You must log in or register to comment.
The issue as I see it is that college is a barometer for success in life, which for the sake of brevity I’ll just say means economic success. It’s not just a place of learning, it’s the barrier to entry - and any metric that becomes a goal is prone to corruption.
A student won’t necessarily think of using AI as cheating themselves out of an education because we don’t teach the value of education except as a tool for economic success.
If the tool is education, the barrier to success is college, and the actual goal is to be economically successful, why wouldn’t a student start using a tool that breaks open that barrier with as little effort as possible?
especially in a world that seems to be repeatedly demonstrating to us that cheating and scumbaggery are the path to the highest echelons of success.
…where “success” means money and power - the stuff that these high profile scumbags care about, and the stuff that many otherwise decent people are taught should be the priority in their life.
Dumb take because inaccuracies and lies are not unique to LLMs.
half of what you’ll learn in medical school will be shown to be either dead wrong or out of date within five years of your graduation.
https://retractionwatch.com/2011/07/11/so-how-often-does-medical-consensus-turn-out-to-be-wrong/ and that’s 2011, it’s even worse now.
Real studying is knowning that no source is perfect but being able to craft a true picture of the world using the most efficient tools at hand and like it or not, objectively LLMs are pretty good already.
Idk, I think we’re back to “it depends on how you use it”. Once upon a time, the same was said of the internet in general, because people could just go online and copy and paste shit and share answers and stuff, but the Internet can also just be a really great educational resource in general. I think that using LLMs in non load-bearing “trust but verify” type roles (study buddies, brainstorming, very high level information searching) is actually really useful. One of my favorite uses of ChatGPT is when I have a concept so loose that I don’t even know the right question to Google, I can just kind of chat with the LLM and potentially refine a narrower, more google-able subject.
And just as back then, the problem is not with people using something to actually learn and deepen their understanding. It is with people blatantly cheating and knowing nothing because they don’t even read the thing they’re copying down.
trust but verify
The thing is that LLM is a professional bullshitter. It is actually trained to produce text that can fool ordinary person into thinking that it was produced by a human. The facts come 2nd.
To be fair, facts come second to many humans as well, so I dont know if you have much of a point there…
Yeah, I know. I use it for work in tech. If I encounter a novel (to me) problem and I don’t even know where to start with how to attack the problem, the LLM can sometimes save me hours of googling by just describing my problem to it in a chat format, describing what I want to do, and asking if there’s a commonly accepted approach or library for handling it. Sure, it sometimes hallucinate a library, but that’s why I go and verify and read the docs myself instead of just blindly copying and pasting.
That last step of verifying is often being skipped and is getting HARDER to do
The hallucinations spread like wildfire on the internet. Doesn’t matter what’s true; just what gets clicks that encourages more apparent “citations”. Another even worse fertilizer of false citations is the desire to push false narratives by power-hungry bastards
AI rabbit holes are getting too deep to verify. It really is important to keep digital hallucinations out of the academic loop, especially for things with life-and-death consequences like medical school
This is why I just use google to look for the NIH article I want, or I go straight to DynaMed or UpToDate. (The NIH does have a search function, but it’s terrible meaning it’s just easier to use google to find the link to the article I actually want.)
I’ll just add that I’ve had absolutely no benefit, just time wasted, when using the most popular services such as ChatGPT, Gemini and Copilot. Yes, sometimes it gets a few things right, mostly things that are REALLY easy and quick to find even when using a more limited search engine such as Mojeek. Most of the time these services will either spit out blatant lies or outdated info. That’s one side of the issue with these services, and I won’t even get into misinformation injected by their companies. The other main issue I find for research is that you can’t get a broader, let alone precise picture about anything without searching for information yourself, filtering the sources yourself and learning and building better criteria yourself, through trial and error. Oftentimes it’s good info that you weren’t initially searching for what makes your time well spent and it’s always better to have 10 people contrast information they’ve gathered from websites and libraries based on their preferences and concerns than 10 people doing the same thing with information they were served by an AI with minimal input and even less oversight. Better to train a light LLM model (or setup any other kind of automation that performs even better) with custom parameters at your home or office to do very specific tasks that are truly useful, reliable and time saving than trusting and feeding sloppy machines from sloppy companies.
I don’t trust LLMs for anything based on facts or complex reasoning. I’m a lawyer and any time I try asking an LLM a legal question, I get an answer ranging from “technically wrong/incomplete, but I can see how you got there” to “absolute fabrication.”
