EricHere's some thoughts and feelings about needing to vibe code to do accessibility design work. https://ericwbailey.website/published/the-case-for-an-accessibility-designer-vibe-coding-when-all-his-coworkers-are-also-vibe-coding/
Drew Mochak@eric I asked the WhatsApp people if my various vibe-coded accessibility fixes to their web app would create thrash for them reviewing baby's first code, and the answer was that basically they were already doing that from everyone else, so, no.
Mark T. Tomczak@eric I hadn't considered AI for accessibility before... It may plug an unfortunate gap I've seen, which is that developers really, really, really don't want to support a11y in most firms and management doesn't care enough to force the issue.
But if an LLM can reduce the problem of identifying issues to one analysis pass... That's not a small thing. And I'm generally seeing LLMs are better at that than traditional heuristic approaches, because they end up trained on real-world examples of what goes wrong.
@mark Unfortunately, my experience has been a high degree of wildly incorrect fixes when it comes to a LLM performing autonomously. It takes a high degree of scoped human intervention and verification, much to my boss’ disappointment.
@eric @mark Very much this--I can get pretty good results out of it, because I (not to toot my own horn,) am a genius. I know precisely how it must function and test it riggorously before submitting If you just say, "here's a thing, fix it" it usually doesn't work the first time. But if you stick with it you'll either get it to a good place or realize your problems were deeper than you thought, which is still useful.
Gerard K Cohen@eric @mark #Facts The amount of manual effort to get AI to do what everyone thinks it can do is severely understated. And even after all the setup, my experience was that it did not accelerate anything. Output was barely equal, but mostly lagging behind humans. (yes, this was actually measured)
And the reality is, thats "ok?" if everyone agrees and understands the effort and output. It's not magic. It's more artificial than it is intelligent. It's definitely not cheaper.
@eric The approach I've seen work best is AI as an automated reviewer in the pull request workflow. It can false positive and false negative, but in practice the false-pos rate I'm seeing is like 1-5% (and much of that corner case stuff, like "yes you're right but that is autogen code that cannot be fixed at this layer").
It needs human in the loop, but it doesn't get bored in the same way humans do.
This effort is being dogfooded, 100% LLM-forward with its approach. Because of this, I also need to vibe code in order to contribute.
I'm sorry, am I misreading or do they actually force the real, code-literate engineers to use the stupid dream machine rather than their own actual knowledge? To design a thing that's user facing? Good grief?
@eric
< I am being structurally compelled to work this way, and also tracked and ranked based on my frequency and volume of token use.
Yikes. Starting to think getting sacked was the best thing for me.
Okay I'll stop now.
@eric Anyway thank you for writing up what was basically my life until recently.
Blake Watson 
@eric Great writeup. I also carry a lot of cognitive dissonance when using AI for development work. I don’t really know what to do with the feelings and feel a bit stuck.
At the same time I was recently able to make a macOS accessibility tool that me and a handful of people have been using daily and it’s made things so much easier for us. It probably would have never existed without the LLMs.
karolina@eric I really appreciate you directly speaking to the conflict of your ethical stance vs work mandates and evaluating if there’s anything in the technology that advances the goals you’ve set. many of us are in this position, so even sharing openly makes me feel less alone ♥️