Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

I asked my autonomous AI to solve AI alignment. Fascinating to watch. Agents: - Task execution: text-davinci-003 - Task creation: text-davinci-003 - Task reprioritization: text-davinci-003

64,053 Aufrufe • vor 3 Jahren •via X (Twitter)

9 Kommentare

Profilbild von Bobcat
Bobcatvor 3 Jahren

Damn you gotta start adding music

Profilbild von Yohei
Yoheivor 3 Jahren

Oh tight. Wait, should I start a Tik tok? Take requests and just post videos of the output.

Profilbild von kache
kachevor 3 Jahren

solved

Profilbild von Bil Dye
Bil Dyevor 3 Jahren

Ugh. I wish we hadn't defined the term "alignment" to cover both "do what I actually want" (human-to-computer), and "we need to agree now on what we want" (humans to each other). Is there a term for just the 1st challenge -- aligning what you ask for with what you truly wanted?

Profilbild von Yohei
Yoheivor 3 Jahren

Funny you say that, because it’s suggesting I take a mortgage on my house. Needs more compute or something - it’s for the greater good.

Profilbild von AI Notkilleveryoneism Memes ⏸️
AI Notkilleveryoneism Memes ⏸️vor 3 Jahren

excellent. yohei, pls use your big brain for this, we have no idea how to solve alignment

Profilbild von Not Mr. Bean
Not Mr. Beanvor 3 Jahren

How do you stay within with 4k token limit as it keeps extrapolating ?

Profilbild von Yohei
Yoheivor 3 Jahren

It’s a balance of figuring out how much past context you want to pull in. If you pull zero, it stays within limit easily but no prev context. If you pull 5, you get decent context but it can get large. So then you can consider adding a summary step.

Profilbild von Yang
Yangvor 3 Jahren

Why text-davinci-003 instead of 3.5-turbo?

Ähnliche Videos