Hacker News Reader

Comment by sky2224 | Hacker News Reader

As a student currently, I'll also throw in this perspective. The colleges themselves make it feel transactional and not about learning even if I'm interested in doing so.

For example, I'm taking a physics course right now (electricity and magnetism). The concepts are difficult for me and I was hoping that the homework would help. So, I go to do the homework, but the homework is online. With the online homework I get five chances to get the problem correct, but there is zero partial credit, zero feedback, and every time I get the answer wrong, it negatively impacts my grade.

I have no chance to make mistakes and learn. At least with homework that was handed out back in the day, there was at least the possibility of partial credit being handed out. So my options are going to office hours (which I try to do), go to tutoring hours (which conflicts with my job's work schedule), or go to ChatGPT and/or Chegg.

Additionally, since students have been cheating, I think it gives professors a skewed perspective on how much time is actually needed to get work done, so the deadlines get moved up. This means I get even more pressure put on me when I'm just trying to learn and be a good student.

View on HN

In the 90s we had the "Plato" system for chemistry. It was a question/answer terminal in the library. Our Chemistry TA advised us to use it to study for exams as it had a lot of sample questions. It was really good because if you got it wrong, it actually gave you a detailed explanation of how to solve it. It was so helpful to have that. When I used the system, I made a bunch of mistakes but ended up learning from them, and it really helped for the exams.

1990, "PLATO reached it's maximum enrollment, with 4,029 course seats and approximately 30 courses and other applications." Plato was decommissioned in 1994.

https://www.umass.edu/it/it-timeline

Honestly as an engineer some of schooling was learning enough just to get by. We always envied the non-engineers who had more freedom to choose classes they were fascinated by.

For me the Masters Degree gives a better chance to dive deep into a single topic.

What I find extremely sad in the whole academic business, is that all the work that is put in creating those tools, curriculum, classes materials etc are just wasted. Systems are decomissioned, professors refuse to share their materials or to even update it when receiving students feedback or when there are new disoveries in the field. And copyright holders are threatening to bite when learning material is put online...

If you want to experience it again, they re-released the final version awhile back; they've got emulators for both the mainframe and terminals.

https://codex.retro1.org/plato:operation.r1:installing

I recently read The Friendly Orange Glow and found it inspirational.

Is the entire PLATO system, including courseware, available? That would be tremendous.

[Edit: it looks like there is a lot of stuff there, so ... maybe??!]

Of course I also want to play Moria[1], etc. ;-)

[1] https://en.wikipedia.org/wiki/Moria_(1975_video_game)

IIRC some of the PLATO courseware was ported to the TI 99/4A as well.

> We always envied the non-engineers who had more freedom to choose classes they were fascinated by.

Personally, I kind of pitied the non-STEM students.

My own problem was there were not enough slots in the schedule for all the classes I wanted to take. I figured the university knew what it was doing with the required classes, and they were right.

> Personally, I kind of pitied the non-STEM students.

The only time I've had this opinion is when I was younger and conceited, holding onto an attitude that they're all wasting their money, probably fueled by envy.

Although there are moments—largely driven by other aspects of bureaucracy—where I wish I'd completed my bachelors, I'm quite happy in retrospect that I instead chose a bunch of random off-topic interesting humanities courses and non-cs-stem stuff, some of which I failed for inane reasons. Aside from a few moments in data structures and algorithms, I barely remember anything from the CS courses, they were unbelievably dull and poorly structured. In one case I believe I got nearly 100% on all the homework but failed both exams because I just kind of zoned out and wasn't driven to write java by hand for 2+ hours, which generally shocked the profs because I was typically the most engaged, personable, and probably older than everyone else by like 4 years.

>> Personally, I kind of pitied the non-STEM students.

Well, as a STEM university graduate myself I admit being guilty of this thought and I think it's just something "hard science" thinks of "humanities" in general. Mostly due to mercantilism, most non-STEM studies have very poor job perspectives.

But ... it all comes down to numbers. The elephant in the room is that THERE ARE TOO MANY COLLEGE SEATS FOR STUDENTS. This has multiple causes but it's a societal problem and a pretty dire one.

Take one of the universities in my city: https://www.ubbcluj.ro . In 1989 (at the fall of the communist regime in Romania) it had 5,619 students. Overall, in all specialties and all years, took 4 years in general to graduate.

