You don't pay your doctors, insurance does, so I mostly expect insurance will replace doctors with chatbots as soon as its feasible. Sadly in a large percentage of scenarios that would actually be an improved outcome. (yes, not all and those are the ones that will probably kill you... but come on, medicine is such a disaster its should be completley rebooted)
>Sadly in a large percentage of scenarios that would actually be an improved outcome.
I predict triage will be converted to AI who will be able to offer minor prescriptions, stuff unavailable to over the counter, without doctor oversight. Like allergies, rashes, insect stuff, fungal or worms. Will likely be able to reduce medical load by treating people who just need to be sent to get over the counter stuff.
> but come on, medicine is such a disaster its should be completley rebooted)
Agreed, but is AI the right move? Do you really want to be the first one to try it?
Asthma inhalers (at least, some types) are available over the counter here.
As far as I know, only "narcotics" (opiates, psychostimulants...) require regular check-ins (due to byzantine legislation) - but even that's been worked around with one-click refill requests and telehealth.
The types available over the counter are unsafe for long term use and frankly worse in every way compared to the one you need a doctors permission to purchase.
You need a established relationship to access one click refill and telehealth, they will tell you to come in after a time.
I once had a doctor google webmd in front of me at an appointment. It was funny, but then I remembered how often I look stuff up at work and thought it’s a bit unrealistic to expect them to have EVERYTHING memorized. I’d rather have them double check a hunch they have instead of holding back because they’re unsure
I had this thought the other day: why do we expect an immediate diagnosis? I can't imagine giving a decent answer to any advanced question without research.
I understand appointments have a time limit, but perhaps it would be beneficial to batch process this.
Because those off-the-shelf solutions only work for very obvious and basic cases, and you don't know what you don't know. Meaning, if you have a more complex medical situation, you wouldn't be able to tell, so you wouldn't know if the off-the-shelf solution is wrong.
Also, we still need doctors to perform physical tasks, like surgery. The doctors diagnosing stuff and the doctors performing surgery aren't different doctors, usually. When I got diagnosed with Testicular cancer, the urologist who felt my balls up and said "yeah, this ain't right" was also the one who removed the testicle. And, he wasn't the first doctor I went to - the other doctor clearly had not felt up enough balls. He said it could be X, could be Y, maybe Z. Not the urologist, he knew right away. So, I think it's more complex.
It's funny. During tax season this year, I was engaging with several "tax professionals" from H&R Block in order to answer some, hopefully elementary, questions about my tax situation. And you know what, I could never ever get a straight or clear or definitive answer from them! It was infuriating. It seemed that my business was too small or insignificant to them to bother. They missed appointments and most of their branches were closed. I believe that they're a significantly understaffed industry.
When I finally got an in-person tête-a-tête, the guy was just Googling stuff -- I mean he was not even looking on IRS.gov, just Google -- and I got so pissed I walked right out.
So ironically, the best advice I found, the best distillation of good information, was through Google's Gemini LLM. I asked Gemini many tax questions, and it was able to cogently answer them with correct terminology, and enough correct advice that I was able to navigate the IRS.gov publications and corroborate them from the horse's mouth.
So at this point I feel like engaging an LLM in particular topics is far more productive than engaging a human on the front-lines. The humans are poorly trained, underpaid, overworked, and unreliable.
H&R tax people are on minimum wage plus commission. Also, for a lot of them the work is seasonal: crazy hours for a couple months every year, then barely any work until next tax season.
Next? Unprescribed stimulants and test banks from Greek orgs has been the path to med school for generations. I guess we just cut to the chase by prescribing ADHD meds to everyone and making cheating tools openly available.
Why is it a moral outrage when students do it, but universities cheat students in many ways and we just grumble and accept it as a necessary evil. Just accept cheating as the same, maybe attack financial aid itself instead of the students. Say "this money is disproportionately being siphoned directly into the organizations coffers and very little of it is being used to teach anyone anything whatsoever. Therefore we are restricting access to financial aid, capping it to $X million total" etc. What's the point of a scholarship if there's no "school" being taught?
The person you responded did not frame this in moral terms, but emotional terms. I'm not here to grandstand about the morality of anyone's behavior, but knowing students can't engage with long-form written content is depressing to me. Knowing that many (most?) will not teach themselves how to engage in critical thinking and work through frustration is also depressing.
Part of this is empathy, just knowing they are depriving themselves of skills they don't even realize the importance of. Part of is selfishness, feeling like our next generation of thinkers and creators are potentially self-sabotaging.
I don't know, I'm out of the prediction game, I don't really want to get drug into utilitarian arguments about the place of AI, or debates on what was wrong with education, etc. Frankly, I'm out of my league there. I'm just sad to see people lose the ability to think for themselves and struggle through hard problems.
Yes, I use generative AI myself, I'm not a luddite or anything.
There's a huge leap (a chasm, really) between depressing and moral outrage, and I am not attacking anyone. This comment is a complete non-sequitur from mine.
So there's something depressing about cheating but not from a moral basis? What's depressing about people who want to minimize work and maximize profits?
My partner's an academic and this tracks with what I've heard from her and her colleagues. It's possible they're all in on a conspiracy to lie to me for shits and giggles, but I doubt it.
I also remember what it was like to be in University and I know several of my classmates would have done this if it had been available. Human nature has not fundamentally changed in the 13 years since I finished school.
I teach at a college in The Netherlands, and this is not my experience at all. We give the students assignments, and when they hand them in, there's usually an individual assessment. Then we walk through the code, and ask them a bunch of questions. If you can't answer these questions, we ask them where they got their code from. And I caught many students.
However not all code matters. If the subject is "design patterns", then we're not going to ask them questions about run-of-the-mill Spring Boot code.
For a number of reasons I got my university degree in Spanish(and Spanish speaking country), where very few people spoke English - I was one of two in total as a matter of fact. I always felt like I was cheating my way through cause I had access to infinitely more and better quality resources and all I had to do was translate them into Spanish and I was instantly 10 steps ahead of everyone else. With all the LLM bollocks these days, I'm suddenly feeling a lot less guilty.
Idk, maybe not cheating but a very unfair advantage - everything I ever did took me a fraction of the time compared to everyone else - while everyone was going nuts last minute, I had long forgotten about anything I had to do and was spending days or even weeks out playing pool in the local pub.
Life in general is full of unfair advantages or disadvantages form birth. Nobody is ever equal, equality is a made up fiction, and in a competitive environment without strictly enforced rules, cheating is sometimes just the norm. If you don't cheat, you'll be outcompeted by those who do. Statistically, the more unscrupulous people behave, the more likely they are to get ahead in life. Look at politicians, CEOs, big business owners, etc.
This is not cheating at all. Professors often tell serious students to read multiple textbooks in parallel to see different pedagogical presentations of the same material. (In the US, syllabi often list “recommended” texts as well as “required” texts.) You just happened to be able to read texts in a different language.
As usual, this article totally conflates "to help with homework" with "to cheat on homework". If you use it like a better Google, learning from its output rather than directly using the text, that’s definitely not cheating by any definition.
ChatGPT and other LLMs just turned the screws on a already broken system.
1. Most people get a degree to pass degree checks in jobs.
2. College is stupid expensive.
3. Unlike with other goods and services, there is no way to get your money back on bad service/goods here
4. Professors can basically do whatever they want
5. Fed student loans are in the same group as criminal judgements of debt.
When I look at all of these, logically, you expect a return from the cost. And the return isn't knowledge, but a diploma. And humans being good tool users, use a tool to raise the chances of getting that (half-assed promised but not really) diploma.
Hopefully, this will destroy universities as diploma mills, and go back to proper academic rigor and study. But one can be ever hopeful, I guess.
This. LLMs have ended the need for white collar junior work. Recent grads at this point are just a legal/tax (sec 174) liability - a great way to get a law suite. Colleges need to pivot fast.
>LLMs have ended the need for white collar junior work.
I think the high interest rates, poor consumer economic situation and remote work with outsourcing did way more to damage that jobs market than LLMs right now.
