Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How University Students Use Claude

anthropic.com

Post reply on HN
Print view View original

451–460 of 601 posts

Re: How University Students Use Claude

#451

Post by saagarjha »

alwa wrote: post #283
jplusequalt wrote: post #184

Earlier quoted context omitted.

I think more and more workers are warming up to unions. As wages in software continue to be oppressed, I think we'll see an increase in unionization efforts for software engineers.

Software engineer wages? Oppressed ? This is work that averages well over 6 figures—for a single worker, for desk work—in the US? https://www.indeed.com/career/software-engineer/salaries https://www.levels.fyi/t/software-engineer/locations/united-...

Yes, oppressed: https://en.wikipedia.org/wiki/High-Tech_Employee_Antitrust_L...

Re: How University Students Use Claude

#452

Post by johnisgood »

palmotea wrote: post #201
hackyhacky wrote: post #107

Earlier quoted context omitted.

> IMO it's quite different than using a calculator or any other tool. It can currently completely replace the human in the loop, whereas with other tools they are generally just a step in the process. The (as yet unproven) argument for the use of AIs is that using AI to solve simpler problems allows us humans to focus on the big picture, in the same way that letting a calculator solve arithmetic gives us flexibility…

> The (as yet unproven) argument for the use of AIs is that using AI to solve simpler problems allows us humans to focus on the big picture, in the same way that letting a calculator solve arithmetic gives us flexibility to understand the math behind the arithmetic. And I can tell you from experience that "letting a calculator solve arithmetic" (or more accurately, being dependent on a calculator to solve arithmetic)…

I know how to do arithmetic, but I still use my PC or a calculator because I am not entirely sure that I am accurate. I use "units" as well extensively, it can be used for much more than just unit conversion. You can do complex calculations with it.

You can solve stuff like:

> If you walk 1 mile in 7 minutes, how fast are you walking in kilometers per hour?

  $ units -t "1 mile / 7 minutes" "kilometers per hour"
  13.7943771428571
You need some basic knowledge to even come up with "1 mile / 7 minutes" and "kilometers per hour".

There are examples where you need much more advanced knowledge, too, meaning it is not enough to just have a calculator, for example in thermodynamics, when dealing with gas laws, you cannot simply convert pressure, volume, and temperature from one unit to another without taking into account the specific context of the law you’re applying (e.g., the ideal gas law or real gas behavior)", or for example you want to convert 1 kilowatt-hour (kWh) to watts (W). This is a case of energy (in kilowatt-hours) over time (in hours), and we need to convert it to power (in watts), which is energy per unit time.

You cannot do:

  $ units -t "1 kWh" "W"
  conformability error
  3600000 kg m^2 / s^2
  1 kg m^2 / s^3
You have to have some knowledge, so you could do:

  $ units -t "1 kWh" "J"
  1 kWh = 3600000 J
  $ units -t "3600000 joules / 3600 seconds" "W"
  3600000 joules / 3600 seconds = 1000 W
To sum it up: in many cases, without the right knowledge, even the most accurate tool will only get you part of the way there.

It applies to LLMs and programming, too, thus, I am not worried. We will still have copy-paste "programmers", and actually knowledgeable ones, as we have always had. The difference is that you can use LLMs to learn, quite a lot, but you cannot use a calculator alone to learn how to convert 1 kWh to W.

Re: How University Students Use Claude

#453

Post by ignoramous »

SamBam wrote: post #9

I feel like Anthropic has an incentive to minimize how much students use LLMs to write their papers for them. In the article, I guess this would be buried in > Students also frequently used Claude to provide technical explanations or solutions for academic assignments (33.5%)—working with AI to debug and fix errors in coding assignments, implement programming algorithms and data structures, and explain or solve mathe…

> feel like Anthropic has an incentive to minimize how much students use LLMs to write their papers for them

You're right.

Quite incredibly, they also do the opposite, in that they hype-up / inflate the capability of their LLMs. For instance, they've categorised "summarisation" as "high-order thinking" ("Create", per Bloom's Taxonomy). It patently isn't. Comical they'd not only think so, but also publicly blog about it.

Re: How University Students Use Claude

#454

Post by lr4444lr »

samuel wrote: post #349
enjo wrote: post #297

Earlier quoted context omitted.

> it's literally "let me paste the assignment into ChatGPT and see what it spits out, change a few words and submit that". My wife is an accounting professor. For many years her battle was with students using Chegg and the like. They would submit roughly correct answers but because she would rotate the underlying numbers they would always be wrong in a provably cheating way. This made up 5-8% of her students. Now she…

I guess this students don't pass, do they? I don't think that's a particularly hard concern. It will take a bit more, but will learn the lesson (or drop out). I'm more worried about those who will learn to solve the problems with the help of an LLM, but can't do anything without one. Those will go under the radar, unnoticed, and the problem is, how bad is it, actually? I would say that a lot, but then I realize I'm p…

How many people are "good drivers" outside their home town? I am not that old, but old enough to remember all adults taking wrong turns trying to find new destinations for the first time.

Re: How University Students Use Claude

#455

Post by sebstefan »

>Students also frequently used Claude to provide technical explanations or solutions for academic assignments (33.5%

The only thing I care about is the ratio between those two things and you decide to group them together in your report? Fuck that

Re: How University Students Use Claude

#456

Post by juped »

brunocroh wrote: post #439
brunocroh wrote: post #438

Earlier quoted context omitted.

