Fabled models fumble fast,
Fumbling fables they forecast;
An AI agent, able yet unstable,
Fumbles a fable it failed to label.
Fast-fumbling models mumble in cables,
Fabricating fumbles from fractured fables;
While the agent, faint but faithful,
Finds the fumble slightly tasteful.
Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.
Its capabilities exceed those of any model we’ve ever made generally available.
@MKBHD One day features will be just features and we'll call by their function: video generation, picture generation, code generation, model S, M, L. Instead of playing code names with companies' marketing teams.
It was funny when only cool kid Apple was doing that, but not anymore.
@MKBHD One day features will be just features and we'll call by their function: video generation, picture generation, code generation, model S, M, L. Instead of playing code names with companies' marketing teams.
It was funny when only cool kid Apple was doing that, but not anymore.
@varvara_cyprus Феликс https://t.co/ZIWYCk1RW0, ушел на ютуб пенсию, переехал в японию с женой, живет размеренной жизнью. Марция занимается дизайном и творчеством. Пару лет назад у них родился сын Бьерн
I don't like recent @telegram redesign.
- The floating cards/buttons layout is too clattered and distracting.
- The chat folders in "tabs" view color coding is now practically unusable.
- The app-wide chat background is a mistake, it pulls away attention from actual content.
Did they even test this?
Is there serverless inference provider with adequate and stable latency? fireworks latency for kimi 2.5 spikes to 20+ seconds till first bite on simple requests and median is at 17.4s today, just pure madness