I am yet to hear jobs being replaced with LLMs but I hear about jobs moved to cheaper locations all the time. LLMs make workers more efficient at easy, low-stakes repetitive shit but don't fully replace them. Yet. They also don't replace accountability. A lot of times managers/corporations just hire to have a scapegoat to throw under the bus if the shit hits the fan. You can fire and intern for doing something wrong, you can't fire a LLM that fucks up, it's still the user to blame.
Couple with the bullwhip effect of too many grads going to University chasing those lucrative white collar jobs from the covid boom days and this is what you get.
>Recent grads at this point are just a legal/tax (sec 174) liability
You don't take a worker and explicitly replace them with an LLM (unless it's a customer service agent or something)... It looks more like, the company just doesn't need as many people for a given job type but it's not a 1:1 replacement at an individual level.
The GGP said, "LLMs have ended the need for white collar junior work". That claim is that LLMs are a "1:1 replacement at an individual level", and this is what the GP was responding to.
I read "ended the need for" in terms of the overall situation. When I say they're not 1:1 replacement, that's a possible reason why it might not look directly like people are being replaced, even though it may still be happening in the aggregate.
They are pretty good at most white collar jobs. Blue collar not so much.
> Colleges need to pivot fast.
The problem is the same everywhere - the management (administration) has grown like a mushroom. We need more profs and instructors because oral exams are the only way to verify knowledge that can't be easily gamed - aside bribes. That's scary because ethics are not what they once were.
Towards other fully WFH white collar careers: OnlyFans and Twitch streaming. We're speed running towards the Idiocracy timeline where humanity is too stupid to do anything but be employed in the ad driven attention economy.
> LLMs have ended the need for white collar junior work.
If you think that juniors are solely for doing the low level work (which is wrong).
The purpose of the junior is to work towards becoming senior (in experience and capability). Doing the the bs work is just an additional use of their time.
Your thinking is akin to the belief that the purpose of the medical residency is to do examinations and not to become a full fledged doctor.
LLMs are currently at the level of a fresh graduate, not even someone with 2 years of real-world experience.
While I expect this to change, I don't know how fast, nor where the ceiling is — always the trouble with S-curves, you can never tell how close you are to the mid-point, and with even a tiny bit of noise you can't even tell which side of the mid-point you are.
Maybe that was true at some point, but they certainly can't anymore.
Speaking with professors around me (at universities on different continents!), they all report that they basically have to give everyone passing grades, they can't tell students off/kick them out of the class for engaging in outwardly rude behavior (eg wearing headphones, obviously playing games/watching videos/etc. for the duration of the class), they have to accept pretty much any request for extension/makeup exams/etc, they can't give a student a 0 when they've obviously used chatGPT (ie cheated) for an assignment.
If you want to go back to proper academic rigor and study, you have to return authority to the professors.
This is a variant of 'kids these days', that we've documented all the way back to Plato.
If I buy a thing at Wal-Mart or Bezosmart, and it doesn't work, I can get my money back.
If I pay for service, like dry-clean my clothes, and its not done right, I can get my money back, or they offer to redo it.
In academics, none of this applies. Garbage professor? Too bad. Professor grades on curve? Too bad. Professor doesn't teach and just reads from book? Too bad. Professor does (bad but not illegal thing)? Too bad.
And failures are always assumed to be the students pure responsibility, and none of the professors' responsibility.
Combined with 'you will pay this and all other classes back up to when you're 50', and 'you need a diploma to be able to pay it back', the situation is set.
As I said, its a mixture of these premises is what leads to this situation. And there's no good answer here. Well, I guess de-capitalizing universities could work, along with jobs no longer demanding a 'degree', would go to that direction as well. But I am not hopeful.
My time in college was 10 years ago, but absolutely none of that was true at the time (I personally seen pretty much everything you referred to as "not allowed"). I find it very hard to believe that much has changed since then.
My spouse is a professor at a highly ranked US university. While these things are not completely impossible, there is major institutional pressure against this sort of authority amongst professors.
In Australia, two friends who are uni professors, one in admin/leadership position for their faculty, and my partner teaches some science undergrad courses in their phd topics. It is absolutely that bad with both subtle pressures and blatant failures aligning to progress fee paying students to completion regardless of competency or pedagogy.
I disagree. I had a number of tenured professors who absolutely did not give AF. They were openly hostile and put in the barest minimum to teaching, because as they told us, they would rather be doing research. I attended a research university, and that's cool and all, but the job of a professor should be to teach, and research should be second. In my humble opinion, that has been turned around. So we do have a lot of tenured professors, essentially immune from their core responsibility, slowly phoning it in as they prioritize research over lecture.
My wife is currently a professor. None of that is true for her. She fails people all the time for cheating, non-attendance or whatever. She gives them a few chances not to dig their own grave, and some still do it.
Re: Do whatever they want. She has a colleague who is a terrible teacher. She hears all the time and students learned nothing in her class. That teacher does run a couple of software businesses on the side, and most likely is just trying to find smart kids and recruit them for cheap coding labor.
There is pressure to admit students from overseas colleges, where it's kind of known they are rich kids who cheated and bribed their way through undergrad programs and often come unprepared for the masters programs they are applying for. These are some the kids who I described at the start of my post, and the most egregious cheaters (turning in coding assignments with other people's names on them, getting caught cheating 5 times, and asking for a call where they ask if there's any way they can pass). It's nuts.
> they can't give a student a 0 when they've obviously used chatGPT (ie cheated) for an assignment.
I used to teach at university level (not in the US). I gave a 0 to a cheating student. Next day, the head of the section came to me in a panic. "omg, what have you done? you can't do that...". Actually, I think he wasn't even worried about legal implications, he genuinely thought this was very unfair to the student and that I was some kind of prick.
Actually, there are very specific procedures to deal with cheating, and it's so time consuming that it's better to avoid them altogether and just close your eyes.
Edit: another anecdote, I refused late students to my class (after warning them several times). Again, I got reprimanded by the head of the department for being too strict.
Do you believe that if you were given a diploma saying you passed med school with flying colors, you could just go be a doctor? Knowledge doesn't matter, it's all about that diploma?
Personally I think it's the exact opposite. The diploma doesn't matter, it's the knowledge you want. CS is a good example of that, plenty of great programmers with no degree. Diploma doesn't matter.
A lot of my peers slacked off in university, since then I've had an extremely easy time getting jobs whereas I know many of my peers have struggled. Most of my job applications have resulted in interviews and all of my interviews have resulted in job offers.
We all have the diploma, difference is knowledge. While my peers were struggling with their assignments I was already done, spending my time helping them and studying further. When they were finally done with the assignment they slacked off until the next thing they were required to do, while I was programming for fun, exploring new things, studying to prepare for every lecture etc.
University is the perfect environment for learning. You have lots of time, guidance, peers, resources etc. The fact that most people completely ignore that opportunity does not change its value. You get what you give.
If you're going into a field where you need a solid technical foundation the knowledge is what matters. But if you're going into an unrelated domain like most college grads who enter non-technical, white-collar jobs the diploma is the return because the coursework has no direct relevance to the profession, but employers will filter out applicants without degrees.
Someone with a bachelors in history has a much higher likelihood of being hired for an entry level sales role for example, than someone without one. But that bachelors isn't conferring any unique knowledge about sales.
I'd bet that the vast majority of office jobs that require a degree do not actually need or even check for any of the educational content that the degree signifies. They purely use "college degree in anything" as a generic hurdle. No way any employer hiring some generic office assistant whose job is to move papers from one filing cabinet to another and help schedule meetings, is going to care what that person's major was. They just want to see "degree."
> 1. Most people get a degree to pass degree checks in jobs.
How many top physicists didn't go to university? How many Nobel Prize winners (aside from Peace or Literature)? How many degree-less engineers do you want to be building your bridges? Or surgeons operating on you?
This is such a software dev centric view. In software it kind of works because you can learn and "practice" with just a computer. This is also why there's a fair amount of car mechanics without degrees. But it's much harder to "practice" many other things. As a software dev who never went to university, I also readily admit that I never did some of the more theoretical foundational stuff. Is that relevant in my day-to-day work? Mostly not. Until it is.