Today just at the admission exam were accepted 16,800 new students. Taking all years that makes about 55,000 students. 50 fucking thousand! That's 10 times the level during commies.

And that's the problem. A lot of people who have no business being in the academic environment are now funneled through it. Reason is first, because "The West" had every man, women and their dog get a college degree so we had to play catch up. Problem, as the article states, they are getting that degree on paper only, de jure, not de facto.

But the even deeper problem is the dissolution of white collar jobs. That happened both in the West and the East. Agriculture used to employ 90, 50, 30% of the population, now there's 2 to 5% working there. Industry used to employ 90, 50, 30% of the population, now there's 5 to 10% working there. And the rest? God have mercy! We're all in "services". We're fucking servicing the shit of each other.

So in order to avoid fixing the hard problem (what the fuck to do with people who don't have the skill and intellect to do academic stuff but are good enough for plowing in agriculture or operating a CNC machine), the powers be have opened the gates of colleges. Get a college degree, that will compensate for the lack of activities to do with it!

I could write more.

Bottom line, what you expecting from peasants going to college? You can take the peasant out of the village but you can't take the village out of the peasant, they say.

> Well, as a STEM university graduate myself I admit being guilty of this thought and I think it's just something "hard science" thinks of "humanities" in general. Mostly due to mercantilism, most non-STEM studies have very poor job perspectives.

What does mercantilism have to do with that? The problem with the job perspectives of students of many non-STEM subject is rather that it is often not easy to find economic applications of the knowledge that is taught in the humanities courses (which is made even more complicated by the often "left", "woke" bias that many humanities faculties have).

So it's a natural progression. From agriculture (primary) to industrial (secondary) to services (tertiary). The AI revolution means services go the same way as industrial. Everybody goes to the 4th level.

<Narrator Voice>: There is no fourth level.

Nobody could envision the next level while being busy in the current one.

The fourth level would be the capitalist class (I.E everyone becomes an entrepreneur/businessman), but we didn't really figure out a reliable way to convert the upper-middle to upper. Funnily enough, the problem isn't lending, money was pretty cheap for the last decade or so, and even China had that whole microloan thing happening, but that not enough of the upper-middle were capable of executing correctly.

Imo, we should go back to education and take a more serious look about the social utility of what we are teaching, not on an individual but a systemic level and whether this really is the best way to allocate subjects. How much of this after all is about actual training or just social signalling?

> Imo, we should go back to education

We tried that before the service revolution took off. That's exactly why we started pushing people into college, thinking that education could build entrepreneurship/businessmenship. We told the kids that if they become those entrepreneurs/businessmen they'd earn more money, live a better life, etc. in hopes to compel them in that direction.

But that's not what the people wanted.

> everyone becomes an entrepreneur/businessman

...this has been "figured out" -- it's brutal, and it looks like the gig economy.

I don't even have to paraphrase this: "Drive with Uber - Be Your Own Boss"

At some point you’d think we’d learn that our economic system, and the way that we manage work, has stopped making the sense it made centuries ago when we developed it.

I personally also found taking upper division CS courses extremely time consuming. Like 80hrs/week outside of class by the end of semesters.

Was it this system? I used these terminals at U of Illinois in the late 80's, but I only remember using them to take my physics tests. I don't remember ever using them for studying or interactive learning.

https://grainger.illinois.edu/news/magazine/plato

They were more beige terminals. For us it was just reviewing material, tests were fill in the dot..(scantron)

I graduated (admittedly many-many years ago) from a good but not top-notch university. I remember a somehow similar situation: obviously learning was considered a good thing, but both the students and the professors realized that it's the diploma what brings most students there, not a pursuit of pure knowledge.

So I quickly realized that, unlike, say, elementary school, a university is not a push system, it's a pull system. If you want to learn, you need to make an effort and extract knowledge from this source. There's still plenty, but nobody is going to force-feed it to you. I read quite a lot beside the required books. I practiced quite a lot beside the lab practice (fortunately wielding a soldering iron or writing programs was a marketable skill; still is, but used to be, too). I asked my professors questions that were not entirely in the books; often that was during a few minutes after a lecture / classes / labs, so I got from them ideas and pointers to new directions to learn by myself.

