Ok… I’m left a bit confused regarding your recommendations.
I’m familiar with this specific Brown “scandal”. I have a former associate who currently instructs there who has no direct contact with this situation, but has conversed with me regarding more “Big Picture” analysis regarding knowledge, practical evaluations, ethical considerations, etc.
I’ve instructed some specialized technical courses at major universities. I do not come out of any university system, but from the Naval Nuclear Power Program. I only note that to admit my “outsider” viewpoint regarding traditional university educational paths and evaluation methodologies.
My own kid is, coincidentally, currently wrapping up her Master’s right now, so I’m not entirely unfamiliar with recent changes in the higher education system.
I’ve heard many snappy self-righteous assessments regarding the character of the students involved in this matter. My younger self (pre-60) might have quickly grabbed the low-hanging fruit of righteous indignation that these students did not knuckle down and engage their exam in a similar manner to that we employed in the late 1970s / early 1980s. Hours of rote memorization, recopying of binders of notes , etc.
As I’ve gotten older and watched my daughter and her fellow student negotiate increasingly absurd educational infrastructures, I no longer leap to judgement of the individual students but attempt to determine how the system has failed.
My first M-Div Chief on our boat had a good analogy… If you’re hammering away and it’s getting you nowhere in your task, you can either seek out a bigger hammer or consider whether you might need an entirely different tool for the job.
Statistically, it is unlikely that the proportion of students who used AI for some or all of that exam suddenly degenerated into amoral schiesters. Given the accomplishments required to even be admitted to Brown (though I understand that there are paths to elite universities that bypass intellectual rigor) it is unlikely that any of those students could be simply dismissed as lazy or a craven opportunist.
I could go on about this, though I’ll just attach the El video discussing this topic. I found her points to resonate with my own. She looks much better and has that intriguing accent.
I do agree with what I feel you’re saying in that “we” as a culture are almost entirely unprepared for the cultural effects of AI. I do not at all feel the technology is approaching (or will approach in any near future scenario) the level of societal impact implied by vapid hucksters such as Amodei and Altman, but our collective systems only barely grasp where we were in the information infrastructure realm PRIOR to LLMs becoming commonplace. Now that everyone’s got their own little Chatbot influencer we need a broad public education program to assist in defining where we are, and more importantly ARE NOT. China seems to have a much better cultural grasp on those two benchmarks than do we here in FreedumLandia at present.
Hey Paul, I agree with most of what you say. The main thing I disagree with is regarding the reason why the students act this way. I don't think resorting to laziness or opportunism is necessary nor would I ascribe all the responsibility to them either. I'd say it's merely "the banality of evil" except instead of following orders, they're following the new status quo. Thanks for the references and the comment
Fair enough. Though I chatted with a group of primarily Law postgrads the other evening who, by the time they actually begin drawing a paycheck will incur somewhere between 0.5 and 1.2 M $$$ in debt. These aren’t even elite schools. Good schools, but by no means the nosebleed sector.
My daughter’s classmates generally carry a lot of cynicism in receiving moral admonitions from “above” in the university system.
They know, in a manner that possibly no generation has before them, that they exist in a stunningly corrupt and self serving educational infrastructure and roll their eyes when a school administrator chastises them for not applying top level ethics as they claw their way through the Cash-for-Credentials machine that they’ve been told since elementary school is the only way for them not to be “Left Behind” and economically irrelevant.
I understand where you’re going with the Arendt quote, though I feel it applies more to the Administators than the Student Body writ large. The folk that run our local university have been distilled to a concentrated essence I can personally verify to be an amoral lot. Unfortunately that has become a prerequisite for elevation within the system. They’ve been my neighbors, our kids grew up together. I know their stories.
Yeah I agree. It's not an easy situation at all. And if I were a student I would have to restrain myself and yet I would always ask myself: "what do I owe them? What do I owe them?" My position is more of a "it's best for you as a student if you take the opportunity of AI without letting it erode your brain." Not even taking revenge on the system is worth betraying yourself, IMO.
I think that knowledge compounds and that one's abnegation of the sole means of acquiring knowledge (i.e., doing things and making conscious formulations that explicitly tie phenomena into one's existing base of knowledge) will contribute to the formation of a generation of weak-willed, irrational humans.