Also you don't need a student loan or huge amount of money to get a degree. There are plenty of places where you can get bachelor/master degree for small tuition fee.
You're talking about rigorous fields where practitioners need a solid technical base. There are lots of white-collar jobs that really don't require a college education but employers will filter out applicants with degrees.
Previously, tests could be administered to determine aptitude, but was found illegal form of discrimination. However, academic degrees somehow werent that.
As for other comments, I find it telling that professions like Medical Doctor, professional engineer, and the like have an apprenticeship built in, but somehow that invalidated my comments? (Hint: it doesn't)
> Previously, tests could be administered to determine aptitude, but was found illegal form of discrimination.
Not strictly true. But you have to be able to show that your test actually measures something important to the task, without at the same time being biased by unrelated issues. Which is hard, and requires domain knowledge, and might mean that some tests couldn't't be written tests, so it scared off the HR types.
And, yeah, of course you're right that those same unrelated issues keep people from getting degrees. But everybody gets to ignore that somehow.
Perhaps the right approach to the "degree" thing would be more regulation, in the form of a requirement that you show that the actual subject matter of the specific degrees you were requiring or advantaging was relevant to job performance. No more "any college degree" BS.
The point of trying to determine aptitude is exactly that it does not measure current ability to perform specific tasks or possession of specific domain knowledge, but rather general ability to learn/flexibility to be given new tasks.
One of the places where software has succeeded is that the industry gets away with quietly trying to test for intelligence while ostensibly making it look work related (algorithm puzzles, "wanting to hear how candidates think"). IBM straight up gave me math problems (like solving systems of equations under time pressure) when applying to my first programming job in addition to some basic whiteboard coding. This is part of why it's possible for smart people without CS degrees to break into the field.
Basically any job where you're not acting as a human automaton benefits from general learning ability, so if degrees stop working or stop being allowed, people will no doubt look for another proxy.
> The point of trying to determine aptitude is exactly that it does not measure current ability to perform specific tasks or possession of specific domain knowledge, but rather general ability to learn/flexibility to be given new tasks.
Great. You can go ahead and do that then... provided that your test actually measures that and not a bunch of unrelated knowledge. Which is a lot of what an "aptitude test" designed by a clueless person will actually tend to measure. And honestly is also a lot of what the ability to get a degree tends to measure.
You might be able to fix the test, but I don't think you can fix the degree.
> IBM straight up gave me math problems (like solving systems of equations under time pressure) when applying to my first programming job in addition to some basic whiteboard coding.
... which is mostly going to measure how recently you've had to do a bunch of similar mathematical tasks, to the point where that overwhelms almost anything else. And has the bonus of not being obviously related your general ability to learn or flexibility. I suppose how you do on that might have some correlation with your ability to code under similar time pressure... which of course only happens in badly managed projects.
That's exactly the sort of selection process you get when you let clueless fools try to come up with ways to measure things like "ability to learn" or "flexibility".
Degrees are a filtering mechanism. In almost all fields/positions they are a weak signal of aptitude, not a strong one. But there are no strong signals of aptitude before the interviewing process at the very least, and most of the time there is no strong signal until the candidate is actually employed and working. So I don't disagree with you, but the person you're replying to is just stating a reality of the job market and the reason people are going to college.
Whether we like it or not, college has become a vending machine, where you insert $X00,000, and out comes a key to allow you access to at least interview for maybe 60+% of (generally higher-paying) jobs out there. Education may happen as a nice side effect, but people aren't spending this kind of money purely for education. I sure would not. I also would not spend that kind of money and then just accept not getting that "key" at the end of the transaction.
If we want universities to function as a place to obtain a rigorous education, we need to drastically reduce the cost, and stop employers from using the diploma to gatekeep nice white collar office jobs.
The more expensive the university, the greater the chances of getting the even higher paying of the higher paying jobs. Its rational for people to pay the premium for name-brand university if the expected payoff is some CEO-track position. Not all will get it of course, but that is the calculation.
The only way that will change is if companies stop requiring college degrees as a "check the box" step to even get considered by some automated hiring system. _That_ is what is driving universities as diploma mills, not the universities themselves.
See what jobs you can get these days with a high school diploma, and you understand the economics of a university degree.
This also ties in with the US no longer manufacturing much of anything -- which created a lot of decent jobs that didn't require a college degree. Those jobs are gone.
It's 200€ per year here in France, not that expensive. I am talking Sorbonne and other prestigious universities.
Then we have chep engineering schools (including the most famous ones) and then we have crazy expensive finance schools (including the best in the world) at about 25k€ pet year. Except if you are poor and then it is basically free.
> When I asked him why he had gone through so much trouble to get to an Ivy League university only to off-load all of the learning to a robot, he said, “It’s the best place to meet your co-founder and your wife.”
Many people understand all these things. They're just not nihilistic assholes who will lie and cheat to any amount to get ahead. That's the difference.
When I read articles like this I don't know whether to think I have some decent job security or not. On the one hand, people graduating having used LLM's for a significant majority of their work are almost certainly deficient at critically thinking about computer systems and programming them (there's a good quote in the article stating this much better than I am here). On the other, if they're cheating so much to get well-paying jobs will I --without cheating, relying on just the merits and knowledge from my career-- be competitive enough to find work in the future? Been wondering if anyone else feels this way.
>> On the other, if they're cheating so much to get well-paying jobs will I --without cheating, relying on just the merits and knowledge from my career-- be competitive enough to find work in the future? Been wondering if anyone else feels this way.
You need to find a place that understand the difference between BS and real work. The next 5 years might bankrupt the ones that embrace the BS.
a lot of the ChatGPT assignments are so incredibly obvious. in the most egregious cases literally copy and pasted from the browser without reading. laziness is all-pervasive.
the problem is an academic cheating penalty can have grievous consequences for someone's life, so to prove it requires almost like a "beyond a reasonable doubt" criminal justice standard. and that's difficult to do. so they get through their classes.
"don't hire someone" is a much lower bar. if there are obvious but unprovable ChatGPT vibes in the application, you probably just don't interview them. if there are weird pauses in their conversation during the interview and they sound like they're reading long sentences, don't advance to next round.
a lot of these cheaters are going to get filtered hard.
there is a smarter way to use it but this still requires some level of thought to disguise your cheating, you have to be able to understand the LLM outputs yourself, and a large portion of the cheaters can't manage that.
i think people who learned how to do a skill prior to AI will have an edge. writing will be a valuable skill. same with people who know computer fundamentals.
I will likely retired within the next 5-7 years and by that point after nearly 45 years in tech I will be damn happy to be gone. I believe the coming AI shift in tech will definitely benefit society, but will devastate many/most tech disciplines as a vocation especially for those average technologists within those discipline. In my experience…that is most of the folks working today.
My guess is there will still be a market for those folks who don’t have to rely on AI to solve problems (but can still utilize those AI tools effectively) and that is where I would be focusing my own growth if I was younger with decades to go. Either that or I would give up tech entirely and try and find a plumber to apprentice under.
Sure it’s a shit job, but let’s face it…AI and automation wouldn’t be able to replace it…probably ever. I don’t see too many people talking about “vibe plumbing”. We will never reach a point in society where people will not pay handsomely to remove backed up shit from their homes and have clean water piped in.
And frankly, I know a few plumbers—most of the ones my age are already semi retired, play golf daily, just work when they want to, and all have a shitton more money than I do.
Now what will said cheaters do when they don't have access to AI? Like if the internet goes down, or if they work in an environment where they can't access it? Or if the company says no, because it's a security risk?
'Sorry boss, I can't work because I don't have internet and don't have access to my AI crutch' isn't exactly a stellar argument. That's someone I would immediately regret hiring.
If you have an architecture office and the power goes out, do you have your workers give up on using AutoCAD and go to the drafting boards and slide rulers till the power comes back up? What's the point of going backwards with technology? You'll be outcompeted by those who embrace modern technology.