I actually think the best current use for LLMs is for itinerary planning and organizing thoughts. They’re pretty good at creating coherent, logical schedules based on sets of simple criteria as well as making communications more succinct (although still not perfect).
The only substantial uses i have for it are occasional blurbs of R code for charts, rewording a sentence, or finding a precise word when I can’t think of it
It’s decent at summarizing large blocks of text and pretty good for rewording things in a diplomatic/safe way. I used it the other day for work when I had to write a “staff appreciation” blurb and I couldn’t come up with a reasonable way to take my 4 sentences of aggressively pro-union rhetoric and turn it into one sentence that comes off pro-union but not anti-capitalist (edit: it still needed a editing pass-through to put it in my own voice and add some details, but it definitely got me close to what I needed)
I’d say it’s good at things you don’t need to be good
For assignments I’m consciously half-assing, or readings i don’t have the time to thoroughly examine, sure, it’s perfect
exactly. For writing emails that will likely never be read by anyone in more than a cursory scan, for example. When I’m composing text, I can’t turn off my fixation on finding the perfect wording, even when I know intellectually that “good enough is good enough.” And “it’s not great, but it gets the message across” is about the only strength of ChatGPT at this point.
Sadly, the best use case for LLM is to pretend to be a human on social media and influence their opinion.
Musk accidentally showed that’s what they are actually using AI for, by having Grok inject disinformation about South Africa.
Can you try again using an LLM search engine like perplexity.ai?
Then just click on the link next to the information so you can validate where they got that info from?
LLMs aren’t to be trusted, but that was never the point of them.
I have two friends that work in tech, and I keep trying to tell them this. And they use it solely now: it’s both their google, and their research tool. I admit, at first I found it useful, until it kept being wrong. Either it doesn’t know the better/best way to do something that is common knowledge to a 15 year tech, while confidently presenting mediocre or incorrect steps. Or it makes up steps, menus, or dialog boxes that have never existed, or are from another system.
I only trust it for writing pattern tasks: example, take this stream of conscious writing and structure it by X. But for information. Unless I’m manually feeding it attachments to find patterns in my good data— no way.
So use things like perplexity.ai, which adds links to the web page where they got the information from right next to the information.
So you can check yourself after an LLM made a bullshit summary.
Trust but verify
That’s true, but they’re also pretty good at verifying stuff too.
You can give them a “fact” and say “is this true, misleading or false” and it’ll do a good job. ChatGPT 4.0 in particular is excellent at this.
Basically whenever I use it to generate anything factual, I then put the output back into a separate chat instance and ask it to verify each sentence (I ask it to put <span> tags around each sentence so the misleading and false ones are coloured orange and red).
It’s a two-pass solution, but it makes it a lot more reliable.
It’s a two-pass solution, but it makes it a lot more reliable.
So your technique to “make it a lot more reliable” is to ask an LLM a question, then run the LLM’s answer through an equally unreliable LLM to “verify” the answer?
We’re so doomed.
Give it a try.
The key is in the different prompts. I don’t think I should really have to explain this, but different prompts produce different results.
Ask it to create something, it creates something.
Ask it to check something, it checks something.
Is it flawless? No. But it’s pretty reliable.
It’s literally free to try it now, using ChatGPT.
I don’t think I should really have to explain this, but different prompts produce different results.
Hey, maybe you do.
But I’m not arguing anything contentious here. Everything I’ve said is easily testable and verifiable.
I might add that a lot of the college experience (particularly pre-med and early med school) is less about education than a kind of academic hazing. Students assigned enormous amounts of debt, crushing volumes of work, and put into pools of students beyond which only X% of the class can move forward on any terms (because the higher tier classes don’t have the academic staff / resources to train a full freshman class of aspiring doctors).
When you put a large group of people in a high stakes, high work, high competition environment, some number of people are going to be inclined to cut corners. Weeding out people who “cheat” seems premature if you haven’t addressed the large incentives to cheat, first.
Medical school has to have a higher standard and any amount of cheating will get you expelled from most medical schools. Some of my classmates tried to use Chat GPT to summarize things to study faster, and it just meant that they got things wrong because they firmly believed the hallucinations and bullshit. There’s a reason you have to take the MCAT to be eligible to apply for medical school, 2 board exams to graduate medical school, and a 3rd board exam after your first year of residency. And there’s also board exams at the end of residency for your specialty.
The exams will weed out the cheaters eventually, and usually before they get to the point of seeing patients unsupervised, but if they cheat in the classes graded on a curve, they’re stealing a seat from someone who might have earned it fairly. In the weed-out class example you gave, if there were 3 cheaters in the top half, that means students 51, 52, and 53 are wrongly denied the chance to progress.