Thanks: again, and as usual, delightful food for thought!
As a matter of fact, we have been brainstorming on the question of teaching and grading students in the time of AI and what strategies could prove useful.
Here, are some conclusions that may (?) tackle the problem in a different light (? again, many question marks)
- Teachers will have to adjust (evidently!) but are not doomed in an AI-society, even though, many institutions want to replace us by “teaching AI”
- Two phases of activity should be distinguished: (1) the process (all the activities and their organization involved in discovering / learning / doing something, “something” being defined by a goal) = “the how to” and (2) the product (more precisely the evaluation of the end result, its adequation to the goal) = “the honestly relevant?”.
NB evidently those two activities should be performed in alternance to verify that the process is still aligned with the goal and that the product is still useful and suited to the goal (and hopefully, as best as possible).
- Teachers should help student to develop proficiency in both, with careful delineation of these two activities and can make use of AI (or not) – and there, the human interaction, with all the unvoiced information, the subtext, the silent communication cannot (yet?) be replaced by AI.
- Grading should not aim at evaluating either the process or the end product since there a many ways to reach a desirable goal, each suited to the person doing the work and what he as to work with. We propose that grading should aim to evaluate how the student is able to justify why this product fulfill the goal, how it is simultaneously suited to the need, the context, the actors, the constraints … ie being capable of situating the product in the (real) world, with (real) actors… Again, the use of AI is neither forbidden nor mandatory but is simply not taken into account: it is up to the student to decide whether AI can help him (or not). This grading strategy can be applied to anything, from fulfilling a MCQ, to writing an essay, to design or do a task or grasp what something means.
We are conscious that this strategy is much more demanding upon the students (and the teacher) and that some won’t reach this level of proficiency. However, we but hope that every student will still gain from this type of teaching / grading strategy. We also think that it helps to develop both critical thinking and (reflexive) uses of AI. Furthermore, we are not only aiming at adapting teaching to the fact that in many professions, AI-proficiency will become a necessity. We are thinking of helping youth adjust to a world where all sorts of AIs are silently creeping in each and every aspect of our lives.
I had a conversation about this on LinkedIn yesterday.
Comment:
I think the solution is *mandatory disclosure* - AI usage is not disallowed, but if used the student must explain how they used it and for what purpose (and ideally what they learned from it)
Like references in a paper. Declare the tools used, the prompts, and how it is reflected in the work.
My reply:
In 2026, this is what kids should be learning in high school: AI usage is not disallowed, but if used the student must explain how they used it and for what purpose (and ideally what they learned from it).
Speaking as a former teacher, I would love to have those insights into their thinking. It gives much richer information to go off of than a multiple choice test does, for example.
And students need to learn how to use AI well and and responsibly. If AI can find the answers: 1) how do you use it well? 2) how do you use it honestly? 3) how do you learn to be the driver and to remain in charge? 4) how can you still learn to think, be honest, be creative, and be accountable?
The comment's solution addresses all of those things very well.
After 15 years as a professor in a graduate program at a prominent Spanish public university, a year ago I finally decided to call it quits. It was an online master degree and I had been getting the clear impression that 90 percent of my graduate students were not being honest about their remote academic assignments.
What was the point of doing in-depth textual remarks and comments on each of their papers or exams when it was clear they had done them using (forbidden) automated tools and technologies instead of their own intellectual effort and applied knowledge?
Year after year, I had to acknowledge the fact that my diligent work as a feedback writer was becoming pointless and even ridiculous, as I acutely felt that I was grading not the effort of our students but the performance of machine translation tools. Only that grading MT engines was not the task either the university had hired me to do.
Also, the frustrating feeling that the university was absolutely lagging behind in terms of verification technologies for their online programs, so there was no way (other than taking in-house exams) to ensure that students' grades were a reliable proof of their actual knowledge and learning. In the end, I felt that students just enrolled in prestigious online masters where, as an exchange for payment, they could cheat to get their diplomas.
According to the students and to the university, I was an exemplary, well-appreciated teacher, but as a result of the collective, unpunished cheating, I felt obliged to remove myself from the job. And so whichever benefit I could provide to the university, or the students, is now gone. I am probably proving your point.