I'm not aware of any modern engineering profession that doesn't use electronic crutches regularly. Nobody knows absolutely everything in their heads. We're paid for critical thinking , not to have everything memorized just in case the internet is out. Even before the internet, SW devs had stacks of books on their desks they'd reffer to often.
Remember when IBM wanted to pay devs per line of code written? Yeah, exactly. I see the skills being problem solving, not writing code.
'Sorry boss, I can't work because I don't have internet and don't have access to GitHub/CI/license server/email/any of our Saas tools/any of my files' happens every time the internet goes down and bosses are quite understanding.
Yes, prompting an LLM is a kind of skill. But relying on it has been shown to hinder development of the most important skill of all--critical, independent thinking. And ultimately that's the only "accomplishment" that matters when it comes to education. Completed homework assignments are not a product, they're a process for making someone think, and that's what's being avoided with LLM usage.
Educators are already adapting. I teach CS and we do verbal assessments of students their code. According to my students, a certain percentage slip through but all in all, it's nowhere the problem that this article makes it out to be (at my college).
Wishful thinking if you read the article. Most of these kids aren't learning anything about how to use AI. They aren't like a Simon Willison or Janus or anything. They're not learning about confabulations, they're not learning about scaling laws or tokenization, they're not learning to use the API, they're not running multiple models to get a feel for how they think differently, they ain't doing none of that shit you wishfully think they are - because they're cheating to minimize effort so they can go do something else like watch Tiktok. (Even Lee, who is by far the most successful and sophisticated user profiled in OP, is just dumping out stuff as fast as he can to try to make a buck.)
They are copy-pasting the assignment from Blackboard to ChatGPT and vice-versa. As the article points out, they usually aren't even reading the outputs to check that the essay doesn't suddenly start bloviating about Aristotle.
And no one needs to pay a college wage premium, or any wage, for that:
> He worries about the long-term consequences of passively allowing 18-year-olds to decide whether to actively engage with their assignments. Would it accelerate the widening soft-skills gap in the workplace? If students rely on AI for their education, what skills would they even bring to the workplace? Lakshya Jain, a computer-science lecturer at the University of California, Berkeley, has been using those questions in an attempt to reason with his students. “If you’re handing in AI work,” he tells them, “you’re not actually anything different than a human assistant to an artificial-intelligence engine, and that makes you very easily replaceable. Why would anyone keep you around?” That’s not theoretical: The COO of a tech research firm recently asked Jain why he needed programmers any longer.
I read the article. Those kids are definitely learning how to think, just not in the way educators wish. One of the girls quoted said she loves writing, but finds it helpful to get structure and initial content from generative AI. The kid at Columbia has already launched two startups. Clever kids always figure out how to leverage the tools they have at hand.
With the advent of new technology, humanity has no choice but to move up the value chain -- in this case, from executing tasks to judging, supervising, and curating AI work.
> One of the girls quoted said she loves writing, but finds it helpful to get structure and initial content from generative AI.
Which shows she's not learning from the AI or even improving on it. All she's doing is filling out the 'structure and initial content' - oh, is that all???
> The kid at Columbia has already launched two startups.
Which doesn't require learning anything much from the LLMs, as he did that on his own before them (note the Leetcode). Also, considering that the second one is fraudulent and if it works, exists largely to abet fraud, this is not a ringing endorsement.
> With the advent of new technology, humanity has no choice but to move up the value chain -- in this case, from executing tasks to judging, supervising, and curating AI work.
ChatGPT couldn't written that ending any better. (I will charitably assume it did not, unless you took the trouble to replace that emdash with a double hyphen just to be cute.) It could have written it about 1000x cheaper, however.
I'm not sure there's a choice. The economic incentives are too strong. Like it or not, those individuals who figure out (i.e., learn) how to leverage AI to do more, faster, and (eventually) better will have an advantage over those individuals who don't. Many kids intuitively seem to understand this.
Believe it or not, I write all my comments the old-fashioned way -- without the help of any AI!
there is on choice. my kid uses about 11 different AI tools and she is about to turn 12. it is on parents to guide this but any kid who is not all over this at the deep level will be at insurrmountable disadvantage
it's really fucking hard to find competent engineers.
i've interviewed about 20 cs majors who claim to be experts at typescript and only 3 of them have been able to explain what the purpose of useState is (not trolling).
(There are at least 2 ways this tweet is embarrassing.)
Maybe they're "learning" how to use AI [1], but in at least some cases they're not learning what they're supposed to be learning from the class they're enrolled in [2]. The point of the assignments was never to produce the the answers as fast as possible by any means necessary [3], but to build strong foundations so that you can later do much more challenging/useful things.
If your assignment is "program X in assembly," and you just copy/paste the prompt into ChatGPT, you aren't learning assembly, you aren't learning important concepts relevant for low-level performance optimization, etc. etc. That was the point, not to produce X.
What would be useful is specific classes with assignments that are designed to be worked on with the assistance of AI. "AI-enabled SWE" etc. Lessons on how to effectively utilize it for research and learning. But those things aren't replacing learning the fundamentals, and probably can't unless AGI is achieved and actual knowledge is just economically irrelevant.
[1]: For a lot of undergrad assignments you can probably just copy/paste the description into ChatGPT, which I don't think qualifies as being difficult enough to merit "learning" how to do).
[2]: I guess this is "fine" if you're in the "AI will imminently replace all white-collar work" camp, but then its just dumb to be enrolled in college in the first place, unless maybe you're going for free.
[3]: In some sense, fully outsourcing your thinking to AI is the same as just paying someone to do your homework, which has always been possible but less accessible.
The words of Lee in this essay highlight the problem. Paraphrasing: "I didn't do it because I did not feel like doing it, because it felt unimportant/irrelevant/dumb."
This is the hubris of youth talking.
While this kind of thinking will teach you some practical skills, ultimately you become an operator, and not an engineer.
CNC machine designers earn more than CNC machine operators.
Car engine designers earn more than car mechanics.
Computer chip engineers earn more than people who put computers together.
Production line engineers earn more than production line workers.
In software development, the titles are unclear, but there is a clear distinction between operators and engineers. Unfortunately we call them all programmers or developers or engineers, as the titles don't truly reflect the kind of work they are doing. When you work with someone though, you know which title they really deserve.
It's hard to put it into words, but going forward, the operators will have a harder time finding work as automated AI-based tools will replace them. The only solution is to learn to become engineers, but with the kind of mindset that Lee flaunts in the above article, that is not going to happen.
My advice: Learn the fundamentals. Don't cheat. Be curious and strive for deep understanding. It's tempting to cut corners, but already today I see a deluge of people who just don't have what it takes to engineer software.
Every example you give, there are 1000s and 10000s of workers to every designer... and the road to a real engineering job is something that frankly, most people going to college could not complete.
Also, this is the kind of crystal clear thinking and analysis that we need more of (and why I come to HN). Thank you for cutting though the BS and highlighting the underlying lessons.
If you look at countries with a lot of corruption, then there's this sense that "everyone uses corruption to get ahead, and I'd be a fool not to". This view is not entirely wrong.
I don't envy students here; aside from the shining examples meritocracy, hard work, and accountability that you mentioned, you're also left with a feeling of "well, everyone else is using ChatGPT, and I'd be a fool not to". I don't entirely blame them, as the choice is between "use ChatGPT for one hour and get a passing grade" and "do my own work for six hours and get a passing grade". Of course the outcomes are not the same as there is value in doing the work yourself as it will give you a much better understanding of things, but taking delayed gratification in account has never been a top quality of young people.
I would be concerned that I'd get out of school and not get a job because I don't actually know anything.
But I wouldn't blame students if they're not preparing for interviews and jobs.
Because in this world having a job seems like a recipe for being poor.
Instead, maybe students are just going to prepare to get ahead, in any way possible.
And they'll go to a university to network and such.
Personally I doubt that's gonna work out well, but I'm old what do I know.
I’ve worked with a lot of people over the years who had degrees and didn’t know much about their field. I’m sure that’s not new, but it may become even more prevalent.
Consider that college tuition is often equal to what an employer would pay you to do similar levels of work. The opportunity cost for attending college is therefore 2x the tuition. The system is exploiting students who need a paper to get a job, so why shouldn’t students exploit it too?
