by Michael Skyba
(Initially written on 2024-07-18)
A few weeks ago now, I finished my twelfth year of public education, and graduated from my high school. At last?
I can still picture myself inside the graduation ceremony. The ceiling fans did little to quell the heat. The graduation gown the board decided on, sold by a company whose X account I now see has 8 followers, stuck to my increasingly sweaty wrists. I tried rotating the position of the sleeve every fifteen minutes, while wondering how much better at negotiating I’d have needed to be to escape my parent/guardian’s command to attend.
The surrounding seats were filled by students I had never spoken to, and now hopefully never will. Throughout all of the speeches and presentations, the one to my left, when not stabbing at the neck of one of his nearby friends in the next row, scrolled through TikTok on his phone. Another victory for ByteDance’s training data?
In one section of the ceremony, every graduating student had their name called sequentially. Each student would walk to the stage, receive a counterfeit piece of paper symbolizing a diploma, pose for a picture, and then walk off to make room for the next student. It took at least an hour to iterate through everybody.
At the start of the section, the announcers clarified that we were to hold our applause until the very end, but people still opted to clap after each name. The more popular the student, the more cheering you could hear.
When it was finally my turn, my thoughts turned to the correct procedure for accepting the paper and shaking the principal’s hand, and my ears failed to make an evaluation. My guess is that the magnitude of applause would have been in 20th percentile or so.
A handful of students would yell out inside jokes or otherwise obscene slogans into the crowd as they saw their friends being called. I wonder how the parents felt when the principal spoke her grandiose words of praise, emphasizing how every student in our school would go on to achieve great things. We, the leaders of the future, comprised the heart of our strong community! I felt sick.
Who was the target audience of this? Is this really what those anime writers were trying to capture, when they put graduations into their stories?
A subset of the announced awards were for genuine accomplishments, and I clapped with respect for the recipients. Still, is this what was motivating them to keep working? Did they feel fulfilled, as they looked around at the restless faces and held up their prize?
For the finale of the event, it was time for the students to throw their caps into the air, as a unitive, sacred tradition!
…I couldn’t force myself to do it.
…
In this post, I will write about some of my experiences throughout schooling. Perhaps there’s a total of at least three or so people whom could find such descriptions to be interesting. Mostly, I’m writing this because I expect to wish to look back on it in the future, and would hope for it to have slightly more polish than my usual notes.
The focus will be on high school. My memories from elementary school are so hideous (near-entirely the fruits of my own shortcomings) that I’d prefer to privately extract the few valuable updates they offered, and toss the rest away. I only intend to lightly touch on them.
Regardless, these recounts should not be taken as serious evidence for an assessment of the school board’s quality, considering how overflowingly biased I am going to be on any relevant topic. I would expect many other students to disagree on anything I paint as supposedly positive or negative.
Let’s start with the social environment. You can guess that I’m not exactly qualified to have a strong opinion on it, but I’ll still describe it as disappointing.
Imagine yourself silently working on your laptop, rubbing your hand across your hoodie’s sleeve because it’s somehow always freezing in your P3 classroom. Suddenly, you hear the distant sound of a crash bar compressing. What do you expect next?
Hmm, someone must have entered the central passageway through the Great Hall’s stairwell. They must be returning to class, so I’ll hear them gently close the door and then calmly walk to their appropriate room.
Ah, what a sweet tale…
…Yeah, that’s not how it goes.
Instead, the first thing you will hear is the sound of the adjacent crash bar, because the second door is opening, to make room for the stampede of 10 students who are traveling together.
Huh? What’s that new, high-pitched sound? Did the fire alarm turn on (someone inevitably has the heroic idea of tripping it as a joke every few months)? Nah, tuned specifically for you, that’s an original, deluxe mix of screeching and cackling, because one kid is lunging at the other as the group pushes each other through the hallway.
Whoa, something’s getting louder? That’s the sound of hard, hurried steps, because they’re fully sprinting now. That echo of skidding collision, reminding you of bent plastic? Ah, that’s nothing to worry about. It’s just them literally tugging the garbage bin off of its hook on the wall and whipping it down the hallway.
