I can’t understand how you think that feeding a spellchecker current events is in any way better than feeding it fiction: our original point is that the program is useless and it’s only saying the things it is because that’s what it’s been fed with. It’s not helpful or hopeful.
And that’s the point of the post. Fortunately it’s been ‘fed’ with ideas that are not fascist in nature but actually half sensible. Which is actually, hopeful. Unlike bots which prove on X, to be angry little far right trolls.
The biggest problem is that it does not check anything for truth, as when Michael Cohen produced a briefing for the court that turned out to have case references that in reality opposed his argument: it knew that it needed case references but had no way to find the right ones so it used any ones.
It hasn’t been fed with ideas that are sensible: it hasn’t been fed with ideas at all. If I told it that Trump crashed the economy and then asked it what happened a couple of days ago, you know what it would say? If a Republican told it that he made America stronger, guess what it would say then?
Yes, I get that. It responds to the person that uses it. As I keep saying, it’s a tool.
Sure: but it responds in a way that only depends on what you feed it, allowing anyone to generate anything they want and claim that AI said it so it’s somehow true. That makes it really pointless to use, and is neither hopeful nor depressing.