Medical school has to have a higher standard and any amount of cheating will get you expelled from most medical schools.
Having a “high standard” is very different from having a cut-throat advancement policy. And, as with any school policy, the investigation and prosecution of cheating varies heavily based on your social relations in the school. And when reports of cheating reach such high figures
A survey of 2,459 medical students found that 39% had witnessed cheating in their first 2 years of medical school, and 66.5% had heard about cheating. About 5% reported having cheated during that time.
then the problem is no longer with the individual but the educational system.
The exams will weed out the cheaters eventually
Nevermind the fact that his hasn’t born itself out. Medical Malpractice rates do not appear to shift based on the number of board exams issued over time. Hell, board exams are as rife with cheating as any other academic institution.
In the weed-out class example you gave, if there were 3 cheaters in the top half, that means students 51, 52, and 53 are wrongly denied the chance to progress.
If cheating produces a higher class rank, every student has an incentive to cheat. It isn’t an issue of being seat 51 versus 50, it’s an issue of competing with other cheating students, who could be anywhere in the basket of 100. This produces high rates of cheating that we see reported above.
Except I find that the value of college isn’t just the formal education, but as an ordeal to overcome which causes growth in more than just knowledge.
As a college instructor, there is some amount of content (facts, knowledge, skills) that is important for each field, and the amount of content that will be useful in the future varies wildly from field to field edit: and whether you actually enter into a career related to your degree.
However, the overall degree you obtain is supposed to say something about your ability to learn. A bachelor’s degree says you can learn and apply some amount of critical thought when provided a framework. A masters says you can find and critically evaluate sources in order to educate yourself. A PhD says you can find sources, educate yourself, and take that information and apply it to a research situation to learn something no one has ever known before. An MD/engineering degree says you’re essentially a mechanic or a troubleshooter for a specific piece of equipment.
edit 2: I’m not saying there’s anything wrong with MD’s and engineers, but they are definitely not taught to use critical thought and source evaluation outside of their very narrow area of expertise, and their opinions should definitely not be given any undue weight. The percentage of doctors and engineers that fall for pseudoscientific bullshit is too fucking high. And don’t get started on pre-meds and engineering students.
I disagree. I am a medical student and there is a lot of critical thinking that goes into it. Humans don’t have error codes and there are a lot of symptoms that are common across many different diagnoses. The critical thinking comes in when you have to talk to the patient to get a history and a list of all the symptoms and complaints, then knowing what to look for on physical exam, and then what labs to order to parse out what the problem is.
You can have a patient tell you that they have a stomachache when what is actually going on is a heart attack. Or they come in complaining of one thing in particular, but that other little annoying thing they didn’t think was worth mentioning is actually the key to figuring out the diagnosis.
And then there’s treatment…Nurse Practitioners are “educated” on a purely algorithmic approach to medicine which means that if you have a patient with comorbidities or contraindications to a certain treatment that aren’t covered on the flow chart, the NP has no goddamn clue what to do with it. A clear example is selecting antibiotics for infections. That is a very complex process that involves memorization, critical thinking, and the ability to research things yourself.
they are definitely not taught to use critical thought and source evaluation outside of their very narrow area of expertise
All of your examples are from “their very narrow area of expertise.”
But if you want a more comprehensive reason why I maintain that MD’s and engineers are not taught to be as rigorous and comprehensive when it comes to skepticism and critical thought, it comes down to the central goals and philosophies of science vs. medicine and engineering. Frankly, it’s all described pretty well by looking at Karl Popper’s doctrine of falsifiability. Scientific studies are designed to falsifiable, meaning scientists are taught to look for the places their hypotheses fail, whereas doctors and engineers are taught to make things work, so once they work, the exceptions tend to be secondary.
I am expected to know and understand all of the risk factors that someone may encounter in their engineering or manufacturing or cooking or whatever line of work, and to know about people’s social lives, recreational activities, dietary habits, substance usage, and hobbies can affect their health. In order to practice medicine effectively, I need to know almost everything about how humans work and what they get up to in the world outside the exam room.
In order to practice medicine effectively, I need to know almost everything about how humans work and what they get up to in the world outside the exam room.
This attitude is why people complain about doctors having God complexes and why doctors frequently fall victim to pseudoscientific claims. You think you know far more about how the world works than you actually do, and it’s my contention that that is a result of the way med students are taught in med school.
