AI Is Dogshit at Design

tomcreighton.com

2 points·speckx··7 comments

The question is --- why would anyone expect otherwise from a probabilistic language prediction engine?

Because, if we ignore the begging the question in your rhetorical question, we've found that probabilistic language prediction engines are quite good at many things, like translation, coding, apparently even theoritical mathematics.

Do you have tips on how to get better at design, now that everything looks like the same AI-slop online?

Like how do you come up with a cohesive looking set of components for example? I use design tokens but they are not enough to get to a coherent design where components look like they belong together and kind of stand out from other designs.

>now that everything looks like the same AI-slop online

Designers have managed to make everything look the same pre-AI too, like the shitty "flat design" era.

Flat design is still pervasive, and has horrible ergonomics. If I'm supposed to click on something, make it look that way, OK? Also, a tiny, grey chevron on the far right of some text is not adequate for marking a pull-down list. Who thought any of "flat design" was a good idea, and did they have to black mail all the human factors people to get it in production?

All that to say, I emphatically agree. Lemmings following a trend is not a good way to distinguish anything or any one.

Do you have a good example of a website that does not follow flat design that you like?

Agree. Systems like shadcn accelerated this moving towards sameness a lot.

Reply on news.ycombinator.com