Live data from Hacker News

AI boosted homework scores, then exam scores dropped: study

economist.com

261–270 of 409 posts

Re: AI boosted homework scores, then exam scores dropped: study

#261

Earlier quoted context omitted.

Stand-alone bottom-up observations (even if valid) obscure the fundamental problem that AI (or "AI" if you like) is inducing systemic change across human civilization. If cohorts are graduating into the workforce with decreasing stand-alone skills, that implies the economic value-add of labour is shifting to AI, which implies broadly falling standards of living, falling political power, and an ongoing transfer of pow…

> Stand-alone bottom-up observations (even if valid) obscure the fundamental problem that AI (or "AI" if you like) is inducing systemic change across human civilization. Students have been cheating and taking shortcuts on their work long before the current generation of AI tools existed (and probably for as long as graded exams have existed). Somehow, civilization has managed to survive.

And civilizations do fall. The bronze age civilizations did collapse. There was something dark about the dark ages. There is no law of nature which says that things will be okay, and plenty of examples where things have fallen apart.

Just to add to this, if AI is as epochal a change as its boosters say (and I am open to that argument), then surely it must have the potential for awful consequences, in proportion to its potential for good.

Re: AI boosted homework scores, then exam scores dropped: study

#263
post #131
post #13

In my professional career, my employers give me a lot more assignments than exams.

Your employers want those assignments done. That’s why they pay you. School assignments belong in the trash can and you pay for someone to look at them before throwing them away.

As an example: my employer doesn't require that I learn how to communicate well. It requires that I already learned to do so.

Re: AI boosted homework scores, then exam scores dropped: study

#264

Earlier quoted context omitted.

All at the cost of ... {List of negatives regarding the construction and powering of AI data centers }

If you added up the negative effect of every single "ai data center" you'd end up with less than a single coal mine. It's good to think about second order impacts but this is not useful.

Citation needed,

Just one 20 GW AI data center which is being built with gas powered turbine generator power plant, will be the largest fossil fuel power plant in the world... How exactly is that the same as one coal mine

Re: AI boosted homework scores, then exam scores dropped: study

#265

Earlier quoted context omitted.

All at the cost of ... {List of negatives regarding the construction and powering of AI data centers }

The social costs are worse.

I dunno, concentrating the environmental impact of 20GW of gas based power generation would be the equivalent of 6-9 Million cars running... All in the space of a corner of a small city

Re: AI boosted homework scores, then exam scores dropped: study

#266
post #26

[flagged]

The important thing is that I could learn to use a flint and steel pretty easily if I wanted to. I don't need to carry out that particular task, but my ability to adapt to new problems means I can carry out more relevant tasks when needed. The most useful part of school and university for me wasn't that I learned the difference between a cirrocumulus and a cirrostratus cloud in geography class. It's that I learned ho…

And imagine the staggering power that already gives OpenAI et al. Hundreds of millions of people _doing whatever they suggest_.

Re: AI boosted homework scores, then exam scores dropped: study

#267
post #254

Earlier quoted context omitted.

I'm way beyond homework, but I do appreciate the compliment on seeming young! I meant: what's going to stop the average lazy student from ignoring whatever harness the university recommends and instead use an unrestricted LLM, thus learning nothing?

> what's going to stop the average lazy student from ignoring whatever harness the university recommends and instead use an unrestricted LLM, thus learning nothing? Nothing until they start failing exams and maybe seeing the error of their ways. My suggestions wasn't for you, it was for the smart student who wants to use AI to enhance their learning but not have it do all the work. But cheating isn't new. People have…

> But cheating isn't new. People have been cheating in school for centuries. It's easier now, but the consequences were always the same. At some point the chickens come home to roost and you pay the piper.

My concern is that with LLMs and above average bullshitting abilities, the chickens might never come home to roost.

Re: AI boosted homework scores, then exam scores dropped: study

#268

Earlier quoted context omitted.

The social costs are worse.

I dunno, concentrating the environmental impact of 20GW of gas based power generation would be the equivalent of 6-9 Million cars running... All in the space of a corner of a small city

That sounds much worse than I had expected. Is this enough consumption to significantly affect the local price of gasoline?

Edit: Why don't they have lower-emission power sources?

Re: AI boosted homework scores, then exam scores dropped: study

#269
post #225
post #157

Earlier quoted context omitted.

It's not so clear that AI is an amplifier.. The paper has some fascinating analysis on this topic: "At the other end of the distribution, AI students who spend more than 65 minutes on their homework receive homework and exam scores similar to those of non-AI students, suggesting that these students do not use generative AI for homework assignments. However, this group consists entirely of students who adopted generat…

Depends who's using it. Like many tools the force multiplier depends on the operator. I'm confident it's an amplifier for people who know how learning works and already do a lot of it, successfully. However the level of "learning fluency" I'm talking about isn't reached for many until late college or grad school, and sometimes not at all. So I'm not surprised by the quoted results for 12-18 year olds.

> Depends who's using it. Like many tools the force multiplier depends on the operator.

How would you prove/disprove this assumption without falling into a True Scotsman fallacy?

Post reply on HN