And this grayness and sameness is what happens when people are forced to "think" as a chorus, either by the authorities or their censorship, or voluntarily by using the same AI's help.
Calculators are uncontroversial now. But when they first became cheap and widely available, they were not allowed in math classes. Then only four function calculators, then graphing calculators. But still today, programmable calculators are prohibited in many academic contexts.
"When will I use this in real life" is a declaration that you have no expectations of learning the next lesson that builds upon this one.
I hadn’t yet learned to program back when I was still using a feature phone, but I have a lot of fond memories of J2ME applications that I installed on my phones. Mostly games, of course.
I encourage anyone that wrote J2ME games and utilities, no matter how small or big, to upload the source to GitHub :)
But normally it depends on the subject and if the automation/machine solves the primary skill being teached or if its just a "secondary/tertiary" skill. Are you in a Calculus 101 class? Calculators like TI-89 are likely to be prohibited when examining for deriving analytical solutions for derivatives and integrals.
Statistics, Physics or any other subjects that needs applied maths? Such a calculator is probably a minimum requirement to take the course.
I vaguely remember thinking that one likely reason shortcuts like mine were not prohibited was because no one in charge suspected that such things were even possible with current technology, or if they were, that a child would be able to exploit it. But as long as I kept to the letter of the rules, I considered myself ethically in the clear.
Yeah, totally, just to be clear I'm not judging.
In fact, if you programmed it to handle those operations, one could argue you had already learned a big chunk of what was going to be measured in the exams.
Kind of similar to the paradox of creating and using cheat sheets, is highly likely you're accidentally learning about the subject matter in the process of writing the sheet, sometimes up to a point where the cheat sheet is not necessary anymore.
The problem is (example from mathematics): even if you are capable of deriving some formula (you thus understood the topic well), it takes a lot of time in the exam. Looking at the cheat sheet is much faster - in particular when the time is somewhat precious in the exam.
The other thing it was awesome for was solving systems of linear equations. I could do the nodal or loop analysis just fine, I'd write down the matrix that represented the system of equations and then just punch that matrix in and invert it.
The day when the relationship between s in a Laplace transform and d/dt finally clicked was also a really excellent day :D
Edit: Automate in the sense of coding it myself, not in the sense of downloading some software.
> Calculators are uncontroversial now. But when they first became cheap and widely available, they were not allowed in math classes.
The author of TFA means specifically for his cohort of students, not in general. He polled his students, and the result was that they thought calculators weren't seen as unethical but they were more skeptical/uncertain about AI. By his current students, now, not in general.
Five years into the availability of a calculator with an arbitrary advanced feature, it was controversial in academic contexts. Some of the author’s students could be grand-children of students from the early days of consumer calculators.
The author is comparing a new technology with an old one. And ignoring programable calculators which are still sometimes banned after fifty years…and many of the author’s students probably have used LLM’s for homework despite their statements that please the author.
Among other things, he's analyzing his students attitudes' re: AI and cheating with AI, and also comparing what they claim to feel vs what they actually do! It's mostly a reflection of what his students feel about the use of AI in English writing, not about calculators vs AI.
It seems as if you're responding to the line you quoted (out of context) through your preconceptions of what the article is about instead of actually reading the article!
> and many of the author’s students probably have used LLM’s for homework despite their statements that please the author.
"Probably have used LLM's"!? Don't take this the wrong way, but it seems you didn't read past the line you quoted, am I right? Because this is explicitly addressed multiple times, and the answer may surprise you.
am I right?
Of course not...and did-you-read-the-article comments are contrary to HN community standards. Such comments wallow in lameness.
Of course yes. My comment was rhetorical!
> The teacher is asking a rhetorical question and getting an et cum spiritu tuo response.
He wasn't asking a rhetorical question! It was two polls, at least one of which was anonymous! Followed by several experiments!
I mean, you completely missed the point of the article and are making comments that are immediately refuted by simply reading it. You are making factually incorrect statements about the article, what am I to think? That you read it, but decided to ignore its contents to make some irrelevant remark about calculators vs LLMs?
The article is all about the tension of what his students initially believe (or claimed to believe, anyway) about AI-assisted writing, and what they come to realize at the end of the class. They also discuss authenticity vs formulaic writing, and are surprised by the results as well! They also discuss the tropes that show up in AI-writing, speculate on what may cause it, and are surprised about some of the results they get. They also discuss AI-assisted teaching. At the end, they revisit whether to be pro or anti AI, and the future of English classes, with or without AI.
All of it has very little to do with how recent LLMs vs calculators are. You focused on an out of context comment which doesn't inform the bulk of the article, and doesn't say what you claimed it said.
The equivalent in maths would be if you handed students a theorem prover or have Wolfram Alpha give you step-by-step solutions and obviously nobody to this day allows this, because like ChatGPT for writing it'd defeat the point, that students think.
When I was in uni we were allowed basic but not programmable calculators during exams and a lot of CS classes even were pen&paper, if the prof was a bit hardcore
In grade 2 i had a teacher who would say "I don't believe in erasers", you know, the things that "undo" pencil. As a ~6 yr old i actually didn't understand this phrase: "Well I have one, they're real!"
A first grader probably would be prohibited from using any kind of calculator on arithmetic tests, 4-fn or not. But 8th graders are usually permitted scientific (non-programmable) calculators.
As you go up in grade level, you "get access to" calculators capable of functionality at the level below you. Because the point is that when we're educating students we want them to actually learn the subject matter, but once we've deemed them to have understood it and we have them move onto the next goal, we give them the tools to make that prior goal easier. We lessen the burden of the little mechanical concepts they already know so that they have an easier time becoming familiar with the next more advanced concepts.
AI systems are so much more advanced than what's capable on a TI programmable calculator. It's hard to draw clean boundaries around the tiers and enforce them by telling the model "help the user with tasks of tier 1-4 but not 5+". That's the issue, that it's really infeasible to strictly use them strictly as learning tools. You can almost do it with a lot of self-discipline and self-reflection to analyze your own workflow, but it's not generalizable across domains.
But reliably restricting output at the semantic level is very much an open problem.
Realistically, the answers the students gave the teacher were probably motivated by the practical benefits that come with giving teachers the answers they want to hear…bullshit questions are likely to produce bullshit answers. It’s not like first year college students haven’t had twelve years of academic standards moralizing talked at them.
Like I said to you in another comment, this is a fundamental misunderstanding of the article.
At least one of the polls was anonymous, too. There were no punishments in the class for using AI, in fact at least a couple of the students later revealed they had been using AI for everything.
This wasn't a "bullshit inducing poll", it was an experiment in perception vs reality, and how they modified those perceptions after the experiment had run its course.
Yes, they are uncontroversially bad. Schools that don't use them have higher scores.
Unfortunately, even SAT/ACT have calculator slop now.
It didn't.
I was in math class when calculators were introduced. At least for high school level and beyond, the curricula were designed to make problems solvable without calculators, and they weren't of much use. This was still the case when I taught an undergrad college math class in 1997. Graphing calculators were allowed, and the kids who tried to use them just screwed themselves up.
I would have gladly changed the curriculum to use calculators and computers from the very beginning. As tools, and not just to administer the same old exercises and quizzes. Give them Jupyter Notebook. Math education has never been a success story.
Education faces a dilemma, which is that it has always used heuristics to guide study and assess performance. Exercises such as the "three paragraph essay" had no use in the real world, even long before AI could generate them on demand. When one of those heuristics is broken, another one has to be found. Even word processing forced teachers to grade papers on content, rather than mechanics.