(This is around the point where three different teachers would have to interrupt their lessons to leave their classrooms and try their best to pacify the crowd.)
Fortunately, the worst of them would skip class entirely and have their “harmless fun” outside, behind some convenience store somewhere. Also fortunately, this type of student will often opt to take the “Applied” versions of courses, making it rare for you to have to handle them directly.
Downstream are some effects on infrastructure, such as the state of the washrooms. Even in elementary school I had a policy of avoiding public washrooms, but high school took it to another level.
If you ever come in, even if only to wash your hands, you’ll see a looming circle of hooded figures, but only barely through the clouds of smoke.
As you reach for the soap, you’ll realize someone has already stolen the entire dispenser off of the wall, and so you have to use the other side, closer to the group. You distribute the foam as quickly as you can through your fingers, hoping to avoid the pieces of gum and mysterious red stains inside the sink, and then scuttle out of the room, before they have a chance to start offering you anything. After finally escaping, you’re left with the sense that your hands are dirtier now than before you entered…
To the detriment of my physical health, I quickly decided to never bring my water bottle to school, and thus only consume liquid in the subset of the day after I return home. In certain cases, I’ve needed to make exceptions, and so in total over my entire high school career, I’ve used the washroom around three times.
It’s common to hear of other students leaving school at lunch, walking all the way to the Metro in the nearby plaza (~5 min each way), using the public washroom there, and then walking back.
Relatedly, almost every school assembly starts and ends with the vice principal devising new strategies to deter students from their favourite ritual: showing up late to class, asking the teacher to use the washroom, and then spending the rest of the period vaping with their friends around the stalls. It never seems to work, though; it’s still common to see these large groups enter the washroom together, and thirty minutes later, for an irritated teacher to storm in (“Out! Now!”).
My guess is that a large minority (~30%?) of students regularly consumed recreational drugs. My few ~acquaintances were not the type to be interested, so I lack solid, empirical evidence on this. Yet I have overheard otherwise seemingly normal people casually admit to, for example, having smoked cannabis while skipping their previous class.
One student, after learning that my weekends consisted of programming or doing other work on my laptop alone in my room, genuinely believed I was lying when I later clarified that I was not relying on any such drugs to support my lifestyle.
Modafinil? The only drug 🍎 you really need is the upbeat, charismatic sight of Sam’s smile as he announces GPT-5.
Fortunately, many reached the threshold of neutrality, but unfortunately, with very few exceptions, I didn’t see students who were actively attempting to acquire knowledge and ability. It seemed like people’s standards only ever matched, if even, whatever the school, or their parents, directly demanded of them. The common priority wasn’t anything amusing like deliberately shortening timelines or Death With Dignity, but instead ensuring an invite to the pool party being planned for next Friday night.
To clarify, I’m not trying to pretend like I’m particularly noble either. I’d describe my skill level as a merely pitiful “capable of performing a ‘web development’ or ‘software engineering’ role at an average tech company”. (Of course, anybody somewhat capable of learning anything, including me, could be ramped up into higher roles; it’s hard to draw an exact boundary on cache.) I believe that this, however, is literally the 100th percentile for my school. Some students clearly exceed me in certain niches and some were close in total breadth, but the supermajority didn’t have any care about progressing in anything.
I suppose you could argue that seeing people avoiding your goals like the plague gives you a reference for behaviour not to copy. Although perhaps that could be beneficial to some, it has only ever felt draining to me. External poor performance only gives me something to point to as an excuse to slack off. Surely I’ve done enough and deserve a break if I’ve passed N fraction of the rest? No. I’ve always been and still am far too slow. Subconsciously, it’s a debilitating anchor, in this case for normative self-expectations.
I much prefer the kind of environment where I’m only in, perhaps, the 5th percentile of ~skill. I’d say this is indeed the case for the circle of Xitter that I frequent, and although much of my time there has certainly been counterproductive, it is consistently motivating to behold so many people miles outside of my figurative reach.
