- Joined
- 19 Oct 14
- Moves
- 155955
@wildgrass
Thanks for this, I have read some of these pieces and I will check out the others. I am always struck with the naïveté - both politically and philosophically - with which these 'tech gurus' approach issues concerning society and the economy. I guess all their focus is on computing and not reading history/ philosophy etc. subjects that have analyzed and discussed these issues for centuries.
In many regards AI (and technology in general) is treated in a very similar fashion to how 'God' was treated in the Middle Ages, namely allowing people to project fantasies of what AI will do to 'improve' society, along with the benefits it holds for an (always) unknown future. In deeply cynical age these delusions seem to be all many people have left to cling on to...
@beardmusic saidWhy would sam Altman read philosophy when he could just ask chat gpt to summarize 1000 years of human thought and regurgitate it in a concise 200 word abstract?
@wildgrass
Thanks for this, I have read some of these pieces and I will check out the others. I am always struck with the naïveté - both politically and philosophically - with which these 'tech gurus' approach issues concerning society and the economy. I guess all their focus is on computing and not reading history/ philosophy etc. subjects that have analyzed and discusse ...[text shortened]... future. In deeply cynical age these delusions seem to be all many people have left to cling on to...
1 edit
@beardmusic saidHere's another worrisome story from a few days back, highlighting that they could find more than 146,000 hallucinated scientific citations that were found in serious scientific papers. This is clear evidence that AI is being used on a massive scale to generate new knowledge, and possibly all that knowledge is just wrong. The problem is not being fixed, or getting better, because the new iterations of the large language models are using the scientific data that was generated by an AI algorithm that was hallucinating and then the data was published. This is quite problematic considering how scientific discoveries build on one another and now we have machines poking non-productive holes in central premises.
@wildgrass
Exactly!! LOL
https://phys.org/news/2026-05-ai-generated-fake-citations-scientific.html
@kmax87 saidalso, occasionally, the teacher pet suggests mass genocide.
This. It’s like a teachers pet that can repeat all the pet phrases on demand, that if you didn’t know might think the teachers pet actually had some insight.
whether it is "solve world hunger by killing all children" or "solve this bug by deleting the entire code base" or "my next chess move is to materialize a queen in the middle of the board, jump over enemy pawns and kill the king directly"
@Zahlanzi saidThose are extreme examples but the real damage is being done much more subtly. People think they don't need to know things anymore. At one point it was just maps. Like, I don't need to know where I'm going because Google Maps will tell me after I'm in my car.
also, occasionally, the teacher pet suggests mass genocide.
whether it is "solve world hunger by killing all children" or "solve this bug by deleting the entire code base" or "my next chess move is to materialize a queen in the middle of the board, jump over enemy pawns and kill the king directly"
But ChatGPT knows "everything" right? Why should we study or learn or read books?
And the damage to knowledge directly is also real. As noted above, scientific journals inundated with fake citations and fake data making unsound conclusions. Peer review catches a lot of it, but what gets through is interpreted by ChatGPT as real informative facts, stuff that trained scientists would know is not real but everyone using the AI algorithms to look it up just think it's real.
Sure, one day the "allocation of world resources" will be done by AI and we'll all live in a dystopian future dictated by the monetary value computers put on human life, but in the meantime we'll all get dumber.