The rigid formatting imposed by graded assignments may not have any use. But on the other hand, having completed a liberal arts/"soft" science degree before my CS degree I've greatly benefited in the workplace from the writing experience it provided. I had to write so many papers that it became more or less effortless to produce long form, well structured writing and those written communication skills have helped me distinguish myself far more than my technical ability at work.
The generic ChatGPT overly formal corporate tone has no nuance or subtlety and is a poor substitute for well crafted, deliberate communication. I am always conscious of how my exact words and phrasing would be perceived by the intended audience, frequently requiring a balancing act between competing interests while maintaining clarity. Due to that I manage to avoid stepping on toes or sabotaging relationships due to inartful phrasing. It's frustrating to receive emails consisting of LLM boilerplate because it has such low information density and is so much more difficult to infer tone and emotion from the other side .
I'm very grateful I completed my education without the temptation to just churn out low effort writing or code and depriving myself of that experience. I'm not confident at all I would have been able to maintain that self-discipline.
Yeah. Overcommitted to partying and skipping class.
Has this author ever been to an average American university?
Between work and school and other responsibilities they have no time to decompress so they burn out
Their test scores show that they don't already know a lot of the material that's covered.
>Isn't it basically a remedial class, or rather a class for people who did the minimum of HS math?
It's College Algebra (or the non-trig part of Precalculus).
Their writing skills were also abysmal. Frankly if they were that bad by college they didn’t seem likely to improve in my opinion.
I went to about the cheapest US school that had a decent math program. It costs $17,500/yr between tuition, rent, books, and rice and beans.
That's a lot of money. It's over $10/hr in pre-tax income, even if you work full-time all year, which isn't an easy bar to clear in the sort of towns with cheaper universities. Wages don't scale well enough with more expensive tuition for there to be substantially better options.
Classes are another 22+ hours each week (you could complete school in 3-4 yrs instead, but that makes it even harder to afford and doesn't really reduce the workload enough to make a difference, however I'll also factor in a 15-hour workload later).
The rule of thumb is that you should study 3 hours for every hour of class. I found that approximately correct. Some classes took a lot less. Some took a little more. Combined with the self study you need in adjacent topics, 3hrs is a fair bit low.
During the school year then, you have something that looks like a 128hr/week schedule, or 100hr if you're finishing in 4yrs, and still 60hr/week even if you're finishing in 4yrs and racking up $70k in debt.
Don't get me wrong; I had free time (I worked more during the summer, less during the school year, allowing loans to cover the slack, which bought extra time here and there), but it wasn't exactly a party either. When I skipped class it was because I had to work, had to study for some other more pressing class, or found it more efficient to study the book than to try to understand that particular lecturer.
lol stop the cap.
Its more like 0.5 hours of grand total work for every hour of scheduled class, since most kids are skipping 50% of the classes and using "chat" (a really fetch name for a new digital drug btw) to make up the rest.
College is just a hoop, remember?
A lot of students didn't do what I did, and they washed out pretty quickly (I had a lot of classmates from HS that didn't last the first quarter). My first quarter was pretty harsh (only got one 4.0, and a 2.7 in a chemistry class I had no reason being in), but I wised up quickly. It was hard going from High school where I could do all my homework in the time between classes, to college where I had to do real actual studying.
You know that's fair, I hadn't considered the generational differences to be this vast.
I could maybe see your point about admittance (I had something like a D average, maybe C- or C+ or something, in high school), but I think my financial estimates were about correct. Is something majorly incorrect?
Even if so, there are a dozen cheaper states in the country with halfway decent programs (if any future internet denizen is reading this, Fayetteville Arkansas is actually great for math right now, both in quality and in cost/jobs). I doubt my observations are too far off-base for a typical student trying to go to school economically.
But I read an article recently about the death of partying in the USA: https://news.ycombinator.com/item?id=44514550
As learning goes, I'd say anybody taking more than two classes at a time is overcommitted. They might manage to get A's but I speak with these students about the courses we took together a year or two after the fact and it's clear to me that taking four or five classes at once is an awful strategy for retention.
Tangentially: I've helped out some college students with mentoring and advice from time to time. One common theme I've noticed is that their class load virtually doesn't matter. They find ways to run out of time no matter how much free time they start with.
We all like to imagine the poor, overburdened college student working 2 jobs and attending classes to make ends meet when reading statements like that. But to be completely honest, the students like that usually have their time management on point. The hardest ones to coach were the students who had no real responsibilities outside of classes, yet who found their free time slipping through their fingers no matter what they did.
Among all of the other problems with easy AI cheating, I wonder how much the availability of these tools will encourage even more procrastination. Feeling like you always have the fallback option of having ChatGPT write the homework for you leaves the door open to procrastinating even longer
> I asked my students to complete a baseline survey registering their agreement with several statements, including “It is unethical to use a calculator in a math class”
Unless there was more to this survey, this wording seems misleading. In a college-level math class, using a calculator is a common expectation depending on the type of class and the problem. The students would probably think of their TI-89, not a magical AI calculator that could solve every freeform problem for them.
I write at about 1200 words per day and I considered each fo the major multi-week assignments in my entry-level English courses to be worth no more than one day of my time apiece. For the finals, I gave them two days apiece, because I wanted an extra day to define the scaffolding for my argument.
My mother indicates that this is how she went through college too; very occasionally, a serious paper would require more effort than this, but for the most part it was “load assignment into brain, study assignment mentally until T-2d, write assignment, submit”. If several essays are due, then they have to be staged at various days numbered T-2d through T-5d for example — and it’s really important to not depend on T-1d existing at all due to courseware/internet/power outages.
I could technically write a worse essay the day it was assigned, but ultimately, I’m turning in A-tier work by this method. The hardest lesson was that I have to try not to wait until T-1d, because there’s a lot of risk encoded in that and it outweighs the value derived from having an extra day to think about it while I do other things.
But it wasn’t about “free time slipping away” — it’s just that I’m writing crap throwaway work that doesn’t matter after it’s done, and so I can barely motivate to care relative to literally anything else in my life that matters. Thus the T-2d compromise: I’m not about to give them precedence over literally anything, but I will concede that I do need to do so one day early, however boring it feels, because I’d rather have a crap day at T-2d than the same crap day at T-1d with the unproductive anxiety of risking a class-retake if my internet drops out.
Notably, when I actually genuinely care about what I’m writing, I’ll spend weeks researching sources and studying arguments and selecting quotes and then assembling it all over a couple days into a work of art — but assembly day is still always as late as possible in the time window assigned, because by then I’m most able to think and write about it efficiently and with a minimum of frustration. Not a zero of frustration, that is — I am a grouchy writer — but I’m healthy-grouchy on T-2d and bitter-grouchy on T-1d, so I do make the effort to put in my writing that day early now.
So: for your coaching efforts, try working with students to construct a working calendar that has non-writing activities in the leadup and then writing activities at the end. ie assuming a 7-day window,
T-7d: Assignment given: Read the assignment. (Seems obvious; is not obvious!)
T-6d: Think about your argument during your free time, while playing games or out at coffee or whatever.
T-5d: Try to construct a very halfass outline on a piece of paper. One sentence per argument you’d like to make, draw arrows to rearrange them. Not complete sentences, not punctuated, doesn’t have any structure at all. Point is that trying will help brain coalesce.