* Mostly applies to degrees where "learning on the job" would have resulted in equivelant knowledge and skills.
The students are still playing along and paying into the system. They might even do more education and get higher degrees than before with LLMs doing the heavy lifting. Everyone wins!
Last semester I gave my students a take home final exam. I told them they could use AI. I just asked that they document their usage of external sources. This was precalculus class.
The first problem was a graph x(t) and y(t). The component functions of a parametric curve. The question was: Estimate the times the particle is at the origin.
I got an email from a student: I put problem 1 into ChatGPT and it couldn’t answer it. What are we supposed to do then?
My response: Think!
Students have mostly been trained that anything that can’t immediately be solved is too hard. They have no academic grit. It’s bad.
This will sound boomer, but I've noticed this with a lot of people: absolutely no determination to figure things out.
It isn't just young people. It really seems like people's determination has just gone to shit.
I'm an engineer and teach people a lot. The first thing I have to do is determine what kind of learner people are, and my gosh, it's frustrating. People give up so fast.
Well as a member of gen-z, I'll happily agree with you if that makes you feel like less of a boomer. And I think it mostly has to do with attention. The capacity for focused attention has been deteriorating at a societal level for quite a long time. It's not a generational thing either—we are all facing similar pressures with the sheer amount of information we've been forced to process on a daily basis. That loss of attentional capacity seems to be affecting really important faculties like learning, creativity, and deep work.
I see this in myself all the time but I find a lot of people my age are quite aware of what is happening to us while many older people are stuck blaming the kids—hell, there are times in public when I see older folks more glued to their screen than your average teenager. It's not just young people. The path forward isn't very clear.
So many professors in college are detached and tone-deaf about their syllabus, their approach to teaching, and their workload, that the smart students adapt.
One of those adaptations is cheating. I have seen really hardwokring students in grad school finally resort to cheating because the workload was insane, and they realized everyone else was cheating.
Controversial opinion - perhaps getting free tutoring from LLMs is a good thing?
Some thoughts:
0. Rich kids (not me) got expensive tutoring. Now that is available to everyone. How is this not great? Was it cheating when rich people paid for it also?
1. At my undergrad (Cornell) about 20% of TAs couldnt speak english (or did not want to.) Some of the group office hours would tangent off into other languages. This was terrible to those who only understood the standard language of the school (English.) ChatGPT is available to everyone.
3. You dont even need to pay exorbitant tuition to get the same level of support as everyone in the world, this is a great democratizer. Isnt this what people have always wanted?
I think these points resonate most with a worldview where college is just a status competition, where it doesn't matter if you actually learn anything.
By analogy, we could replace the human players of a sport with identical robots and it would level the playing field, but it would also completely defeat the purpose of sport. It is an activity done by humans for the benefit and enjoyment of humans.
Rich students have long cheated themselves out of an education by using expensive tutors to do their homework. Now that's available to everyone, yes, and that somewhat levels the playing field, yes, but it's not a good thing if you care about education. It's more like, narcotics used to be only for rich people, but now we have powerful cheap ones that everyone can get easily.
>> Rich students have long cheated themselves out of an education by using expensive tutors to do their homework.
Students in a crunch or those who are short-sighted use it to do homework. There have always been ways to cheat.
Students who want to learn use it for guidance, insight, examples, detail. There have not always been ways to do this unless you were wealthy. So much of education is just mediocre. You are lucky (or rich) if you get access to great education.
I think the difference between us here is, I don't think current LLMs are well-suited to actually help people learn. In my assessment, they lack the bedrock understanding of the material to resolve student confusion, and their understanding of student needs is very poor. It's probably not impossible to learn from them, but I think they strongly encourage shortcuts that only give the impression of learning.
I do hope this changes. LLMs are of course an amazing technology, but sometimes being amazing isn't enough to actually solve human problems.
I am certain that there are some students using LLMs to learn more effectively. This does not appear to be the norm. Instead, my spouse gets a large portion of the class turning in essays that are very obviously just the prompt sent to ChatGPT and then the entire response copy-pasted. I'm sure that there are some rich kids whose parents paid somebody to do all their homework for them. But that's not something we should be seeking to do more.
"Tell me about your thought process being the argument in this essay" is met with blank stares, despite them insisting up and down that there is no way that they cheated.
> perhaps getting free tutoring from LLMs is a good thing?
Yes it is, and I have seen it work. A friend has a hard time reading through long material. They have showed me a log of step by step questions they asked their LLM AI. That was impressive. In essence a long private teacher interview going on and on on the general topic they needed help with. With each answer manageable for them, and the long succession fitting their research style. Well done. There was still a risk of garbage coming from the LLM, but mitigated by the long sequence of step by step advancement (and space for further progress as the tech continues to improve.)
Definitely for them a better teacher and fitting their style of attention.
ChatGPT hit the mainstream market in my final year of undergrad.
I was indeed guilty of using it for one assignment wholesale, and a sizable portion of my final practicum. However, in the article it mentions something lightly that teachers use to distinguish LLM work from human work, which also rubbed me the wrong way.
The arguments and counterarguments were given equal weighting, unless a command were to be given to the LLM to spit out partiality to one, whereby it is overwhelming in substance (if not in language) towards that thesis. Now, finalizing my grad school time, I've not used it, and have actively discouraged group members from using it, as I feel there is an advantage in searching painstakingly for new, obscure ideas - where LLMs tend to give the same advice for recommendations to anyone who sets them the same problem. I used to do the same thing with Bing, in lieu of Google search, for more 'quirky' ideas to implement in my arguments.
There, I believe, lies an advantage for the semi-industrious knowledge student. Travel not the beaten path, but the one that takes slightly more effort for proportionally greater rewards.
> “There might have been people complaining about machinery replacing blacksmiths in, like, the 1600s or 1800s, but now it’s just accepted that it’s useless to learn how to blacksmith.”
I feel a large part of this is the definition of "cheating". For example, I give take home exams and I allow students to use whatever resources they want.
They can use chat GPT, the textbook, Stack Overflow, etc. as much as they want for as much time as they need. Access to these resources during an exam is typically considered "cheating" because using them would confer an unfair advantage, since tests are designed without them in mind. So then just give them access to the resources and make the test harder.
Instead of a question like: "Here is some code find the bug", ask them to implement a large system and to document the bugs they encounter along the way, and how they used ChatGPT to fix them. Instead of "what's the definition of TCP" ask them to measure TCP vs. UDP using iperf.
If all the students are cheating, then you're not testing the right thing. Make the test fit better to the current information landscape, and professors / students alike will have a better time. Stop testing information recall, start testing for information literacy, analysis, and integration. Stop asking students to answer multiple choice questions, start asking them to build systems.
This is why it is vitally important to provide a rich, varied educational experience.
If it is possible to pass a course using AI, it's partially the student's fault. But it is just as much the EDUCATOR's fault.
get these students into groups. Make them spontaneously interact. Use gamification principles. Count these interactions in grading. Prove you have internalized this material.
individual attention and personalized instruction were a vital part of my own education (college in the 80s and 90s). There are a thousand excuses why schools think they can't provide it. It often reduces to budget shortfalls. But this can be done with existing tools and personnel. It's more a matter of teaching more engagingly, more creatively.
175 comments
[ 4.4 ms ] story [ 238 ms ] threadDo you have some stats to back that up?
I predict triage will be converted to AI who will be able to offer minor prescriptions, stuff unavailable to over the counter, without doctor oversight. Like allergies, rashes, insect stuff, fungal or worms. Will likely be able to reduce medical load by treating people who just need to be sent to get over the counter stuff.
> but come on, medicine is such a disaster its should be completley rebooted)
Agreed, but is AI the right move? Do you really want to be the first one to try it?
In Europe you can just buy those sort of things outright in many cases. Things like inhalers can just be had without getting a doctor on board.
I believe it's a uniquely US thing where the doctor needs to collect a well visit fee on you treating yourself for a common or chronic illness.
"You still have asthma? OK, here is permission from my caste to breathe, pay the fee on your way out." Is bonkers when you think about it.
