[deaux]

Yes and no [0]. There's no chance I'm the only one. And no, it's not a chatbot or automation tool or anything else that's "selling shovels", it's an end product. I've had multiple people reach out to me organically with how much it has helped them, reviews are very good and so on.

But really, you don't even need this counterexample because it's trivial. It's like a C fanatic saying "No useful software can be made using Python", and then asking for a counterexample. Take all useful small applications created. Here's one, Maccy [1]. There's zero reason every line of its code has to have been written by hand rather than prompted. Maybe some of it in fact was. It's a nifty little app, does its job well.

[0] https://news.ycombinator.com/item?id=47477440

[1] https://maccy.app/

[thinkharderdev]

Are you saying Maccy was vibe-coded or that it was written in Python? I don't think either are true. I've definitely been using it (you're right, it's great!) since before vibe-coding was a thing. And looking at the GitHub it seems to be 100% in Swift.

> It's like a C fanatic saying "No useful software can be made using Python", and then asking for a counterexample

At which point you could provide them many, many counterexamples?

I like AI coding assistants as much as the next red-blooded SWE and find them incredibly useful and a genuine productivity booster, but I think the claims of 10/100/1000x productivity boosts are unsupported by evidence AFAICT. And I certainly know I'm not 10x as productive nor do any of my teammates who have embraced AI seem to be 10x more productive.