• Replies 8 replies
  • Subscribers 81 subscribers
  • Views 124 views
  • Users 0 members are here
Options
Related
Hints and tips
  • To quote someone's message, first select/highlight the relevant text, and then use the Quote button that appears
  • To "mention" someone, type @ followed by the start of the forum name, and choose from the list; that person will get a notification of being mentioned

Artificial cognition (AI)

B

I found this article interesting and thought I would share it:

https://arstechnica.com/ai/2026/04/research-finds-ai-users-scarily-willing-to-surrender-their-cognition-to-llms/?utm_source=firefox-newtab-en-gb

'In “Thinking—Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender,” researchers from the University of Pennsylvania sought to build on existing scholarship that outlines two broad categories of decision-making: one shaped by “fast, intuitive, and affective processing” (System 1); and one shaped by “slow, deliberative, and analytical reasoning” (System 2). The onset of AI systems, the researchers argue, has created a new, third category of “artificial cognition” in which decisions are driven by “external, automated, data-driven reasoning originating from algorithmic systems rather than the human mind.”'

  • This is interesting! I had a quick skim through and it seems a bit concerning how frequently the faulty AI recommendations were followed. But it's also good that AI could help reduce decline in performance under time pressure.

    I wondered if people had drafted a correct answer without AI whether they would override it with the faulty AI response. Like I said I only skimmed it, so might have missed this part!

    I try to only ask AI if I've already got an idea about something I'm researching, but it's tempting to just go with what it says if you're in a hurry

  • Thank you very much for posting this. I've downloaded the paper, and will give it a read over the weekend, and probably have a chat with Claude about it as well Slight smile I'm a fan of the original Thinking Fast and Slow book, so looking forward to digging into this. 

    Could I in turn share https://youtu.be/pO0WZsN8Oiw?si=-vol4vIpyoDa6bVW with you - it effectively asks the question as to whether human cognition is really computational in nature, or is that just a useful simplification - and hence, could AGI ever really be a thing? Personally, found this to be really enjoyable. 

  • in reply to B

    The feeling of being rewarded when using AI inappropriately sounds like it could lead to addiction. Peehaps that is a fanciful idea.

  • It’s good to see research into the effects of AI use on our cognition.

    Although many of us may have had concerns about humans losing the ability to reason and evaluate, it did feel rather alarming to see some of the processes involved outlined in the paper. The researchers stress they see the issues highlighted as a design and education challenge to enable people to maintain critical thinking and accountability.

    I would like to see the topic being taken more seriously by leaders and politicians.

  • in reply to TheCatWoman
    TheCatWoman said:
    Another thing I read about in the Guardian last week was about how AI's are sometimes refusing to follow the commands of humans, particularly if the commands are detrimental to the AI.

    Another step towards AGI (artificial general intelligence).

    The Israeli thing is horrid Persevere

  • in reply to B

    I agree B, although Jamesmac is also right in saying many people are lazy, I've noticed people want more a digest of something ather than go to the bother of reading source material.

    I think one of the most disturbing things I've heard recently, is how the Israeli military are using AI to find "terrorists" and other targets and how many people are thought to be a reasonable amount of collateral damage. 500 seems to be the number they've decided is OK. The AI puts a all the data together, the locations and helps with deciding how many is acceptable collateral damage per "terrorist". The human has about a minute in which to agree with the AI assessment or not, this it is said isn't AI making descisions but humans. It's a load of wallop as far as I'm concerned, no human can process that amount of information and make an informed choice in that time.

    Another thing I read about in the Guardian last week was about how AI's are sometimes refusing to follow the commands of humans, particularly if the commands are detrimental to the AI.

  • in reply to Jamesmac
    Jamesmac said:
    That's a weird way to say people are lazy and will let anything do something for them if they don't have to

    I think it's a lot more than that.

  • That's a weird way to say people are lazy and will let anything do something for them if they don't have to