domain:noahpinion.blog
Pasha
Defend Kebab
2m ago
Undergrads here are typically 3 years. I remember having 8 crunch periods per year for my degree. 4 quarters with one review exam in the middle and one big one at the end. Besides this, nobody stopped you from taking extra courses to graduate earlier and many did.
By no means continental European university system is great. Most countries have their own pathologies and the Bologna Process makes everything typically shittier but it’s pretty strange to claim students got to goof around because they don’t have constant deadlines to write parroting essays about queer indigenous history.
I fail to see how almost any of my education would had been disrupted by the LLMs the way the New York mag article is describing. The only courses I can imagine are currently swamped with LLM problems are the bunch of liberal arts inspired nonsense courses I had to follow for credits.
2rafa
5m ago
As @Pasha said I think the classic Anglo humanities model where you do essays at home for practice (but which count nothing toward your final grade) and then have a combination of (hand)written exams in class and an oral exam seems like the best. That was true even before AI, but it’s especially true now.
In this case there is really no real to use AI to do homework since it doesn’t affect your grade and is pointless as a learning tool (which is not to say AI can’t be a useful learning tool, but ‘write 1200 words on x that I will never read but will email to the TA’ isn’t it).
Owlify
5m ago
The reason housewives aren't higher status is because they have no money and they do unpaid work. They have no independent ability to take care of themselves and make life decisions either, and are financially dependent on someone else. Unless that changes (due to AGI/UBI), things will remain the same.
roystgnr
9m ago
assertion that many people on the Motte hold tech jobs of the kind found in Silicon Valley
doesn't accurately describe
most of the Motte works in Sillicon Valley
but I'm more astonished by the contrast between
uncontroversial
versus
you and ten other people angrily surge out of the woodwork
It looks like you might have a habit of choosing words inaccurately for the purposes of hyperbole, and that's a bit rough when you're some place like this where Aspergery people like me and your challengers upthread are welcomed, but it's easy enough to fix: when you get called out on it just admit the error, compose an accurate rewording (without pretending it's just an equivalent paraphrase) instead, and you're done. Everybody makes mistakes. Doubling down is just digging the hole deeper.
veqq
11m ago
Has there ever been a golden age of critical thinkers, schooled to think well, untainted by the technology of its era, or the character of its students?
Those growing up around 1900-1914 seemed closest, the great flower of our civilization, who died under flowering shrapnel on the French frontier.
I base this belief on reading historical (highschool/Gymnasium/lycée) exams (and submitted answers, with sample bias, of course) and cheat books (with more detail than modern academic treatments of the same... This is how I originally came to this.) They should write poetry on different topics in a certain style (movement or author), find problems in economic or business data, articulate various thinkers' contrasting beliefs about a topic etc. Transposed to modern times, have a student write a memoization macro, calculate some vector angles, write an essay on LaRoche, McKinley and and Teddy Roosevelt's views on tariffs, another on leadership (why the most popular kid's popular and what prevents the writer from taking his place, and to what extent the (chosen group/nation/state/movement) needs good leaders vs. institutions) ...written in Mandarin. The Overton window was far wider in those days, with multiple popular opinions about why x or y state was illegitimate with justifications from Renaissance, Classical and Biblical sources, advocating for paganism/atheism/state controlled religion etc. etc. Unfortunately, the war seemed to invalidate the whole framework and civilization behind this and mass education never recovered.
That's still a far cry from our ideal, but...
ChickenOverlord
13m ago
That's exactly my point. The kids will just goof off for 90% of the semester and just cram it all in before the exam. If so when what's the point of even having the rest of the semester.
You should look into Western Governor's University and their model. Basically you can finish your degree as fast as you can finish testing for each course.
roystgnr
15m ago
A lottery is very similar to insurance.
This is true only in the same sense that negative ten is similar to ten. They're both numbers, right? But they're opposite number. Likewise, here one gamble increases volatility (because the payoff is the only random event), and the other reduces it (because the payoff happens only when it cancels out a random expense; the net change from the random outcome is reduced).
But one of the two is supposedly justified while the other breaks their model and makes no sense whatsoever.