T-4d: Research references for fun. End up with far too many. Start highlighting quotes to yourself using highlighter or digital tools. If you’re going to experiment with a new tool, get it working and productive in 2 hours or discard it and do something shittier.
T-3d: Bind quotes to your argument phrases from that halfass outline. This may force reorg of outline; cool. Compile Works Cited from whatever you end up using so that you don’t have to fuck around with it tomorrow.
T-2d: Write paper, referring to outline / phrases handwritten note. Do one paragraph at a time. Plan to spend your entire day on this with 1 hour away from desk handling bio/sanity needs for every 2 hours at desk. Enforce that upon yourself.
T-1d: Finish whatever writing you didn’t feel like you were prepared to write on T-2d. Ideally try to do this earlier in the day than later, since that every hour you let this slip towards midnight l measurably increases your chances of a life outage causing you to fail the class.
The point of this schedule is to bake in the daydreaming / slow cooker aspect of the creative process but to keep it on the rails. I play video games extensively when I’m thinking about a paper because I can feed my literary brain the assignment to simmer and then go occupy my reflex brain with the game. I usually end up having to use some T-1d time but I’m getting better at managing my life’s dependencies ie. Food and Water and Sleep so that I’m more reliably at T-2d completion :)
I was thinking of exams rather than day to day learning of basics, seeing tests as teaching rather than assessing. Point being you can't cheat at learning.
The calculator can't fake that you understand e.g. how to differentiate. The working is the bulk of available exam marks but the calculator only gives answers.
Is it unambiguous cheating if it doesn't help?
Knowing what the answer is supposed to be makes it much easier to reverse engineer the workings, and it lets you double check that your workings are correct. Also many of the more advanced graphing calculators (don't know the TI-89 specifically) have build in CAS systems and can do symbolic differentiation and as such help with showing your workings as well.
Plus the fact that these calculators let you store arbitrary data, so you can have your entire textbook stored in memory if you wanted.
I want to say something like "who learns advanced calculator functions solo but not the taught subject?" but of course shortcuts-by-rote spread like lice. I suppose that's the thrust of the problem with AI in schools
Ludditism is the human death drive externalized for the modern age. Reject it in all of its incarnations.
There's objectively no value with learning how to perform calculations by hand that CAS's can perform better automatically.
We have well over the number of human beings on earth amounts of calculator capable calculating devices. I will not "be in a situation where I don't have a calculator".
Time spent not teaching folks how to use computational aids of math and teaching them to do it the "hard way" is time robbed from them.
Those who invent systems which subvert or short circuit the attempts to enslave individuals into doing hard work for hard works sake are ontologically good. Those who try to defend work for works sake are ontologically evil and I hope that they reincarnate into a durian fruit in the next life.
Your attitude makes sense in engineering but not in a math class for example.
By that logic, you must also permit students another extremely well-established labor-saving technique: Hiring someone else do the homework/exams on their behalf!
After all, for students from certain families, forcing them to use the calculator personally would likewise be "robbing time from them", denying them valuable experience honing their skills in managing and subcontracting. They'll never be in a situation where they don't have a calc^H^H^H^H lackey.
I trust we can agree that (A) permitting outsourcing is absurd, (B) that cheating is not "ontologically good" even when reduces the total human labor, and (C) prohibiting the practice should not put a teacher into a state of stinky-fruit damnation.
I submit that we do care about how students do it and which skills they use, even if we disagree on what those are. It is not as simple as saying that good education means spurring them to supply Valid Answers By Any Contemporary Efficient Means.
We definitely should allow that for MBA students, since that will most closely mirror what they'll be doing once they get into the work force.
Frankly, I'm quite glad to be able to do basic arithmetic in my head -- having to pick up a calculator every time I need to multiply two numbers would burn through far more time over the course of my life than what I spent in Kindergarten (or whenever) learning to do it myself.
All you want is for people to be better slaves for your capitalist machine. Never learning how to do things themselves: only consuming, buying what could otherwise have been achieved by the simple (but non-GDP-increasing) exercise of their own mind. Your rhetoric betrays your own class.
You can't even imagine a reality any different than "Capitalist Realism"
Go read Mark Fischer.
I imagine some guy several thousand years ago: "Once we finish teaching this animal to pull the plough for us, descendants of The Tribe Between The Two Rivers shall have lives of pure leisure!"
We were not permitted to use a TI-89 in math courses since we were expected to learn the underlying concepts, which is much more than learning how to use a tool.
The reasoning was similar in physics. The instructors couldn't care less whether we used the CAS functionality of the TI-89 because that wasn't a part of their curriculum, but they were concerned about students downloading solver programs. (Most of the instructors would agree that creating our own programs to solve problems would be a valuable learning tool, but they had no way to determine whether we created those programs or someone else.)
Academic courses tend to be biased towards laying the foundations so that we can build upon our own knowledge. There are other types of courses that are purely concerned about applications and the use of tools.
You're not learning how to perform calculations in most calculus courses. Perhaps not even most algebra courses.
A given expression can be simplified/factored in multiple ways. That TI-89 is going to do it only one way. When working with a typical physics/engineering problem, the way you decide to arrange the terms can help tremendously in understanding the physics of the situation.
I hear this take only from people who've not gone ahead and done advanced (graduate) level work.
The tool itself is not questionable or acceptable, it becomes questionable or acceptable depending on the usage. A pencil and paper can be questionable if the test is designed and expected to be completed without it.
You can design tests where an LLM spitting out python is an expected tool, but what are you testing for then? I doubt there are classes that teach whatever that test would be for yet.
It has been quite an adjustment, the hardest part isn't really the content, there have just been assessment tasks that are more demanding in time than cognition. I think a big part has been these are introductory subjects.
You have much less agency over your schedule outside university itself than you would have when you're working as well, which has been a massive PITA having previously worked as a software engineer for quite a while. As an institution it just doesn't respect peoples time, which is fine when you're much younger and you're just coming out of school (another institution that doesn't care for your time outside of it).
I think my problems have been compounded by the fact this is a undergraduate degree not a masters. There was only a masters for applied Econ in my area (which is much less data and math focused than econometrics), I've had this conversation 100 times, not looking to repeat that, trust me the moment I see an option to change to a masters in my area of interest in the city I live in, I would jump at it. But my choice to go back to uni to some extent is an as much an act of consumption as it is getting a piece of paper to saying "you can trust this man with econ".
Anyways in these first year subjects, you'll have assessment tasks like "make a recording of you demonstrating your understanding of <basic function in excel>, explain the value of <basic function in excel>". They are easy subjects but they are also really time consuming so to some extent it feels degrading. I would take steps to skip the subject based on my prior experience but the focus of these subjects isn't even Excel, its just thrown in there because they anticipate you have little experience with these things and it'll be useful in later subjects when it becomes assumed knowledge.
Edit: phrasing
12 credits is normally a minimum. That's roughly 12 hours a week in classroom, give or take. You need 3x that number of hours out of the classroom--that's 36 hours.
So, you're at 48 hours of academic work every single week. A 15 credit load means a 60 hour week.
Most people working jobs would start to complain about burnout at 50+ hours per week for 4 years running. They would almost certainly complain at 60 hours per week.
You also have to consider that a semester is 16 weeks. The first week or 2 of each semester is very light.
So we’re talking 30/52 weeks a year for most people.
For most people, you’ll never have that much free time again in your life.
College feels like a lot of work because you aren’t good at time management yet. And you remember the last few weeks of each semester where you are actually extremely busy.
Please tell that to the undergrad STEM professors, please. Almost all of mine had an assignment first class that was due by the third class.