For a long time, I was dead-set on this necessarily being the case in all
possible settings: whenever I would witness any student achieving anything
remotely nontrivial, I would immediately create a long backstory in my head of
how infinitely talented they must be and how much of a living waste of space I
am compared to them. I later came to realize that this isn’t an amazing practice
on an epistemic level, but it still obviously has instrumental value in certain
cases.
The staff were usually not too much more impressive. I don’t remember meeting a single teacher since Grade 1 that seemed to actively pursue further study in whatever subject they were teaching. (In particularly broad cases, you could say that gym teachers are “exercising” in their spare time, for instance, but it’s not the same.)
However, some at least seemed to like their subject, and on the whole, there was a nonzero level of personal interest in their students learning and succeeding.
A couple were even willing to invest additional time to support students in related pursuits outside of class. The CS teacher, for instance, went out of his way to allocate part of the tech department budget to purchasing Raspberry Pis. I wrote some software to run on them, and we ended up deploying them around the TV monitors, as an additional display of school information.
(The NVIDIA presentation was a test deployment that I soon after had to remove.
For other tests, I played a few of my Advent of Code and Touhou 10 recordings.)
I’ve had some other, mostly regrettable social experiences that could be seen as “interesting”, but I won’t be able to share them publicly.
Sure, anyone can get unlucky with harm by association, but what about the education itself?
For some context, here is a final screenshot of my user account, on the web platform the board employs for archiving course marks:
(You can tell that the 100 in ENG4U1 was given by mistake, at least, judging
from the slop of this post so far. I suppose the expectation is for students to
fail even to string simple sentences together.)
It makes me feel ill, thinking back on how much time I wasted, how much brain damage I coerced myself into, all for these tiny numbers. Judging from my comparisons with other students, the university programs I was accepted into would likely have made the same decisions even if my average was, for instance, 5-10 points lower, which could have saved hundreds of hours.
The picture’s only significance now is that it can lend a few drops of credibility to my opinion about the processes that determined it. I only know a handful of students who would consistently earn higher grades than me across our shared courses.
It’d be a safe summary to say that as a whole, my experience with the content of classes was of dislike. The material suggested by the curriculum to be covered, while almost exclusively shallow, was occasionally interesting as information. The bulk of your engagement with a course, however, is not the lessons. It’s… the homework assignments…
Since Grade 1, I don’t remember hearing a single student ever unironically express fun or enjoyment at work they’ve been given to complete. I can’t speak on their behalf for what they value spending time on, but in my case, a large component of the drag was that the instructed tasks would be laughably inefficient ways to become familiar with the relevant topic.
Even if the content itself was something I was genuinely interested in, I was left with a constant sense that anything I was doing was not for my own benefit, but to satisfy a teacher running out of ideas on how to “creatively” design an activity.
As a minor example, part of the defined curriculum for Grade 9 MPM1D1 is that students should grasp the basics of linear relations and analytic geometry. If the course was designed with so much as a grain of respect for the student, a relevant task might involve thinking through something like the following:
- a) Given a real number r between 0 and 1 and two points P = (x1, y1) and Q = (x2, y2) in the plane, find the coordinates in terms of r of the point T on segment PQ such that PT/PQ = r.
- b) What happens if you allow r to be greater than 1 in your formula in part (a)? Where on the line connecting P and Q is the resulting T?
- c) What happens if you allow r to be negative in your formula in part (a)? Where on the line connecting P and Q is the resulting T?
(This is problem 8.59 from AoPS Intro to Algebra.)
Instead, in an actual MPM1D1 course, a breathing adult will look at you with a straight face as they toss you leftover vomit:
(This exact image is only of a worksheet; a marked, summative assignment in
MPM1D1 would be similar except with 10x more of the same questions, and no
option to submit work digitally.)
Mathematics makes for a clear comparison, but in this respect, the most subjectively wearisome assignments would often come from humanities courses. They would demand a large amount of thought about and effort into something backwards and inconsequential, mostly unrelated to the actual course subject.
