Mohammad Norouzi

And then we released it and we were really surprised.

It's just so many people were so excited about text generation.

And then we realized, oh, actually, that's the whole graphic design and storytelling industry.

Text is a very important part of image generation.

And that became a very important part of our brand.

So if you search Idogram, people talk about the quality of typography, the quality of text.

We are known for typography.

really stylized typography for logo, t-shirt design, graphic design in general.

And so we continue to push forward.

I think our previous model wasn't really beating the state of the art in text generation, but we continue to focus on that.

And we had a bunch of research breakthroughs.

And with this model, despite the fact that it's very tiny, the text generation is very, very accurate.

Yeah, it's kind of difficult to exactly describe what resulted in such an amazing model.

I think a lot of it is focus and evaluation.

Evaluating image models is actually a very difficult thing to do.

There are lots of benchmarks out there, but

People look at them and they're like, okay, this doesn't correlate with pixel fidelity that I care about or realism.

You don't really want Novice users to judge the quality of these models because they may be looking at small monitors that aren't really adjusted for color accuracy.

And they always care so much about quality for the realism and

Again, text accuracy.