you are viewing a single comment's thread
view the rest of the comments
[–] 55 points 8 months ago* (last edited 8 months ago) (21 children)

In my experience, people who use LLMs as educational tools... don't actually learn very well. They think they are, but they don't retain the knowledge nor do they seem able to infer from or apply the knowledge very well. There are even some early studies that are showing that using LLMs decreases cognitive ability, and considering how many kids and young people are using it to get their way though school and even higher education... I think we're using AI to raise a generation of stunted minds. That's going to be a bigger issue as time goes on and with the state of the world and who owns the LLMs... it looks like a grim, sad future thanks to this tech.

  • source
  • hideshow 21 child comments
  • [+] [S] -13 points 8 months ago (20 children)
  • [–] 26 points 8 months ago* (11 children)

    Here ya go.

    I'll give you a quick summary. It turns out spending time thinking improves your ability to think and vice versa. When you rely on LLMs to do your thinking for you, you become less skilled at thinking.

    It's important to remember that it doesn't really matter how you, personally, use their product or think it should be used, it matters how it is used by large swaths of society. Don't get fooled into promoting some billionaire's tool to shift wealth further upward and further denigrate the working class in your quest to get out of spending 15 minutes searching for the right D&D character picture.

  • source
  • parent
  • hideshow 11 child comments
  • [+] [S] -4 points 8 months ago (10 children)
  • [–] 5 points 8 months ago (3 children)

    The study you linked doesn’t just show a positive impact on education. That’s only half the study. The other half is about the negative impacts. That study is giving a full picture of ai use in the classroom, about where it helps and where it hurts. They created 6 categories for how ai is getting used in the classroom and explained the positive and negatives found in those studies for each category. Some categories see more benefits or more harm than others

  • source
  • parent
  • hideshow 3 child comments
  • [+] [S] -3 points 8 months ago (2 children)
  • [–] 6 points 8 months ago (1 child)

    The key issue here is how it’s being used and regulation. Ai has caused a lot of harm bc it is unregulated. Like, people have committed suicide or killed others bc of conversations with chatbots. And yes, in many of these cases there are pre-existing mental health concerns, but it’s still causing someone who is unhealthy but non violent to become violent. That’s really bad.

    Currently, ai not being used in positive ways, when we’re looking at the broader use of ai. Sure, some people or small organizations may use ai for specific uses which it’s good for, but that’s not how it is for most ai use. A lot of ai is taking people’s jobs/promising employers that they will be able to fire half their workforce. Even in the positive example you gave of getting a character portrait, sure you could use an ai, but there are a lot of artists that are losing commissions because it’s cheaper for people to just use ai. So the artists aren’t technically losing their job; their work is just being devalued, which is very unfortunate bc—as people have said—ai generated images don’t have intention and care put into it. Ai literally can’t do that. True art, no matter the skill level it’s made at, is made to evoke emotions, to communicate something to the viewer/reader/audience. Ai can’t create true art bc it cannot think or feel; it cannot be deliberate

    I hate ai bc currently ai is a horrible thing bc of how it’s being used the majority of the time. I think after all the ai hype has died down and companies look at how these ai tools can actually be used effectively then it will be more tolerable. But right now it’s just an investment bubble and an unregulated technology that has caused severe harm

  • source
  • parent
  • hideshow 1 child comment
  • load more comments (1 reply)
  • [–] 3 points 8 months ago (5 children)

    I think it's genuinely too early to say definitively one way or another, but anecdotally, the people I see educated with LLMs are a lot less capable than people who aren't.

    Only time will tell how badly we've allowed ourselves to get fucked over to benefit a few ultra rich companies.

  • source
  • parent
  • hideshow 5 child comments
  • [+] [S] -3 points 8 months ago (4 children)
  • [–] 3 points 8 months ago* (last edited 8 months ago) (3 children)

    I've watched people get stupider in the past few years using LLMs to replace thinking for themselves, so, yeah I guess so.

    You can say it's just a problem with capitalism, and, yes, capitalism is a problem, but AI is the current tool de jure that capitalists are using to crush workers, so, yes, I oppose it.

    It's pointless to argue about how it could be used in a post capitalist world, we don't live in that world, we live in a world where LLMs are actively being used to harm the working class. This applies to a myriad of things, of course.

    You want to use and develop AI tools for use in a post-capitalist world? Alright, let's get to the post capitalist part first. Until then, I'll oppose tools used to harm the underprivileged.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 14 points 8 months ago* (6 children)

    Here's another fun piece you can read.

    The incoming AI apocalypse isn't about Skynet drones or malicious AGI, it's about creating generations of vastly less educated and cognitively deficient lower classes and restricting traditional education to the wealthier echelons of society, gatekeeping the poor out by cost alone, undoing decades of hard-fought socialist efforts to bring education to the masses.

  • source
  • parent
  • hideshow 6 child comments
  • [+] [S] -1 points 8 months ago (5 children)
  • [–] 11 points 8 months ago (2 children)

    That's the thing.

    Your specific opinions on how LLMs should be used or not use don't affect how they are being used.

    A system is what it does, plain and simple, and LLMs look like they're doing serious damage to our societies to really just benefit the wealthy.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 4 points 8 months ago (1 child)

    I fully believe that the US government plans on replacing teachers with AI. It is all part of a grand plan to eliminate the Department of Education and defund schools nationwide. Once this crisis comes into full view we will be presented with this plan and we will no longer have a choice unless you have enough money to send your kid to private school.

  • source
  • parent
  • hideshow 1 child comment
  • load more comments (1 reply)
  • [–] 14 points 8 months ago

    , I don’t think it can hurt any more than a calculator hurts your ability to do math.

    Because using AI atrophies the part of your brain that handles critical thinking...

    The more you use it, the less you notice how you can't do things without it.

    If AI worked, that would be normal. The problem is it's just good at conning people into believing it.

    That's why you can't realize if it ever takes off and people start using it, they're going to make it shittier and more expensive.

    But again, the people already relying on AI have lost the critical thinking to see that coming. It's like a bus driver closing their eyes because a bridge is closed. The bridge is still closed, they didn't solve any problems. They just don't see it coming now.

    What you're doing is asking all the passengers why they're still screaming if all they need to do is close their eyes...

  • source
  • parent