samcodes

well this really takes me back to fiddling with FruityLoops in all my spare time in my teens. very fun, and I love the code

I think alcohol and tobacco use are primarily symptoms, not causes, of problems. They, along with depression, anxiety, ADHD, and other addictions, are strongly correlated with ACE (adverse childhood experience) score. As smoking has become taboo, other stimulants became more popular (adult ADD). I'd expect something similar if we restricted alcohol without solving the problems that lead to dependency.

if you use a library for Chinese/Japanese tokenization (which is harder because the lack of space), it seems like the rest of the code would work?

having the laity meditate at all is, AFAIK, a pretty modern development

it's possible to do NLP at scale much better than they do. they just don't care

do you have a repo? this sounds cool

"deep state" meaning... information you didn't know already?

I don’t understand how someone can say “this is the worst software I use daily” and then you say “you are wrong.” Maybe you use worse software than them? Also, you seem to implicitly default to comparing each piece of software to others in it's class; you’re grading on a curve. The parent comment is saying “of all software, these are the worst”, and you are saying “but other PM/VC/OS software is worse” which is different.

For example, git is the best version control software; it is also a disaster of a UX.

Your appraisal of Docker is also too generous, I honestly can’t tell if it’s sarcasm

Not to nitpick, but wouldn’t Etherium Be the right cryptocurrency? I try to ignore it, but I remember that that’s the “smart contracts” one, so it could make the job contracts in such a way that if they are completed, payment is automatic or something?

> the mask averse amongst us

Oh can we coin a term like this for drunk driving? You know, when it’s more convenient to drive home drunk than leave your car out? That’s just being taxi-averse

They model a system which they know to be constrained by a closed-form equation called the Hamiltonian. They (cleverly, IMO) force the network’s predictions to be constrained by the Hamiltonian, by choosing the right output and loss function.

I don’t see a way to generalize this to the procedural rule-based systems you describe, unless they too are governed by a fairly simple continuous function Like the Hamiltonian.

I don’t know if it was “dramatic”, but it made me really happy.

I think software engineering is depressingly, suboptimally stable. The space of all possible ways to write programs is so, so large and almost completely unexplored. Lisps, Hindley-Milner type systems, Cox, neural networks with autodifferentiation ... whenever we substantially depart from the local optima around C-like programming, we find cool shit.

However, the vast majority of programming is done to build nearly indistinguishable CRUD SaaS apps. If we could just spend half our time doing real, interesting exploration of programming-space to satisfy our curiosity, maybe we could all agree to build apps with a more boring tech stack. However, that would require a large scale change in incentives that is hard to imagine shy of a Space Race style event.

... or Jonathan Blow was right and C was more or less the global max, and everything else was a mistake. I really hope that isn’t the case.