As far as I know, only "narcotics" (opiates, psychostimulants...) require regular check-ins (due to byzantine legislation) - but even that's been worked around with one-click refill requests and telehealth.
You need a established relationship to access one click refill and telehealth, they will tell you to come in after a time.
-Aristotle, Civilization V (2010)
I understand appointments have a time limit, but perhaps it would be beneficial to batch process this.
What I'm really shocked about is that new AI tools are now fast enough—and regulated enough—to be used live at the point of care.
So why do I need to pay doctors as much money, when they themselves are using an off the shelf solution?
Also, we still need doctors to perform physical tasks, like surgery. The doctors diagnosing stuff and the doctors performing surgery aren't different doctors, usually. When I got diagnosed with Testicular cancer, the urologist who felt my balls up and said "yeah, this ain't right" was also the one who removed the testicle. And, he wasn't the first doctor I went to - the other doctor clearly had not felt up enough balls. He said it could be X, could be Y, maybe Z. Not the urologist, he knew right away. So, I think it's more complex.
When I finally got an in-person tête-a-tête, the guy was just Googling stuff -- I mean he was not even looking on IRS.gov, just Google -- and I got so pissed I walked right out.
So ironically, the best advice I found, the best distillation of good information, was through Google's Gemini LLM. I asked Gemini many tax questions, and it was able to cogently answer them with correct terminology, and enough correct advice that I was able to navigate the IRS.gov publications and corroborate them from the horse's mouth.
So at this point I feel like engaging an LLM in particular topics is far more productive than engaging a human on the front-lines. The humans are poorly trained, underpaid, overworked, and unreliable.
Part of this is empathy, just knowing they are depriving themselves of skills they don't even realize the importance of. Part of is selfishness, feeling like our next generation of thinkers and creators are potentially self-sabotaging.
I don't know, I'm out of the prediction game, I don't really want to get drug into utilitarian arguments about the place of AI, or debates on what was wrong with education, etc. Frankly, I'm out of my league there. I'm just sad to see people lose the ability to think for themselves and struggle through hard problems.
Yes, I use generative AI myself, I'm not a luddite or anything.
I also remember what it was like to be in University and I know several of my classmates would have done this if it had been available. Human nature has not fundamentally changed in the 13 years since I finished school.
However not all code matters. If the subject is "design patterns", then we're not going to ask them questions about run-of-the-mill Spring Boot code.
Not sure why you keep talking about cheating. What OP described doing emphatically and unequivocally is not cheating.
An actual homework question - in whole or part - is (anecdotally) definitely considered cheating from a policy standpoint.
On the other hand, "How do I [atomic thing]" (say, "reverse iterate over a list in Python") seems to be fine.
1. Most people get a degree to pass degree checks in jobs.
2. College is stupid expensive.
3. Unlike with other goods and services, there is no way to get your money back on bad service/goods here
4. Professors can basically do whatever they want
5. Fed student loans are in the same group as criminal judgements of debt.
When I look at all of these, logically, you expect a return from the cost. And the return isn't knowledge, but a diploma. And humans being good tool users, use a tool to raise the chances of getting that (half-assed promised but not really) diploma.
Hopefully, this will destroy universities as diploma mills, and go back to proper academic rigor and study. But one can be ever hopeful, I guess.
I think the high interest rates, poor consumer economic situation and remote work with outsourcing did way more to damage that jobs market than LLMs right now.
I am yet to hear jobs being replaced with LLMs but I hear about jobs moved to cheaper locations all the time. LLMs make workers more efficient at easy, low-stakes repetitive shit but don't fully replace them. Yet. They also don't replace accountability. A lot of times managers/corporations just hire to have a scapegoat to throw under the bus if the shit hits the fan. You can fire and intern for doing something wrong, you can't fire a LLM that fucks up, it's still the user to blame.
Couple with the bullwhip effect of too many grads going to University chasing those lucrative white collar jobs from the covid boom days and this is what you get.
>Recent grads at this point are just a legal/tax (sec 174) liability
Sounds like a US specific self inflicted problem.
> Colleges need to pivot fast.
The problem is the same everywhere - the management (administration) has grown like a mushroom. We need more profs and instructors because oral exams are the only way to verify knowledge that can't be easily gamed - aside bribes. That's scary because ethics are not what they once were.
The opposite of a liability is a deduction
If you think that juniors are solely for doing the low level work (which is wrong).
The purpose of the junior is to work towards becoming senior (in experience and capability). Doing the the bs work is just an additional use of their time.
Your thinking is akin to the belief that the purpose of the medical residency is to do examinations and not to become a full fledged doctor.
While I expect this to change, I don't know how fast, nor where the ceiling is — always the trouble with S-curves, you can never tell how close you are to the mid-point, and with even a tiny bit of noise you can't even tell which side of the mid-point you are.
Maybe that was true at some point, but they certainly can't anymore.
Speaking with professors around me (at universities on different continents!), they all report that they basically have to give everyone passing grades, they can't tell students off/kick them out of the class for engaging in outwardly rude behavior (eg wearing headphones, obviously playing games/watching videos/etc. for the duration of the class), they have to accept pretty much any request for extension/makeup exams/etc, they can't give a student a 0 when they've obviously used chatGPT (ie cheated) for an assignment.
If you want to go back to proper academic rigor and study, you have to return authority to the professors.
If I buy a thing at Wal-Mart or Bezosmart, and it doesn't work, I can get my money back.
If I pay for service, like dry-clean my clothes, and its not done right, I can get my money back, or they offer to redo it.
In academics, none of this applies. Garbage professor? Too bad. Professor grades on curve? Too bad. Professor doesn't teach and just reads from book? Too bad. Professor does (bad but not illegal thing)? Too bad.
And failures are always assumed to be the students pure responsibility, and none of the professors' responsibility.
Combined with 'you will pay this and all other classes back up to when you're 50', and 'you need a diploma to be able to pay it back', the situation is set.
As I said, its a mixture of these premises is what leads to this situation. And there's no good answer here. Well, I guess de-capitalizing universities could work, along with jobs no longer demanding a 'degree', would go to that direction as well. But I am not hopeful.
Things really have changed over the last decade.
Re: Do whatever they want. She has a colleague who is a terrible teacher. She hears all the time and students learned nothing in her class. That teacher does run a couple of software businesses on the side, and most likely is just trying to find smart kids and recruit them for cheap coding labor.
There is pressure to admit students from overseas colleges, where it's kind of known they are rich kids who cheated and bribed their way through undergrad programs and often come unprepared for the masters programs they are applying for. These are some the kids who I described at the start of my post, and the most egregious cheaters (turning in coding assignments with other people's names on them, getting caught cheating 5 times, and asking for a call where they ask if there's any way they can pass). It's nuts.
I used to teach at university level (not in the US). I gave a 0 to a cheating student. Next day, the head of the section came to me in a panic. "omg, what have you done? you can't do that...". Actually, I think he wasn't even worried about legal implications, he genuinely thought this was very unfair to the student and that I was some kind of prick.
Actually, there are very specific procedures to deal with cheating, and it's so time consuming that it's better to avoid them altogether and just close your eyes.
Edit: another anecdote, I refused late students to my class (after warning them several times). Again, I got reprimanded by the head of the department for being too strict.
Do you believe that if you were given a diploma saying you passed med school with flying colors, you could just go be a doctor? Knowledge doesn't matter, it's all about that diploma?
Personally I think it's the exact opposite. The diploma doesn't matter, it's the knowledge you want. CS is a good example of that, plenty of great programmers with no degree. Diploma doesn't matter.
A lot of my peers slacked off in university, since then I've had an extremely easy time getting jobs whereas I know many of my peers have struggled. Most of my job applications have resulted in interviews and all of my interviews have resulted in job offers.
We all have the diploma, difference is knowledge. While my peers were struggling with their assignments I was already done, spending my time helping them and studying further. When they were finally done with the assignment they slacked off until the next thing they were required to do, while I was programming for fun, exploring new things, studying to prepare for every lecture etc.
