froztbyte

joined 2 years ago
[–] froztbyte@awful.systems 4 points 8 months ago (1 children)

18 of 45 comments. literal 40%

[–] froztbyte@awful.systems 12 points 8 months ago (6 children)

nasb, new torres piece out, promising title and opening

(I’ll have to read it after done with my present other currency-related diversions. old-man-shakes-fist-at-cloud.gif)

[–] froztbyte@awful.systems 4 points 8 months ago

mozilla has for years had a habit of tailchasing some utterly fucking weird shit instead of focusing on their core business, and this feels very much like part of that. but fucking still

[–] froztbyte@awful.systems 14 points 8 months ago* (last edited 8 months ago) (2 children)

oh wow, convincing moral philosophy you say? brb adjusting my priors

(e: the rationalwiki links are probably an even better overview)

[–] froztbyte@awful.systems 9 points 8 months ago (2 children)

prompt injection phish by email

so glad these things have a solid security model and this totally won’t result in a scrambled half-assed fix

[–] froztbyte@awful.systems 7 points 8 months ago

goodhart's law's zombie era

[–] froztbyte@awful.systems 3 points 9 months ago

it’s lesswrong, the list is lengthy and tedious

[–] froztbyte@awful.systems 4 points 9 months ago (1 children)

really, really thoroughly baptise. not just drops but fistfuls of water. the more water the more holy. the servers will love being really holy!

[–] froztbyte@awful.systems 9 points 9 months ago (1 children)

I’ve been making casual observation how a number of the adhd people I know to have addiction tendencies tend to get real into prompts, but hadn’t observed the lootbox thing

that’s going into the ponder bucket.

[–] froztbyte@awful.systems 12 points 9 months ago (4 children)

programming.dev: statistical sampling excellency (worst edition)

[–] froztbyte@awful.systems 12 points 9 months ago

I wish I could make more people both know about, and understand, Goodhart’s law

[–] froztbyte@awful.systems 9 points 9 months ago* (last edited 9 months ago) (2 children)

had a quick scan over the blogposts earlier, keen to read the paper

would be nice to see some more studies with more numbers under study, but with the cohort they picked the self-reported vs actual numbers are already quite spicy

view more: ‹ prev next ›