(auxiliary) OpenAI updates, timelines, transcript, etc

[mishka_discord]

(auxiliary post, not one of the 500 words HalfHaven posts)


One thing that happened is that they did not publish yesterday livestream on YouTube. Instead it is published on Vimeo: vimeo.com/1131297184

It's annoyingly difficult to copy the Transcript from the Vimeo site, so I am going to gradually copy-paste the Vimeo autogenerated transcript to the comments (the quality is...oh well... the idea was to ask an LLM for a summary based on that transcript, but I am not sure this would work well with this quality of transcript).

So I have copied the audio track (see the first comment for details) and I am going to see if I can autogenerate a better transcript.

***

Anyway, first 15 min include timelines with caveats:

Currently can do 5 hour long tasks (that corresponds to my feelings)

Expected timeline (with all kinds of caveats, but they do organize around these timelines:

September 2026: Automated AI research intern
March 2028: Automated AI research

They also give more details about their approach to AI safety during these first 15 min.

***

finally, I see a YouTube version: https://www.youtube.com/watch?v=ngDCxlZcecw

also Sam's summary: https://x.com/sama/status/1983584366547829073

Comments

[mishka_discord]

Transcript

Hello, I'm Sam. This is our chief scientist, Yoko, and we have a bunch of updates to share today about OpenAI.
00:00:06

Um, obviously the news of today is our new structure. We're gonna get to that near the end, but there's a lot of other important
00:00:12

context we would like to share. First, given the importance of a lot of this, we're going to go into, uh, an unusual level of transparency about some
00:00:16

of our specific research goals and infrastructure plans and product. But we think it's, uh, you know, sort of very much in the public interest at
00:00:24

this point to cover all of this. So, our mission at OpenAI in both the nonprofit and our new PBC is to ensure
00:00:30

that artificial general intelligence benefits all of humanity. As we get closer to building this, we have new insights into
00:00:36

what that is going to mean. There was a time earlier on in open AI where we thought that AI would, or a GI would be sort
00:00:42

of this oracular thing in the sky, and it would make all these wonderful things for us. And we now have a sharper view of that, which is
00:00:49

we wanna create tools and then we want people to use them to create the future. We wanna empower people with AI as much as possible,
00:00:56

and then trust that the process that has been working for human history of people building, um, better and better things with newer
00:01:03

and better tools will continue to go on. We, we can now see a vision where we help build a personal a GI that people can use anywhere
00:01:09

with all of these different tools, access to all these different services and systems to help with work and personal life and their personal life.
00:01:18

And as AI gets better and better is AI can even do things like discover or help discover new science.
00:01:25

What people will be able to create with that, um, to make all of society better in their own lives,
00:01:30

more fulfilled, we think should be quite incredible. There are three core pillars we think about for OpenAI research, product and infrastructure.
00:01:35

We have to succeed at the research required to build a GI. We have to build a platform that makes it easy
00:01:44

and powerful to use, and we have to build enough infrastructure such that people can use at a low cost, all
00:01:49

of this amazing AI that they'd like. Here's a little like cartoon of how we think about our world. So at the bottom layer here, we have chips racks
00:01:55

and the systems around them, the data centers that these go into, and the energy. Uh, we'll talk more about the first
00:02:03

three today and energy another time. Then we train models on top of these. Then we have an open air account on top of that.
00:02:08

(difficult to copy, and the quality is really rather awful, pausing)

I should see if I can generate a better transcript

At least the following works to download the audio track:

$ ./yt-dlp -x --audio-format m4a --audio-quality 0 --embed-metadata --embed-thumbnail --extractor-args "generic:impersonate" "https://vimeo.com/1131297184"