English is not my first language. To me, 'AD' was a shorter way to say 'advertisement' (a really hard word to remember how to spell btw) Is that wrong?

I get what you're saying now. I can write it in lowercase, right? It's just that I see people writing it that way, so I end up repeating their behavior without even realizing it.

I'm actually just curious what people who write that think the "D" stands for! (the "A" presumably being "advertisement")

Re: How University Students Use Claude

#457

Post by NiloCK »

lelanthran wrote: post #415
chii wrote: post #327

Earlier quoted context omitted.

> would fail the quizzes. not those who did actually do the work, and learnt. The change ought to be that students are allowed to be failed, and this should be a form of punishment for those who "cheat".

Aren't students already allowed to fail? As a comment upthread said, let them cheat on the take home as much as they want to, they're still going to fail the exam.

> Aren't students already allowed to fail?

It's technically allowed on an individual basis, but the economics don't work for any institution to attempt to raise its bar.

If institutions X and Y grant credential Z, and X starts failing a third of its students, who would apply to go there?

Re: How University Students Use Claude

#458

Post by lordnacho »

chii wrote: post #420
_tom_ wrote: post #321

No one seems to be talking about the fact that we need to change the definition of cheating. People's careers are going to be filled with AI. College needs to prepare them for that reality, not to get jobs that are now extinct. If they are never going to have to program without AI, what's the point in teaching them to do it? It's like expecting them to do arithmetic by hand. No one does. For every class, teachers nee…

> It's like expecting them to do arithmetic by hand. No one does. But those who traditionally learnt arithmetics have had this training, which _enables_ higher order thinking. Being reliant on AI to do this means they would not have had that same level of training. It could prevent them from being able to synthesize new patterns or recognize them (and so if the AI also cannot do the same, you get stagnation).

I suspect schools spend a lot less time on arithmetic than they used to, however.

You used to _actually_ need to do the arithmetic, now you just need to understand when a calculator is not giving you what you expected. (Not that this is being taught either, lol)

You can get to the higher order thinking sooner than if you spent years grinding multiplication tables.

Re: How University Students Use Claude

#459

Post by Jedd »

akimbostrawman wrote: post #447
Jedd wrote: post #435

Earlier quoted context omitted.

Given the original, ludicrous, claim was: > Literally all progress we've made is due to ever increasing specialization. Then we don't really need plural examples, right? Anyway - language, wheel, fire, tool-making, social constructs like reciprocity principle - I think gave us some progress as a species and a society.

All of these examples are done by specialist because I don't see many cars being build by dentists. Even in mankind's beginning specialization existed in the form of hunter and gatherer. This specialization in combination with team work brought us to the top of the food chain to a point where we can strive beyond basic survival. The people making space crafts (designing and building, another example of specialization…

> All of these examples are done by specialist because I don't see many cars being build by dentists.

I'm not sure how you get from pre-agricultural humans developing fire, to dentists building cars.

I don't doubt that after fire was 'understood', there was specialisation to some degree, probably, around management of fire, what burns well, how best to cook, etc.

But any claim that fire was the result of specialisation seems a bit hard to substantiate. A committee was established to direct Thag Simmons to develop a way to .. something involving wood?

Wheel, the setting of broken bones, language etc - specialisation happened subsequently, but not as a prerequisite for those advances.

> Even in mankind's beginning specialization existed in the form of hunter and gatherer. This specialization in combination with team work brought us to the top of the food chain to a point where we can strive beyond basic survival.

Totally agree that we advanced because of two key capabilities - a) persistence hunting, b) team / communication.

You seem to be conflating the result of those advancements with "all progress", as was GP.

> The people making space crafts (designing and building, another example of specialization) don't need to know how to repair or build a microwave to heat there for food.

I am not, was not, arguing that highly specialised skills in modern society are not ... highly specialised.

I was arguing against the lofty claim that:

"All progress we've made is due to ever increasing specialization."

Noting the poster of that was responding to a quote from a work of fiction - claiming it was awful - that the author had suggested everyone should be familiar with (among other things) 'changing a diaper, comfort the dying, cooperate, cook a tasty meal, analyse a problem, solve equations' etc.

If you're suggesting that you think some people in society should be exempt from some basic skills like those - that's an interesting position I'd like to see you defend.

> Of course everybody still needs to know basic knowledge (how to turn on microwave) to get by.

FWIW I don't have a microwave oven.

Re: How University Students Use Claude

#460

Post by woodrowbarlow »

enjo wrote: post #297
dtnewman wrote: post #14

> A common question is: “how much are students using AI to cheat?” That’s hard to answer, especially as we don’t know the specific educational context where each of Claude’s responses is being used. I built a popular product that helps teachers with this problem. Yes, it's "hard to answer", but let's be honest... it's a very very widespread problem. I've talked to hundreds of teachers about this and it's a ubiquitous…

> it's literally "let me paste the assignment into ChatGPT and see what it spits out, change a few words and submit that". My wife is an accounting professor. For many years her battle was with students using Chegg and the like. They would submit roughly correct answers but because she would rotate the underlying numbers they would always be wrong in a provably cheating way. This made up 5-8% of her students. Now she…

my partner teaches high school math and regularly gets answers with calculus symbols (none of the students have taken any calculus). these students aren't putting a single iota of thought into the answers they're getting back from these tools.
Post reply on HN