lose it 😈 the streaks feature is the only thing standing between this company and the full consequences of their profound enshittification

somewhere in there is a Manufacturing Consent joke

good god

ah, I see

en.wikipedia.org/wiki/Flehmen... this?

Nobody in modern tech is satisfied with building A Thing that Works and then just... letting it be.

So close to calling it a disease at this point.

"I have refused to articulate my position" is a phrase that will be in the Democratic Party's obituary

this is an instance where overhyping leads to people disregarding something useful: as Tao points out, being able to find the research papers is nontrivial mathstodon.xyz/@tao/1153850...

LLM coding assistants, like so much else in the current age, feel like a superficial solution for what's really an incentive alignment problem, and I just don't anticipate it making anyone's lives better. I'm much more interested in addressing the root of the problem.

I get that certain things (operating systems, video games, etc.) need to be big. But IMO most run-of-the-mill web apps could be pretty small if not distorted by organizational incentives.

So from that perspective, I'm not terribly interested in software velocity. I think most web tools that I'm interested in using could be built by one or a handful of competent engineers in a relatively short time, and the reason that doesn't happen is incentives rather than velocity.

I really just want people to decide what software they're going to make, make it, and then call it a day other than basic maintenance. But it seems like that would require a different economic model where it was fine for a project to simply exist with minimal intervention instead of grow.