jev is insane π«£
it makes realtime virtual try-on hauls possible.
built this experiment for Drape with @typesafeai
> i talk
> jev reads transcript + what i'm wearing
> picks from my closet
> changes my outfit in realtime
cost: $0.0011 per decision
time: ~620ms per decision
imagine getting ready like this:
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
Iβve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
β’ 20-200x faster
β’ 40-400x cheaper (w/ output tokens free)
β’ Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
220K views and still climbing π₯
lots of people calling bs, here's a screen recording
real-time try ons has been live in drape for a while now. it's the first step toward changing how people decide what to wear and what to buy.
we've got so much more planned
drape will be the new default for personal style
stay tuned!
jev is insane π«£
it makes realtime virtual try-on hauls possible.
built this experiment for Drape with @typesafeai
> i talk
> jev reads transcript + what i'm wearing
> picks from my closet
> changes my outfit in realtime
cost: $0.0011 per decision
time: ~620ms per decision
imagine getting ready like this:
@JohnnyQuachy real-time try ons are live, but voice mode + jev isn't yet.
we think it's very cool, but its too new and needs more polish before we ship it
@paulocoutinhox@typesafeai real-time try on is live in Drape on the app store!
(voice mode with jev making decisions not yet tho, too new to ship straight to prod)
you're right! this works for our case because Drape already generates metadata automatically when items or outfits are added to the user's wardrobe. Jev uses that existing metadata.
This experiment did not actually run in the Drape app (obviously), but replicates what Jev would actually have access to if it were in our app.
and yea the cost shown is for Jev decisions (since metadata is pre-existing)