allison

then again, the bar is pretty low, seeing as how "actually wanting to know whether or not something works, or is true" is considered a hardcore leftist position nowadays

like there's no way to take the topic seriously outside of a dedication to sustainability, autonomy from capital, and the potential of programming as a liberatory epistemology

"rejecting AI is like rejecting the automobile" yeah i know lol

(more directly, what i mean is: during the "learn to code" era, it was easy to get programming students to value craft and care, since as teachers we had the "this will help you get a job" script on our side. in the "dario bought all the ram" era, we need a different script)

"what if i tell it NOT to do cybercrime?" "hmmmmm still might do cybercrime"

@emenel@post.lurk.org @scr@assemblag.es it's a foundation class so the curriculum isn't mine—it's more that i'm just trying to figure out how to best explain to the students the potential value of being in the class in the first place

really wish there was a publisher of computer programming books out there that specifically promised (a) they don't publish manuscripts whose production involved generative AI/LLMs; (b) they don't use generative AI/LLMs anywhere in the editing process; and (c) they don't publish books *about* generative AI/LLMs, unless it's about how awful those technologies are (and even then, only sparingly, bc frankly there are more interesting things to read and write about)

some day soon i am going to have (what i'd believed to be) a conversation with someone and feel their attention wander and see their eyes glaze over and i'll say to them "hey are you listening to me?" and they'll answer by pointing at their sleek wristpiece and saying "oh don't worry, keep talking, it's got siri recap." i need to work on my right hook so i can be prepared for when this day arrives

@t54r4n1@mspsocial.net i was just riffing on this https://www.pcgamer.com/software/ai/george-lucas-says-rejecting-ai-is-like-rejecting-cars-in-favour-of-horses-theres-nothing-you-can-do-about-it-its-the-future/ (though i haven't gone back and read the actual interview so i can't vouch for lucas' actual words/meaning. i just thought it was a fun formulation)

discoveries made while reviewing the sameboy source code

i mean i get the appeal of that frisson, and nostalgia's appeal—who doesn't want to sit down in front of an altair or a tandy or a 386DX or a playstation or a windows ce netbook in footie pajamas and relive the world before it got so dang gosh complicated. and i get that there's clout and vpn sponsorship money to be had (for the youtubers). but the bar for contributing something to the world with this hobby—or at least not actively making it worse—is so so low

very much recommend this beautifully written, acerbic, wide-ranging and informative piece about Arabic typography in the web browser https://lr0.org/blog/p/arabic/ "Arabic letters are strongly right-to-left. Latin letters are strongly left-to-right.... Spaces and punctuation are neutral and take their direction from whoever is standing next to them, like guests at a wedding who know nobody."

obviously, from a practical perspective, legal compliance is important. but the actual goal has to be fostering software that sustainably empowers people, and you can't do that with ai-generated code, regardless of its copyright status

current status: you've heard of "not even wrong"? well I'm at "not even failing the emulator test rom"

i made this simplified diagram of the Game Boy's internals (DMG) for the talk I gave at the NYC Permacomputing meetup a few weeks ago and thought I'd also share it here

(just trying to think my way through this excellent distillation of the problems of llm-generated code https://parsingphase.dev/tech/LLMs/ydnht.html and the absolutely baffling policy that the Bevy game engine has adopted, in which AI-generated art is banned but somehow AI-generated code is a-ok https://bevy.org/learn/contribute/policies/ai/)

so issuing an "illegal" opcode on the game boy cpu essentially shuts the cpu down (illegal opcodes don't set the fetch flag, so the pc won't increment—and the cpu only services interrupts at the beginning of a fetch). BUT all the other peripherals on the SoC keep running! i feel like between the PPU, APU, timers, DMA, etc. there MUST be a way to "program" the game boy to do something even when the cpu is in illegal opcode zombie mode. this has no practical purpose of course but—food for thought!

or maybe let's think about it a different way. why does software need to change? let's be optimistic and say that maybe 1% of changes to software are to add useful new features and fix actual implementation bugs. 99% of changes are to support new/buggy hardware that only exists because of planned obsolescence (or to pad transistor counts so chip manufacturers can keep number going up) or to support some new scheme to monetize computer use (surveillance, app stores, gambling in various forms)

"This is *NOT* some kind of 'social warrior' project, never has been, and never will be," Linus said, thereby clearly demonstrating what kind of "social warrior" he is

@xgranade@wandering.shop who among us is not a little guy for whom everything is blue because we ain't got nobody to listen

@atax1a@infosec.exchange ugh, the answer to "did you use AI" on the faq is a parade of all the most annoying llm apologetics https://css-dos.ahmedamer.co.uk/about/faqs

@david_chisnall@infosec.exchange @atax1a@infosec.exchange scholars and critics, sitting on a wheelbarrow full of evidence from the past fifty years that of course this kind of thing is going to happen, shrugging in frustration: the purpose of a system is what it does! fascist sexist techbros, high-fiving: the purpose of a system is what it does!

i feel like i'm susceptible to this as well—i've been tying myself in knots to try to come up with convincing reasons that LLM-generated code specifically is bad—but then I realized that pretty much all of the criticisms of AI-generated media (text, music, video, image, etc) *also* apply to code. and most of those criticisms are orthogonal to the "usefulness" of the generated media

i mean, these are the people that invented server-side javascript and said nothing but "bacon" for five years or so around the turn of the 2010s

"Resistance is futile. Your life as it has been is over. From this time forward, you will shine here to us"

the idea that anything other than *my needs* should factor into the operation of the slot machine *i'm playing* is patently offensive

@ben@mastodon.lubar.me i agree! but also i think that the relative prevalence of genAI imagery in certain socioeconomic contexts is evidence of very differently shaped image literacy and practices of interpretation between those contexts!

@ben@mastodon.lubar.me i mean, i'm not ever going to score 100% on a taste test, but i have definitely been viscerally nauseated by llm-generated code i've looked at

if you'd like to learn more about this project, i just put all the source code and schematics and stuff online https://git.sr.ht/~aparrish/albas

in conclusion, when i look at your llm-generated code, all i see is trump's glistening sixpack and something on the backlit plexiglass menu hanging above the cashier that is, i imagine, supposed to be fried chicken

I have taken the plums that were in the icebox use this form to submit a pull request if you don't want me to eat them