anders

@patriksvensson lgtm!

Will it get better as models evolve? I don't think so, because it's a systematic problem. Every little inconsistency or other piece of desinformation in a code base will mislead the llm and generate even more noise. It's like thermodynamics. The only way to reduce the noise is by doing work/spending energy on it (a.k.a. refactoring). And that doesn't seem to fit with the "black box" approach.

@sop@unstable.systems 100%. I can't believe that this ux nightmare is the default setting.

@ez@chaos.social it sounds reasonable to expect one's fellow leads to know their way about meeting invitations and how to prepare for meetings. It sorta comes with the role. But then again, I don't know anything about the context your team(s) operate in. In any case: sorry to hear about the frustration

@sop@unstable.systems @dave@basilisk.gallery The life saver button isn't easy to find, but exists: code completion can be switched on/off via the copilot button in the status bar. I set it to "off" unless there is actually something I want it to help with. Much nicer that way.

We may not be able to stop relying on fossil fuel from one day to the next -- our technology and our society has run on fossil fuel since the industrial revolution. But what we (the humans) CAN do, and what our politicians and legislators SHOULD do, is to differentiate between important uses of fossil fuel and not-so-important uses, and make the latter ones unfeasible. Economically or otherwise.

That is: unless a use of fossil fuel is important for society (not necessarily individuals), tax it hard, or add regulatory friction, so that the alternatives become more attractive. Because the alternatives exist -- or can develop -- they just have a hard time in the shadow of the current go-to solution, which is fossil.

An interesting read, in particular the part about the "too big to fail" kind of situation that execs may find themselves in with the AI hype all around them. That was a novel perspective at least for me. https://ludic.mataroa.blog/blog/ai-mania-is-eviscerating-global-decision-making/

@hn100@social.lansky.name Nice article, and sure, it does make sense for a device that will be used in isolated areas to not be dependent on a high bandwidth internet connection. I just wish they could call it "machine learning", when that is what they do, instead of "ai", which most people seem to think means "llms".