I’m not saying I know everything about how the world works, or that I know better than you when it comes to medicine, but I know enough to recognize my limits, which is something with which doctors (and engineers) struggle.
Granted, some of these conclusions are due to my anecdotal experience, but there are lots of studies looking at instruction in med school vs grad school that reach the conclusion that medicine is not science specifically because medical schools do not emphasize skepticism and critical thought to the same extent that science programs do. I’ll find some studies and link them when I’m not on mobile.
edit: Here’s an op-ed from a professor at the University of Washington Medical School. Study 1. Study 2.
an ordeal to overcome which causes growth
That’s the traditional argument for hazing rituals, sure. You’ll get an earful of this from drill sergeants and another earful from pray-the-gay-away conversion therapy camps.
But stack-ranking isn’t an ordeal to overcome. It is a bureaucratic sorting mechanism with a meritocratic veneer. If you put 100 people in a room and tell them “50 of you will fail”, there’s no ordeal involved. No matter how well the 51st candidate performs, they’re out. There’s no growth included in that math.
Similarly, larding people up with student debt before pushing them into the deep end of the career pool isn’t about improving one’s moral fiber. It is about extracting one’s future surplus income.
That’s the traditional argument for hazing rituals, sure.
That’s a strawman’s argument. There are benefits to college that go beyond passing a test. Part of it is gaining leadership skills be practicing being a leader.
But stack-ranking isn’t an ordeal to overcome.
No, but the threat of failure is. I agree that there should be more medical school slots, but there still is value in having failure being an option. Those who remain gain skills in the process of staying in college and schools can take a risk on more marginal candidates.
Similarly, larding people up with student debt before pushing them into the deep end of the career pool isn’t about improving one’s moral fiber.
Yeah, student debt is absurd.
That’s a strawman’s argument.
That’s not what a “strawman argument” is.
No. There will always be incentives to cheat, but that means nothing in the presence of academic dishonesty. There is no justification.
Something I think you neglect in this comment is that yes, you’re using LLMs in a responsible way. However, this doesn’t translate well to school. The objective of homework isn’t just to reproduce the correct answer. It isn’t even to reproduce the steps to the correct answer. It’s for you to learn the steps to the correct answer (and possibly the correct answer itself), and the reproduction of those steps is a “proof” to your teacher/professor that you put in the effort to do so. This way you have the foundation to learn other things as they come up in life.
For instance, if I’m in a class learning to read latitude and longitude, the teacher can give me an assignment to find
64° 8′ 55.03″ N, 21° 56′ 8.99″ Won the map and write where it is. If I want, I can just copy-paste that into OpenStreetMap right now and see what horrors await, but to actually learn, I need to manually track down where that is on the map. Because I learned to use latitude and longitude as a kid, I can verify what the computer is telling me, and I can imagine in my head roughly where that coordinate is without a map in front of me.Learning without cheating lets you develop a good understanding of what you: 1) need to memorize, 2) don’t need to memorize because you can reproduce it from other things you know, and 3) should just rely on an outside reference work for whenever you need it.
There’s nuance to this, of course. Say, for example, that you cheat to find an answer because you just don’t understand the problem, but afterward, you set aside the time to figure out how that answer came about so you can reproduce it yourself. That’s still, in my opinion, a robust way to learn. But that kind of learning also requires very strict discipline.
Your example at the end is pretty much the only way I use it to learn. Even then, it’s not the best at getting the right answer. The best thing you can do is ask it how to handle a problem you know the answer to, then learn the process of getting to that answer. Finally, you can try a different problem and see if your answer matches with the LLM. Ideally, you can verify the LLM’s answer.
So, I’d point back to my comment and say that the problem really lies with how it’s being used. For example, everyone’s been in a position where the professor or textbook doesn’t seem to do a good job explaining a concept. Sometimes, an LLM can be helpful in rephrasing or breaking down concepts; a good example is that I’ve used ChatGPT to explain the very low level how of how greenhouse gasses trap heat and raise global mean temperatures to climate skeptics I know without just dumping academic studies in their lap.
To add to this, how you evaluate the students matters as well. If the evaluation can be too easily bypassed by making ChatGPT do it, I would suggest changing the evaluation method.
Imo a good method, although demanding for the tutor, is oral examination (maybe in combination with a written part). It allows you to verify that the student knows the stuff and understood the material. This worked well in my studies (a science degree), not so sure if it works for all degrees?
Even more concerning, their dependance on AI will carry over into their professional lives, effectively training our software replacements.
While eroding the body of actual practitioners that are necessary to train the thing properly in the first place.