Freshman engineering is generally Calc I, Physics I, Chem I, and English Composition/Writing and often some random engineering/computer thing. I assure you that schedule sucks even harder that it looks like it sucks, and it gets more time consuming as the years progress.
While there were lots of Party Hardy(tm) types in the College of Arts, the ones I knew who were taking their degrees seriously were working every bit as hard as the STEM folks. Possibly, they were working harder as they needed a lot more extracurricular work and achievement since what they were doing didn't have nice, clean objective measures like STEM does. They spent a lot of time being unpaid labor at functions and networking like crazy.
By contrast, no matter how many hours they worked me at my summer internship, it simply never compared to the grind at school.
We both agree that we had far more time in undergrad than at anytime since.
We frequently had small assignments the first week, but they were universally not worth much because many people aren’t even in the class yet since drop/add ran through the whole first week. They were also not very much work.
The point is there is no way you were spending 12-16 hours per class the first week.
I never spent that much time in Physics I, Calculus I or any English Class. The only classes I spent that much time on or more were higher level project based CS classes.
And even then I wasn’t spending that much time on them till closer to midway through the semester.
Then you get nearly the whole month of December off—Spring Break, Thanksgiving, and Summer.
No one expects anything from you at a summer internship. Companies don’t expect anything from experienced employees in the first couple months. Harder than a summer internship doesn’t say much.
On top of that you’ll never have fewer responsibilities than when you were at school.
It just seems like you had no free time because it’s the time in your life when you haven’t learned time management yet, and you remember the last few weeks leading up to final exams and semester project due dates.
You are so far removed from the ability of the average student that your personal observations about college simply don't generalize.
> I never spent that much time in Physics I, Calculus I or any English Class.
Do you understand just how far out of the norm that is?
Not everybody is coming from elite high schools and can blow through a college 101 class. The vast majority of engineers fight through all three of those--especially an English composition class. I had friends who did poorly in Calc I, dropped it but stayed in the class just so they plowed through it next semester. These aren't people "fooling around" with bad "time management". They were bog standard state school students who needed to get through, get out, and start making money. They were first college generation who didn't have rich parents backing them. They were motivated and got out in 4 years--something that most college students regard as difficult.
> No one expects anything from you at a summer internship.
Seriously? As a summer intern I always had deliverables. When I became a manager instead, we always had deliverables for co-ops and interns.
I dropped out the first time—2 years into a history degree—because I was working full time.
Eventually I moved home, and started over with CS. Despite CS being a lot harder, I had plenty of free time to work on a startup, build side projects, and play video games.
The reason was because a few years of experience made me much better at time management and prioritization.
I’m not saying you or anyone else was bad at time management as an insult. It’s just that college the time in your adult life when you have the absolute least experience at time management, so most people are very bad at it.
But also when you average it over the whole semester, none of my friends, even the ones who were bad at their classes spent 3 hours per credit hour outside of class. The ones who were bad at it tended to just skate by with Cs.
> As a summer intern I always had deliverables.
No one cares about those deliverables though. They aren’t trusting summer interns to do anything that really needs to get done.
IMO you’re observing Parkinson’s law, “work expands to fill the allotted time”. The students who take a million classes look like they have time management in check, and I’m sure some do, but they also are benefiting from the inverse of Parkinson’s law — if they can _only_ allot X hours a week to a task, they’re going to make the most of those X hours. This practically holds regardless of the student’s time management skills. I should know, I have horrible time management, and only succeed by overcommitting and rising to the challenge.
[1] https://en.wikipedia.org/wiki/Parkinson's_law?wprov=sfti1
Despite how much I wish it weren't, this is exactly me. I seem to be able to only work under pressure, so as I got later into my degree I would start my work later and later, until it was standard that I'd start at 4am on the day it was due. A terrible way to function or to learn
But at some point, you either stop procrastinating as you find the absolute limit, or fail outright (which is why survivorship bias exists if you look at the student body and see the amount of procrastination).
[...] I expected them to laugh, but they sat in silence. When they did finally speak, I am happy to say that it bothered them. They didn’t like hearing how their AI-generated submissions, in which they’d clearly felt some personal stake, amounted to a big bowl of bland, flavorless word salad.
But that's what it has been trained on - almost all academic writing is bland flavorless word salad, and this is extremely noticeable in title fads. I have a nearly decade-long game running with my friend where me make up absolutely bullshit concepts that could nevertheless be plausibly published in a journal, and the process has been going on long before that.
'Verbing the noun: towards a genericization theory of expressivity in high-entropy counter-heterogeneity' describes an ongoing problem in academic writing where novelty is deprioritized in favor of acceptability by an evermore tightly circumscribed set of peer professionals whose socioeconomic interests favor the establishment of intellectual stasis that maximally conserves positionality in a quais-Simmelian network space parameterized by income, tenure proximity, and citation count.
Or put more clearly, the more academics write to impress each other instead of to reach the public, the more generic their titles and language will be. Being able to parse and regurgitate wordy titles and abstracts constitutes table stakes in academia, so the incentives tilt toward burying the lede any original proposals as deeply as possible so as the minimize the career-damaging possibility of rejection on technical/syntactical grounds.
I understand why much of the discussion about AI and university education has focused on first-year writing classes in the U.S. Some of my own first experiments with ChatGPT in December 2022 were having it write school-like essays [1].
Over the past few months, I cotaught a university class in which we also had first-year students use and reflect on their use of AI in their classwork. But the context was different: the class was a seminar on science communication (how to make science engaging to children and the general public), and most of the time was spent with the students doing group projects. Also, the class was at a university in Tokyo, and we taught it in Japanese.
We have just started analyzing the feedback from the students, but my impression is that they were less conflicted about the use of AI for their group work than they might have been if they had been doing their projects as individuals.
Meanwhile, as the semester progressed, agentic frameworks started to mature. I spent a lot of time on my own experimenting with Claude Code and Gemini CLI. While none of the students in that class seemed to use them, it became clear to me that such higher-level cognitive tools will pose an even greater challenge to higher education than essay-writing chatbots do now.
[1] https://www.gally.net/temp/202212chatgpt/defaultessay.html
Should education systems aim for that for all of their students? Certainly, because AI alone is not sufficient to raise the bar. As impressive an AI is when it seems to invent a new molecule, it is still only possible because of the original works of many people.
Has it sparked that? I'm doubtful but I'd be very interested if anyone has a references about such introspection.
I grew up reading Douglas Hofstadter back in the 70s and what I appreciated with his ideas was using AI to illuminate what is human. His wave of AI failed, of course. But still, it's disheartening how little of that kind of inquiry seems missing from the current wave of AI.
For example, in an English class with a lot of essay-writing assignments, the assignments could simply be worth 0% of the final mark. There would still be deadlines as usual, and they would be marked as usual, but the students would be free to do them or not as they pleased. The catch would be that the *proctored, for-credit* exams would demand that they write similar essays, which would then be graded based on the knowledge/skills the students would have been expected to gain if they'd done the assignments.
Advantages:
- No more issues with cheating.
- Students get to manage (or learn to manage) their own time and priorities, as is expected of adults, without being whipped as much with the cane of class grades.
- The advanced students who can already write clearly, concisely and convincingly (or whatever the objectives are of the writing exercises) don't have to waste time with unneeded assignments.
- If students skip the assignments, learn to write on their own time using ChatGPT and friends, and can demonstrate their skills in exam conditions, then it's a win-win.