University is the perfect environment for learning. You have lots of time, guidance, peers, resources etc. The fact that most people completely ignore that opportunity does not change its value. You get what you give.
Someone with a bachelors in history has a much higher likelihood of being hired for an entry level sales role for example, than someone without one. But that bachelors isn't conferring any unique knowledge about sales.
How many top physicists didn't go to university? How many Nobel Prize winners (aside from Peace or Literature)? How many degree-less engineers do you want to be building your bridges? Or surgeons operating on you?
This is such a software dev centric view. In software it kind of works because you can learn and "practice" with just a computer. This is also why there's a fair amount of car mechanics without degrees. But it's much harder to "practice" many other things. As a software dev who never went to university, I also readily admit that I never did some of the more theoretical foundational stuff. Is that relevant in my day-to-day work? Mostly not. Until it is.
not in America, not anymore
https://supreme.justia.com/cases/federal/us/401/424/
Previously, tests could be administered to determine aptitude, but was found illegal form of discrimination. However, academic degrees somehow werent that.
This Forbes article goes I to it much better than I. https://www.forbes.com/sites/georgeleef/2014/11/06/thank-or-...
As for other comments, I find it telling that professions like Medical Doctor, professional engineer, and the like have an apprenticeship built in, but somehow that invalidated my comments? (Hint: it doesn't)
Not strictly true. But you have to be able to show that your test actually measures something important to the task, without at the same time being biased by unrelated issues. Which is hard, and requires domain knowledge, and might mean that some tests couldn't't be written tests, so it scared off the HR types.
And, yeah, of course you're right that those same unrelated issues keep people from getting degrees. But everybody gets to ignore that somehow.
Perhaps the right approach to the "degree" thing would be more regulation, in the form of a requirement that you show that the actual subject matter of the specific degrees you were requiring or advantaging was relevant to job performance. No more "any college degree" BS.
One of the places where software has succeeded is that the industry gets away with quietly trying to test for intelligence while ostensibly making it look work related (algorithm puzzles, "wanting to hear how candidates think"). IBM straight up gave me math problems (like solving systems of equations under time pressure) when applying to my first programming job in addition to some basic whiteboard coding. This is part of why it's possible for smart people without CS degrees to break into the field.
Basically any job where you're not acting as a human automaton benefits from general learning ability, so if degrees stop working or stop being allowed, people will no doubt look for another proxy.
Great. You can go ahead and do that then... provided that your test actually measures that and not a bunch of unrelated knowledge. Which is a lot of what an "aptitude test" designed by a clueless person will actually tend to measure. And honestly is also a lot of what the ability to get a degree tends to measure.
You might be able to fix the test, but I don't think you can fix the degree.
> IBM straight up gave me math problems (like solving systems of equations under time pressure) when applying to my first programming job in addition to some basic whiteboard coding.
... which is mostly going to measure how recently you've had to do a bunch of similar mathematical tasks, to the point where that overwhelms almost anything else. And has the bonus of not being obviously related your general ability to learn or flexibility. I suppose how you do on that might have some correlation with your ability to code under similar time pressure... which of course only happens in badly managed projects.
That's exactly the sort of selection process you get when you let clueless fools try to come up with ways to measure things like "ability to learn" or "flexibility".
If we want universities to function as a place to obtain a rigorous education, we need to drastically reduce the cost, and stop employers from using the diploma to gatekeep nice white collar office jobs.
See what jobs you can get these days with a high school diploma, and you understand the economics of a university degree.
This also ties in with the US no longer manufacturing much of anything -- which created a lot of decent jobs that didn't require a college degree. Those jobs are gone.
Then we have chep engineering schools (including the most famous ones) and then we have crazy expensive finance schools (including the best in the world) at about 25k€ pet year. Except if you are poor and then it is basically free.
Roy understands society better than most people.
You're confusing "hack" with "cheat" in the same way that some YouTube dickheads confuse "prank" with "being a dick for laughs".
If writing is now table stakes with GPT perhaps colleges need to adapt? Nobody can afford to waste time given how hyper-competitive everything is.
You need to find a place that understand the difference between BS and real work. The next 5 years might bankrupt the ones that embrace the BS.
the problem is an academic cheating penalty can have grievous consequences for someone's life, so to prove it requires almost like a "beyond a reasonable doubt" criminal justice standard. and that's difficult to do. so they get through their classes.
"don't hire someone" is a much lower bar. if there are obvious but unprovable ChatGPT vibes in the application, you probably just don't interview them. if there are weird pauses in their conversation during the interview and they sound like they're reading long sentences, don't advance to next round.
a lot of these cheaters are going to get filtered hard.
there is a smarter way to use it but this still requires some level of thought to disguise your cheating, you have to be able to understand the LLM outputs yourself, and a large portion of the cheaters can't manage that.
My guess is there will still be a market for those folks who don’t have to rely on AI to solve problems (but can still utilize those AI tools effectively) and that is where I would be focusing my own growth if I was younger with decades to go. Either that or I would give up tech entirely and try and find a plumber to apprentice under.
And frankly, I know a few plumbers—most of the ones my age are already semi retired, play golf daily, just work when they want to, and all have a shitton more money than I do.
They are learning how to accomplish a lot more, a lot faster, with AI.
That's a valuable skill -- one of the most valuable skills that students can learn today.
It's not the only valuable skill they should learn, but it's certainly one of the most valuable.
Educators will have to adapt.
I'm not aware of any modern engineering profession that doesn't use electronic crutches regularly. Nobody knows absolutely everything in their heads. We're paid for critical thinking , not to have everything memorized just in case the internet is out. Even before the internet, SW devs had stacks of books on their desks they'd reffer to often.
Remember when IBM wanted to pay devs per line of code written? Yeah, exactly. I see the skills being problem solving, not writing code.
They are copy-pasting the assignment from Blackboard to ChatGPT and vice-versa. As the article points out, they usually aren't even reading the outputs to check that the essay doesn't suddenly start bloviating about Aristotle.
And no one needs to pay a college wage premium, or any wage, for that:
> He worries about the long-term consequences of passively allowing 18-year-olds to decide whether to actively engage with their assignments. Would it accelerate the widening soft-skills gap in the workplace? If students rely on AI for their education, what skills would they even bring to the workplace? Lakshya Jain, a computer-science lecturer at the University of California, Berkeley, has been using those questions in an attempt to reason with his students. “If you’re handing in AI work,” he tells them, “you’re not actually anything different than a human assistant to an artificial-intelligence engine, and that makes you very easily replaceable. Why would anyone keep you around?” That’s not theoretical: The COO of a tech research firm recently asked Jain why he needed programmers any longer.
With the advent of new technology, humanity has no choice but to move up the value chain -- in this case, from executing tasks to judging, supervising, and curating AI work.
Which shows she's not learning from the AI or even improving on it. All she's doing is filling out the 'structure and initial content' - oh, is that all???
> The kid at Columbia has already launched two startups.
Which doesn't require learning anything much from the LLMs, as he did that on his own before them (note the Leetcode). Also, considering that the second one is fraudulent and if it works, exists largely to abet fraud, this is not a ringing endorsement.
> With the advent of new technology, humanity has no choice but to move up the value chain -- in this case, from executing tasks to judging, supervising, and curating AI work.
ChatGPT couldn't written that ending any better. (I will charitably assume it did not, unless you took the trouble to replace that emdash with a double hyphen just to be cute.) It could have written it about 1000x cheaper, however.
Believe it or not, I write all my comments the old-fashioned way -- without the help of any AI!
If your assignment is "program X in assembly," and you just copy/paste the prompt into ChatGPT, you aren't learning assembly, you aren't learning important concepts relevant for low-level performance optimization, etc. etc. That was the point, not to produce X.
What would be useful is specific classes with assignments that are designed to be worked on with the assistance of AI. "AI-enabled SWE" etc. Lessons on how to effectively utilize it for research and learning. But those things aren't replacing learning the fundamentals, and probably can't unless AGI is achieved and actual knowledge is just economically irrelevant.