I have since learned that there are some universities which use complex verification technologies to monitor students' remote activity online, checking their bodies and eye-movements through cameras, doing background surveillance of their computers to detect which apps or programs they are running, etc. But the battle is lost with uploaded remote homework which GPT does in their stead--- and to our face.
I have read this week that a colleague is not accepting typed or printed assignments any longer. She is requesting handwritten papers instead. That way, she reasons, at least they will have to copy, word by word, whatever text ChatGPT or IA burps out, and so there's a chance that something will linger in their minds as a result of transcribing. What a jump forward to the Middle Ages!
Yes, it is (or soon will be) the end of university as we know it. I foresee that new universities will emerge, much resembling the Free University of Varsovia (where Marie Curie clandestinely studied during the war) or to the Bologna University, the very first created in Europe... by eager students passionate about learning who actively recruited teachers from whom to learn.
The tragedy of the commons framing is exactly right. Every individual student had a rational reason to cheat and collectively they destroyed the thing that made the degree valuable. The irony of cheating on a welfare economics exam is almost too perfect.
This is true but is it relevant? With even modest expectation of AI improvement all undergraduate degrees will likely soon be worthless anyway and so students may be rationally looking to exploit the vanishing window of opportunity that exists for them. Better to have put away a few years of good but undeserved income before AI comes for your job. Of course, if you are just studying for your own personal development then that argument does not work but, in that case, it's hard to justify the expense of going to university at all.
This is based on a few fallacious premises. For one, artificial intelligence will not "come for your job." The substrate that gives structure to a job will merely migrate to another role; the doom-mongering predictions have not and will not hold up. Secondly, Ayn Rand was right when she said that one needs to constantly flex and augment one's ability to focus. University used to fulfill that role, as lads had to sit down, study, and synthesise. Now that Claude exists, people are abnegating their mind, not focusing, and relegating themselves to brute-like existences. This fucks the way that they solve their personal problems. As a young twenty-something, 99.99% of my peers are unfocused neurotics; I'm certain that their usage of artificial-intelligence tools is a big reason why this is the case.
Ok… I’m left a bit confused regarding your recommendations.
I’m familiar with this specific Brown “scandal”. I have a former associate who currently instructs there who has no direct contact with this situation, but has conversed with me regarding more “Big Picture” analysis regarding knowledge, practical evaluations, ethical considerations, etc.
I’ve instructed some specialized technical courses at major universities. I do not come out of any university system, but from the Naval Nuclear Power Program. I only note that to admit my “outsider” viewpoint regarding traditional university educational paths and evaluation methodologies.
My own kid is, coincidentally, currently wrapping up her Master’s right now, so I’m not entirely unfamiliar with recent changes in the higher education system.
I’ve heard many snappy self-righteous assessments regarding the character of the students involved in this matter. My younger self (pre-60) might have quickly grabbed the low-hanging fruit of righteous indignation that these students did not knuckle down and engage their exam in a similar manner to that we employed in the late 1970s / early 1980s. Hours of rote memorization, recopying of binders of notes , etc.
As I’ve gotten older and watched my daughter and her fellow student negotiate increasingly absurd educational infrastructures, I no longer leap to judgement of the individual students but attempt to determine how the system has failed.
My first M-Div Chief on our boat had a good analogy… If you’re hammering away and it’s getting you nowhere in your task, you can either seek out a bigger hammer or consider whether you might need an entirely different tool for the job.
Statistically, it is unlikely that the proportion of students who used AI for some or all of that exam suddenly degenerated into amoral schiesters. Given the accomplishments required to even be admitted to Brown (though I understand that there are paths to elite universities that bypass intellectual rigor) it is unlikely that any of those students could be simply dismissed as lazy or a craven opportunist.
I could go on about this, though I’ll just attach the El video discussing this topic. I found her points to resonate with my own. She looks much better and has that intriguing accent.
I do agree with what I feel you’re saying in that “we” as a culture are almost entirely unprepared for the cultural effects of AI. I do not at all feel the technology is approaching (or will approach in any near future scenario) the level of societal impact implied by vapid hucksters such as Amodei and Altman, but our collective systems only barely grasp where we were in the information infrastructure realm PRIOR to LLMs becoming commonplace. Now that everyone’s got their own little Chatbot influencer we need a broad public education program to assist in defining where we are, and more importantly ARE NOT. China seems to have a much better cultural grasp on those two benchmarks than do we here in FreedumLandia at present.