Depending on the severity, it’s common to hear other students complain about the magnitude of work, or a feeble “sigh …Why do we have to do this?” You can witness the looming discomfort usually convert to them ducking under their desk and opening Snapchat.
For most students, I think this is a natural form of Yudkowskian procrastination. When actually engaged in the work, they have, at most, a mild frown on their faces.
For me, I don’t think this standard model fully applies. As an analogy to my perspective, imagine sitting in front of a small button. Though its glossy surface may appear innocent, if pushed, an activated mechanism will proceed to slowly cut your entire left arm into little slices of meat and bone. Staring at the button, the inevitably induced procrastination may leave you compulsively squirming in your seat.
In this scenario, the pain in procrastinating is not out of shame at yourself for not having started, but out of genuine fear. The inherent cost in switching tasks itself is trivial, and any temporal discounting will be of the future pain of your entire body going through the same chopping process (the theoretical price if you ultimately fail to press the button at all).
I’m only moderately exaggerating. The same functional dynamic would apply to my experience with many such assignments. Every minute spent working on them would feel like being not simply annoyed but actively injured, with each internal scream of "Please… please stop" only delaying its end.
In exceptional cases, I was simply incapable of forcing myself to flood the searing pain through the entirety of my conscious attention. Despite the accompanying nosedive in productivity per unit of time, I’d need to primarily focus on a neutral, unrelated activity, and only occasionally jot another idea or line down on the course task in the background.
The harshest such assignment was definitely the culminating of Grade 11 English. By itself, it would already have been nearly unbearable, but it also compounded with some severe, external issues I happened to be managing at the time. I don’t know if I could have pulled through if I wasn’t rhythmically chanting to myself all throughout:
Based on my current academic, professional, and personal trajectory, it is highly probable that the pain inflicted on me in this moment will have been the most excruciating out of the entire duration of my lifetime. If I can somehow survive these next few days, it’ll all be downhill from here…
(For a number of reasons, I’m not going to describe the actual work involved in the culminating.)
Perhaps in part related to these assignments is that over time, a growing fraction of my habits, routines and overall mental infrastructure have developed solely around this meta problem of “coping with dissatisfaction”.
Fortunately, unlike summative assignments, in-class informational lessons are easy to ignore. I started doing so the second I had the maturity to assign a hint of thought to what I was spending my time on. It quickly became apparent that even if I made my best effort to pay attention, I wasn’t going to be learning much of anything.
One fairly early example is that during Grade 10 GLC2O1, we students were supposed to be learning about personal finance, so in class, the teacher (“H”) showed us sample, hypothetical student budgets. In these budgets, a small fraction (e.g. 10%) of your income was allotted to a “savings” bucket, and then you would actively and deliberately seek out random expenses to fill the remaining 90%.
I was fully expecting to hear a lecture on why this is a terrible plan to follow, but H instead described it as perfectly reasonable, going so far as to affirm that 10% was atypically high. Upon witnessing this, I was deeply disturbed. A rare sense of gratitude appeared towards my past self for at least having learned about relevant basic concepts like the compound growth of market indexes.
After class, thinking that I must have misunderstood something, I asked H why the examples weren’t trying to minimize expenses and reach, for instance, an 80% saving rate, rather than a 10%. H’s response was that lower spending is conceivable in theory but will involve far too much willpower than what can be implemented in practice. I was told that over the course of my future career (implied: many decades), 10% is already sufficient for eventual retirement.
(At the time, I had already decided to skip common memes like going out to eat during lunch or purchasing Clash Royale gems. My only expenses were of buying a Touhou or Danganronpa game once every six months: without being particularly diligent of a student or a person, my rate was around 95%.)
This pattern repeatedly applied across any topic covered that I felt even somewhat comfortable with my prior understanding of. It wasn’t limited to disagreements in philosophical stances: I often heard teachers explaining ideas in ways that were inconsistent and self-contradictory. They assumed you were incapable of following the standard theory, and thus would rely on leaky, oversimplified analogies that discarded critical nuances in the original concept.