Was it helpful in my career? Certainly yes, I started doing contract jobs three years before graduation, and then joined a bunch of interesting companies where that knowledge was somehow useful, mostly as a foundation of more specific skills.

I was certainly not alone; I knew (and often was friends with) a bunch of other students who craved knowledge and skills, and we helped each other shake these out of the university, past the transactional bounds. It wasn't all that hard, but it required a conscious effort.

Very certainly a large number of other students did more coasting than knowledge-mining. They got their diplomas, got some white-collar jobs that did not require such deep knowledge of engineering, I suppose, or started unrelated businesses.

> If you want to learn, you need to make an effort and extract knowledge from this source.

Oh I'm 100% aware of this, and actually think it's better than the push system that school prior to college follows. The issue is that the content is significantly worse now.

There ends up being a lot of guesswork today of finding resources that are good. I always have to question: "does this person actually know what they're talking about or am I wasting my time?" I'm sure you had to do this back in your day, but with the overwhelming amount of information available, it becomes difficult to parse.

I would kill for a class where the professor just said, "Everything you need is in that book." Now we get, "The book doesn't talk about this, but you should know..." It's infuriating.

Yes, the first encounter with a professor not knowing well the thing he was teaching was a mild shock. But, thought I, that professor had been a student back in the day, he also knows how to quickly prepare to the basics, and then just wing it, hoping that nobody is going to dig deeper.

After that I just started believing books more than some professors, as long as the books cross-checked with one another.

Recently a friend of mine, herself a professor, watched how some other professor, invited to give a special lecture, was obviously out of their depth in certain questions that they should know like the back of their hand in order to give such lectures. She was pretty depressed by that, and especially by the fact that her students might be fed incomplete or even wrong information. So the problem is there, is known, and is not an illusion :(

While at it, no one book contains all that you should know about a subject, if you want to know it well. Not the Feynman Lectures. Not Code Complete. Not even the Mahabharata. You always get to read more. (I'm not talking about the formal exam questions here, of course.)

> While at it, no one book contains all that you should know about a subject, if you want to know it well. Not the Feynman Lectures. Not Code Complete. Not even the Mahabharata. You always get to read more. (I'm not talking about the formal exam questions here, of course.)

Absolutely, and I'd like to clarify, I'm not expecting a single physics book to cover all there is to know about electricity and magnetism. I just mean for a particular course (where the purpose of the course is to expose me to the topic) to be centered around a book properly, in which new topics that aren't in the book aren't introduced (within reason of course)

It depends on the subject, of course. By the time someone has written a book on "programming language du jour" (say, Rust right now), and gotten it published and printed, it will be 1-2 years out of date. And students will complain that "all the information is online for free". Except, it's really hard to point at a specific website that is not in your (the professor's) control to say "everything you need is in here" when it could be taken offline tomorrow. Or reorganized and re-written in such a way that content is added or removed.

The course I think I did the best in teaching was to say "here is the textbook" (on databases) and then when a specific solution / technique came up, to point out that "this is how mysql does it", or "this one is used by postgres", etc.

I started college at the CC level (having no HS diploma) to get into a State school. And from a series of poor choices and ignorance on my part needed to take a several years gap before returning to finish up.

I don't think in my experience students have changed all that much.

CC students have always felt more motivated in my opinion. But good Lord the quality of the education at the State level is abysmal. I am not saying there aren't quality professors and classes. There are.

There is however an alarming high number of poorly designed classes, nearly broken technology, poorly edited and badly written assignments, and questionable instruction.

I have to compare the quality and price with what I experienced in CC and it just makes me sad and depressed.

The higher the level of education, the less attention paid to pedagogy. I had better teachers in high school than in college.

This was my general experience as well. The very best quality education I ever received was in my 7th grade math class, followed closely by my 7th grade English class.

I did have some very excellent university classes (including ones that were so good that I audited them without receiving credit), but I also had a lot that were positively abysmal, taught by professors who were experiencing severe mental health issues (one who'd had a stroke and could no longer comprehend the material, another who was going through a mental break and stopped teaching us altogether, etc.) or extremely stressed grad students who were not fluent in English and spent class time trying to catch up with their PhD workload.

