calflegal

I've started to have a sneaking suspicion that claude code has already reached a key level of what I will call 'quite amazing'. When Google released Gemini 3 and Antigravity, I tried them out. They're good, great even. But, as a software dev, I'm back to trusty Claude Code. I really like it. More than anything, I really hope it can stay where it's at with performance and price and become profitable so that I'm allowed to keep it!

I don't particularly care about SWE bench or how much higher the next model scores on it. Perhaps that's not very futurist of me.

How do you feel about this?

I bet there are some good ones.

I think I've realized that most (all?) of the software I pay for offers a native iOS client. Sometimes, the stuff I pay for has no web client at all. What web-first software do you pay for?

Like many, I'm a little frustrated by the relentless push into AI (though I use and love cursor), specifically agent patterns. Is there no home in the future for simple systems with button interfaces and function calls? Do you expect those to be built on the fly by some agent?

There's so much python slop being slung around, but I guess one day these things will write their own bytecode, I don't know.

It seems clear to me that AIs are going to write more source code going forward. At the moment I review a lot of it, it seems necessary. But it occurs to me the same may have been true of early compilers. Is it the case that early engineers would review and change resulting assembly? That certainly isn't how most of us operate today, so I'm wondering if the same is bound to happen at this new abstraction level. Will we stop reviewing the output and write...english? Why or why not?

Update based on comments: Perhaps by using 'LLM' I was too specific to today's non-deterministic LLM systems..I suppose I meant AI systems in general.

I'm thinking if shipping some little web music games. If I were ever to monetize it, should I go with free to play and one-time purchases or should I try to pursue some subscription model?

At this point I personally feel tired of everybody wanting some number of bucks per month, though I get that it aligns business costs with customer payments. How do you feel about this issue and what would you do in the case of this project?

What piece of software has brought about the most human happiness? Let's ignore examples that lack counterfactual support, e.g. 'the software that ended WWII prevented x'. My intuition points to medicine but I have no concrete example.

Like a lot of us, I think about trying to work for myself, but then I wonder if I'd be happier running marketing AND software dev? Sometimes I think I'd rather be an in-person worker rather than a remote one. I'm interested in hearing some stories of people making big changes and being surprised in both negative and / or positive directions.