Captaining IMO Gold, Deep Think, On-Policy RL, Feeling the AGI in Singapore — Yi Tay

**Yi Tay** (0:00)
The thing that I find the most useful about these models in general is when I have these big spreadsheets of a lot of results, and I just don't even need plots of it. I think models can quite go to the screenshot, and then you can plot all this. I hate making this method-like stuff about it. It's so annoying. There were so many moments this year where AI suddenly crossed that emergent thing. Why did AI call this one? Well, let me just discuss. I think Nano Banana also got to the point where I usually make these images just for fun, just troll your friend or something like that. But Nano Banana actually got so good.

**SPEAKER_2** (0:36)
Welcome back.

**Yi Tay** (0:37)
How are you? I'm good. Great to be back.

**SPEAKER_2** (0:40)
It's been one and a half years.
Feels like a long time. So last time we talked, you were at RECA. And then you joined GDM again, working for QoQ again. Yeah. And more recently, you've started GDM Singapore. Yeah. Is it GDM Singapore or Gemini Singapore? I don't know if you've named the team.

**Yi Tay** (0:59)
I think we have a Gemini team in Singapore.

**SPEAKER_2** (1:00)
Yeah, I think in Singapore. It's called Reasoning and AGI.

**Yi Tay** (1:03)
Yeah, Reasoning and AGI.

**SPEAKER_2** (1:04)
Is it important to have AGI in the name?

**Yi Tay** (1:08)
It was like a white thing that we put AGI in. Yeah, I think that like one reason why we work on this model is that we want to get to AGI. And it was a white thing that we added AGI to the job posting. Yeah, there is no like formal name of the team yet. But it's basically the Gemini team in Singapore.

**SPEAKER_2** (1:25)
I mean, I think people are like trying to triangulate Amazon as an AGI team. You guys have AGI team. And then let's say Meta now has a super intelligence team. What are people signaling when they choose these names for their teams? Do they have, oh, we have a plan. Or is it just vibes?

**Yi Tay** (1:41)
Are you trying to fish out hot takes on the team? No.

**SPEAKER_2** (1:44)
You have officially AGI in your job title.

**Yi Tay** (1:46)
No, it's not a team name. This is not a team name. Yeah, it's just, you know, we just want to signal the north star of we're bringing these models to AGI. Yeah.

**SPEAKER_2** (1:54)
Yeah. No, I wasn't really fishing for hot takes. Okay. So you rejoined GDM. Yeah. And I think last time we talked about, I listened back to the whole thing. It was an amazing episode last time. You were talking about how it's like externally, we're in brain and came out and now you're back in GDM. Yeah. I wonder what's your general reflections, just plugging back into the Google infrastructure.

**Yi Tay** (2:15)
Oh yeah. So I guess coming back is very interesting because it felt like, when we return to Google, everything, including your LDAP, your username is all the same. It's like you play Pokémon, you leave it aside and then you go back and you click continue. It's a safe game. Yeah, it's a safe game and a continue game. It's like that. Obviously the last 1.5 years, while I was away, many things have changed. Brain is now part of GDM and stuff. So I think that obviously a lot of things have changed, but I think overall the coming back has been pretty seamless. Obviously, I love Google infrastructure and I think DBUs are great and stuff like that. Yeah. And I'm very glad to be back to Google Infra.

**SPEAKER_2** (2:52)
And was the intention always that you were going to work on Deep Think?

**Yi Tay** (2:56)
No, not really. I think I miss research a lot, like doing research, not like super fundamental research, but like close to model research. But I really miss being at the frontier and trying to go beyond that. So I really miss that a lot. And I think when I came back, Big Think wasn't a thing. And I don't think there was any plans, actually. It was just like, I'm just going to work on research and see what happens. Yeah.

**SPEAKER_2** (3:19)
I'm sure, I guess, there was some inclination that reasoning is the next frontier. And that's like, obviously, the most rewarding research path, especially this year.

**Yi Tay** (3:30)
Yeah, I think reasoning, these days, reasoning and RL is like, probably quite, it's RL reasoning, because I spent a lot of my past life, I call it the past art, working on like architectures and pre-training. But I think now I more, I have like transitioned more into RL research. I'm not like old school RL, but the games RL and the old school RL. And to be honest, I had almost no RL background coming back, but I think like, like RL is the main means of modeling these days. And yeah, so I think it was pretty easy to jump back in. And I think a lot of fundamental skills in research is for general purpose and universal, and it's quite easy to innovate, even in a toolset that you're not super used to. And yeah, so I think RL is basically the main modeling toolset that we play around with these days.