My best university-level education actually came after I graduated and got a job working in a lab at my university. During that time, I worked closely with the professor and grad students, and it was such an amazing learning opportunity that I will never forget for the rest of my life — sadly cut short by the 2008 financial crisis.

If you look at how most professors and adjuncts are rewarded and paid, it makes sense. You can't get quality instruction from a adjunct who is only a half-step away from sleeping in their car, especially when they know they might be gone mid-semester due to a budget cut. Even the full professors are trying to bring in enough grants, oversee enough RAs and TAs to do the work for the grants, get some of their own research done, and barely have time to teach. Teachers in high school have a high teaching load relative to colleges and universities, but they are doing a job and generally are paid at least middle class wages.

I had a math prof who didn't speak English.

Didn't matter, one of the best courses I ever took. He made the math look beautiful on the old fashion chalk board. Absolutely wonderful and enlightening equations.

Better than the one who didn't speak much english and complained how chinese is so much easier because you don't have to learn a new word for beef, you just combine the symbols for "cow" and "meat", and still didn't know the material well.

> I don't think in my experience students have changed all that much.

I'm actually somewhat inclined to agree with this. I also started at a community college first, so I got to see a lot of adults trying to do career switches into tech.

Many of them were frankly the same if not worse than many of the young students you see in college today. One could definitely attribute some of that to the fact that they have more responsibilities to handle impacting them, but even just overall demeanor was noticeably worse. I frequently had adults 15-20 years older than me throwing their hands up and asking me to just give them the answers to what were ultimately very simple programming problems.

It was great for me because I took it as an opportunity to reinforce the material we were learning, but I knew I was doing them a disservice, so at some point I would stop enabling the poor behavior.

Honestly, to me the biggest thing impacting everyone is the inundation of information from technology today. I know it sounds cliche, but it's making academics a billion times harder than it needs to be. It's also making it less enjoyable and satisfying, thus people take the shortcut to get the grade they're looking for.

What the college tuition debate overlooks is that for costs to go down, so does the quality of the experience. this means college is more barebones and less handholding, like in Europe.

Maybe for the costs to go down, the administrative fat could be trimmed instead.

Especially when you consider that it's gotten to the point where at many schools with the worst and most extreme administrative bloat, there is one non-teaching, bureaucratic administrator for every student.

In the absence of UBI can you blame them? Not everybody can, or wants to, be productive.

Why should those who don't want to be productive be supported by those who do?

Taxpayers end up with bill either way, I think, either directly through UBI and welfare or indirectly through losses due to crime, policing costs.

Europe has a good approach by using subsidies as a more palatable UBI initiative to keep their startups afloat.

I'm saying the opposite has been my experience.

There's plenty of ways for Universities to lower costs without hurting quality.

There are plenty of open resources nowadays. Many of the paid online only textbooks with inconsistent tooling and accessibility for starters could be eliminated.

Expensive corporate software contracts is another. Access to MS office is nice in theory but in practice many students rely on Google instead AND there are open alternatives that don't have unreliable authentication problems.

University tuition here is state sponsored, the difference is not that great (everything over 10k a year is a cashgrab anyway) You have the "institution rate" aka real cost + milk foreign students tax, and the European cost aka affordable. The university gets a yearly allowance per European student plus a lump sum on graduation (yes this brings graduation rates up and quality down).

The end result is that the average money received is about 60% (give or take 5-10 percent depending on left or right wing politics) of the institution rate so you now know the foreign "tax". If you are over 32 you should also pay the institution rate but it is almost fully tax deductible.

Costs have been increasingly hard to justify given the wealth of information the internet provides, for some time now. Often, a sufficiently motivated person can piece together a lot of material themselves, given a general "structure" of topics/material.

Lots of textbooks floating around out there too.

LLMs add another layer to this. In many cases, the whole thing is looking a bit silly (at least at the state level)

This is really unfortunate, and I think your instructor should read it.

It sounds like your instructor has confused homework with quizzes, and the cheating issue demands some rethinking of the course pace and assessment system.

In physics and related fields, I have found fully worked problems to be very valuable. If your textbook includes some of these, I recommend reviewing them and working similar practice problems if possible. I wonder if things like supplementary texts, khan academy, or tutorials on youtube might help as well.