It makes sense, for the reasons above. Does it make sense to you too, now? If not, I'm afraid that's probably the best I can do. I've taught grad school math classes, which says good things about my math ability but bad things about my teaching ability...
PutAHelmetOn
Recovering Quokka
18m ago
I notice a parallel between the Christian's love for God and his faith in God. Your post is about the tension between loving an object for its properties, versus loving an object inherently (the latter I still maintain is quite meaningless). In faith, there is tension between believing a proposition because of evidence, versus believing a proposition inherently.
It's an old idea around these parts that Christians do not believe their religion. The Christian's behavior here is not really confusing. Professions of faith are tribal signals of group loyalty, not beliefs. But it would be wrong to ruminate on "the contradiction of belief" and ask about "is belief based on evidence" or "do people believe inherently?"
Likewise, "loving things for their properties" is just a different kind of thing than "tribal signals of loyalty." You're damned right I am loyal to my wife, what of it?
If you ever get a chance, do a self-driven review compare / contrast of ancient human-sacrifice rituals for different religions with different stakes in humans harm.
If you had good directions of where to start, I might just do that. It sounds interesting, and I expect some free time later this year. However, it is a bit harder to find structured reviews of them than, say, pointing three distinct but overlapping types of professionals.
Why don't you provide two good sources for the Aztecs and Carthaginians ethics? Good as in effectively and analytically characterizes their ethical systems. A bad work would be one that simply relegates Aztec morality to 'they conducted human sacrifice to keep the world from ending.' Yes, that is a utilitarian justification. It is not an ethical system.
And then, once you've read that, presumably you will somehow have changed your mind and believe human sacrifice is a good thing instead of a senseless waste of human life. You will probably even want to sacrifice your own children to Moloch, when the time comes. I know I haven't provided any reason why that should be the case, but apparently that's how this works now.
If that was what you took away from my post, then congratulations- you demonstrated a point by missing it.
The value of studying different forms of professional ethics isn't to change your own mind on ethics. The value is understanding what others want, or expect, the ethics of a professional to be. This has relevant insights when it comes to dealing with specific professions in isolation, when multiple professions with different professional ethics engage each other, or even how the same profession's ethics across different cultures.
Understanding other people's ethical frameworks has never been endorsement, or required conversion, unless you subscribe to some universal morality theory. Which most people don't. Which is what makes understanding their value systems useful, even if you think their values are wrong. Even the most hard-core anti-moral relativist is going to be surrounded by professionals whose value systems they cannot force to change.
Pasha
Defend Kebab
26m ago
I wonder if AI will make us all polyglots because it’s an incredibly useful tool of language learning or eliminate any learning altogether because it’s also really good at translating. Or perhaps it does both at the same time simultaneously so we have a bunch of conscious smart learners mastering new languages in 6 months and everyone else loses any motivation at all because any digital content they ever encounter is instantly perfectly translated (desire to access broader internet/gaming/tv shows was the reason I got good at English as a teenager, school instruction was useless)
Jiro
26m ago
I wonder if you could have a new university that initially paid students to come.
They would have to pay the first students an amount equivalent to the increase in lifetime earnings from going to a regular university (minus the cost of a regular university). This would be cost prohibitive.
07mk
31m ago
I absolutely think they should be. Now, maybe it's not practical to check each student's individual political preferences and assign bespoke assignments for them on that basis (which could be gamed anyway). Rather, humanities-based courses should test students on their ability to defend a wide variety of different, highly offensive and ideally "dangerous" ideas in whatever topics are at hand, to stimulate actually learning how to think versus what to think.
Hard to say if that will work, though; teaching students how to think seems to be one of those things that people in education have been trying to do for ever, without there being any sort of noticeable progress whatsoever. I just know that that was how I was educated, and it seemed to work for me and my classmates (but of course I'd think that, and so my belief that it seemed to work should count for approximately nothing), but even if it did, that doesn't mean that it's generalizable.
Tree
35m ago
That is the high brow justification for it, but I disagree with how they model people’s utility functions, and besides, people’s utility functions are neither set in stone nor economically justified.
