I think the obvious counter to an influx in bug reports / pull requests is to have the first layer of review of bug reports / pull requests to be AI as well.

Final remains humans.

I dunno, if that's how people vote, that's how it should be.

Though there is a big question of how they plan to run a vote with multiple options. I hope it's a Condorcet system to avoid the risk of watering down one side with similar competing options...

Oh, but surely they'll win, they indicted [strike]a ham sandwich[/strike] this man! And the Trump administration has such a spectacular record with high-profile prosecutions thusfar...

He's got a valid point. For example, if you use wolfram alpha after you do a math problem to check the result you came up with you can actually learn MORE than just doing the problem yourself, and may even add to the time the work takes, as if your result is wrong, you'll spend extra time working out where you went wrong and fixing it. And ultimately learning more in the process.

OTOH - If you use wolfram alpha to DO the question for you, you will learn nothing at all.

But further to that, you can even use tools to save time and learn more.

The other day, my daughter and I chatted about docker build scripts while she was doing her homework (which used docker containers), and that conversation clarified a lot of details and subtleties. (I learned a few things in the course of that conversation myself.) Nothing we talked about wasn't in the documentation, strictly speaking, but the conversation was a more natural way to learn and address the specific details we were talking about, with examples, and use cases that we came up with on the flay to make sense of and highlight those specific nuances. She probably could have read and reread the docs and easily spent more time on it, while learning less.

Now, in this case, she was learning 'from me', but you could swap "Dad" out for an "LLM", and get the same benefit.

All that said, I don't think the AI systems we have right now are likely a good idea for homework. The temptation to use it self-destructively is just going to be too high for a large number of students. Probably the majority of them. But it definitely can be used to learn more, and even learn more faster in the hands of someone who uses it for that.

 

That is not what they found. What they found is that if you spend 60 minutes doing the homework with AI, you learn approximately as much as someone who spends 60 minutes doing it without AI. If you spend 20 minutes doing it with AI, you learn as much as someone who spends 20 minutes doing it without AI. Etc. The only real difference is that the AI people get way better homework scores, and as a result, tend to stop spending as much time, and thus tend to learn less.

That's what this preprint study found.

100% this. Even a short 5-minute quiz with just a couple questions at the start of each day will do the trick. But you simply can't rely on homework scores to be a proxy of learning or time spent studying anymore. Brief daily quizzes on the previous day's material will force them to.

Also, It's not like students have the excuse of, "I couldn't study, I didn't have anyone to explain X to me" anymore. The very AI that they might otherwise use to cheat on their homework can also explain to them the things they missed in last year's algebra class that are blocking them from learning this year's calculus, or whatnot. Congrats - you now have an infinitely patient personal tutor who will spend hours dumbing things down for you until you get it. Now you have no excuse - learn the material. It becomes all about actually putting the time in.

There's a lot wrong with your question.
Firstly, China's gun laws are among the strictest in the world. Private ownership of firearms is generally prohibited. You'll get better returns on matching China's homicide rates by matching their firearms laws than you will by catching up with them on surveillance.

Are you wiling to do that? I somehow doubt it. And to be a safe as china, you could do both -- give up your guns AND submit to a police surveillance state. What could go wrong?

China has done other things too - absolutely massive government investments in infrastructure and construction, creating jobs and reducing poverty in the process, uplifting people who would otherwise be more desperate and more likely to turn to crime etc.

And speaking of Crime, China is pretty busy with state executions too. And is quite adept at artificially making those numbers look smaller, frequently recording the execution of multiple people as a single execution. That may play a part in the lower homicide rate too.

Secondly, you imply the tradeoff is simple prudishness - we give up peoples lives for the ability to "go to a strip club secretly".
That's not that half of it, and you should know that. It's also the ability to go, for example, to a political party meeting secretly.

