OpenAI's abandonment of robotics is an implicit admission that they don't think deep learning can solve AI.
-
Nah, it’s admission that there are paths to AGI that are more promising than the robotics/embodiment path to AGI.
-
-
Why not. Robot is just an agent to control/change its surroundings. The response/manipulation is purely abstract.
-
Which doesn't actually alter his point. It hasn't been solved at the abstract level either AFAIK. When/if it is, it should be conceptually easy enough to map to the "real" world if the right abstraction was used.
Keskustelun loppu
Uusi keskustelu -
-
-
@pmddomingos somehow I hold the same intuition. Can you please elaborate more or point to your favorite references for your belief ?Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
you could generate the same challenges physical robots face inside a simulation. e.g. "apply AI to procedural animation",etc. rather than literal embodiment, wouldn't it just require a certain level of interactivity?
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Definitionally you may be correct, artificial intelligence implies computational ability to learn through sensory perception, but functionally, algorithmic machine learning may be the more promising route to harness computational power for good purposes.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Is your idea of an AGI Data from Star Trek or HAL ?
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
In terms of AGI research, at this moment investing in generative models has more return than investing in robotics. Robotics will come later.
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.