People would be far better off with a flatter, less utility-diminishing curve, given that they spend near half their income to slightly reduce lifetime income volatility, and if the last century is any guide, they want to spend even more, no future loss is too small to be tolerated, should it cost half of gdp.
There’s no real difference between the TV insurance and home insurance, it all depends on the assumed steepness of the diminishing utility curve.
People’s utility gets modeled as a steadily diminishing curve. In reality there should be one huge drop in utility when you go from from starving to non-starving income, and then very very flat. Because the only way to lose all future utility, to get wiped out in the kelly sense, is to die irl.
And on that subject, people also gamble. A lottery is very similar to insurance. You pay a small sum, and after a random event you sometimes get paid a multiple. It’s a negative EV transaction because the losses in the pipes are large. Any rational man with a sufficiently flat utility curve would reject them.
But one of the two is supposedly justified while the other breaks their model and makes no sense whatsoever. Gambling people are spending good, high utility money, then losing some in the pipes, and for what? To get low utility money.
07mk
45m ago
I can't be 100% sure, but I think even if I hadn't been told, I would have pegged this as LLM-produced. It has the exact sort of "how do you do, fellow human kids?" energy that I'd expect from an LLM that was prompted to create a post that sounded casual, especially the very first paragraph.
I saw this article and was saving it to write an effort post, and now you beat me to it. A shame, but I guess I should put the outline to use anyway.
My intent would have been to use this article to highlight my concern about the AI revolution, and share my perspective on a topic I've never really gone into.
I am on record on being a skeptic / doubter on AI singularity fears (or hopes). I broadly think the 'the winner of AI is the winner of all' is overstated due to other required dynamics for such a monopoly of power/influence to occur. I think other technology dynamics matter more in different ways- for example, I think the drone revolution matters more than the AI revolution for shaping geopolitical contexts in the decades to come. I think that AI technologies under human control are more likely to do something irrevocably stupid than AI-controlled technologies deciding to paperclip everything and somehow having the unique ability to compel all other AI to align with that.
I do think it's fine to characterize AI as a significant disruptive technology, even if I think the inherent limits of LLM are more relevant to certain fields (especially anything novel/emerging without substantial successful learning material) than is commonly appreciated. Something doesn't have to be world ending to be a major disruption. I just think it's one of many, many major disruptions in the decade to come, and not even necessarily the worst. (Though disruptions do compound.)
What scares me isn't the AI singularity, but the AI-educated youth.
Specifically, I fear for- and fear from- people who might otherwise have learned critical thinking skills in how to not only search for answers, but organize and retain answers, to things they didn't know at the start. The example in the articles covered people using AI not only in lieu of finding a solution, but even knowing what the solution was. (The students who didn't know their own essay's response.) I don't think AI is bad for students because the answers AI provide are bad, necessarily. Getting an answer from AI isn't that different from getting an answer from a first-few pages search of google. (Even before they were the same thing.) It's more that if you don't even know how to do a tailored good search, or you don't know where other alternative answers are, you can't compare even that result. And if you're not retaining the solution- if you don't understand 'why' the solution is correct- the student is missing the opportunity. What's the point of passing a test if you, the student, haven't learned?
And I think the process of learning is important. In fact, I think learning the process of learning is among the most important things to learn at all. How to find an answer you don't know. How to distinguish good answers from bad answers. How to detect and distinguish bias from error from manipulation. How to generate a new solution to a complex problem when there isn't a proven solution at hand, or if the old solutions aren't accessible because [reasons]. And finally, how to both organize and communicate that in a way that other people can use. 'Knowing' a lot is not enough. 'Communicating' it can be just as important. All of these are skills that have to be practiced to be developed.
AI can compromise critical thinking and skill development. AI can compromise learning how to look for answers. AI can compromise how to retain the answers. AI can compromise the ability of people to respond to unclear situations with incomplete information or no baselines. AI can compromise the ability of people to convey their ideas to other people.
I had a great big screed on how I think AI is ruining youth... and then I looked back to that first mention of google, and asked myself 'what is so different?'
