AI is going to hack us.

[R_J_K75]

Apparently long term usage of AI causes people to lose cognitive function.

I wonder if this is why every time I get fast food my order is wrong?

 

[Retros_x]

Apparently long term usage of AI causes people to lose cognitive function. (I'm so scared of it, I don't even want to use spellchecking n_n) It is a silver lining ar least that in the future people cheating through AI will not be smart enough to hide their cheating.

Are there credible studies that claim this? AI is not so long on the market, I would doubt short-term studies at least a bit.

 

[Ulorian - Agent of Chaos]

Are there credible studies that claim this? AI is not so long on the market, I would doubt short-term studies at least a bit.

I linked to a discussion of one on phys.org in the reply to the thread you're replying to.

 

[Retros_x]

eventually it WILL be run by someone evil. Or it may do those things on it's own.

It already is run by someone evil. I won't go into details because of forum rules, but as far as my morals go I consider many of the leaders of the tech companies running current AIs as evil due to their involvement in politics.

 

[Retros_x]

I linked to a discussion of one on phys.org in the reply to the thread you're replying to.

I read the studies summary, but this is not conclusive evidence of AI making users loose critical thinking ability. The author of the summary has some good sceptical points already but I want to add that the correlation could be explained vice-versa: It could be that people with less critical-thinking ability are relying more heavily on AI. Or there is some third factor as causation.

We are in need for a long term study design that has a group using AI tools measuing their cognitive abilities over time and a control group who doesn't use AI at all. But I would bet somewhere something like this is already in research.

 

[MarkB]

Yeah, some people have always been self-deluding, and other people have always exploited that. I think this gen AI stuff just cuts out the "middlemen" so to speak, ie the cult-leader or propagandist or whatever. Now I can just fire up an app to get a friendly voice to feed my confirmation biases - and all without the pesky constraints of experience or training or a moral compass!

Yeah, you've gotta wonder how much cult propaganda was out there publicly-accessible and got picked up as part of these AIs' training data. If someone asks an LLM AI a question that would be typically asked of a cult leader, does the AI then just generate a suitably cult-leaderish response because that's what comes up as most heavily weighted in its training data?

 

[Ulorian - Agent of Chaos]

I read the studies summary, but this is not conclusive evidence of AI making users loose critical thinking ability. The author of the summary has some good sceptical points already but I want to add that the correlation could be explained vice-versa: It could be that people with less critical-thinking ability are relying more heavily on AI. Or there is some third factor as causation.

Exactly! I didn't say it was a flawless study; I agree that its findings could be explained in other ways. Part of the reason I linked to a discussion of the study rather than to the study itself.

We are in need for a long term study design that has a group using AI tools measuing their cognitive abilities over time and a control group who doesn't use AI at all. But I would bet somewhere something like this is already in research.

Yup.

 

[GMMichael]

skip to the end with a TLDR.

Aw! I was hoping for something short after "TLDR."

I wonder if this is why every time I get fast food my order is wrong?

No, that's just the universe telling you not to order fast food.

Did Rolling Stone throw in the "turn off her phone . . . surveillance concerns" bit as if cellphones aren't personal tracking devices? I expected better, RS.

 

[Ruin Explorer]

I do not think it is just AI telling people things, but some believe it to be THE answer and not just a tool to help look at all the sides. It will get worse once kids lean on it and grow with it- same as cell phones and social media. We are not teaching people how to think and it is easier to just ask someone to skip to the end with a TLDR.

We were already not teaching people how to think - that's the key problem and part of why AI has been such a big hit.

Across the West (and in the East too), we've seen a decline in the value put on critical thinking and analysis skills, and it was never super-high, even when it peaked in probably the early 1970s. And that decline isn't just in educational curriculums and so on, but across the culture as a whole. News, particularly, has found that people would rather be told how to think about an issue than encouraged to think about an issue critically. Hell, people will literally change channel or newspaper to find one that tells them how to think (c.f. The Times in the UK for example, possibly the most contemptuous "This is how you must think" paper in Britain, and I'm including the Mail in that!). And worse, completely fake and dishonest "balance" has lead even well-meaning news channels into just pushing absolute drivel as if it were on-par with science or logic. Laws and political systems are treated as if they're inconveniences to emotion-based goals, and if anyone tries to have a reasoned opinion or show the slightest nuance or complex, they're given short shrift, because emotion-based stuff, often completely irrational and obviously false, gets more views, and most interviewers seem to prefer worthless soundbites over substantive interviews/comments.

Frankly it's shocking we've lasted this long when much of our own media are basically trying to make us as stupid and emotional as possible, and refuse to hold politicians and corporate spokespeople and so on to account, because they might require some effort and would get fewer views than just letting them spew nonsense.

AI obviously has the potential to compound this issue, but I'm not sure it's as big of a change as you're suggesting. I think it's even possible it'll actually turn out that future generations use it more responsibly, having grown up with it, than current ones do. Not that there won't be problems in the meantime.

 

[Ruin Explorer]

I wonder if this is why every time I get fast food my order is wrong?

I actually have a comment on this! I order fast food all the time in the UK (less so recently but that's a different story), and I haven't had a wrong order for literally several years! My wife is working in the US atm, and gets wrong orders from fast food regularly. We order stuff pretty much the same way, but it's very clear that there's some kind of operational difference in the US that's leading to a much, much higher error rate, including her being given entirely the wrong order occasionally. I think somehow the US doing more takeout and less delivery, particularly via drive-thrus factors into this.

But I might be absolutely talking nonsense.