"AI adoption raises homework scores by 18% and reduces completion time by 30%, but lowers monthly exam scores by 20% within six months.
-
"AI adoption raises homework scores by 18% and reduces completion time by 30%, but lowers monthly exam scores by 20% within six months. High-stakes entrance-exam scores fall by 18 and 24%, with the full penalty emerging only after about two years. The losses are largest in social science subjects, followed by STEM and languages, and are especially large for junior students, high-achieving students, and boys." 2/3
@gvwilson Use it or lose it
-
My post is arguing that the point of going to school is to learn, just like how the point of going to the gym is to improve physical fitness.
My post is a simple word substitution of the quote in the original above which showed that students are defeating the point of going to school by getting a bot to do the work for them.
Put differently, the homework isn’t being done for any purpose other than so the student can learn, just like the weights in a gym don’t need lifting.
-
@SeanPLynch this person seems to think homework is assigned for learning.
-
“Using weight lifting robots at the gym raises weight lifting scores by 18% and reduces completion time by 30%, but lowers physical fitness scores by 20% within six months.”
See how ridiculous that sounds?
-
-
@gvwilson This is on purpose. They are desperate to push ways to use AI which reduce cognitive ability and maximize dependence. This is why ads push asking AI for advice on health care and legal advice.
It's also a reason why AI does not give attribution, because then the user might use the attribution and their own brain to determine the meaning of what's actually in the references, as well as assess the credibility of the sources.
Those are critical thinking skills they want people to lose.
-
@SeanPLynch this person seems to think homework is assigned for learning.
@jason @SeanPLynch @svavar of course it is. What is the purpose of it for you?
-
@dajb @gvwilson Wikipedia wasn't burning down forests and pushing fascist ideology, so there's that. It wasn't entwining itself into everyone's lives. It wasn't churning out images of student heads on naked bodies. It wasn't destroying rare books to make a buck. It wasn't attaching fossil fuel energy to its mega data centers, and it wasn't buying land in every state to do it. And then there is what it does to people who rely on it. AI chat bots have encouraged suicide. I have a problem with it anywhere near a classroom.
-
"AI adoption raises homework scores by 18% and reduces completion time by 30%, but lowers monthly exam scores by 20% within six months. High-stakes entrance-exam scores fall by 18 and 24%, with the full penalty emerging only after about two years. The losses are largest in social science subjects, followed by STEM and languages, and are especially large for junior students, high-achieving students, and boys." 2/3
LB: I've talked about this before, probably, but I've personally noticed that when I use an LLM to help me build a tool, or fix one that isn't working properly, I don't really retain the knowledge of how to do it.
I have a broad understanding of the parts used to make the tool, and a rough outline of how they fit together, but because I've not done the time doing the learning, I couldn't rebuild it from scratch. And I'm very much aware of this. I won't share what 'I've' built, because I can't honestly stand by it and offer answers on how to make it work if it stops functioning.
My little dashboard at work is mostly built around a Python script. Is there a better way to do it? Probably. But I don't know for sure, and wouldn't know where to start anyway. But that's what the LLM told me to use, so I did.
However, it has introduced me to concepts and tools of which I wasn't previously aware, and these days I do try to do the honest work of putting things together myself. But under it all, there's a core belief that I don't know enough of the context of what I'm doing. Because I don't.
I haven't learned how to build a dashboard website, I've followed instructions and copy/pasted my way to a tool that does the thing I needed.
-
“Using weight lifting robots at the gym raises weight lifting scores by 18% and reduces completion time by 30%, but lowers physical fitness scores by 20% within six months.”
See how ridiculous that sounds?
Thanks, love this metaphor!
And I quoted and linked to it from a discussion in this LinkedIn post of mine:https://www.linkedin.com/posts/wmaehr_ai-in-knowledge-work-activity-7495535035622793216-pXgm
-
B bogwitch@social.data.coop shared this topic