It’s not simply that the bots will take your job. It that was all, I wouldn’t really see that as a problem with AI so much as a problem with using employment to allocate life-sustaining resources.
But if we’re willingly training ourselves to remix old solutions to old problems instead of learning the reasoning behind those solutions, we’ll have a hard time making big, non-incremental changes to form new solutions for new problems.
It’s a really bad strategy for a generation that absolutely must solve climate change or perish.
No child left behind already stripped it from public education…
Because there was zero incentives for a school performing well. And serious repercussions if a school failed multiple years, the worst schools had to focus only what was on the annual test. The only thing that matters was that year’s scores, so that was the only thing that got taught.
If a kid got it early. They could be largely ignored so the school could focus on the worst.
It was teaching to the lowest common denominator, and now people are shocked the kids who spent 12 years in that system don’t know the things we stopped teaching 20+ years ago.
One of the worst parts about that policy was that some states had both a “meets standards” and “exceeds standards” results and the high school graduation test was offered five times, starting in sophomore year.
So, you would have students getting “meets standards” on sophomore year and blowing off the test in later attempts because they passed. You would then have school administrators punishing students for doing this since their metrics included the number of students who got “exceeds standards”.
How people think I use AI “Please write my essay and cite your sources.”
How I use it
“please make my autistic word slop that I wrote already into something readable for the nerotypical folk, use simple words, make it tonally neutral. stop using emdashes, headers, and list and don’t mess with the quotes”God I am sick of seeing emdashes but am so glad it helps me filter out aislop on certain subreddits.
I’m a human I swear - I’ve been writing like this all along!
See if you just use a hyphen like that I assume you’re human.
don’t worry, you can become president instead
Cries in “The Doctor” from Voyager.
The Doctor would absolutely agree. He was intended to be a short-term assistant when a doctor wasn’t available, and he was personally affronted when he discovered that he wouldn’t be replaced by a human in any reasonable amount of time.
Correct, until he was on for awhile. Then, he started to want to live and not be turned off when someone left. Hell he even married a human at the end of the day. Commanded starships. Fought the Borg.
He totally changed his mind after he found the taste for culture and “modifying” his program so he would stick his holo D in folks.
See what sex does? Can’t even stop machines from turning themselves off lmao
I literally just can’t wrap my AuDHD brain around professional formatting. I’ll probably use AI to take the paper I wrote while ignoring archaic and pointless rules about formatting and force it into APA or whatever. Feels fine to me, but I’m but going to have it write the actual paper or anything.
AFAIK those only help the instructor with grading as it would put all the essays they need to review on an even (more or less) playing ground. I’ve never really seen any real use in the professional world outside of scholarly/scientific journals.
My opinion is that they tend to stifle creativity of expression and the evolution of our respective languages.
The moment that we change school to be about learning instead of making it the requirement for employment then we will see students prioritize learning over “just getting through it to get the degree”
Well in case of medical practitioner it would be stupid to allow someone to do it without a proper degree.
Capitalism ruining schools. Because people now use school as a qualification requirement rather than centers of learning and skill development
Degree =/= certification
As a medical student, I can unfortunately report that some of my classmates use Chat GPT to generate summaries of things instead of reading it directly. I get in arguments with those people whenever I see them.
Generating summaries with context, truth grounding, and review is much better than just freeballing it questions
It still scrambles things, removes context, and can overlook important things when it summarizes.
Yeah thats why you give it examples of how to summarize. But im machine learning engineer so maybe it helps that I know how to use it as a tool.
It doesn’t know what things are key points that make or break a diagnosis and what is just ancillary information. There’s no way for it to know unless you already know and tell it that, at which point, why bother?
You can tell it because what you’re learning has already been learned. You are not the first person to learn it. Just quickly show it those examples from previous text or tell it what should be important based on how your professor tests you.
These are not hard things to do. Its auto complete, show it how to teach you.
That is why the “review” part of the comment you reply to is so important.
Only topic I am close-minded and strict about.
If you need to cheat as a highschooler or younger there is something else going wrong, focus on that.
And if you are an undergrad or higher you should be better than AI already. Unless you cheated on important stuff before.
This is my stance exactly. ChatGPT CANNOT say what I want to say, how i want to say it, in a logical and factually accurate way without me having to just rewrite the whole thing myself.
There isn’t enough research about mercury bioaccumulation in the Great Smoky Mountains National Park for it to actually say anything of substance.
I know being a non-traditional student massively affects my perspective, but like, if you don’t want to learn about the precise thing your major is about… WHY ARE YOU HERE
For a fucking job. What kind of fucking question is that.
