[일반] 샘알트만 발언은 범용ai의 영역이 이제막 시작되었다는 의미임

[익명(223.38)]

agi는 다들 알듯이 모든 영역에서 단일 ai 모델이 인간 이상의 지능 수준을 갖는 거고

현 gpt3는 단일 영역만 다룰수있던 좁은 ai에서 벗어나 약간의 범용성을 가지기 시작했다는거임7

딱히 새로운 떡밥도 아닌데 과잉반응 ㄴ

- dc official App

[익명(223.62)]

ㅡ선ㅡ - dc App

[익명(121.163)]

아하하하하하하하하 ㅂㄷㅂㄷㅂㄷㅂㄷㅂㄷ

[익명(1.239)]

내가 읽어봤을땐 그것보단 급진적인 얘기였음

[익명(1.239)]

In a very deep sense, I think the biggest miracle that we need to create the super powerful A.I. is already behind us. It’s already in the rearview mirror.

[익명(1.239)]

We just — we need an algorithm that could learn and a network architecture that could somehow encode the knowledge and the capability to reason.

[익명(1.239)]

then once we have that, however imperfect it is, if you apply it to bigger and bigger systems with more and more computing resources, there’s no obvious upper limit.

[익명(1.239)]

Once you have a system that can take in observations about the world, learn to make sense of them-and one way to do that is to predict what’s gonna happen next-I think that is very near intelligence.

[익명(1.239)]

we have the ability to build systems that can learn, update, remember, create. We have the ability to build agents that try to accomplish a goal with that knowledge.

[익명(1.239)]

And we’ve gotten surprisingly far just putting those few simple ideas together.