I’m left to assume that in courses whose content was entirely new to me (e.g. Physics), a more knowledgeable third party would similarly be unimpressed with the presentation.
Most students, of course, only care about extracting the password that will later need to be regurgitated on the test, in which case any mutually perpetuated errors aren’t a particular issue. The few who have genuine interests in deeply grasping the material, however, will essentially have to start from scratch and relearn everything on their own.
The final notable component of your experience as a student will be the iron insistence with which teachers enforce their personal intuitions about optimal learning and work systems onto you.
As an example of what this can apply to, being physically within the walls of the school has always felt threatening to me, regardless of where I was. Some sense of composure and subjective “safety” is important for me to focus, so this spatial discomfort was an added layer of resistance I had to endure when completing the already draining assignments.
Hence, it was always most efficient for me to leave the particularly painful work for when I arrive home and have more control over my environment, in my room. During school hours, I preferred to complete tasks that were either cognitively light, like reading through entries in my RSS feeds, or intrinsically motivating, like solving K&R exercises.
Teachers, of course, have zero sympathy for such approaches: when you’re physically present for the specified 75 minutes of class, they demand that you also be mentally present for 75 minutes.
A common variation of this I’ve noticed in other students is that they sometimes have an important assessment scheduled for their next period. If we’re not engaged in anything particularly urgent, a student may wish, quite reasonably, to neglect whichever frivolous time sink we were handed, and instead spend the current period studying. You can guess what a teacher’s reaction to this is if they learn of it.
Among myself and many others, the result is not a magical loyalty to the teacher’s whim. I continued trying my best to optimize my schedule, to work on what I most believed in at the moment. Yet within this state, much of my mental effort had to be spent maintaining a dynamic mental model of the position of my seat relative to other entities in the classroom. I needed to be ready to switch the active desktop of my laptop’s window manager, depending on the teacher’s viewing angle.
Such practices were as greatly irritating as they were distracting. Occasionally, teachers enjoyed wandering the room so relentlessly that I had no choice but to work as they intended. Every time this happened, as expected, my pace of progress was a crawl compared to when I’m at home. I was initially reluctant to disobey official instructions in Grade 9, but I gradually discovered that faithfully pursuing my goals was incompatible with external honesty.
A more asynchronous example of this is that in mathematics courses, teachers assign pages of mindless homework and genuinely expect you to complete every exercise. (These were similar in style to the graphing worksheet I displayed earlier, but would be designed to consume around 45 minutes of your spare time per day.) In four out of five of my math courses, teachers regularly performed “homework checks”, sequentially visiting each student to confirm that your notebook contained solutions, while marking down notes on a list of names.
Halfway through Grade 9 MPM1D1, I became aware that these checks only influence the “Learning Skills” section of my report card and not my actual grades. Shocked, I asked the teacher whether there was any other possible motivator for students to heed them. After receiving a blurry answer about how I couldn’t possibly be expected to comprehend the material otherwise, I stopped all homework, and haven’t completed a single question since.
I understand that for certain students, this sort of micromanagement is entirely necessary, but there is literally nobody to whom they are willing to extend trust. If a student has a 100.0% in your course, maybe they know a thing or two about effective work and study habits? Perhaps at least consider treating them as something close to an autonomous agent?
Grade 12 MCV4U1 was an outlier: the teacher seemed to realize that my time was being wasted by the forced in-class practice questions at the white boards, and then chose to exclude me from the student pool when randomly compiling work groups. From that point onwards, each day, I would simply show up to class, sit with my back bent on the awkwardly-shaped bar stools, and type away on my laptop.
For that one course, I was completely free to handle my own studying for examinations, to not have any pretense of care for whichever mockery was being discussed in class. (Most lessons would involve extensively describing an algorithm to solve a supposedly new “special case” of a problem type. Anybody should have been able to derive the steps independently within a few minutes of thinking, but no, something apparently justified spending weeks on them.)
My resultant lack of class participation resulted in an S (“Satisfactory”; the second-worst evaluation) for “Collaboration” on my report card.