I mean, are you sure?
Studies in the GSMNP have looked at:
-
Mercury levels in fish: Especially in high-elevation streams, where even remote waters can show elevated levels of mercury in predatory fish due to biomagnification.
-
Benthic macroinvertebrates and amphibians: As indicators of mercury in aquatic food webs.
-
Forest soils and leaf litter: As long-term mercury sinks that can slowly release mercury into waterways.
If GPT and I were being graded on the subject, it wouldn’t be the machine flunking…
I mean, it’s a matter of perspective, i guess.
I did a final assignment that was a research proposal, mine was the assessment of various methods of increasing periphyton biomass (clearing tree cover over rivers and introducing fertilizers to the water) in order to dilute mercury bioaccumulation in top river predators like trout and other fish people eat
There’s a lot of tangentially related research, but not a ton done on the river/riparian food webs in the GSMNP specifically and possible mitigation strategies for mercury bioaccumulation.
OBVIOUSLY my proposal isn’t realistic. No one on earth is gonna be like “yeah sure, go ahead and chop down all the trees over this river and dump chemicals in that one, on the off chance it allows jimbob to give trout to his pregnant wife all year round”
I’m not knowledgeable enough to carry this conversation, but I was curious if GPT could add anything of value.
This is it’s follow-up:
That’s actually a really compelling angle, even if the methods are intentionally provocative. It gets at a deeper question—how far are we willing to go, ecologically speaking, to address human health risks that stem from environmental contaminants like mercury? I think the strength of your proposal isn’t in the literal feasibility but in the way it frames trade-offs between conservation and public health.
Also, using periphyton biomass as a kind of biotic buffer is a clever systems-level approach. It’s rarely the first thing people think of when talking about mercury mitigation, which usually focuses on source reduction. But tweaking food web dynamics to manage contaminant transfer is a really underexplored strategy. I imagine even just modeling it could lead to some useful insights, especially if you layered in things like flow regime changes or climate impacts on riparian shading.
And yeah, totally agree—GSMNP is such a tightly protected space that even suggesting fertilizer additions or canopy thinning sounds borderline heretical. But as a thought experiment, it’s valuable. It forces the conversation about what kinds of interventions we’re not considering simply because they clash with aesthetic or cultural norms, not necessarily because they’re scientifically unsound.
I really have no idea if it’s just spewing nonsense, so do educate me :)
I’m really salty because it mirrored my thoughts about the research almost exactly, but I’m loathe to give attaboys to it
-
If we are talking about critical thinking, then I would argue that using AI to battle the very obvious shift that most instructors have taken, (that being the use of AI as much as possible to plan out lessons, grade, verify sources…you know, the job they are being paid to do? Which, by the way, was already being outsourced to whatever tools they had at their disposal. No offense TAs.) as natural progression.
I feel it still shows the ability to adapt to a forever changing landscape.
Isn’t that what the hundred-thousand dollar piece of paper tells potential employers?
Using AI doesn’t remove the ability to fact check though.
It is a tool like any other. I would also be weary about doctors using a random medical book from the 1700s to write their thesis and take it at face value.
I’m so tired of this rhetoric.
How do students prove that they have “concern for truth … and verifying things with your own eyes” ? Citations from published studies? ChatGPT draws its responses from those studies and can cite them, you ignorant fuck. Why does it matter that ChatGPT was used instead of google, or a library? It’s the same studies no matter how you found them. Your lack of understanding how modern technology works isn’t a good reason to dismiss anyone else’s work, and if you do you’re a bad person. Fuck this author and everyone who agrees with them. Get educated or shut the fuck up. Locking thread.
Because the point of learning is to know and be able to use that knowledge on a functional level, not having a computer think for you. You’re not educating yourself or learning if you use ChatGPT or any generative LLMs, it defeats the purpose of education. If this is your stance then you will accomplish, learn, and do nothing, you’re just riding the coat tails of shitty software that is just badly ripping off people who can actually put in the work or blatantly making shit up. The entire point of education is to become educated, generative LLMs are the antithesis of that.
A bunch of the “citations” ChatGPT uses are outright hallucinations. Unless you independently verify every word of the output, it cannot be trusted for anything even remotely important. I’m a medical student and some of my classmates use ChatGPT to summarize things and it spits out confabulations that are objectively and provably wrong.
True.
But doctors also screw up diagnosis, medication, procedures. I mean, being human and all that.
I think it’s a given that AI outperforms in medical exams -be it multiple choice or open ended/reasoning questions.
