i write everything on a whim which means what i write tends to be based on something i read recently somewhere or whatever topic i’m interested in at the time, which shifts every month or two. did you know, for example, that several months ago i went through a brief light novel phase during which i read and reviewed like four light novels in the span of a month. the evidence is right there on this site if you click the “literature” link up top, the one in quotation marks. anyways since then i have not touched a light novel, in fact i have gotten quite negligent in studying japanese lately, i barely even look up kanji or words i do not recognize anymore and my stack of anki cards to be reviewed is akin to mt. fuji. anyways if you just wait a month i’ll probably get into a new phase and start writing about something else, although i can’t guarantee it will be to your taste.
ah yes, for my website i use hugo and tailwind CSS, you can find the source code on github here... HA just kidding i do almost everything manually. you can see the source code by right clicking and selecting “view source”. pretty much all of my pages use the same 20-line template which i start every new html file with. i write most of the actual "content" using an microsoft word (with an expired license) and then copy and paste it into the online editor, seasoning it with html tags where appropriate. sometimes, especially for the blog posts, i just rawdog it and write everything directly in the editor. don’t ask me about how the side notes or margin images work because frankly i’m not really sure. fun fact, even though i’ve gotten almost perfect grades in every cs class i’ve taken, i have never bothered learning how to use github “properly”. pretty much every commit i’ve ever made to github is just copy and pasting stuff into the github web editor. i will still be listing “github” as a skill on my resume however.
also, about the description list thing: do not, i repeat, do NOT look at the source for this page.
ah yes the topic du jour, all that ai bullcrap. everyone thought they would be automating stuff like driving and flipping burgers first so that people would be free to pursue art but instead they ended up automating art first, whoops! guess all the artists now have time to take up driving and burger flipping. i personally was considering getting into something difficult to automate like landscaping, specifically cactus cultivation, but then i saw some of the chatgpt output and i was like, really? most chatgpt responses i've seen are unbelievably dry and sound like they were written by a pr department to be as milquetoast and inoffensive as possible, sounding confident and saying a lot of words while not really saying anything, if not outright lying. most people seem to have caught on to this and so they’ve turned chatgpt into a kind of game where you try to trick it into saying something racist or otherwise offensive.
i kind of get the feeling like chatgpt is just spitting out the “average” response based on the terabytes of text training data it’s been fed, which is naturally going to end up fairly wishy-washy. i’m reminded of this picture i saw in national geographic as a kid of a composite picture they put together of the “typical person” based upon all the science and statistics etc., which ended up looking like a very fuzzy chinese guy. i looked this up just now to verify i didn’t imagine it and found a hilarious article where cbs went in search of the mythical “typical person” according to all the metrics set out by national geographic, and found one in new york. when they informed this guy he was the world's most typical person, he didn’t have anything interesting to say about it.
the problem with all this machine learning ai stuff is that it’s all a numbers game where you want as much training data as possible. unfortunately, bad writers outnumber the good, especially on the internet where they tend to retrieve training data from. there just might not be enough good writing out there yet to brute force train an ai that can write well. plus, basically all the recent ai advancements are from using larger and larger training datasets and not due to any significant changes in the underlying theory or technology. the problem is, they have kind of already used up everything that’s ever been written on the internet, and there’s nowhere else to get more training data, so using current techniques gpt won't get much better than this.
i estimate that in the past year i have written one million words (disclaimer i pulled this number from thin air). this is a dramatic increase compared to the previous who knows how many years, where i’d probably get out 10k words at the absolute max, most of them of low quality for classes. sometimes these things just happen, all you have to do is sit down and start one year. it has come at a cost, however. i seem to have permanently destroyed my ability to crap out quick mindless bullshit responses to equally-insipid school assignments. i would say it is an acceptable trade.
as for tricks, well, everything i put up is usually very light edited, so there’s that. i also have a lot of time to devote to writing because i don’t really play any video games or watch many tv or movies. mostly i do a lot of reading, which can be converted into writing at a roughly 100:1 ratio. but at the end of the day i’m not really sure there’s anything you can do if you don’t have some fundamental urge to write that can only be relieved by getting words out on the page, like peeing. you can always try to force it, but remember that a bored writer creates boring reading.
i believe there is a mounting crisis of complexity in the developed world. the baseline complexity of living a “normal” life is gradually increasing, and not for any particularly good reason. an example: individual taxes in the us are so outrageously complicated that there is a massive consumer tax preparation industry. but if you mess up your taxes, the IRS will send you a bill or a check for the difference, which implies that they already know how much you owe without needing you to fill out a tax return. in fact that is exactly the case, and the only reason individual tax returns are still required is due to lobbying by tax preparation firms. now multiply that sort of unnecessary complexity into almost every facet of life, and you’re in a situation in which even smart people are starting to struggle to keep up. if this keeps going the minimum iq required to function in society is going to get so absurdly high that most people are going to find themselves completely buried beneath bureaucratic burdens. no wonder many people these days are so stressed and wanting to just "drop out".
i don’t know for sure, i’m horribly fickle. one thing that’s been on my mind lately is doing a Good Old-Fashioned book burning to kick off the new year. “but wait you read so many books and talk about literature all the time why would you possible want to do such a thing, also muh fahrenheit 451”. first off i have not actually read fahrenheit 451 (which probably makes two of us) but i am absolutely convinced it is overrated based off of a graphic novel adaptation i thumbed through in the middle school library. if you want a book that jerks off books and reading in general that isn’t obnoxious or preachy may i recommend “if on a winter’s night a traveller”. second, once upon a time i was indeed an “all books are valid”-type romantic but of late it has become increasingly clear to me that infohazards are out there actively poisoning the collective unconscious, some of which come in book form. i am speaking, of course, about vile optimalist propaganda, countless lists of which you can pull up on any social media (youtube, twitter, etc.) immediately by searching things like “Top 10 books that will sextuple your productivity, make you a millionaire, cure your depression, help you find God, and improve your love life”. most of the books on those lists will be the same, probably because the "content creators” making them don’t actually read them and just copy each other. anyways i’m sure you will agree that most of the books on those lists are ok to burn, not just because they are readily replaceable at any airport's bookstand.
i have also been thinking about digging a shimeji simulation tribute hole. maybe i can do the book burning IN the hole. wouldn’t that be great.
that is because that’s where i got the color scheme from. now allow me to guilt-trip you about this blatant theft. i’m tremendously colorblind and thus i don’t really do colors for fear of accidentally horrifying or offending people. in high school geography class we had these awful, pedagogically-dubious busywork assignments of coloring topographic maps of various continents, the kind of assignment that must exist solely so that they can identify ideal candidates for the future élite, namely the students who meekly accept the assignment and put lots of effort into neatly and accurately completing such a worthless, arbitrary task. anyways the topographic maps we were copying were absolute hell for me because the colors in them went from green to red. not only that, one time i even managed to mess up even the easiest part and accidentally color the entire ocean purple, to the displeasure of my teacher when i tried to turn it in. so now i try to outsource color decisions to other people whenever i can.
incidentally, i once read somewhere that colorblind people are more likely to be depressed. i think i know why. for my birthday one year a distant relative (my dad, who was living on the opposite side of the country at the time) sent me a pair of “colorblindness-correcting glasses”. when i put them on it was like entering the land of oz. everything looked so bright, saturated, and saccharine, like being on psychedelics or something. “you people really live like this?” i exclaimed in disgust as i tore them away from my face. some of my friends tried them on and it seems to have the opposite effect on non-colorblind people, allowing them to see like a colorblind person. “wow, everything looks so gloomy and depressing,” they commented. so there's that.
i think we're done here.