What made you laugh today - Fifth Edition

Users who are viewing this thread

I think there was a time when people considered the pursuit of it, at least, to be something valuable and worthwhile. The trick was not to eliminate truth as a concept but to convince the world at large that it's not worth the effort, and that universality is a lie.
We are already past the post-truth era, and well into the post-reality era, with AI becoming omnipresent.
We used to see or hear something and assumed it to be real unless something gave us pause.
Soon we'll assume anything to be 'fake' unless we have good reason to trust it.

Imagine a news story about an oil spill on a beach. The media show videos or pictures from the beach. Except It's of course all AI generated. It is that beach and it's how the oil spill will look. Today we'll say it's fake but soon no one will care because it's just as 'real' as all the other edited media we see around us. Soon 90+% of anything on the Internet is AI generated so we reach a point where reality is irrelevant. It's all just a matter of whether it's believable or not. No one can tell the AI from reality anyway (soon). Remember close too all images we look at today are 'fake'/photoshopped/filtered and we all accept it. And everyone looking at the perfect lives of influencers and youtubers etc. believe it/buy into it.

Most importantly actual humans will be less and less involved in generation of news and other 'content'. So lies will not actually exist - just anomalies in the algorithms. Like when AI news create nonsense. No one lied - and no one is responsible:

When we've successfully removed humans from everything - economic transactions, media, news, and anything run by algorithms - no one is responsible for anything and no one lies. A self driving car that kills a child is a tragedy. No one is at fault. The company behind the car will say sorry and that they're working to improve the algorithm. The news outlet posting imaginary stories ('lies') will say the algorithm needs adjusting.
Fakeness has been a problem in news media since its inception. Doctored images were more ubiquitous in the 1930s than they are now. I don't think generative AI is going to radically change anything on that front because people under the age of 60 are already skeptical of most images anyway. Facebook for example is full of this stuff and its mostly just generic feelgood stories with 100k botted likes.

What has already changed however is the amount of autogenerated text online. This has been a serious issue for almost 15 years. Now that prompt-based text generators are free its exploded, but the slop being generated isnt necessarily better or worse in quality than the crap real people produced en masse before.
Surely the viability for mass production of believable AI videos impersonating real people is going to change things though? From what I read in the news that is essentially what has already begun to happen (I think there was some sort of prank video of Starmer a few months back that had people fooled), and people are very ready to believe what they're seeing. It's bad enough that someone like me, a right minded person with all the correct preconceptions, believes something if it correlates with their fears and hopes, but what if the cretins who believe Trump et al are the second coming of Christ were to see a video of 'Biden' admitting to buggering children several times a day, and that he is in fact a 30 year old man who would be full of vigour were it not for the draining effect of near constant buggery, which is why he appears to be three steps from the grave and thus quite unsuitable to be president? It's easy for me to believe that such a video could prompt a similar event to the storming of the Capitol, and to cap-it-all would not be surprised if the majority of the Republican party would publically support such an occurrence if they felt it was a vote winner. That's going off topic somewhat, but I do think that the time will come (whether in five years or twenty) when confidence in any footage relevant to public debate will be virtually nil because of the possibility of it having been created by AI. You say that people are generally sceptical of images already, but that doesn't sound reassuring - they have to believe something, and if they disbelieve an image or a statement written by a journalist (the derisive 'mainstream media' seems to have a common meaning across many age groups, they think it is all lies and so they just believe whatever those within their small social circle believe) then they will only be basing their views on less evidence than someone who is willing to give credence to what they see or hear. I don't know, one can speculate a great deal on a subject like this and swing from one extreme to the other. It's a good point you make about doctoring and misrepresentation of images etc. being nothing new though, it is easy to forget that the will to deceive and ease with which people can be swayed - sometimes with terrible results - isn't something unique to recent decades.
Top Bottom