This all requires that whoever is in charge of the class have clear and testable learning goals in mind -- which, alas, they all-too-often do not.
Mind you, I don't really have any alternative suggestions.
Let's get real here. I know why these nonsensical memes keep propagating but dear god. People will just believe anything these days, including that gas stoves cause asthma or whatever other bullshit is being peddled.
But I'm not particularly brilliant, in fact I wouldn't be terribly surprised if I have undiagnosed ADHD. My test taking performance trick, which I freely told everybody to their annoyance, was very simple. I knew the material! Read the assigned texts, do the optional homework, pay attention in class. If you know the material you don't have to try to cram it into your brain in the last half hour before the test. If you know the material you don't have to try to reason it out from first principles during the test. You just go in, fill out the easy answers straight away, go back and do a second pass for the tricky questions, and that's it. If you have to sit there wracking your brain for 30 minutes on a single problem it's because you already fucked up with how you approached the course weeks ago.
Again, I'm not special for this. There were a handful of other students who were as fast as me. We'd sit in the hall waiting for our friends, look at each other and say "you knew all this stuff too, huh?" "yeah of course"
Bullshitting though tests is a learnable / trainable skill, but schools generally do not teach it very coherently or well and most students do not deliberately practice it. It generally doesn't have that much to do with the content or other skills intentionally taught by any particular course or by schools in general (there's decent overlap with the skills involved in competitive debate and extemporaneous speech, which some students participate in as an extracurricular activity). Rating students on how good they are at bullshitting their way through exams is sadly a significant part of the way our education system is focused and organized, but in my opinion it is not a valuable or particularly valid approach. There are certain professional contexts/tasks where this kind of skill is useful, but developing it per se shouldn't be the focus of the education system.
Sometimes this and related skills are summarized as "intelligence" ("oh she aced the test without studying, she must just be really smart", etc.), but in my opinion it's quite a misleading use of the word.
This is thing.
If this choice is between:
1. A gameable system that will be gamed by most students.
2. An ungameable system that will unfairly punish those bad under pressure and time constraints.
There isn't really a choice at all.
One option would be a school-provided proctoring system, allowing teachers to outsource the actual test-taking times. It could be done outside of class time, at the student's convenience, and they could have 3-4 hours if they chose.
Given modern communication technology it’s still gameable
And rates of cheating of "well under 1%" vs "well over 50%".
Even if we allow for less rigorous proctoring standards, we're still probably talking about "2-3%" vs "well over 50%".
Learning to perform under pressure is the main purpose of attending college.
But we could do what I'll call a "monastic exam".
You've got a week, not an hour, but it's in a little monastery and you don't have your phone or other unapproved tools.
I will say it's not practical to have exams that long. In this case, the dorm required me to move out immediately after the exam and my parents were waiting to pick me up, so I decided to leave after 4 hours to avoid unnecessary panic or having to drive overnight. In hindsight, the professor probably would have let me make a phone call, but that didn't occur to me at the time.
Look at any list of advice for new college students and almost every one of them includes "go to class". Simply attending class is way easier than homework and yet, when there's no short term consequences for not doing it, plenty of students will just not do it.
Cheating is another great example. Cheating in college is rampant because kids don't want to do the work they're assigned. I don't understand the logic behind the idea that if you tell all the kids currently using ChatGPT to write their essays, "Hey, you don't actually have to write that essay at all" that you think they will somehow choose to write it anyway. They're already choosing to ignore the long term benefits of homework even when there are short term consequences, so I don't see how removing those short term consequences will make things better.
If you tell kids there are no immediate consequences for not doing homework, many of them just won't do it and they will fail because they haven't learned anything.
Maybe you're okay with that. Honestly, I'm not actually trying to convince you that it's a bad idea. I just think if your proposal is based on the idea that kids will choose to something boring that they don't have to do in the short term because it benefits them in the long term you're overestimating a lot of kids (and adults for that matter).
I unironically believe if you tell all the kids they don't have to write the essay at all, much more will choose to write it.
Kids cheat not just because they're lazy. Cheating makes people feel smart. The fact you can get credits by doing very little while others work their asses off is rewarding and self-validating.
The big issue of exam-only approach is that a one-hour exam is not enough to evaluate a student's performance, unless your educational goal is just to make students memorize stuff by rote. I'd consider a 3-hour open-book exam bare minimal. But if every class does that it'll be too exhausting.
Fun prank: set all of the clocks in your dorm neighbor’s room to different wrong times. Guy across the hall knew we were messing with him, trusted his watch - which had the correct time, but wrong alarm time. Realized he had a problem when he had hot water in the shower and no one was around. He was only 45 min late to the exam. Good times.
He did fine in his exam. 3 hrs was overkill. Sometimes you can be your own worst enemy.
It was the 80’s. I guess kids these days are soft.
There needs to be a reward for doing essays. That reward can be emotional eg. "the teacher I respect liked my essay" or "my essay was read in class" or "the teacher gives feedback that makes me feel a sense of growth". In that case, maybe kids will do it.
However, I think it's hard for a teacher to inspire respect to a classroom and the difficulty scales with the number of people in the class, so grades are used as a hack.
You want me to don't know something? I better make sure to get to know everything about that. You push me to do stuff? Why should I care, if you already do.
I agree that open-book exams, or at least a closed-book portion followed by an open-book portion, is important to actually gauge the student's abilities rather than his/her capability to cram.
I am 100% certain quite a lot of people cheat because they procrastinated and don't have time to learn. Or because they indeed were lazy to learn. Or because they cant learn, because they course is too hard for them.
Or because video games and youtube are more fun.
So, you have 2 kids who are equally bright, and you tell one "you don't have to do these assignments but there is a test at the end" and the other "you have an 80% chance if failing if you don't do these assignments. Analyse each assignments and feedback for shiboleths like the way they ask you to structure your introduction and optimize for demonstrating you know these shiboleths over everything else"
University is a wonderful petri dish for growing into who you want to be. You have access to expertise and resources abs a certain kind of institutional credibility. Few students actually use these fully and the ones who do were told to. You need some idea who you want to be and why, and this is developed in you by other people. Children don't just know stuff.
I think these are positive changes if and only if we accompany them with systematic study skills and self management courses and bridge this gap.
I seriously doubt that. In my experience many students won't do anything that doesn't directly contribute to their grade.
We actually had this exact thing at my university. One sophomore level weed out class was a "self paced" electrical engineering class. It was called self paced because you were given a textbook and were free to work through it at your own pace. But to finish the class by the end of the semester you had to average 2 chapters completed per week, and completing a chapter not only included finishing a problem set and taking a test which you had to score 90%+ on (and were required to finish another problem set and retake it otherwise), but on occasion also demoing some skill in the lab.
It was brutal, but one of the most educational classes I've ever taken - and obviously not just because of what I learned about electrical engineering. Of course it seems modern universities have just become profit-driven degree treadmills. Weeding out students? That's reducing profit! And yeah looking back at my uni's page it seems this class is no longer self paced. Lol. And that's at a top 10 school. The enshittification of education.
Perhaps that begs the question, if those kids can’t handle self-directed education, why are we putting them there in the first place, but that’s definitely a grey area, and there are hundreds of thousands of students who are smart enough to do well in higher education and skilled work, but weren’t disciplined enough to handle what you’re describing as freshmen.
A teacher who thinks this way is probably in the wrong profession. A university that operates this way is failing to educate the students it admitted.