Thanks for your efforts. All the best.
https://youtu.be/Hw8twdIjMJc?is=sYDqrl4clv_IOauF
Hey Paul, I agree with most of what you say. The main thing I disagree with is regarding the reason why the students act this way. I don't think resorting to laziness or opportunism is necessary nor would I ascribe all the responsibility to them either. I'd say it's merely "the banality of evil" except instead of following orders, they're following the new status quo. Thanks for the references and the comment
Fair enough. Though I chatted with a group of primarily Law postgrads the other evening who, by the time they actually begin drawing a paycheck will incur somewhere between 0.5 and 1.2 M $$$ in debt. These aren’t even elite schools. Good schools, but by no means the nosebleed sector.
My daughter’s classmates generally carry a lot of cynicism in receiving moral admonitions from “above” in the university system.
They know, in a manner that possibly no generation has before them, that they exist in a stunningly corrupt and self serving educational infrastructure and roll their eyes when a school administrator chastises them for not applying top level ethics as they claw their way through the Cash-for-Credentials machine that they’ve been told since elementary school is the only way for them not to be “Left Behind” and economically irrelevant.
I understand where you’re going with the Arendt quote, though I feel it applies more to the Administators than the Student Body writ large. The folk that run our local university have been distilled to a concentrated essence I can personally verify to be an amoral lot. Unfortunately that has become a prerequisite for elevation within the system. They’ve been my neighbors, our kids grew up together. I know their stories.
Just my opinion.
Thanks for engaging. All the best.
Yeah I agree. It's not an easy situation at all. And if I were a student I would have to restrain myself and yet I would always ask myself: "what do I owe them? What do I owe them?" My position is more of a "it's best for you as a student if you take the opportunity of AI without letting it erode your brain." Not even taking revenge on the system is worth betraying yourself, IMO.
I think that knowledge compounds and that one's abnegation of the sole means of acquiring knowledge (i.e., doing things and making conscious formulations that explicitly tie phenomena into one's existing base of knowledge) will contribute to the formation of a generation of weak-willed, irrational humans.
Thanks: again, and as usual, delightful food for thought!
As a matter of fact, we have been brainstorming on the question of teaching and grading students in the time of AI and what strategies could prove useful.
Here, are some conclusions that may (?) tackle the problem in a different light (? again, many question marks)
- Teachers will have to adjust (evidently!) but are not doomed in an AI-society, even though, many institutions want to replace us by “teaching AI”
- Two phases of activity should be distinguished: (1) the process (all the activities and their organization involved in discovering / learning / doing something, “something” being defined by a goal) = “the how to” and (2) the product (more precisely the evaluation of the end result, its adequation to the goal) = “the honestly relevant?”.
NB evidently those two activities should be performed in alternance to verify that the process is still aligned with the goal and that the product is still useful and suited to the goal (and hopefully, as best as possible).
- Teachers should help student to develop proficiency in both, with careful delineation of these two activities and can make use of AI (or not) – and there, the human interaction, with all the unvoiced information, the subtext, the silent communication cannot (yet?) be replaced by AI.
- Grading should not aim at evaluating either the process or the end product since there a many ways to reach a desirable goal, each suited to the person doing the work and what he as to work with. We propose that grading should aim to evaluate how the student is able to justify why this product fulfill the goal, how it is simultaneously suited to the need, the context, the actors, the constraints … ie being capable of situating the product in the (real) world, with (real) actors… Again, the use of AI is neither forbidden nor mandatory but is simply not taken into account: it is up to the student to decide whether AI can help him (or not). This grading strategy can be applied to anything, from fulfilling a MCQ, to writing an essay, to design or do a task or grasp what something means.
We are conscious that this strategy is much more demanding upon the students (and the teacher) and that some won’t reach this level of proficiency. However, we but hope that every student will still gain from this type of teaching / grading strategy. We also think that it helps to develop both critical thinking and (reflexive) uses of AI. Furthermore, we are not only aiming at adapting teaching to the fact that in many professions, AI-proficiency will become a necessity. We are thinking of helping youth adjust to a world where all sorts of AIs are silently creeping in each and every aspect of our lives.
