logoalt Hacker News

jmward01today at 7:57 AM2 repliesview on HN

I'm glad to see things like this get built. I hate to admit it but I rarely consider impaired usecases when building things. I wonder how technology is changing this usecase lately both on the user end and the design end. (I know, AI) I imagine an LLM could help discover inadequate UI and build alternative workflows into products more easily.


Replies

shaknatoday at 9:47 AM

Whispir is a much better TTS than almost anything else. However, when it gets it wrong, oh boy does it get it wrong.

For everything else? Not really. JS thrashing the DOM is as much a pain as ever. Using ico files instead of either emoji or... Text... Makes UIs painful and inconsistent.

Everyone using Electron and its broken [0] accessibility, including core Windows features...

These aren't things that can be reasoned away with an LLM. An interface is not just text - its a reasoned nodegraph. And when I'm blind (comes and goes), I need the nodegraph. Not an image of the screen reinterpreted.

[0] https://github.com/electron/electron/issues/45856

show 2 replies
zersiaxtoday at 12:05 PM

'Honestly this is a people problem more than a tech problem. We have the tech. We're just not using it.

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