From my experience studying electrical and computer engineering, I definitely prefer that they chose to put hard electrical engineering courses in the first semesters because I knew immediately not to focus on them because I didn't like them.
Some students also just don't have the aptitude for an Engineering or Computer Science degree. It's better for everyone if this is figured out early. I know plenty of people that dropped out of a Computer Science degree because they hated it or thought it would be a great way to make money and were in over their head.
We had classes that were for 'weeding out' students in Computer Science. They involved calculus because if you couldn't pass this class, you wouldn't be able to handle the 5 or so classes after this class that required it.
With lecture notes/slides available online, well prepared books and study forums readily available - in-person attendance can feel archaic.
We may be experiencing a similar dynamic in education with AI. In a world where we can create individualized curriculum’s for each student encompassing the entire tree of knowledge - Perhaps it’s time to rethink how we educate students rather than push them into lecture halls designed for the Middle Ages.
People thrive under regularity, and young people (especially) tend not to understand that. Similarly, being able to focus on a single thing is a kind of super-power, while multi-tasking generally hurts performance on tasks.
Going to class (and paying attention) means that you've got a regular period of focus on the class topic. That combination of regularity and focus translates into long-term learning and better performance.
In a few of my classes, I have final projects that teams work on. I also have presentations. I used to require them of all students; and quickly learned this is a good way to waste valuable time.
Now, all my presentations are completely optional for NO CREDIT. You don't get penalized if you don't do them, and perhaps more importantly, I give ZERO EXTRA CREDIT for doing them.
As you can imagine, every single presentation I've gotten from this has been absolutely worth it.
Others had some pretty cool things that ended up in real life; I believe the official timers for the Florida Supreme Court testimony things came from one of my classes.
You can take that one step further. What kind of signal does “I can afford to go to University and not worry about credentials” send? I’d argue that’s realistic only for people who are willing to admit that they belong to a leisure class. In the US at least, we like to flatter the leisure class with the pretense that they worked hard to get there.
Maybe this is true of liberal arts or business degrees? I don't know, but I don't think this is the opinion of anyone who went to engineering school.
So this isn't all that crazy.
I think optional homework works for classes that are obscure enough only somewhat intrinsically motivated students would consider taking them, but in mandatory classes or trendy majors, there's going to be many people who need a bit more external motivation to study.
> AFAIK this was because under the old system, there had been too many students registering for exams despite being woefully unprepared, and then predictably failing as a result.
True. That's the real Dunning-Kruger problem: incompetent people do not know how much help they need to get competent. It is our job to show them their weaknesses as early as possible so that they can effectively work on them.
(I believe that state-funded universities (as in Germany) have some obligation to not only educate the self-motivated top 1% but also offer a solid education even for less perfect students - at least if there is a societal need for their competences.)
Another, more important, reason is that written exams are not good tests of programming competence - especially as tasks and frameworks get more complex. We want to assign good grades to students who are competent at developing software in realistic settings, not in highly artificial exam settings.
Of course, the paper by 'Dunning-Kruger' never showed anything like that.
From the abstract:
"Paradoxically, improving the skills of participants, and thus increasing their metacognitive competence, helped them recognize the limitations of their abilities."
All they found was a statistical artefact.
You can just have a look at the big picture at the top of the Wikipedia article: https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
> Relation between average self-perceived performance and average actual performance on a college exam.[1] The red area shows the tendency of low performers to overestimate their abilities. Nevertheless, low performers' self-assessment is lower than that of high performers.
So regression towards the mean explain the entire effect.
See also https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect#...
This is still very much compatible with my claim that unskilled people profit from being reminded (repeatedly, not just in the exam at the end of the semester) that they know less than they think. I will avoid conflating this with the Dunning-Kruger effect in the future (Thanks!).
An recent study found that medical students' estimate of their own intelligence gets lower right after taking an IQ test (confirming the better-than-average effect). But one week later, their self-estimated intelligence returns back to their pre-test levels. To me this suggests that students (and all others) will overestimate their abilities - and invest less time in learning - if they are not constantly given feedback.
https://www.sciencedirect.com/science/article/abs/pii/S01918...
many kids fail to make the transition from spoon-feeding to self-learning, but those who do then begin to realize that they can go as fast as they are able and need not follow the herd. they also develop a strong sense of whether they’ve understood each booklet or not. it leads to a competition for learning fast AND well because there are also traditional proctored checkpoint exams from time to time plus kids do the ordinary standardized tests to calibrate.
i feel it’s an excellent system that prioritizes learning over conformity though it is obviously not a candidate for mass adoption because many kids wash out after making no progress for a while.
You have to get a special blue book with a couple blank pages and then write an essay with the prompt that's given at exam time. Then you turn in the book at the end of the exam. I think it's a great idea and was surprised more classes didn't work that way but I guess it's like you say: grading written assignments like this is a lot of work.
He wasn’t worried about llms, they were not around, but plagiarism. It worked well.
It’s also a disadvantage for people with test anxiety.
We run university programs at my company, and arrived at this bit of insight as well. That said, some of your points are incorrect or incomplete:
- You can't build systems assuming responsible individuals. These systems are guaranteed to fail. Instead, assume individuals are mould-able, and build a system which nurtures discipline towards goals. This works. - There are still issues with cheating, but it's more of an older way of thinking, that we developed methods to reset. - Advanced students need to be given more challenging assignments - quantum of assignments should be the same no matter the capability of students. This solution was unworkable until GenAI came about.
Looked from a pure individual skill-building perspective your ideas are alluring, but if one looks at completion rates of any online courses (Udemy/Coursera - under 4%), then one understands why physical cohort-led education system can work.
Happy to chat with anyone who'd like to delve deeper on this.
As someone with a 96+% 'failure' rate on Udemy/Coursera I honestly don't see the relevance of this statistic. Most people going to University are there primarily because they want/need the degree. That piece of paper is really valuable, perhaps even more so than the knowledge gained. The piece of 'paper' offered by Coursera/Udemy etc. has basically zero value, so the people taking those courses are doing it almost exclusively for the knowledge they offer. Once you've learned what you wanted to learn from the course there is very little incentive to go the extra mile and go for the 'completion'.
I understand how from an individual's pov what you said makes sense. Similarly I hope you understand why from the system's perspective: it's the effort that's mandated and not just the proficiency.
Employers and others (higher education orgs, etc) care a lot about sustained effort, alongside proficiency. Only proficiency-focused systems (like Udemy/Coursera/Youtube) are not respected as credentials, since they do not showcase this.
The difference I notice with AI is that the bell curve is nearly inverted. You have the good students, who use AI to support their learning. You have the students who let AI do their assignments, and then fail miserably on the exam. And there is hardly anyone left in the middle.
A lot of intro programming builds directly on previous lessons, much more so than, e.g., maths. If you missed how variables work (off sick, just didn't get it, whatever), you're still stuck when it comes to functions and anything else following and then you're going to fail - it was quite predictable. We studied other university courses and nothing came close to the pattern we were seeing, except "computing for chemistry" or something, which was basically the same sort of course just in a different department.
So we added explicit catch-up classes a few days after a topic was covered so if you missed it, you could get quite personal help on getting back up to speed. This really shifted the distribution to the right, then the people who failed were either those who just didn't care, or those under more extreme circumstances where this couldn't help (or those who just could not learn programming for love nor money but that was rare ime.)
I think you meant "bimodal", not "binomial". They mean roughly opposite things here. :)
In all seriousness, I don't see the value in this at all. Why would I want to know a statistically likely essay? Wouldn't I rather know what the student thinks?