Again, thanks a lot for your wonderful blog.
Anne
I had a conversation about this on LinkedIn yesterday.
Comment:
I think the solution is *mandatory disclosure* - AI usage is not disallowed, but if used the student must explain how they used it and for what purpose (and ideally what they learned from it)
Like references in a paper. Declare the tools used, the prompts, and how it is reflected in the work.
My reply:
In 2026, this is what kids should be learning in high school: AI usage is not disallowed, but if used the student must explain how they used it and for what purpose (and ideally what they learned from it).
Speaking as a former teacher, I would love to have those insights into their thinking. It gives much richer information to go off of than a multiple choice test does, for example.
And students need to learn how to use AI well and and responsibly. If AI can find the answers: 1) how do you use it well? 2) how do you use it honestly? 3) how do you learn to be the driver and to remain in charge? 4) how can you still learn to think, be honest, be creative, and be accountable?
The comment's solution addresses all of those things very well.
After 15 years as a professor in a graduate program at a prominent Spanish public university, a year ago I finally decided to call it quits. It was an online master degree and I had been getting the clear impression that 90 percent of my graduate students were not being honest about their remote academic assignments.
What was the point of doing in-depth textual remarks and comments on each of their papers or exams when it was clear they had done them using (forbidden) automated tools and technologies instead of their own intellectual effort and applied knowledge?
Year after year, I had to acknowledge the fact that my diligent work as a feedback writer was becoming pointless and even ridiculous, as I acutely felt that I was grading not the effort of our students but the performance of machine translation tools. Only that grading MT engines was not the task either the university had hired me to do.
Also, the frustrating feeling that the university was absolutely lagging behind in terms of verification technologies for their online programs, so there was no way (other than taking in-house exams) to ensure that students' grades were a reliable proof of their actual knowledge and learning. In the end, I felt that students just enrolled in prestigious online masters where, as an exchange for payment, they could cheat to get their diplomas.
According to the students and to the university, I was an exemplary, well-appreciated teacher, but as a result of the collective, unpunished cheating, I felt obliged to remove myself from the job. And so whichever benefit I could provide to the university, or the students, is now gone. I am probably proving your point.
I have since learned that there are some universities which use complex verification technologies to monitor students' remote activity online, checking their bodies and eye-movements through cameras, doing background surveillance of their computers to detect which apps or programs they are running, etc. But the battle is lost with uploaded remote homework which GPT does in their stead--- and to our face.
I have read this week that a colleague is not accepting typed or printed assignments any longer. She is requesting handwritten papers instead. That way, she reasons, at least they will have to copy, word by word, whatever text ChatGPT or IA burps out, and so there's a chance that something will linger in their minds as a result of transcribing. What a jump forward to the Middle Ages!
Yes, it is (or soon will be) the end of university as we know it. I foresee that new universities will emerge, much resembling the Free University of Varsovia (where Marie Curie clandestinely studied during the war) or to the Bologna University, the very first created in Europe... by eager students passionate about learning who actively recruited teachers from whom to learn.
The tragedy of the commons framing is exactly right. Every individual student had a rational reason to cheat and collectively they destroyed the thing that made the degree valuable. The irony of cheating on a welfare economics exam is almost too perfect.
This is true but is it relevant? With even modest expectation of AI improvement all undergraduate degrees will likely soon be worthless anyway and so students may be rationally looking to exploit the vanishing window of opportunity that exists for them. Better to have put away a few years of good but undeserved income before AI comes for your job. Of course, if you are just studying for your own personal development then that argument does not work but, in that case, it's hard to justify the expense of going to university at all.
This is based on a few fallacious premises. For one, artificial intelligence will not "come for your job." The substrate that gives structure to a job will merely migrate to another role; the doom-mongering predictions have not and will not hold up. Secondly, Ayn Rand was right when she said that one needs to constantly flex and augment one's ability to focus. University used to fulfill that role, as lads had to sit down, study, and synthesise. Now that Claude exists, people are abnegating their mind, not focusing, and relegating themselves to brute-like existences. This fucks the way that they solve their personal problems. As a young twenty-something, 99.99% of my peers are unfocused neurotics; I'm certain that their usage of artificial-intelligence tools is a big reason why this is the case.