JB: Why did you say, “Creativity is not only for humans,” implying that AI was autonomously making the work, even when you knew that was a false statement?

What about your narrative that “creativity isn’t only for humans”? Were you playing up the machines and now saying that is not what you meant?

HC: Yeah. Exactly. I think that's what happens when you're doing something and nobody cares, then you’re just goofing around and doing really clumsy stuff. And then when everybody has this view, then they go back to what you did before and then you have to justify it. We kept justifying, because we still think that this part of the GAN operator that creates the images is really interesting and there is some form of creativity there … and we just thought it was cool to just do it like this. For us, it was just a funny way to talk about it. 

JB: You didn't know you were going to be under the microscope.

HC: If we knew we were going to have to 400 press articles on what we do, we most definitely would have done that. But at [that] moment we were like, 'Yeah, it’s silly, okay, whatever, let's put this.' But retrospectively, when we see that, we are like, 'That's a big mistake.'

JB: All you can do is admit the mistake. What creative behavior do GANs exhibit? Many feel they don’t exhibit creative behavior.

HC: For me, the fact that you give it a certain number of examples and then you can continue to see results in the latent space, for me, the gap has to be [bridged]. So necessarily, there's some kind of, like, inventing something. So I guess there is some kind of creativity for me… because creativity is a really broad term, so it can be misunderstood, because creativity is something really related to humans. But at the basic, low level, it was given a set of images, it can create images that does not belong to the training set. So that's something that is transformed by the model, and there's some kind of creativity. So it's just a way you interpret the word "creativity." Maybe from certain perspectives you can say it's creativity. 

JB: So it sounds like you believe it is dependent upon your personal definition of creativity? Some people say GANs are just are approximate distributions and that is not really creative - but it sounds like you think it is creative?

HC: Yeah. It's like, whatever you think creativity is, if we fit on the same definition, we are obliged to agree on something. So if we go to the same definition that creativity is something like, let's say, this ‘Concept A,’ then GANs will fit this concept. Or not? It's just a point of view thing, I guess -- and I understand that people can argue that [it’s] not great, we understand that, but it's just a point of view.

If it was AI then the IQ was very low.