What is the difference from the traditional model building experience?

Yeah, so when you're doing real time, one of the underappreciated aspects of it is that latency matters a lot.

And when latency matters, there are both physics-based and software-based constraints that oftentimes are overlooked and people don't optimize for them.

So, but if you take real time and no latency seriously, then a lot of your decisions around what hardware you choose, where you choose it, how you build the whole stack become very important and you also are going to do it very differently from what has been done so far with respect to say text generation, image generation and video generation.

So I suppose that in terms of inference, effectively the whole stack looks pretty different from what other companies have built today.

For example, a together AI or a file even.

And so you have to make certain decisions that minimize latency actually instead of maximizing throughput.

So in two years, what share of inference runs somewhere else other than hyperscalers?

Like what does that mean for someone and why is it important?

Yeah, I think your definition is very, very good.

And why is it important? Because effectively, a lot of the revolution that we're seeing in world models has actually been enabled by the progress that as an industry we made with video models.

A lot of the learnings that you said that a world model has around, for example, if I tilt a plane and the ball falls down on it, come, not all of them, but a lot of them come from the fact that there was a video model that was pre-trained, it's called, on just a lot of internet video data.

So it has learned concepts, generic concepts around, for example, like, let's say gravity, although the model doesn't think of gravity, or say... water dynamics.

So that kind of work has been made in order to generate pretty pixels with video models.

And then it's leveraged to make world models, to your point.

Is that something that they're doing kind of desk side or client side coming up to you, you know, banging the API? Do you provide anything in that middle layer that helps them kind of achieve success? anything else other than just access to the model?

To be the developer platform in that industry, leading in that way, what does that mean?

Yeah, so it means that you need to build a developer platform that is markedly different from existing gen AI developer platforms.

And really, that stems from thinking logically about the nature of these models.

Like I said, they are, first of all, real time.

And they're also interactive, second of all.

You cannot just give them a prompt and wait for them to do something because that kind of defeats the purpose of them being interactive, right? And so their real usefulness can only be brought forward by enabling interactions and interactivity.

Existing platforms in general have been built for file generation.

yeah