
Lead Dev β @officiallevi_c0 | yuzuruu29
π€πΎπ€πΎπ€
Joined June 2026
- Tweets1.4K
- Following860
- Followers783
- Likes702
ai explaining an error > ai fixing an error.
the first one makes you better. the second one just makes the error go away this time.
building with a small team is underrated.
more ideas, someone to catch your mistakes, and someone to say "no, that's a bad feature" π
ai helping someone ship a working app in an afternoon still feels like a magic trick.
syntax got cheap. knowing what "good" looks like didn't.
honest question for builders:
when you're trying to make money on x, what actually moves the needle more for you?
creator payouts, or turning the audience into users who pay π
actual footage of my grok bot after seeing my usage. @elonmusk how do we make the bot stop crying?
π€ Made with AI
name your branches and commits like ai will read them later.
because it will. clear history = better context = better answers.
building in two worlds:
fitness pace: an android fitness app with the levi co team, in testing.
the last moon: my game.
apps or games, i want to know more people doing the actual building. let's connect π€
the model is rarely the bottleneck anymore.
it's the harness around it: memory, checks, when to stop, how to recover.
most "ai coding" fails i see are harness fails.
builders, drop what you're working on π
app, game, or a tiny tool nobody asked for. doesn't matter.
using ai doesn't make you less of a dev.
shipping nothing because you were too proud to use it might.
weekend plan: fix one bug.
real plan: fix one bug, find three more, refactor something that was fine.
weekend question for builders:
when you want to meet more people shipping stuff on x, what actually works for you?
i've been doing more replies lately and still figuring it out π
ship a v1 that does one thing well.
every feature you add before launch is a feature nobody asked for yet.




