upvote
I keep seeing this question asked, and I don't understand. Have you tried it? It's so easy to build useful things for myself now.

I've been able to build things that I otherwise would not have been able to build, in the free time that I have: - a VST audio plugin

- a wedding website with RSVP functionality

- a relaxing game for my wife

At work, I've been able to build much more than I would have been capable of in the past. I'm a backend eng, and it allows me to build much much nicer frontends than I've ever been able to do in the past.

And before you tell me that the code is crap - it doesn't matter! It may or may not be good code, but it works and serves it's purpose very well. Anyways, I'm I'm not launching a rocket, or putting software into cars.

reply
> Have you tried it?

I've used agentic tooling at work for almost a year now, to great affect.

Tools are helpful, but in my experience they are not to a point where building things useful software is easy or simple.

Most of the things I have seen or heard described are neat demos that someone might play with a few times before forgetting about.

Of the projects you mentioned above, how long did you spend building them relative to how long they've been regularly used? Are you using them on a regular basis? If so, are they published somewhere for other people to use as well?

reply
I think he's talking about professionally, not various hobby ideas.
reply
Professionally, even if GP was to break NDA, it would still take an essay to even begin to describe where they are in order to describe what they're working on, in order to say this thing I'm building, it's being built faster with the help of AI. So the two things are one, hobby projects. Those are greenfield, starting from 0 lines of code. AI is good at that, everyone should try it! The unknown unproven question, thus, is limited to "I have this decades old codebase, can AI actually help?" I am not here to answer that question right now, merely to pose the more specific question you're implying.
reply
I mean, Claude itself has been built by people using Claude. Is that not enough of an example?
reply
Assuming you mean claude code, it is developed by a large team of people being paid among the highest salaries in the industry. ~$500-750k/year according to levels.fyi

I wouldn't count that as evidence that AI tooling makes it easy to build useful software without much effort.

reply
I don't know what those people are doing, but I built a personal day-to-day notes manager at work, for work, for under $20 in credits. Yes, I could have just used text files but this is less friction which means I'll actually stick to it. Nothing exactly like this already existed. It was built in under a week in small portions of spare time, and it probably would have been more like a month if I had to choose all the libraries and write the whole thing myself.
reply
You have build "a stuff". But no one knows what the "stuff" is, and your comment is not exactly describing it.

That's why it's so hard to converse with the AI proponents. They don't want to talk about what they've built and why it's awesome (at least for them). They just want you to take it at face value.

If you're sharing something, just give enough details so that others can contrast to their own experience and learn something from it.

reply
Search Ask HN threads, there are lots asking the same question and it looks like people do make useful stuff, it's just personal software not distributed to the public. I find that nice actually, because it's nice to make something just for yourself and is essentially what software is good for, solving your own problem.
reply
Yup, that's a sentiment I wrote about last year: building software for personal use that doesn't need to scale
reply
I do think people are seeing useful results, but it’s at least several orders of magnitude less than what is claimed.

My metric is: how much code that I create today do I depend on in 3 months (incl learnings and one off data pulls etc.) and its waaay less than what I generate. A lot of it is frankly noise.

reply
Oh, we're building it. It's already out there.
reply
The results are in existing products being built out faster, in more internal tooling that would have previously not been cost-effective. Almost every software engineering shop is just doing the same thing as before but faster, as AI doesn't really give you an edge in the product side of things.
reply