Do you really trust your government so much that you'd give them that power over you? That some fuckwit won't get into power and leverage this against the people? The best protection the civilian population has from government abuse is to resist giving them the tools to abuse you in the first place. The government doesn't need to know where every person is at all times, this will inevitably be used against you. The supposed 'safety' gains are not worth living in a police surveillance state.

Yeah. There's been widespread misrepresentation of this (preprint!) in the media. What they actually found: test scores correspond to how much time you spend doing homework, regardless of whether you use AI or not. Which should not be a surprising result! Of course, the AI people, esp. those who had been using AI for more than a few months, stopped spending much time on their homework, and then their test scores fell to match the non-AI users who didn't spend much time in their homework either.

The TL/DR is, in today's age, you cannot use test scores as a proxy of how much time a user is studying or knows the material.

There are of course some caveats, beyond the fact that it's just a preprint - notably that there's nothing to separate why the user started using AI to do their homework, so there's potential confounders in there. For example, if a student is in a poor family, maybe they had to get a job to help pay the bills. But working a job itself can lower test scores - for example, having to work late the night before an exam. So there are some confounder problems with the study. That said, I'm sure the general conclusion is correct: time spent doing the thing = amount of learning the thing (relatively independent of what tools you're using in the process).

I had to add minimum paragraph and sentence counts to the prompts it uses to keep it from trying to write a few terse sentences each time.

I noticed this in Claude (from DDG). It invariably tries to shorten sentences or make them more compact and concise. Sometimes that can be good, but most times long-winded sentences are what are needed.

and there is no way I can find to make it stop popping up giant intrusive ads for their other services every single time I open it.

Just like Firefox. There is no easy way to turn off being harassed by an intrusive pop-up telling me about a new update.

There used to be. It was a simple checkbox. But somewhere along the way someone had to justify their existence and removed that simple checkbox in favor of harassment.

Laughs in Californian.

Would like to add one tidbit because I've had this same conversation at work. The professor stated they used AI to help edit only after hundred of hours of writing. In other words, they did the heavy lifting and only used AI at the end to help refine what they had already done.

Is this any different than a human editor? If I write a book, do what I can to edit it down, then go to an editor who makes suggests, removes or reorders parts, and so on, is that any different than having AI do the same thing? In both cases you wrote the words. In both cases you went somewhere else to help edit.

If people claim that using AI to edit a work is "cheating" because it's no longer your work, then the same should apply to a human editor.

I feel like I'm in a constant race to flee enshittification : Used to use Cursor (it's still my only option at work), but then Musk bought it, and they instantly went in on hard-sell tactics. Switched to vscode + copilot + OpenRouter, but was getting concerned about API costs. Just yesterday I set up a Tencent TokenHub account with their Max package, which required that I switch off copilot if I wanted to be able to use some models like Kimi K3 (had a temperature=1.0 requirement, copilot doesn't let you set temperatures and hardcodes them to very low), so now it's vscode + Pi + TokenHub.

Sigh. Please don't make me move again...

(Not that Tencent is in any way "the good guys"... but they're also providing a very generic product)

Ah, the good 'ol "No True Communism" argument :)

"It suggests" if you ignore that their confounder control is terrible. First, sugar wasn't the only restricted food in post-war Britain - there was also rationing of eggs, butter, cheese and meats that ended during the same timeperiod. So you're not actually testing the effects of sugar consumption, now are you? Also, the early 1950s was a period of rapid change in the US, including the maturation and expansion of the NHS (founded in 1948), the introduction of new childhood vaccines, major environmental shifts (for example, the Great Smog of London in 1952 vs. the 1956 Clean Air Act). Saying "all changes = sugar" just based on birthdate seems absurd.

It's also worth mentioning that there's also survivorship bias - anyone who died before the study couldn't be enrolled, and there's also no actual linking to any maternal / childhood dietary logs. Also, the ratios of cancer differences they're describing aren't small, they're massive. Are we to attribute a lung cancer difference on the order of smoking vs. nonsmoking households as being to sugar, or due to, you know, smoking (or pollution) trends not accounted for in their cohort?