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

Trying to teach in the age of the AI homework machine

solarshades.club

Post reply on HN
Print view View original

561–570 of 672 posts

Re: Trying to teach in the age of the AI homework machine

#561

Post by yazantapuz »

squigz wrote: post #472
xtracto wrote: post #470

Earlier quoted context omitted.

Back when I was doing my BSc in Software Engineering, we had a teacher who did her Data Structure and Algorithms exams with pen and paper. On one of them, she basically wrote 4 coding problems (which would be solved in 4 short ~30 LOC). We had to write the answer with pen and paper, writing the whole program in C. And the teacher would score it by transcribing the verbatim text in her computer, and if it had one sing…

Yeah, this is absurd. And if you have poor handwriting, the chances of "syntax errors" goes up. My above comment is getting downvoted, and it's honestly a bit baffling. I'd be furious if I were paying tens of thousands of dollars to receive a university-level education in software engineering in 2025... and I had to write programs with pen and paper. It is so far detached from the reality of, not only the industry, b…

No, you don't write code by hand. Maybe pseudo-code, analize some given code or you have to specify the general architecture for a system. But in other courses, for example operating systems, networks, distributed systems you have to answer questions like "when udp is the right choice over tcp?", "what kind of problems are associated with pagination?", "what are vector clocks?", etc., using pen and paper.

Re: Trying to teach in the age of the AI homework machine

#562

Post by armchairhacker »

RHSeeger wrote: post #540
armchairhacker wrote: post #423

Earlier quoted context omitted.

My take: - Make “homework” ungraded. Many college classes already do this, and it has been easy to cheat on it way before AI by sharing solutions. Knowledge is better measured in exams and competence in projects. My understanding is that homework is essentially just practice for exams, and it’s only graded so students don’t skip it then fail the exams; but presumably now students cheat on it then fail exams, and for…

> My understanding is that homework is essentially just practice for exams There are a LOT of people that don't take exams well. When you combine that with the fact that the real world doesn't work like exams in 90% of cases, it makes a lot of sense for grades to _not_ based on exams (as much as possible). Going the other direction (based on nothing _but_ exams) is going to be very painful to a lot of people; people…

I made another comment on this thread about that. Exams should be test important knowledge (not computation or trick questions) so they should be easy for students who learned the material, even those who traditionally have trouble with exams. Most of the grade should be frequent in-class assignments or long take-home projects, which test almost if not the same skills students would use professionally (e.g. debug a simulated server failure in-class; develop a small filesystem with a novel feature at home).

The in-class assignments should also be easier than the take-home projects (although not as easy as the exams). In-class assignments and exams would be more common in earlier classes, and long projects would be more common in later classes.

Re: Trying to teach in the age of the AI homework machine

#563

Post by rwyinuse »

intended wrote: post #389
CalRobert wrote: post #329

Earlier quoted context omitted.

Then there's Europe, where making it easy to get a master's degree just let to jobs requiring people to waste time getting yet another unneeded degree.

This entire situation is something that is predictable, and I have personally called it out years ago - not because of some unique ability, but because this is what happened in India and China decades upon decades ago. There’s only so many jobs which have you a good salary. So everyone had to become a doctor lawyer or engineer. Business degrees were seen as washouts. Even for the job of a peon, you had to be educated…

Yep, it's a job market problem. Only degrees that are somehow limited in their supply will continue to hold value, the rest approach worthlessness. Neither the state nor universities have any interest to limit the supply.

In my country doctors earn huge salaries and have 100% job security, because their powerful interest groups have successfully lobbied to limit the number of grads below job market's demand. Other degrees don't come even close.

Re: Trying to teach in the age of the AI homework machine

#564

Post by Balgair »

rwyinuse wrote: post #560
Balgair wrote: post #455

Earlier quoted context omitted.

I think you've identified the main issue here: LLMs aren't destroying the University or the essay. LLMs are destroying the cheap University or essay. Cheap can mean a lot of things, like money or time or distance. But, if Universities want to maintain a standard, then they are going to have to work for it again. No more 300+ person freshman lectures (where everyone cheated anyways). No more take-home zoom exams. No m…

Over here in Finland, higher education is state funded, and the funding is allocated to universities mostly based on how many degrees they churn out yearly. Whether the grads actually find employment or know anything is irrelevant. So, it's pretty hard for universities over here to maintain standards in this GenAI world, when the paying customer only cares about quantity, and not quality. I'm feeling bad for the stud…

Gosh, I'm so myopic here. I'm mostly talking about US based systems.

But, of course, LLMs are affecting the whole world.

Yeah, I'd love to hear more about how other countries are affected by this tool. For Finland, I'd imagine that the feedback loop is the voters, but that's a bit too long and the incentives and desires of the voting public get a bit too condensed into a few choice to matter [0].

What are you seeing out there as to how students feel about LLMs?

[0] funnily enough, like how the nodes in the neural net of an LLM get too saturated if they don't have enough parameters.

Re: Trying to teach in the age of the AI homework machine

#565

Post by amelius »

math_dandy wrote: post #28

I teach math at a large university (30,000 students) and have also gone “back to the earth”, to pen-and-paper, proctored and exams. Students don’t seem to mind this reversion. The administration, however, doesn’t like this trend. They want all evaluation to be remote-friendly, so that the same course with the same evaluations can be given to students learning in person or enrolled online. Online enrollment is a huge…

Can't we use AI to monitor the students?

