What experiments have not yet been done.
After all, no GPT-4, no matter how many billions of parameters, could tell you what the best battery is in the world if it wasn’t for that one human dude in his garage in Denmark, or Latvia, or wherever, who actually tested them all.