zlacker

>>davidb+(OP)
Sam implied OpenAI had a major breakthrough a few weeks ago in a panel yesterday:

"Like 4 times now in the history of OpenAI, the most recent time was just in the last couple of weeks, I've gotten to be in the room when we sort of like, pushed the veil of ignorance back and the frontier of discovery forward. And getting to do that is like the professional honor of a lifetime".

https://www.youtube.com/watch?v=ZFFvqRemDv8#t=13m22s

This is going to sound terrible, but I really hope this is a financial or ethical scandal about Sam Altman personally and he did something terribly wrong, because the alternative is that this is about how close we are to true AGI.

Superhuman intelligence could be a wonderful thing if done right, but the world is not ready for a fast take-off, and the governance structure of OpenAI certainly wouldn't be ready for it either it seems.

>>ademeu+wp
No, we are not close to AGI. And AGIs can't leave machines yet, so humans will still be humans. This paranoia about a parroting machine is unwarranted.

>>csomar+VJ1
you're right. agi has been here since GPT-3 at the least.

it's honestly sad when people who have clearly not use gpt4 would call it a parroting machine. that is incredibly ignorant.

>>aidama+622
Let me know when GPT can even play chess without making invalid moves, then we can talk about how capable it is of logical thinking.

>>ric2b+Ec2
Let me know when you can prove that "logical" and "intelligent" were ever stored on the same shelf, much less being meaningfully equivalent. If anything, we know that making a general intelligence (the only natural example of it we know) emulate logic is crazily inefficient and susceptive to biases that are entirely non-existent (save for bugs) in much simpler (and energy-efficient) implementations of said logic.

>>kliber+oZ2
An AGI that can't even play a game of chess, a game that children learn to play, without making an invalid move doesn't really sound like an AGI.