Transcript
Podcast Host / Narrator (0:02)
Tetragrammaton.
Greg Brockman (0:23)
One thing that's really changed over the course of 2025 is, was people started to use ChatGPT for much more personal, very intimate applications. For example, so my wife has complex medical conditions, including hypermobile Ehlers Danlos syndrome, which took many years for us to get diagnosed. And as we put those symptoms into ChatGPT, it would be able to figure out pretty immediately. But the thing is that every doctor has their own specialty or rather than one doctor who can see across everything, and she uses ChatGPT to manage her health all the time.
Interviewer / Podcast Host (1:00)
Great.
Greg Brockman (1:00)
For me, the funny thing is I'm actually a late adopter of our own technologies usually, and I usually test them and I stress test them in all sorts of ways. And it's funny, for early versions of our models, I would usually like try to break some of their filters. And so I would swear at them and yell at them. And my wife is always like, he's just kidding, you know, telling me to be nice to the AI. And I think that for me, I actually have been someone who's almost very set in my ways. And the first time this has changed is very recently with Codex and really starting in December, like I've been a curmudgeon, I've been, I've got my way of doing things, I use my terminal, I use my emacs, like all these like tools that I grew up with and I've just abandon all of that now.
Interviewer / Podcast Host (1:45)
Wow, just using Codex, that's revolutionary.
Greg Brockman (1:49)
It really is.
Interviewer / Podcast Host (1:49)
You sound like you were set in your ways.
Greg Brockman (1:51)
I really was, yes.
Interviewer / Podcast Host (1:53)
What changes with each new model?
Greg Brockman (1:56)
Everything. And the way to think about it is that we, from the outside perception is, oh, you're just scaling up the models, you're just doing this kind of dumb thing on the inside. Every single part of the process, we are always up leveling. Like the thing that works in machine learning and that machine learning rewards is attention to detail. And so you really want to make sure that all of the scaling is right. You want to make sure that the systems, for example, GPUs failing, that happens. And so how do you detect if you have a run of 100,000 GPUs, how do you detect which GPU is the broken one? It's not easy, right? You can't just be like that one. Right. So you need this almost. There's a physical process to it, there's the software process to it, there's understanding if your data is any good and that there's so much reward to just actually Looking at the data to understand what's in there and to make sure that you format it correctly and it's tokenized properly. And so there's just every single part of the input we are constantly improving and we pay attention a lot to the output too. We're trying to see how does improvement here connect to an eval shifting. And one observation I have across many years of OpenAI is that if we have some signs of life, some application that kind of works right now, one year from now, you should expect it to be excellent. And so we are on this exponential and able to make these like very, very sophisticated improvements over time.