For example, in the UK, it was shown that biasing course results towards exam marks caused woman to perform worse than men. But when results included assignments, women generally performed better.
This is obviously a generalisation but it is one of the reasons why so many courses now take assignments into account for their final grade.
However what you mention about different people being better in different circumstances reminds of what our maths courses typically did, it was called "plussage" IIRC. Basically, the scores were calculated, and you got the best score from a 40% exam/60% assignment weighting or a 60%/40% (or something, the exact values are lost to time.) So if you were bad at exams but had done the work through the semester, you got a boost. Or if you were bad at deadlines but had still studied, you weren't (too) penalised.
However, when some colleagues tried homework as 0% for introductory courses, most students omitted the homework, then failed the exams. Modern students seem to require explicit incentive to work, otherwise the usual: scrolling upon flat screen devices, hedonism, and so forth.
In this case, who has failed: the student, or the professor?
And to pass an exam students have to prepare for the exam. Homework will only help there if it is similar to the exam.
One time I had to evaluate a written exam where the professor had set up a trap. There was a question that looked like a standard question from homework, but if you used the standard-techniques from the course your calculations didn't work - it was a nasty special case. Most people that started with that question just burned 30 minutes without getting anywhere... a lot of students failed, but at least they learned something about life...
And Oral exams are different. Giving a quick well prepared answer and being able to solve difficult tasks over a few days are completely different skills. Students there prepare for the professor. There are transcripts of previous oral exams. And professors change over the years - the final tough question of an excellent student will a few years later become a starting question. People that didn't know that game and didn't have access to any transcripts were in serious trouble... None of the Homework would have helped in the oral exam.
And to pass an exam students have to prepare for the exam. Homework will only help there if it is similar to the exam."
That's not learning really. I can confidently say that because I was the one who unfortunately regressed to this during the uni, and the same story with my peers. One simply can't prepare to the multiple exams sufficiently in a few weeks time (or less). So the only path left is hysterical rote memorization of as much material as possible to squeeze in at a passing grade, and then immediately forget all of the materials in a few months time. Burst of "learning" twice a year for a short time doesn't translate into real learning.
And that's for some simple courses during first few years. Specialist courses later in the program sometimes are impossible to rush "learn". When I tried to pull this off for Probability Theory course, I've failed spectacularly to get even lowest passing mark at first try. And others failed the same way.
If good results are important, it's most important to know what happens in the exam - and adapt professionally. Learning is fun. I myself always did a lot of homework. But I wish I had been more professional - constant challenges are fun, but most of it is not very time-efficient.
- People need to learn how to write. The quality of student writing was one of the biggest criticisms of students when I was in university, and that was 30 years ago. Writing will only improve with practice and someone to evaluate it. Very few people will be able to learn how to write properly by reading about it, and even fewer people will even realize that you can learn how to write by reading the work of other people (which is important for learning about style in a particular field). For most students, even well meaning ones, no grade means no work done.
- A certain segment of the student population will find ways to cheat anyhow. All you have done is raised the bar so that, hopefully, fewer people will cheat. Quite frankly, I don't know how helpful that is if the "top" of the class moves on since the top of the class tends to be defined by their GPA.
- Test anxiety is a real thing. Different people go to school for different reasons, not all of which lead to high pressure careers. Do we really want to limit who can effectively access an education because of that?
There is no easy solution to this problem. Likely the best solution would be to remove traditional assignments and exams from the loop altogether and having students work directly with their instructors. Yet this has it's own set of problems (it assumes both parties are honest, it is difficult to ensure consistency in the delivery of curriculum, etc.).
But the point of the university is not only teach English grammar and math operations, but also to work in teams, manage yourself, etc. The social stuff. And I suspect a significant number of students benefit from it. And I also suspect that by doing this at scale, the whole society benefits on average.
Removing all control and only checking the knowledge during the exam would lead to a lot of students never catching up. It is likely that it will also lead to the top students being more and more lax and eventually also falling behind.
The whole idea hinges on the base motivation - why do we need primary/secondary etc. education at all? To produce a dozen elite self motivated geniuses per year per country? Then your proposal would work perfectly. Or maybe motivation is different?..
AI can do the full assignment and do it faster and better
The school thinks that the student joined to learn and the school acts this way up to a certain extent. But in the end grading a failing student isn't helping him learn. The school ends up fostering competition and stress internally and uses GPA as a ranking system.
Why a school does all this to a student when the student pays them? Like why give students bad grades when the student gives them money? Because of integrity. Jobs trust the school if the school has prestige and integrity. In order to maintain that prestige and integrity the school needs to pressure and rank the students. Then when the school is sufficiently prestigious companies know and students pay tuition to get the integrity and thus get the job.
The more prestigious the school the more is invested into "ranking". Only at schools with lower prestige do you get more of a "learning environment". But companies prefer candidates who can make it through a ranking system over ones that have to be coddled in an environment that fosters learning.
It is not an illogical concept. Students that make it into these prestigious schools generally have much higher IQ. This is a data driven fact along with the fact that entrance to many schools involves the equivalent of an IQ test. People with higher IQ perform better. So it's the most rational choice for companies to want to pick students from schools that are bad at teaching and better at ranking.
Some students go to higher ed for learning. Most go to run the gauntlet of a ranking system and end up with a degree that bestows prestige and excellence.
There's my ultra real take on your "why don't students want to learn" bs. But it's not like I said anything you didn't already know.
Oh yeah, Marlboro is out bidness.
That's too broad to be of any use, if the math class is teaching you to calculate in your head, then using any calculator is cheating. If the math class is teaching some algebra equation solving skills, then using a programmable calculator that auto solves them is cheating.
That's the similar issue with such experiments - they unfortunately aren't rigorous to provide any insight into education
When LLMs became widely usable, I was one of the people who really liked much of the writing that they did. I found it was relatively close to my writing style, which I consider to be good, despite the disagreement from those in the humanities. It was close enough to my own writing that I have even had people on Discord accuse me of using LLMs to write my messages for me, when I had not.
The linked article was clearly written by an English teacher. He criticizes AI-generated texts as “a big bowl of bland, flavorless word salad“. Now, there are many cases where LLMs output nonsense, but in cases where the writing logically flows, does not self-contradict in any way and avoids unnecessary repetition, “bland” and “flavorless” are good. The goal of writing is to convey information across space-time; writing that is “bland” and “flavorless” is the best way of conveying information.
I can see a number of things he did in his writing to avoid being “bland” and “flavorless”, and I consider them to be examples of poor writing:
He used dozens of idioms that make the text difficult for non-native speakers and unpleasant for native readers. He used a number of colloquialisms, including some that are inappropriate in professional contexts (although I will not repeat them since I refuse to write them). He used a word whose only definition is provided by Urban Dictionary and therefore is not even an official word:
https://www.urbandictionary.com/define.php?term=cyborging
He also brought politics into an apolitical topic. The injection of politics is a great way to derail any form of productive dialogue and should be avoided.
He used a story format of the kind that has infected journalism. It is very rare that the process by which something was learned is useful to readers and presenting it for dramatics wastes their time. That is with the exception of stories on the topic of security, where hearing the process is often genuinely informative. My aversion to this writing style is so severe that I have a standing policy to stop reading a news article the moment that I see that it uses this style for a topic that is not security-related. After reading his article, I will extend my policy to apply to essays by academics in the humanities too.
