Poor form and scientific modeling

I was trying to think of an analogy for a problem that has become very widespread these days. 

The basic idea is that when you play a sport at an amateur level, you can get away with a lot of mistakes / sloppiness, and win by athleticism alone. But as you scale it up, and reach higher levels of difficulty, these same mistakes will cripple you and prevent you from advancing: you'll lose to people who may be less talented, but know how to do things properly.

And if you go on for too long that way, you'll be in a situation where you'll have to choose between being stuck at whatever point you max out at, and never progressing in the future, or being forced to go back to fundamentals and relearn everything from scratch, a very expensive and time-consuming process. 

The reason why a good instructor is so harsh on these things isn't because they're mean or out to get you specifically (at least, not in most cases) but because weeding these mistakes out early saves a ton of time later, and at a small cost, compared to allowing them to fester.

The same is true for language acquisition, and really most types of learning in general.

Ironically enough, this can mean that sometimes the worst players at higher levels are the most gifted or talented ones, at the amateur level-- their raw talent allows them to avoid getting punished for mistakes, which means they never bother learning, and this in turn cripples them when they move up to a new level of difficulty where these trivial things actually do matter. (See the whole gifted student -> pizza deliverer pipeline for proof in another subject.)

The way that I see it being relevant for science is that often people with high IQs but no actual knowledge can experience a ton of success in life, without bothering to formally study anything. Unfortunately they then assume that the same combination of 'intuition + raw intelligence' will help them in building mental models or solving problems at much higher levels. But they are actually crippled by naive assumptions and sloppy habits which-- due to their lack of experience-- they cannot even detect (though a trained professional could), and so they are doomed to spend their time crashing into predictable walls, and getting stuck there forever. 

Look at the ongoing and probably never-ending inflationary meltdown, or the AI bubble-- ultra-expensive models are constructed by pseudo-experts who congratulate themselves for their visionary genius, only to fail in the most basic and predictable ways. 

More music I enjoy

I wanted to post some of the music I've been listening to lately, in order to lower the pretentiousness of the blog-- not everything has to be political, or deranged, or whatever.

'You're being a Boomer / normie bro,' -> probably the reaction I'd get on just about every social media platform. (God forbid someone not be into garbage.)

I said this in a somewhat emo in a way in a previous post, but the bitter haters that make up most of the society / the Internet will try to cancel you, shame you, and laugh at you for anything, regardless of what you may or may not have done-- anything to score a W at someone's expense. And really, because of how warped today's society is, they'll target you harder for your good deeds, compared to deviant or evil behavior (which they overlook, or even encourage!). 

That's why I don't really mind being myself at this point, however against-the-grain not being into gay slop is-- why bother faking it? If you're going to be punished regardless, you might as stand up for yourself, and be honest about your personality.

AI, negative applications?

(I'm trying to go with a different approach, where I structure / edit / polish my posts less going forward. I think there's a certain point where being perfectionistic just restricts your output, while making you sound pretentious.)

I haven't talked much about AI, since I don't find it very interesting or useful, but a thought that has been in the back of my head is that to the extent that it is useful, it's more of a question of "negative" rather than "positive" applications.

Fundamentally, and despite trillions of dollars being thrown at it, it cannot create great art, introduce new scientific paradigms, or even really come up with interesting or insightful comments during a conversation (AI's speech sounds like a cross between an HR person and a used car salesman). It thus has no real positive contributions to make-- but I think can be used as a tool to reveal fakeness within the economy, as well as the broader society. 

Namely if a certain sector is vulnerable to what amounts to an industrial-scale Slop Machine, it means there wasn't much real work being done in the first place. AI can be used as a detector, to test for and reveal the presence of bullshit and fake work: if it successfully replaces human labor in this or that area, it's because the humans were not doing much in the first place.

So the fact that AI is having extraordinary success within academia should not be taken as a sign of its genius, but rather as an indication of the deep structural flaws and decline of academic culture. The fact that they are incapable of distinguishing between hallucinated meaningless gibberish and real honest / insightful work, reveals how low standards have fallen.

In that sense, AI is like a species, and academia is the environment. Its exponential growth within the niche reveals that the environment positively selects for garbage, which it can produce at a much faster rate than any human can (notwithstanding the valiant efforts of academics in recent decades). 

I say the same thing for everyday tasks: anything that you're doing which AI can do better, is not worth doing in the first place. Whatever it is can probably be consigned to meaningless / filler work, and cut out of your routine. 

So for example, I've heard that AI is great at writing emails-- what we can infer is that, obviously, most emails are a waste of time, and ultimately meaningless, and so perfectly suited for a machine that can only generate meaningless slop. 

Other insightful people have already noticed this, and it seems to be a general theme: instead of heading to a techno-dystopia, where the all-powerful machines subjugate humanity, what we're really headed for is a kind of absurdist Philip K Dick scenario, where the laughably incompetent machines break all kinds of things, and we all have to act like babysitters cleaning up after an immature, reckless, ADHD child.