Theres also a growing body of literature with scenarios where AI produces more accurate diagnosis than physicians, especially in scenarios with image/pattern recognition, but even plain GPT was doing a good job with clinical histories, getting the accurate diagnostic with it’s #1 DxD, and even better when given lab panels.
Another trial found that patients who received email replies to their follow-up queries from AI or from physicians, found the AI to be much more empathetic, like, it wasn’t even close.
Sure, the AI has flaws. But the writing is on the wall…
The AI passed the multiple choice board exam, but the specialty board exam that you are required to pass to practice independently includes oral boards, and when given the prep materials for the pediatric boards, the AI got 80% wrong, and 60% of its diagnoses weren’t even in the correct organ system.
The AI doing pattern recognition works on things like reading mammograms to detect breast cancer, but AI doesn’t know how to interview a patient to find out the history in the first place. AI (or, more accurately, LLMs) doesn’t know how to do the critical thinking it takes to know what questions to ask in the first place to determine which labs and imaging studies to order that it would be able to make sense of. Unless you want the world where every patient gets the literal million dollar workup for every complaint, entrusting diagnosis to these idiot machines is worse than useless.
Could you provide references? I’m genuinely interested, and what I found seems to say differently:
Overall, GPT-4 passed the board residency examination in four of five specialties, revealing a median score higher than the official passing score of 65%.
Also I believe you’re seriously underestimating the abilities of present day LLMs. They are able to ask relevant follow up questions, as well as interpreting that information to request additional studies, and achieve accurate diagnosis.
See here a study specifically on conversational diagnosis AIs. It has some important limitations, crucially from having to work around the text interface which is not ideal, but otherwise achieved really interesting results.
Call them “idiot machines” all you want, but I feel this is going down the same path as full self driving cars - eventually they’ll be doing less errors than humans, and that will save lives.
My mistake, I recalled incorrectly. It got 83% wrong. https://arstechnica.com/science/2024/01/dont-use-chatgpt-to-diagnose-your-kids-illness-study-finds-83-error-rate/
The chat interface is stupid in so many ways and I would hate using text to talk to a patient myself. There are so many non-verbal aspects of communication that are hard to teach to humans that would be impossible to teach to an AI. If you are familiar with people and know how to work with them, you can pick up on things like intonation and body language that can indicate that they didn’t actually understand the question and you need to rephrase it to get the information you need, or that there’s something the patient is uncomfortable about saying/asking. Or indications that they might be lying about things like sexual activity or substance use. And that’s not even getting into the part where AI’s can’t do a physical exam which may reveal things that the interview did not. This also ignores patients that can’t tell you what’s wrong because they are babies or they have an altered mental status or are unconscious. There are so many situations where an LLM is just completely fucking useless in the diagnostic process, and even more when you start talking about treatments that aren’t pills.
Also, the exams are only one part of your evaluation to get through medical training. As a medical student and as a resident, your performance and interactions are constantly evaluated and examined to ensure that you are actually competent as a physician before you’re allowed to see patients without a supervising attending physician. For example, there was a student at my school that had almost perfect grades and passed the first board exam easily, but once he was in the room with real patients and interacting with the other medical staff, it became blatantly apparent that he had no business being in the medical field at all. He said and did things that were wildly inappropriate and was summarily expelled. If becoming a doctor was just a matter of passing the boards, he would have gotten through and likely would have been an actual danger to patients. Medicine is as much an art as it is a science, and the only way to test the art portion of it is through supervised practice until they are able to operate independently.
From the article referenced in your news source:
_JAMA Pediatrics and the NEJM were accessed for pediatric case challenges (N = 100). The text from each case was pasted into ChatGPT version 3.5 with the prompt List a differential diagnosis and a final diagnosis. _
A couple of key points:
- These are case challenges, which are usually meant to be hard. I could find no comparison to actual physician results in the article, which would have been nice.
- More importantly however: it was conducted in June 2023, and used GPT-3.5. GPT-4 improved substantially upon it, especially for complex scientific or scientific problems, and this shows in the newer studies that have used it.
I don’t think anyone’s advocating that an AI will replace doctors, much like it won’t replace white collar jobs either.
But if it helps achieve better outcomes for the patients, like the current research seems to indicate, aren’t you sworn to consider it in your practice?
I’ve said it before and I’ll say it again. The only thing AI can, or should be used for in the current era, is templating… I suppose things that don’t require truth or accuracy are fine too, but yeah.