[1]: For a lot of undergrad assignments you can probably just copy/paste the description into ChatGPT, which I don't think qualifies as being difficult enough to merit "learning" how to do).
[2]: I guess this is "fine" if you're in the "AI will imminently replace all white-collar work" camp, but then its just dumb to be enrolled in college in the first place, unless maybe you're going for free.
[3]: In some sense, fully outsourcing your thinking to AI is the same as just paying someone to do your homework, which has always been possible but less accessible.
This is the hubris of youth talking.
While this kind of thinking will teach you some practical skills, ultimately you become an operator, and not an engineer.
CNC machine designers earn more than CNC machine operators.
Car engine designers earn more than car mechanics.
Computer chip engineers earn more than people who put computers together.
Production line engineers earn more than production line workers.
In software development, the titles are unclear, but there is a clear distinction between operators and engineers. Unfortunately we call them all programmers or developers or engineers, as the titles don't truly reflect the kind of work they are doing. When you work with someone though, you know which title they really deserve.
It's hard to put it into words, but going forward, the operators will have a harder time finding work as automated AI-based tools will replace them. The only solution is to learn to become engineers, but with the kind of mindset that Lee flaunts in the above article, that is not going to happen.
My advice: Learn the fundamentals. Don't cheat. Be curious and strive for deep understanding. It's tempting to cut corners, but already today I see a deluge of people who just don't have what it takes to engineer software.
Maybe. Or maybe
Every example you give, there are 1000s and 10000s of workers to every designer... and the road to a real engineering job is something that frankly, most people going to college could not complete.
Also, this is the kind of crystal clear thinking and analysis that we need more of (and why I come to HN). Thank you for cutting though the BS and highlighting the underlying lessons.
*With hard work and a good environment maybe we can extend this to 15%. Certainly not 50%.
Average people should not become engineers. Average people should not be creating software.
Think of the bridges, planes, and skyscrapers average people would build.
You can't reasonably expect that people WONT lie and cheat their asses off.
I don't envy students here; aside from the shining examples meritocracy, hard work, and accountability that you mentioned, you're also left with a feeling of "well, everyone else is using ChatGPT, and I'd be a fool not to". I don't entirely blame them, as the choice is between "use ChatGPT for one hour and get a passing grade" and "do my own work for six hours and get a passing grade". Of course the outcomes are not the same as there is value in doing the work yourself as it will give you a much better understanding of things, but taking delayed gratification in account has never been a top quality of young people.
But I wouldn't blame students if they're not preparing for interviews and jobs. Because in this world having a job seems like a recipe for being poor. Instead, maybe students are just going to prepare to get ahead, in any way possible. And they'll go to a university to network and such.
Personally I doubt that's gonna work out well, but I'm old what do I know.
* Mostly applies to degrees where "learning on the job" would have resulted in equivelant knowledge and skills.
https://muratbuffalo.blogspot.com/2023/03/getting-schooled-b...
The first problem was a graph x(t) and y(t). The component functions of a parametric curve. The question was: Estimate the times the particle is at the origin.
I got an email from a student: I put problem 1 into ChatGPT and it couldn’t answer it. What are we supposed to do then?
My response: Think!
Students have mostly been trained that anything that can’t immediately be solved is too hard. They have no academic grit. It’s bad.
It isn't just young people. It really seems like people's determination has just gone to shit.
I'm an engineer and teach people a lot. The first thing I have to do is determine what kind of learner people are, and my gosh, it's frustrating. People give up so fast.
I see this in myself all the time but I find a lot of people my age are quite aware of what is happening to us while many older people are stuck blaming the kids—hell, there are times in public when I see older folks more glued to their screen than your average teenager. It's not just young people. The path forward isn't very clear.
I know a lot of older folks who are horrible about this. It's a problem across the board.
Thanks for being civil about it. This is a topic that many people get very upset about when you mention it.
Kids aren't going to email profs like this out of the blue (mostly). This tactic has worked for them in the past.
Still, good on you! Hold the line. The students will thank you in the future.
One of those adaptations is cheating. I have seen really hardwokring students in grad school finally resort to cheating because the workload was insane, and they realized everyone else was cheating.
Some thoughts:
0. Rich kids (not me) got expensive tutoring. Now that is available to everyone. How is this not great? Was it cheating when rich people paid for it also?
1. At my undergrad (Cornell) about 20% of TAs couldnt speak english (or did not want to.) Some of the group office hours would tangent off into other languages. This was terrible to those who only understood the standard language of the school (English.) ChatGPT is available to everyone.
3. You dont even need to pay exorbitant tuition to get the same level of support as everyone in the world, this is a great democratizer. Isnt this what people have always wanted?
By analogy, we could replace the human players of a sport with identical robots and it would level the playing field, but it would also completely defeat the purpose of sport. It is an activity done by humans for the benefit and enjoyment of humans.
Rich students have long cheated themselves out of an education by using expensive tutors to do their homework. Now that's available to everyone, yes, and that somewhat levels the playing field, yes, but it's not a good thing if you care about education. It's more like, narcotics used to be only for rich people, but now we have powerful cheap ones that everyone can get easily.
Students in a crunch or those who are short-sighted use it to do homework. There have always been ways to cheat.
Students who want to learn use it for guidance, insight, examples, detail. There have not always been ways to do this unless you were wealthy. So much of education is just mediocre. You are lucky (or rich) if you get access to great education.
I do hope this changes. LLMs are of course an amazing technology, but sometimes being amazing isn't enough to actually solve human problems.
"Tell me about your thought process being the argument in this essay" is met with blank stares, despite them insisting up and down that there is no way that they cheated.
Yes it is, and I have seen it work. A friend has a hard time reading through long material. They have showed me a log of step by step questions they asked their LLM AI. That was impressive. In essence a long private teacher interview going on and on on the general topic they needed help with. With each answer manageable for them, and the long succession fitting their research style. Well done. There was still a risk of garbage coming from the LLM, but mitigated by the long sequence of step by step advancement (and space for further progress as the tech continues to improve.)
Definitely for them a better teacher and fitting their style of attention.
I was indeed guilty of using it for one assignment wholesale, and a sizable portion of my final practicum. However, in the article it mentions something lightly that teachers use to distinguish LLM work from human work, which also rubbed me the wrong way.
The arguments and counterarguments were given equal weighting, unless a command were to be given to the LLM to spit out partiality to one, whereby it is overwhelming in substance (if not in language) towards that thesis. Now, finalizing my grad school time, I've not used it, and have actively discouraged group members from using it, as I feel there is an advantage in searching painstakingly for new, obscure ideas - where LLMs tend to give the same advice for recommendations to anyone who sets them the same problem. I used to do the same thing with Bing, in lieu of Google search, for more 'quirky' ideas to implement in my arguments.
There, I believe, lies an advantage for the semi-industrious knowledge student. Travel not the beaten path, but the one that takes slightly more effort for proportionally greater rewards.
Kids, this is why you need to go to school.
They can use chat GPT, the textbook, Stack Overflow, etc. as much as they want for as much time as they need. Access to these resources during an exam is typically considered "cheating" because using them would confer an unfair advantage, since tests are designed without them in mind. So then just give them access to the resources and make the test harder.
Instead of a question like: "Here is some code find the bug", ask them to implement a large system and to document the bugs they encounter along the way, and how they used ChatGPT to fix them. Instead of "what's the definition of TCP" ask them to measure TCP vs. UDP using iperf.
If all the students are cheating, then you're not testing the right thing. Make the test fit better to the current information landscape, and professors / students alike will have a better time. Stop testing information recall, start testing for information literacy, analysis, and integration. Stop asking students to answer multiple choice questions, start asking them to build systems.
If it is possible to pass a course using AI, it's partially the student's fault. But it is just as much the EDUCATOR's fault.
get these students into groups. Make them spontaneously interact. Use gamification principles. Count these interactions in grading. Prove you have internalized this material.
individual attention and personalized instruction were a vital part of my own education (college in the 80s and 90s). There are a thousand excuses why schools think they can't provide it. It often reduces to budget shortfalls. But this can be done with existing tools and personnel. It's more a matter of teaching more engagingly, more creatively.