No plans for other languages at the moment. Right now my business model is to lose a bunch of money. The $5 plan might get close to breaking even. If it gets too expensive and I don't want to make a business I'll open source it all with instructions to install for those interested.
Exactly. I haven't reached the "let 1000 agents bloom" mode yet, so currently I'm spending real headspace managing agents doing work, and that work is all important, so why "settle" for sub-frontier models for that work? Maybe I'll get there for non-coding work.
i've said this before on Hn but I called the sheriff to discuss these in our neighborhood and asked if they could search for our whereabouts and he said "only if there's an active crime for your plate", so I said "what if the governor imposes a 'bird flu curfew'?" and after some kerfuffling said in theory yes they could then use it to track offendors.
Ya my memories are similar: ~12 years old building BBS's late into the night, then after college first startup, programming from midnight to 5am "in the flow" - nobody around or online. Just me and the problem at hand.
that's 100% happening to me right now not just with a specific codebase but with the fun/engineering parts of coding in general.... it makes me sad, slightly concerned, but optimistic that I can build a trillion dollar company with no employees. I'm kidding. Who knows. I'm enjoying building so I guess there's that.
I haven't used them but I'm sure they're good and I'm sure they're even using LLMs themselves now as a non-deterministic pass. For my use cases I'm just trying to keep vendors/costs to a minimum.
I don't have a scientific answer to this but I'm just assuming a different model, and more specifically, the frontier latest models like Fable/5.6 would catch things the author model might skip over.
I find myself doing this but then I worry that the slop will just compound and 3, 6, 12 months from now as my services scale I'll have a harder time operating them. Maybe I'm wrong.
I continually run codebases through different models to have them look for bad code smells like repeated code. That's been pretty effective. You do have to maintain over time or else you end up with a sloppy mess which I can only imagine compounds.
"I don't need a receipt for the donut, I'll just give you some money, you give me the donut. End of transaction. We don't need to bring ink and paper into this."
I missed the fact that Meta was developing and releasing closed-weights models... bummer. Would be great to see some more progress with American open-weights models.
there seems to be a lot of burnout from folks who have been in tech for a decent stretch, not just devrel. AI is an accelerant but not the only reason. probably the "power swap" as well as money has stopped pouring into hiring and jobs are more scare now and for the first time in, ever (?), employers have more leverage.