Re: Trying to teach in the age of the AI homework machine

#566

Post by danhodgins »

Fight fire with fire.

Use AI to determine potential essay topics that are as close to 'AI-proof' as possible.

Here is an example prompt:

"Describe examples of possible high school essay topics where students cannot use AI engines such as perplexity or ChatGPT to help complete the assignment. In other words - AI-proof topics, assignments or projects"

Re: Trying to teach in the age of the AI homework machine

#567

Post by hollandheese »

Brybry wrote: post #256
BrenBarn wrote: post #232

Earlier quoted context omitted.

I see that pressure as well. I find that a lot of the problems we have with AI are in fact AI exposing problems in other aspects of our society. In this case, one problem is that the people who do the teaching and know what needs to be learned are the faculty, but the decisions about how to teach are made by administrators. And another problem is that colleges are treating "make money" as a goal. These problems exist…

Instructors and professors are required to be subject matter experts but many are not required to have a teaching certification or education-related degree. So they know what students should be taught but I don't know that they necessarily know how any better than the administrators. I've always found it weird that you need teaching certification to teach basic concepts to kindergartners but not to teach calculus to…

>don't know that they necessarily know how any better than the administrators.

If someone is doing something day in and day out, they do gain knowledge on what works and doesn't work. So just by doing that the professors typically know much more about how people should be taught than the administrators. Further, the administrators' incentives are not aligned towards insuring proper instruction. They are aligned with increasing student enrollment and then cashing out whenever they personally can.

Re: Trying to teach in the age of the AI homework machine

#568

Post by greenavocado »

bongodongobob wrote: post #172
eru wrote: post #167

Earlier quoted context omitted.

> I don't see why this is so hard, other than the usual intergenerational whining / a heaping pile of student entitlement. You know that grading paper exams is a lot more hassle _for the teachers_? Your overall point might or might not still stand. I'm just responding to your 'I don't see why this is so hard'. Show some imagination for why other people hold their positions. (I'm sure there's lots of other factors tha…

Why can't the teachers use LLMs to grade?

Ah, the eternal dream of offloading all human labor to machines. Why can't teachers just let an LLM grade? Because, of course, nothing says "educational integrity" like a glorified autocomplete deciding whether little Timmy's essay on Shakespeare adequately captures the existential dread of Hamlet. Sure, let's trust a model that hallucinates citations. But fine, if we're really committed to stripping all nuance from education, why stop there? Let's just plug students into Anki's FSRS algorithm and call it a day. Just assign grades based on how fast their retention decays, because nothing says "holistic assessment" like reducing a human being to a set of coefficients in a spaced repetition formula. Never mind that actual learning involves things like critical thinking or, heaven forbid, creativity. No, no, we'll just reduce the entire process to a forgetting curve. Because nothing inspires a love of knowledge like treating human minds as poorly optimized flashcard decks, mechanically processed and discarded the moment their retention scores dip below acceptable thresholds.

Re: Trying to teach in the age of the AI homework machine

#569

Post by johnea »

pugio wrote: post #182
johnea wrote: post #58

One of the most offensive words in the anthropomophization of LLMs is: hallucinate. It's not only an anthropomorphism, it's also a euphemism. A correct interpretation of the word would imply that the LLM has some fantastical vision that it mistakes for reality. What utter bullsh1t. Let's just use the correct word for this type of output: wrong. When the LLM generates a sequence of words, that may or may not be gramma…

When a person hallucinates a dragon coming for them, they are wrong, but we still use a different word to more precisely indicate the class of error. Not all llm errors are hallucinations - if an llm tells me that 3 + 5 is 7, It's just wrong. If it tells me that the source for 3 + 5 being 7 is a seminal paper entitled "On the relative accuracy of summing numbers to a region +-1 from the fourth prime", we would call t…

I think your defense of reusing terms for new phenomenon is fair.

But in this specific case, I would say the reuse of this particular word, to apply to this particular error, is still incorrect.

A person hallucinating is based on a many leveled experience of consciousness.

The LLM has nothing of the sort.

It doesn't have a hierarchy of knowledge which it is sorting to determine what is correct and what is not. It doesn't have a "world view" based on a lifetime of that knowledge sorting.

In fact, it doesn't have any representation of knowledge at all. Much less a concept of whether that knowledge is correct or not.

What it has is a model of what words came in what order, in the training set on which it was "trained" (another, and somewhat more accurate, anthropomorphism).

So without anything resembling conscious thought, it's not possible for an LLM to do anything even slightly resembling human hallucination.

As such, when the text generated by an LLM is not factually correct, it's not an hallucination, it's just wrong.

Re: Trying to teach in the age of the AI homework machine

#570

Post by lowbloodsugar »

sshine wrote: post #69

I teach computer science / programming, and I don't know what a good AI policy is. On the one hand, I use AI extensively for my own learning, and it's helping me a lot. On the other hand, it gets work done quickly and poorly. Students mistake mandatory assignments for something they have to overcome as effortlessly as possible. Once they're past this hurdle, they can mind their own business again. To them, AI is not…

Programming with AI is the job now. That’s what you need to be teaching if you want your graduates to get a job programming.

What’s changed is that “some working code” is no longer proof that a student understands the material.

You’re going to need a new way to identify students that understand the material.

Post reply on HN