I grew up in an era where the pre-AI internet promised unparalleled information access. An era where seemingly infinite libraries of fiction (fan or otherwise) were open to anyone with an internet, with more to read than a lifetime of book purchases. Access to other people's opinions would break people out of their small-minded closed-worlds. The truth was out there, and the internet would help you reach it. In one of the earlier versions of Civilization, the Internet was considered a world wonder, and would give the civilization that developed it first (eventual) access to any technology that at least two other states knew.
But I also grew up in an era where people bemoaned that google was ruining the ability of people to find anything not on the internet. Documents that were never digitized, people who never wrote down their thoughts, the subtext that comes from investing things in person rather than from a distance. You can think you know how hilly a hike is from reading it, but a picture of it is worth a thousand words, and actually hiking it yourself in the heat and humidity and while carrying dozens of pounds of equipment is something else. It's hard to capture the sublime beauty of nature, and thus understand why people would value nature preservation for its own sake, if you don't go out to it.
(Then again, I did go into it. I also didn't like it. My sympathies were never exactly with anti-industrial environmentalism after that.)
And it's not like the pre-AI google-internet wasn't directly facilitating cheating. Who here was ever introduced to SparkNotes? The best friend of anyone who didn't want to actually do the required reading, but still needs a talking point or essay about a famous book. It advertises itself as a 'study guide' site these days. It condensed hundreds of pages into a few small pages of summary, and that was Good Enough.
Similar points could be made about cheating. I remember when facebook was not only young, but mostly a college student thing. And I remember how schools wrestled with students sharing answer sheets to quizzes, past essays, and so on. Even if I didn't partake, I know people did. Were they getting substantially more critical thinking skills than the modern AI exploiter just because their cheating methods were a bit more taxing on time or effort?
Maybe. But then, what's so different between the pre-AI/post-internet student cheating, and the pre-internet student cheating?
Were cheating circles any less of a thing in eras where colleges had notorious stories of famous historical figures basically fooling around until last-minute cramming? Were those cramming sessions really imparting the value of critical thinking not only to the Great Figures of History, but their less memorable peers?
Or information. If you're getting all your politics from AI, that was pretty dumb. But then, I remember when it was (and still is) a common expression of contempt to dismiss people who watched [bad political TV station], or read [biased partisan news paper], or listened to [objectional radio figure] rather than the other alternatives.
But were the people who were turning into [good political TV station] being any more critical thinking for listening to the 'correct' opinion shows? Or was it just 'my noble voters know I speak truth through their own critical thinking, yours are misled by propaganda that critical thinking would negate'? Were radio listeners decades prior any less mono-tuned for having even fewer alternative stations to listen to? Were regional or municipal newspapers any partisan when there was less competition outside the influence of political machines? Were their readers any more objective critical thinkers when there were fewer easy alternative options?
Has there ever been a golden age of critical thinkers, schooled to think well, untainted by the technology of its era, or the character of its students?
Or has critical thinking been consistent across history, with most students of any era doing the least possible to get through any required courses, and missing the point along the way?
And- by implication- some minority of critical thinkers existing and emerging regardless of the excuses of the era? And often out-competing their contemporaries by the advantages that come with critical thinking?
The more I think of it, the more convinced of the later. Most people in history wouldn't have been great critical thinkers if only they had access to more or even better information. They'd still have taken the easiest way to meet the immediate social pressure. Similarly, I doubt that the Great Critical Thinkers of History would have been ruined by AI. Not as a class, at least. They already had their alternative off-ramps, and didn't.
Critical thinking can always be encouraged, but never forced. The people who do so are the sort of people who are naturally inclined to question, to think, or to recognize the value of critical thinking in a competitive or personal sense. The people who actually do so... they were always a minority. They will probably always be a minority.
So on reflection, my fear about bad students isn't really warranted by AI. There has always been [things degrading critical thinking] that the learners of the era could defer to, or cheat with. If I'd been born generations earlier, I'd have had an equivalent instinct 'warranted' by something else. My fear is/was more about the idea of 'losing' something- an expectation of the critical thinking of others- that probably never existed.
