One lesson I wish tech folks would grasp is to stop building and releasing frameworks out of academic theory without testing it in... you know, real world scenarios.

Compulsary annual post: Another great #ffConf yesterday! Had no signal in the venue, but even if I did, too busy thinking and chatting to people to be posting.
It felt good to bring some of our team again. Sad I didn't get to talk to most people I wanted, but energy and time were not on my side.

Please ask your employers to support these scholarships!
If you're able to, please consider chipping in (the ticket platform allows you to throw in even a few quid 😉, if I'm not mistaken).

Your marginalised/future junior colleagues will thank you, speaking from experience.

We need to teach people the difference between being uncomfortable and being unsafe. Because the discomfort of the safest people is endangering the most vulnerable, all day every day.

I wonder if LLMs would have even taken off had Google not degraded search. For years they used questionable ranking metrics which rewarded SEO slop over useful content. It's not surprising fast but inaccurate information was more attractive than sifting through a 10,000 word essay to find a recipe.

“ICE protestors […] are turning what the tech companies see as an asset into a vulnerability, and they’re exploiting it. There is no one around to stop them […]
They have turned these indifferent and extractive technologies into instruments of protest. They are weaponizing the accountability sink.”

hey! Haymarket is making our book No Cop City, No Cop World: Lessons from the Movement FREE in ebook form for the next 12 days!

hopefully it's very clear this book is not about making money (& it won't 😂) lol so please tell everyone to download it if they want!

one of my favorite chatgpt commentaries is "you dont need AI i can also drink a bottle of water and lie to you"

“I just don’t know how you go around, asking everyone first. I just don’t see how that would work”

It would work, if your lot weren't greedy rapists.

How did we get from "Do not use Wikipedia as a source" to "ask hallucinating chatbot everything"?

Using accessibility applications to legitimate generative AI is seriously gross.

"But an AI that describes images helps the blind"

Someone chose that image and put it where people can see it. *They* can describe it.

OpenAI has to be the most insufferable company in the world. They can steal from the whole world & guzzle all resources. But no one can give them a taste of their own medicine even a little bit.

How long till they use this to say give us even more resources & this is why we can't release anything.

Some requirements of a civilised society:
✊🏾Healthcare for all
✊🏼Education for all
✊🏽Accessibility for all

As a disabled web accessibility practitioner. This pisses me off. Extremely.

GENERATIVE AI DOES NOT AND CANNOT EVER REPRESENT DISABLED PEOPLE.

Just fucking hire disabled people like me who know what they're doing.

Believing AI over disabled people is ABLEISM and DEHUMANIZING. For fucks sake.

Publisher: Do you want to put an LLM clause in the copyright statement?
Me: Definitely. How specific can we get?
Publisher: As specific as you like.

This. Good reasons for using alt text.

Also I'd add: not everyone can immediately understand what you mean if you only use an image to convey that information.

The amount of times I stared at an image, wondering WTF does this mean... It could be as harmless as repurposing a meme I don't know.

Thiiiiiiis. A checklist doesn't always prove your site is accessible. Test with users! Learn what tools they're using. Find the gaps that can lead to creative solutions. I always rail against the idea that meeting accessibility needs means stifling creativity; I think it actually unlocks it.

Hey Blue Sky people! There are so many #ffconf people on here that I had to make an account.

Also, as usual, the conference on Friday was an absolute banger and a joy. Can't wait for next year already ❤️