Tilton RaccoonOh boy! We added a “Gamified Tracker” for AI use at work. The future is AMAZING and I LOVE IT! I’m being VERY SARCASTIC!
senna.apk
Kay Ohtie 🔜 FWA@tilton Reading that, someone next to me would've been able to hear the multitude of screams from within my head
Luna Lactea@tilton Score it like golf!
@jackemled I'm just gonna blast Claude with questions until I use a million dollars in tokens, I guess.
FayeSoftware development pre-AI: “output and activity based metrics are easy to track but counter productive because, e.g. LoC count is stupid because every line is code is a liability. Outcome case metrics are hard to track but the only thing that’s actually worth while.”
Software development post AI: “I PRODUCED TWENTY MILLION LINES OF CODE AND THEN PROD EXPLODED, BUT I HIT THE METRIC SO I GOT PROMOTED”
It’s one of the really galling parts of the whole thing. The industry was _just_ starting to wake up to how to actually build good value and now suddenly everyone is drooling about having created two hundred PRs in a day.
@FayeDrake @tilton Want to maximize lines of code? Write in a simple, high level language, then crosscompile to assembly & say you wrote the assembly. Remember to turn off compiler optimizations!
@tilton I have suggested shell scripts running in a loop sending messages to LLM interfaces as a method of wasting tokens to others.
FlatFootFox@tilton Yikes. Y’all have a leaderboard for who’s burning through the most tokens?
@FlatFootFox Yup! It’s wild