The point is, literally nobody reacts to subway malfunctions with, “and we call this progress???” as if returning to previous modes of transport is somehow the right answer to problems with far less drastic solutions than throwing the baby out with the bathwater.

LLMs are a new technology that people are still figuring out how to use effectively. Part of that process is becoming reliant upon “the new way of doing things” to prove that one can rely on it. Clearly, there’s more work to be done. (My dayjob includes working on this same reliability problem.)

One can argue the wisdom of being an early adopter in any new technology. Some thirty years ago, I was told I was insane for going all-in on Linux. The times change. The sanctimoniousness of the peanut gallery hasn’t. The lunatics betting the farm on all that wacky open source stuff three decades ago turned out to have been largely right, despite the numerous failed ventures involved in getting to here.

This is just how the new technology cycle works. With every new tech, a whole lot of people discover all of the ways it doesn’t work before somebody figures out the way to make it work more reliably than any alternative.

Funny how nobody seems to use this argument every time there’s a problem with the NYC subway.

Nice. Good for him.

I first met Chris around 2006 or so. Always been a decent person and very dedicated OSS contributor.

Oh good. Next wave counter-culture is starting to name itself. This is a fun period in any movement.

I was present for the early copyleft era that corporate software development exploited then squashed with more “business-friendly” OSS licenses. Now we’re seeing the mega-scale tech companies enshittifying their products necessitating a new wave of open-information counter-culture to fight the Big Tech birthed from the post-dot-com OSS movement.

History doesn’t repeat itself. It often rhymes.

That’s an improvement over humans. Humans violate ethical constraints due to KPI pressures far more often.