Realizing that made me fear the effects of AI a bit less. As silly as it sounds to put my updated prior in this way, and the sillyness is the point here, there was no golden age of critical thinking and enlightened education that just so happened to be when I was maturing. Just as [current year] wasn't the first time in human history moralistic college students felt ideal social morality was obviously achievable, a downgrade of critical thinking didn't start after I left college either.
So when I read that article about the south korean kid who viewed Ivy League not as a chance to learn in an environment of unparalled access to quality minds and material, but as a change to meet his wife and co-founder of some company, I shouldn't- don't- despair. Instead, I shrug. As it was before, so it shall be again.
Two centuries ago, his mindset would have been right at home in his home country. He would probably only have cared about the material the nominally-meritocratic gwageo civil service exams assessed (including classical literature) to the degree it let out-compete other would-be competitors and join the yangban, a relatively comfortable aristocratic-social class. If he had the ability to cheat at the civil service exam and get away with it, I imagine he would have.
I doubt the social sanctity of meritocrat exams would have bothered him anymore than the espoused value of critical thinking in a progressive academic institution.
What is so different?
hydroacetylene
59m ago
Infantry privates don’t sleep in barracks on base either, not in the U.S. military. They live in dorms that aren’t as nice as on college campuses but are pretty similar to conditions for 90’s dorm rooms.
Kind of. As I understand it, Timmy is more about "dumb" big flashy stuff, Johnny more about "brainy" subtle off-meta strategies. Similar to Spikes, Johnnies still play for a challenge, but the challenge is about making some weird game mechanic work, not straightforward winning. In my experience, Timmy is the most derogatory term in practice, basically saying someone plays like a five year old or at best "just for fun" with no effort whatsoever, Spike is in the middle, sometimes used negatively for tryhards, sometimes positively for straightforward good playing, and Johnny is the most positively connotated, the kind of person who doesn't "netdeck" but still wins often enough due to their good deck building & playing.
MadMonzer
Temporarily embarrassed liberal elite
1hr ago
Of course the worst parts of Western leftism come from Christianity - almost everything about Western civilisation, good and bad, comes from Christianity. (And it is fair to say that the bad bits of Western leftism don't come from Ancient Rome).
Thoroughlygruntled
1hr ago
I wonder how many billions of dollars (and how many lives) have been lost due to a "eh, I skimmed it and got the gist" approach to analysis fostered by forcing students to read large quantities of horribly overwritten, low content academic work.
sarker
It isn't happening, and if it is, it's a bad thing
1hr ago
Anyways doing coursework is a huge ass chesterton's fence right there.
Pretty sure coursework is the newcomer to the University education scene. I don't think that professors in 17th century Oxford were grading homework.
NewCharlesInCharge
1hr ago
It wouldn’t eliminate the problem, but the proctored exams could filter out many of the people who wouldn’t be able to hack it but for cheating.
Or maybe even force them to learn how not to cheat.
Chrisprattalpharaptr
Ave Imperaptor
1hr ago
Once upon a time, this is what I got out of Wheel of Time. It didn’t matter if the prose was florid or the plotting glacial. The sprawl was the point. I wasn’t reading it to find out what happened in each finale, but to watch the setting evolve, further selling the illusion of another world.
That's curious; I'd find myself skipping Perrin, Egwene, Elayne, Nynaeve POV chapters out of boredom to get to the climax with Rand. Almost all the moments from the series that stick with me a decade or so later are with Rand: Picking up Callandor (and trying to revive the dead child), Rhuidean, cleansing saidin, using the True power against Semirhage for the first time, his epiphany on dragonmount.
I suppose as a teen I was even more of an uncultured swine than I am now.
phailyoor
1hr ago
What I have seen around Europe typically was that learning happens during exam crunch time and coursework is either just recommended or has relatively little effect on your end grade.
That's exactly my point. The kids will just goof off for 90% of the semester and just cram it all in before the exam. If so when what's the point of even having the rest of the semester.
using chatgpt to write personal introduction
Even if it's busy work, using chatgpt to do something that would take literally 5 minutes is just a retarded move.
My problem with all the people calling it rigged is this: I've been a poker player for half my life. And in poker you very quickly learn that 1.8% odds of something happening is not the same as 0%.
More options
Context Copy link