All non-trivial abstractions, to some degree, are leaky.
Every layer we stack between ourselves and the metal eventually breaks. The fix isn't a better abstraction—it's knowing what's underneath when it does.
Save this - and 500 more like it - to your library.
Linkwise remembers everything you read, and lets you chat with it.
Get Linkwise — free on iPhoneEngineering schedules often slip even with predictable technologies, and the non-deterministic nature of LLMs makes their future inherently unpredictable.
Gruber's Apple Intelligence post-mortem: Cupertino shipped a Siri demo it couldn't deliver. The hard AI bet isn't capability, it's reliability.
A practitioner's case for treating distributed-systems thinking as a default skill, not a specialist one, as single-machine assumptions keep breaking in production.
youtube.comDwarkesh sits down with two Anthropic researchers - one scaling RL, one doing mechanistic interpretability - to trace how modern LLMs actually think. The "auditing game" section, where the interp team catches a deliberately-planted evil model in 90 minutes, is worth the whole thing.