AI has been great for UX prototypes.

Get something stood up quickly to react to.

It's not complete, it's not correct, it's not maintainable. But it's literal minutes to go from a blank page to seeing something clickable-ish.

We do that for a few rounds, set a direction and then throw it in the trash and start building.

In that sense, AI can be incredibly powerful, useful and has saved tons of time developing the wrong thing.

I can't see the future, but it's definitely not generating useful applications out of whole cloth at this point in time.

▲

SeasonalEnnui 3 days ago | parent | next [-]

Yes, totally agree. The 2nd thing I found it great for was to explain errors, it either finds the exact solution, or sparked a thought that lead to the answer.

	▲	mexicocitinluez 3 days ago \| parent [-]
		It's the height of absurdity to me that this is possible and devs will still say outrageous shit like "These tools have no use"

▲

criddell 3 days ago | parent | prev | next [-]

For me it's useful in those areas I don't venture into very often. For example I needed a powershell script recently that would create a little report of some registry settings. Claude banged out something that worked perfectly for me and saved me an hour of messing around.

	▲	humpty-d 3 days ago \| parent \| next [-]
		It’s useful for almost any one-off script I write. It can do the work much faster than me and produce nicer looking output than I’d ever bother to spend time to write myself. It can also generate cli args and docs I’d never "waste time" on myself, which I’d later waste even more time fumbling without They’re insanely useful. I don’t get why people pretend otherwise, just because they aren't fulfilling the prophesies of blowhards and salesmen
	▲	chasd00 3 days ago \| parent \| prev [-]
		yeah they work pretty well for scripts. I use claude to create scripts i need for .csv transforms and things. Like read all these files in a directory, merge them together on some key and convert this to that and all kinds of things and then output a csv with the following header row. For things like that they work pretty well.

▲

phba 3 days ago | parent | prev | next [-]

> We do that for a few rounds, set a direction and then throw it in the trash and start building.

Unfortunately PMs tend to forget the throw-it-in-the-trash part, so the prototype still ends up in prod.

But good for you, if you found a way to make it work.

▲

cck9672 3 days ago | parent | prev [-]

Can you elaborate on your process and tools here? This use case may actually be valuable for me and my team.

	▲	waterproof 3 days ago \| parent [-]
		Tools that can build you a quick clickable prototype are everywhere. Replit, claude code, cursor, ChatGPT Pro, v0.app, they're all totally capable. From there it's the important part: discussing, documenting, and making sure you're on the same page about what to actually build. Ideally, get input from your actual customers on the mockup (or multiple mockups) so you know what resonates and what doesn't.