He made numerous attempts to evoke emotional responses to elicit agreement, rather than to make clear arguments based on facts. This is great for propaganda, but not so great for making points. Every one of these appeals to emotions is poor writing.
Beyond those things, he also did not properly cite sources in multiple places. To name a few, the quotes from Sam Altman and Annalee Newitz are uncited. As an academic, he should know better.
Some of these things might actually have places in certain types of writing. They certainly have places in propaganda. They also have places in fictional literature. However, they do not have places in attempts to argue a point.
I imagine if he corrected all of my criticisms, he would find the result to be “bland” and “flavorless”. That is how an attempt to argue a point should be.
I was there. We had been given slide rules, and a decent chunk of my 11-yr-old maths classes were devoted to teaching us how to use them. Calculators were banned because they "didn't teach us anything" (but somehow slide rules did? It didn't make any sense at the time either).
Over the course of the next few years, calculators became more acceptable, and by my 18-yr-old Maths A-level class, we were being advised on which scientific calculators to buy.
It's an interesting analogy, as TFA says.
Then I have had many bland curriculum reciters matching a semi decent youtube lecture, and some bad communicators where you were better of just reading the book they tried to teach.
There was also a few that couldn't really teach, but whose class was more like a standup comedy performance.
I'd say with the exception of the first and last group, ChatGPT would probably be a good if not better replacement for day to day teaching and mentoring.
Somehow this hits hard
I just hope "free market fails" people realise one day that the most overpriced industries (healthcare and education) are the ones free competition is not allowed in.
Other countries don’t have populations chomping at the bit to allow Amazon to dropship healthcare. They aren’t perfect, but the US system is singularly broken.
It's not a Technically oriented college, though it does teach the sciences, but I wouldn't go there (and didn't) for Engineering. That's what VA Tech is for, though the weather is worse, the campus not as nice, and not quite as prestigious.
The generally Liberal Arts system in the US is a strong contrast to the European system, where you often start focusing on a few subjects at the high school level, and then apply to a degree program where you have very few "outside the major" choices. My wife didn't take math or science after she was 12 (but she took languages), Oldest 2 kids are down to 3 or 4 subjects at the high school level (One is doing the Bio/Chem/Physics A levels, one did Phys/Math/Further Math/Geography). The last incidence of anything not STEMish was GCSE/Junior Cert/~15yr old level.
The kings/men who fight over a beauty (which can be a land/crown, or a woman) overcoming a monster/evil opponent or a series of opponents (The Illiad/Troy, Beowulf, King Arthur, ...)
The orphan/abandoned kid/extremely poor that conquers the kingdom, either through marriage or by conquering a (series of) challenges (Moses, Aladdin, Oedipus, Theseus, Heracles, Cinderella, the foundation story of islam/mohamed, ...)
The vizier/prime minister who decides he'll be king and becomes incredibly evil to achieve it. (robin hood, paradise lost (ie. the story of lucifer/the devil), queen esther, aladdin, MacBeth, ...)
The not-evil-but-quite-evil mother who sees her king/husband make a child, either with another woman or sees/fears she or her child will be put aside because of the other child and ... (Hera, Snow White, Medea, biblical story of Abraham, ...)
And then there's stories like Game of Thrones that are in large part a combination and integration of a lot of such stories: Circe and her children. John Snow, after being rejected, climbing up and up and up. Bran becoming king. Arya living through a sort of Herculanean heroic epic. And the kings fighting constantly.
Humans are clearly still a big step up even from state-of-the-art AI, but we are not infinitely creative like we like to think we are. It's a difference in scale, not a fundamental difference.
Then I was asked to review the questions for the final exam and I noticed some pretty weird constructions in the sample texts and sentences. Nothing completely grammatically wrong just unusual or semantically off. The example I recall was "... disproportionately affects men and women" but there were other worse examples. You can guess where this is going. But I knew that another native speaker had written the listening texts so I didn't want to directly criticise these, and I thought they might have been pulled from magazines or some corpus. But, of course, the course leader had generated them with chatgpt.
Some examples:
> Indeed, the recent cases of hyperinflation in Brazil, Argentina, and Poland illustrate that although hyperinflation is harmful to savers and disproportionately affects the poor (The Independent Review)
> A hearing is set Thursday on the new version of a legislative bill to eliminate scheduled pay increases for state employees that nixes a section that disproportionately affects rural legislative information offices (USA Today)
> Suicide is a key mental health issue which disproportionately affects men. (london.gov.uk)
"disproportionately affects men" makes sense.
I guess nobody tried to adjust the temperature via the API then...
You can make something really unbland then!
Eg. "Write a 3 sentence story" Temperature=1
The old cat, Bartholomew, stretched languidly in a patch of afternoon sun, his purr a rumbling motor. A mischievous bluebird, emboldened by the cat's sleepy demeanor, swooped down to steal a whisker for its nest. Bartholomew's eyes snapped open, a silent promise that their game was far from over.
Temperature=2:
The old clockmaker found a gear he didn't recognize, its teeth shimmering with an otherworldly light. He fitted it into the grand clock tower, and as the hands struck midnight, the town square was bathed in a soft, silver glow. From that day on, no one in the town ever seemed to age.
> includes both a human and a computer—and “surprisingly,
Today, many university students struggle with basic calculations. I'm not talking about long division of 8-digit numbers. I'm talking about things like expressing 2/10 as a percentage or knowing how many zeros to use when writing 1/1000 in decimal form. Many very bright students, at both undergraduate and postgraduate level, struggle with such things. It's heartbreaking.
I suspect this was just a vocal minority. This kind of vocal minority is common online to.
In its application that would mean for a LLM you define a sentence with one blank, restructure the sentence so the blank gets extracted and then enter it into the LLM to get exactly one word for your blank. On the oppoisite you wouldn't use the calculator for solving equations, but to also formulate them based on a abstract definition.
Furthermore "math" and "text" are two fundamentaly distinct categories. In math we express mathematical observations/context, which can be proven in its correctnes. We can conclude that 1 Apple + 2 Apples won't result in 5 Apples, but 3. Text may express emotions, toughts, directives, information or observations in context of its author. Not in every case it may be proven as math and we may attribute other values to it. Like honesty, authenticity, information and effort. If it was only for the information, you could filter out hallucinations and call it a day. But for other values you literally can not outsource the work to a tool. Except you live in a South Park episode.
> Navigating the Digital Age: A Personal Reflection on Technology
> Navigating the Digital Age: A Personal and Peer Perspective on Technology’s Role in Our Lives
> Navigating Connection: An Exploration of Personal Relationships with Technology
> From Connection to Disconnection: How Technology Shapes Our Social Lives
> From Connection to Distraction: How Technology Shapes Our Social and Academic Lives
> From Connection to Distraction: Navigating a Love-Hate Relationship with Technology
> Between Connection and Distraction: Navigating the Role of Technology in Our Lives
The immediate and obvious response to the above is: turn up the temperature.
As an aside: many of the critiques of genAI in this piece are about the quality of AI right now and won't hold if and when quality increases.
I assume the students were using the ChatGPT web app (as ~99% of teenagers would); there is no way to adjust the temperature in the web app.
This instantly paints the author as someone out of touch with their students. For the record, it comes from streamers/Twitch a la "Chat, is this real?"
I'm not saying everyone needs to know every meme, but starting off with this does taint the rest of the piece a little.
LeftHandPath•6mo ago
English teachers seem especially prone to that friendly and sporting demeanor the author has. Professors from the engineering schools are far more prescriptive, probably due to the nature of the material.