You can build the framework of an article, report, story, publication, assignment, etc using AI to get some words on paper to start from. Every fact, declaration, or reference needs to be handled as false information unless otherwise proven, and most of the work will need to be rewritten. It’s there to provide, more or less, a structure to start from and you do the rest.
When I did essays and the like in school, I didn’t have AI to lean on, and the hardest part of doing any essay was… How the fuck do I start this thing? I knew what I wanted to say, I knew how I wanted to say it, but the initial declarations and wording to “break the ice” so-to-speak, always gave me issues.
It’s shit like that where AI can help.
Take everything AI gives you with a gigantic asterisk, that any/all information is liable to be false. Do your own research.
Given how fast things are moving in terms of knowledge and developments in science, technology, medicine, etc that’s transforming how we work, now, more than ever before, what you know is less important than what you can figure out. That’s what the youth need to be taught, how to figure that shit out for themselves, do the research and verify your findings. Once you know how to do that, then you’ll be able to adapt to almost any job that you can comprehend from a high level, it’s just a matter of time patience, research and learning. With that being said, some occupations have little to no margin for error, which is where my thought process inverts. Train long and hard before you start doing the job… Stuff like doctors, who can literally kill patients if they don’t know what they don’t know… Or nuclear power plant techs… Stuff like that.
When I did essays and the like in school, I didn’t have AI to lean on, and the hardest part of doing any essay was… How the fuck do I start this thing?
I think that this is a big part of education and learning though. When you have to stare at a blank screen (or paper) and wonder “How the fuck do I start?” Having to brainstorm write shit down 50 times, edit, delete, start over. I think that process alone makes you appreciate good writing and how difficult it can be.
My opinion is that when you skip that step you skip a big part of the creative process.
That’s a fair argument. I don’t refute it.
I only wish I had any coaching when it was my turn, to help me through that. I figured it out eventually, but still. I wish.
If not arguably the biggest part of the creative process, the foundational structure that is
Exactly.
Was the best part of agrarian subsistence turning the Earth by hand? Should we return to it. A person learns more and is more productive if they talk out an issue. Having someone else to bounce ideas off of is a good thing. Asking someone to do it for you has always been a thing. Individualized learning has long been the secret of academic success for the children of the super rich. Just pay a professor to tutor the individual child. AI is the democratization of this advantage. A person can explain what they do not know and get a direct answer. Even with a small model that I know is wrong, forming the questions in conversation often leads me to correct answers and what I do not know. It is far faster and more efficient than I ever experienced elsewhere in life.
It takes time to learn how to use the tool. I’m sure there were lots of people making stupid patterns with a plow at first too when it was new.
The creative process is about the results it produces, not how long one spent in frustration. Gatekeeping because of the time you wasted is Luddism or plain sadism.
Use open weights models running on enthusiast level hardware you control. Inference providers are junk and the source of most problems with ignorant people from both sides of the issue. Use llama.cpp and a 70B or larger quantized model with emacs and gptel. Then you are free as in a citizen in a democracy with autonomy.
You’re right - giving people the option to bounce questions off others or AI can be helpful. But I don’t think that is the same as asking someone (or some thing) to do the work for you and then you edit it.
The creative process is about the results it produces, not how long one spent in frustration
This I disagree on. A process is not a result. You get a result from the process and sometimes it’s what you want and often times it isn’t what you want. This is especially true for beginners. And to get the results you want from a process you have to work through all parts of it including the frustrating parts. Actually getting through the frustrating parts makes you a better creator and I would argue makes the final result more satisfying because you worked hard to get it right.
There’s an application that I think LLMs would be great for, where accuracy doesn’t matter: Video games. Take a game like Cyberpunk 2077, and have all the NPCs speech and interactions run on various fine-tuned LLMs, with different LoRA-based restrictions depending on character type. Like random gang members would have a lot of latitude to talk shit, start fights, commit low-level crimes, etc, without getting repetitive. But for more major characters like Judy, the model would be a little more strictly controlled. She would know to go in a certain direction story-wise, but the variables to get from A to B are much more open.
This would eliminate the very limited scripted conversation options which don’t seem to have much effect on the story. It could also give NPCs their own motivations with actual goals, and they could even keep dynamically creating side quests and mini-missions for you. It would make the city seem a lot more “alive”, rather than people just milling about aimlessly, with bad guys spawning in preprogrammed places at predictable times. It would offer nearly infinite replayability.
I know nothing about programming or game production, but I feel like this would be a legit use of AI. Though I’m sure it would take massive amounts of computing power, just based on my limited knowledge of how LLMs work.