As you note, systems like ChatGPT could be helpful for explaining or working through problems, but obviously you won't learn anything if you rely on them for doing your own problem sets.

Typically these online homework systems are pushed by the department much to the chagrin of the professors, but the students are required to pay ~$100/ea for the privilege of doing automated homework, and the department gets some nice kickbacks.

Student abuse. Bundling textbooks with single-use codes for these systems is also malicious.

Interesting. When I was an undegraduate we had textbooks which were at least 25% problems and solutions, allowing for near endless self practice.

I am currently holding my copy of "Introduction to Electrodynamics" by Griffiths in my hands; somehow it is rarely more than a metre from where I work!

Are such textbooks still popular and used (i.e. mandatory to purchase) in courses like this?

I've noticed specifically with undergraduate physics courses that the textbook will usually get bundled with the online homework service that the school is using, and that part you're required to purchase if you want to do your homework.

Nowadays though, if a professor states that a textbook is required and there's no online service involved, 80% of the time the professor is exaggerating and probably hasn't even read the book themselves. The other 20% of the time, students will generally just find the pdf for free online somewhere (As a CS student, I tend to find my books on GitHub).

I know I'm digressing from what you asked a bit here, but I just really need to take a moment to highlight that the textbooks that are "required" today are not nearly as good as the textbooks you likely went to school with (I'm making a bit of an assumption about the era of your schooling here, so correct me if you only just recently graduated). There are way to many instances of some no-name authors getting the shot at publishing with O'Reilly or Pearson. The content will be mostly correct, but they're never truly illuminating.

On that note, the only textbook I still have is a 1980s Physics 100 book, and I still pull it out occasionally for reference.

The best homework system i ever experienced was high school calculus.

4 points per assignment. Pass the assignment to a peer for grading. If they wrote down the problem and attempted it: 3 points. One more point if the logic and steps were followable, even if wrong.

The answers were in the back of the book. The homework grade should reflect attempts and practice, not mastery as that is what exams are for.

My high school calculus teacher would assign homework every day, anywhere from a couple problems, to a dozen or more, depending on the work. 5 points for the homework, no matter how many problems. One point off for each problem missed. Sometimes broken down by steps if it was very-few problems (but still with way more opportunities to lose points, than there were points)

Assigned every. Fucking. Night.

You could spend two hours on it, get a majority of the problems right and get extremely close on the rest but make some identical mistake on each of them at the end, and still get not just an F, but a zero. It could be an insanely large assignment, and also you have other classes and other things happening in life, so you only get half of it done because these were not small assignments, but at least get all the ones you did right. Zero points, because you missed 5 or more.

To say it was demoralizing would be an understatement.

You are right, what degraded is not simply the students attention and motivation, it is the whole institution. They keep pushing ineffective approaches all over. You are right to blame LMSes, they are absolute disasters, poorly designed and ineffective at anything except save time for professors (and let's be real here, they also do social media and unrelated to their work activities so they are trading their teaching opportunities for leisure). Those LMSes are probably as detrimental as PowerPoint has been for communicating to an audience. It is as if everyone is trying to avoid doing what they are here to do. They replace thought, exchanges and discovery with miserable tools just so they can go waste their time on something else.

> my options are going to office hours (which I try to do), go to tutoring hours (which conflicts with my job's work schedule), or go to ChatGPT and/or Chegg.

Have you considered trying to do the problems yourself, away from the computer, then checking your work with ChatGPT or the like?

I should probably do that more, however, one big issue is I can't always trust that I'm not being led astray with some of the posts on Chegg or the responses from ChatGPT and other AI model interfaces. I will frequently get responses from ChatGPT and/or Chegg that are just flat out wrong. So I could compare my answers against these services (which takes a long time), but I run the risk now of potentially learning the wrong approach when I go and find out that the walkthrough I was following was incorrect.

Which granted, that's just part of the work, but these weird hoops to jump through just shouldn't be there in the first place, and they used to not be there as well. So my gripe is why on Earth did we add them?

Yeah, the new interface sounds ridiculously bad. Sorry you have to go through that.

o1-pro is quite reliable for elementary problems, FWIW. I use it extensively for reading AI research papers.

I'm lucky enough to teach in a school that has small classes; I get to be very accessible to the students. There is some auto-grading, but most homework I grade by hand and give partial credit.

