I'd say LLMs COULD make it easier to implement accessibility, it also couldn't, always a coinflip with those, but I'd say LLMs actually succeeding is probably unlikely given how much shitty code is probably in its training data :P

This is true for many accessibility projects actually (game mods, third-party UIs for inaccessible services/platforms, etc.). These are generally really meant as short-term patches while the problem gets fixed, except ... the problem often doesn't get fixed because the platforms in question figure it's been solved now and they don't need to care about it anymore.

Accessibility really only works when it's an ongoing, first-class process within an app/platform's design, and we can absolutely do that; the standards and guidelines have existed for decades. People working in cybersecurity, localization, general UX should recognize this song and dance, which is amusing because a lot of the tools of those trades have atrocious accessibility and require all sorts of workarounds, ask me how I know.

People just ... aren't including it in this way, which means people like myself (screen reader user and accessibility professional) essentially have to keep reminding people that we exist and that it's kinda shitty to keep forgetting about that fact or to decide the least amount of effort possible (LLM, unpaid volunteer, send in a PR LMAO) is enough to cater to people who have very real, very annoying and very constant UX issues we either crash into or crash through on literally an hourly basis.

What are you basing that on? Screen readers tend to not pick those up at least on interactive elements by default, you need to do a bit of "wiggling" to get those to be announced. Disclaimer: screen reader user

Is this going to work with screen readers, magnification, speech recognition etc? I guess a more abstract version of that question is: Does Ladybird intend to offer some kind of feature parity with existing solutions where integration with OS-specific accessibility architectures (UIA, AT-SPI2, etc.) are concerned? If not, it's a non-starter for quite a few people, and I'd rather know so I know to even keep up with this project or add it to the "user first but oh not actually all users first" pile :)

I'd say for smaller teams it is doubly important to think about this topic early to minimize the overhead of having to fix a whole bunch of bad moves you figure out you've made when someone complains months or even years after you've made them. As for doing this now and again, I think some of the QA firms like APplause and such do offer accessibility testing services and such now, and there's social media like HN, Reddit, Mastodon or, if one must, X, where people can be found who'll be happy to do this provided they can be fairly compensated. Not exactly structural or centraalized but there you have it :)

- Wayland didn't think it was important to immediately include this, and now we have the majority of linux distributions having serious issues with screenreaders and other assistive tech. Fedora's shipped with this broken for almsot a decade. Calamares didn't think it'd be important to fix and has been broken for about as long. - Particularly now, with devs grabbing a component library on top of React with a generous helping of CSS frameworks and third-party NPM-based extra bits that are all tangled together, and what have you, if you don't vet this stuff beforehand you'll have to retrofit half your UI to fix things after the fact. That, right there, is why accessibility seems so hard to implement.

Fixing a native HTML select for accessibility is easy; it already is. Fixing some componentized overengineered monstrosity that figured they'd get to it later and as a result doesn't speak with screen readers, doesn't work on phones, doesn't work when zoomed in, doesn't respond to speech recognition software, goes absolutely nuts when the user scrolls, and doesn't let you use it properly with a keyboard... yeah that is harder :)