The dead internet theory is fairly rapidly happening. More and more of the content has been at least significantly produced by AI and its only going to get worse.

▲

mr_mitm 2 hours ago | parent | next [-]

A corollary of the dead internet theory is the phenomenon where people suspect any content to be AI generated. Sometimes one em dash is enough to spark such suspicions and allegations. Not only is fake content falsely labeled as real, real content is increasingly falsely labeled as fake.

▲

oblio 3 hours ago | parent | prev | next [-]

Amusingly, after a lot of pain this might push us back to the real world :-))

▲

matwood 3 hours ago | parent | next [-]

I was wondering about this. Maybe we were not really meant to spend so much time communicating through screens. And if all we do is communicate through screens, does it even matter if it’s AI, a dog, or a person? I know people will jump in and say yes it matters, but if I was never going to meet the person on the other side of a comment it’s hard to get worked up about it.

	▲	benterix an hour ago \| parent [-]
		A good point. I noticed that every time I see a condescending comment like "The war in Ukraine is totally the fault of the West, NATO should have stopped expanding..." etc., some neurons in my brain get activated and I feel obliged to correct this obvious crap using several reasonable and researched argument. But if these are all bots, who cares... It makes me think if people en masse realize most of their online interactions are with LLMs, they might as well stop using these social platforms for engagement and just switch to totally passive consumption, which gives even less satisfaction and more frustration IMO.

▲

outime 3 hours ago | parent | prev | next [-]

At least when it comes to human interaction (like irl forums etc), I think it has a good chance of happening.

▲

bpavuk 43 minutes ago | parent | prev [-]

damn you are getting downvoted so hard! folks really don't want to leave screens

▲

zombot an hour ago | parent | prev [-]

And it will be increasingly hard to keep the slop of preceding generations out of the training data. The race to the bottom is inevitable.

	▲	simonw an hour ago \| parent [-]
		I don't think that's particularly hard. AI labs are very selective about what they use for training. The days of indiscriminately scraping the entire internet and dumping that in, unfiltered, are long behind us.