But if my classes were 300 people, I couldn't do that.

I also have relaxed deadlines so students can take more time if they need it, and request it in advance.

The object is to learn stuff. That's where I'm aiming.

Can I ask what school? I have kids that will be ready to start college in a few years and I’m always on the looking for places like what you’re describing

I can only vouch for my classes (which are CS undergrad). Other instructors have other policies, but most programs have small class sizes.

The campus is Oregon State University-Cascades in Bend. We're a new campus so we're looking for more students.

I remember this frustration clearly. It's valuable to be challenged and struggle and overcome, but the value of a perfect GPA is a lot more salient.

From your numerous comments on this topic, it seems that you are remarkably self-aware (for a college student) about your own learning process. That is kind of amazing. I hope you really know just how broken the system that you're describing is and that it is absolutely worth fighting to figure out how to really learn something hard.

Also know that there's a yin and yang here. You're in a broken system--but the system used to be broken in other ways. Your point about there being too many resources strikes me as fascinating and true--and yet we have efforts like Three Blue One Brown taking teaching to a whole new level. People who figure out how to learn are always in a golden age.

First, I appreciate the kind words, they really do mean a lot.

I will say, part of the reason I'm likely more aware is because I am an older undergrad (currently 25), but additionally, I've seen all sides of the education system, so I've been exposed to quite a lot.

> People who figure out how to learn are always in a golden age.

100% agree on this. The ability to be able to pick anything up and just go with it opens life up to a wealth of opportunities.

i’m assuming this is a system that changes the numbers in the question each time? often there is a separate “practice” button where you can practice and (maybe) get feedback, figure out the process, and then do the questions?

even if not, try putting the textbook and the question into an AI (I use the paid version of gemini, the $20/ month is the best money you’ll ever spend at college), then as it to explain how to answer the question. Then ask it to generate a similar question, and then give you feedback to you as you try to answer it. Then try and answer it step by step, repeat as many times as you like until you understand and keep getting the right answer, then answer the actual homework question.Feel free to dm me if you want to discuss!

Using AI is risky since it's trained on data from sites like Chegg, StudySmarter (now known as Vaia), Brainly, etc.

Since those are user submitted answers, they sometimes get them wrong and so the AI models just regurgitate the wrong answer, especially if it's for a problem that's infrequently assigned by professors.

Nevertheless, I still use the AI models to try and aid in solving these problems, there's just always this gut feeling that I'm being guided in the wrong direction, but since I'm not versed in the subject I'm learning, I can't identify the hallucinations or inaccuracies.

I agree to some extent. A wild one to me is how schools will limit the number of classes you can take and/or pressure you to graduate on time even if it means choosing a major before you're really ready. That seems like a surefire way to kill any interest in exploring knowledge and replace it with a focus on mechanical box-checking.

They have to make room for incoming freshmen.

That homework setup sounds awful. I wonder how much it was shaped by educators, and how much by techbros.

Unfortunately, the setup I'm talking about is the norm these days. It's the MyLab and Mastering software offered by Pearson.

It doesn't just stop at physics. I've seen it used in math courses and economics courses as well.

Sometimes professors will setup the system so that you can redo the problem with different numerical inputs for the variables, and that approach is fine to me as it lets me learn through trial and error without negatively impacting my grade.

Yeah the outsourcing of homework and grading to publishers is a sorry state of affairs.

It's on educators. It started well before tech.

They want kids to do homework, so they give a few free marks for doing it. Kids cheat and ignore anything not marked because that's how people respond to incentives. Then the lecturers wonder why no-one is doing stuff that isn't marked, and try to fill in the gaps by marking more things until they run out of capacity to police it all.

There are upsides to continuous assessment, but it's effectively micromanagement and has all the predictable downsides. "I applied hard incentives to make people give me X and Y, so why don't they they game the system, and why don't they also do Z, shocked picachu face."

I agree. I took linear algebra through an online course. I can't imagine a worse way to learn the topic. I happily cheated on tests to pass the class.

Mind you, I already apply linear algebra daily—I simply refuse to waste my time computing basic operations on matrices by hand when that's why we have computers and partial credit isn't available.

Reply on news.ycombinator.com