> Even hobby projects are dull because they're done in mere days.

This is only true if your definition of "hobby project" is limited to "initial prototype with basic functionality and a generic corporate-looking landing page".

Yeah, a lot of tools that would have made fun weekend projects can now be one-shot by Claude and aren't even worth sharing, because someone who needs that functionality can just use Claude too. But you can accomplish so much more now. Someone just did a hobby project to reverse engineer NVidia drivers and make them work on Windows XP. Dream bigger!

I agree dream bigger!

If it gets done too fast with LLM you need to think bigger.

Indeed, your half-broken prototypes that are not worth sharing with the world can be now 10x the scale.

But seriously, I have not seen anyone get far with vibecoded hobby projects. The problem is that removing the pain, the frustration, and the triumph also removes any emotional attachment you have to the project to keep it alive. If it's not making you money, if you're not learning anything, and if everyone can vibecode the same thing... well, it still gives you some short-term rush - as if by magic, your words birthed a piece of software - but then what? Why keep going? Why fix bugs, why keep improving it? If you look back at the past several months of "Show HN"-style posts, it's not pretty.