These twelve years had nothing close to a positive ROI on their own terms, but a nontrivial silver lining exists. Any somewhat valuable knowledge or ability I currently possess has developed downstream of the following chain of realizations:
If the school’s quality was a mere 20% higher, if a few teachers were just slightly more willing to engage with my clarification questions… then maybe #1 would have stayed diluted, I would never have been conscious of #2, and everything would have went wrong. Perhaps I would still be peacefully gliding to university right now, rather than bitterly complaining inside a markdown file.
(Being glad that a process accidentally changed your utility function is a highly questionable form of praise…)
In hindsight, George Hotz’s yearbook quote is remarkably wise:
“If something’s worth doing, then it’s worth doing well. Most things are just not worth doing.” (paraphrased)
If I had to relive school again, I’d make the absolute minimum time investment required for a 50.0% average (barely earning every course credit) and then muster as much focus as I can into other, actually useful pursuits.
(Besides the parental pressure, a barrier to such an approach is the increased difficulty of self-assessment: “Sure, I tell myself that I’m strategically underperforming on these narrow tasks, but could I really have done any better if I tried? Am I just incapable of anything?” I suppose in this respect, projects like AoC 2015-19-2 have been far more directly cognitively demanding than any school assignment, and therefore can provide a better reference.)
My only experience with private education was in a Russian-language Montessori program that I attended, from ages ~4-6, instead of the junior and senior kindergarten years that public elementary schools offer. Judging from my memories of it, the competence of the single teacher, paired with the smaller class size of 5-10 students, resolved almost every issue I’ve described, although I’m not sure how well it would scale to higher years.
There are a few other scraps of use that high school has produced, such as my newly diminished fear of “work” in general: I now have little expectation that anything will come close in discomfort to experiences like the Grade 11 English culminating.
However, it’s also evident that I would have been many years ahead of my current self by now if my time was being used even somewhat optimally. Simply spending all those hours talking to GPT-4 about the topics in the curriculum, for example, would already be a massive improvement over dying inside a classroom. I’m bullish on similarly self-paced alternatives to the traditional environment, such as AoPS, Math Academy, and Andrej’s upcoming Eureka Labs.
As a final summary of my school board, here’s a real email I received near the end of Grade 12, regarding a supposed cyber attack that crashed the school’s attendance tracking infrastructure for a few weeks in 2023:
Oh no! The scary hacker entered the default password on a public route that the server admins didn’t bother to configure! My, how lucky we are to have the brave and “diligent” YRDSB to protect us…
In the couple of weeks since graduating, I’ve gained something roughly adjacent to work experience. I can now attest to the fact that adding bells and whistles to the web application of a generic media service does not feel intrinsically valuable.
One moral is that your work circumstances are a reflection of yourself and that I’m in a situation of trading time in exchange for nothing because I’m stupid. I happen to have some amazingly convenient excuses related to my family and living situation, but still, it’s true that I need to be planning more carefully and working much harder.
In a reasonable setting, you’d at least divert your direct monkeytyping time into dev infrastructure time, setting up more efficient feedback loops with LLMs. Unfortunately, this doesn’t work particularly well when it takes two weeks for the company to approve your installation of Git, and only after wrestling with three different security officers whose job it is to protect the stock price from you.
In contrast, the small bit of work I’ve done for OpenAI’s bug bounty program, while frustrating in certain social aspects, felt meaningful enough to greatly spike my standards. I’ve grown annoyed with anything whose only return is sustaining basic needs for survival, or even worse, generic “career advancement”.
It’s hard to predict the short-term impact of AGI, so I’m not sure to what extent any life decisions that are presently regarded as serious are going to be mattering soon, though.
I don’t expect any solving of the alignment problem (sorry, Mira, but my strict policy is to copy-paste Eliezer’s conclusion without reading a word of the theory). Regardless, in a supposedly successful post-ASI world, these issues may make even less of a difference. I’m not sure how anyone is going to pretend to extract meaning from their original status or trait T=1.5 being higher than the average T=0.9, once everyone (except Zy) has already self-augmented into T=1000.
…
