I’m not here to hype. I build, I scar, I share. Mostly about #AI, #billing, and the messy bits of running #startups.
Building @CredytAI and other things
@claudeai the Claude Design migration guide only refers to moving Design Systems. What about projects that aren't design systems? What's going to happen to the CD canvas. Artefacts doesn't work this way. People going to be pissed, just sayin'
My recommendation if you can't or don't want to work for someone else. Find something your passionate about _outside_ of tech. Solve the problems you experience first hand, not to make money, but because they're real problems. Then you'll probably make money.
Becoming a Software Engineer went from being a secure job to being a skill that with AI, anyone can have.
Daily I find myself amazed at how much I can achieve with AI, then reality dawns that so can everyone else. It's like when no-one gave a shit that you know MS Office
Most projects will fail but it's okay because they're cheap. The ones that will win will be driven by having a good sales channel/network/people skills. The shit engineers are notoriously bad at.
So many gen-AI founders launching into the same pit right now. Never been easier to launch a website. Never been harder for it to be seen.
It's heartbreaking to tell these founders that as beautiful as their new site is, no-one will see it without distribution and an evolving content strategy.
Anyone can now build something with AI. But not everything. In most cases it doesn't matter, but sometimes it does.
Vibe coding gets you so far. Engineering gets you further. Good system design and architecture is what enables your software to evolve and scale. If you need to.
@levelsio Given the almost zero cost, I'd also throw Cloudflare Pages in the ring for hosting. In fact most of Cloudflare's services can be used for free.
I think @claudeai is intentionally throttling usage now that we've had a flavour of how fast work _could_ be. Default will likely continue to get slower, premium will become what default used to be.
I don't think this is placebo. I think there has been regression on existing models. I've been happily using Sonnet for most things for a long time and now it's just plain stupid
I don't know if it's placebo but using Fable for those few days it felt it just never gave up on problems and kept trying crazy ways to get whatever you wanted done. Now back on Opus and it's kinda lazy, thinks things are too daunting and keeps asking if you sure