NVIDIA GauGAN artificial intelligence drawing image

The Most Amazing Thing Launched at NVIDIA’s GPU Technology Conference

Amazing is a subjective term. For instance, while there were many incredible improvements to word processing products in the 1980s, the most amazing to me was the spell checker because I’ve never been able to spell consistently to save my life. If it wasn’t for the spell checker, I’d never make it as a writer, so had it not been invented I’d likely be an accountant or an engineer and nowhere near as much fun to be around.

Anyway, I can’t draw worth a damn either, which is pretty embarrassing given my mother and first step-mother were both commercial artists. I even took classes in art and, apparently, I was really good at abstract art because no one could tell what I’d drawn. Oh, and being color blind didn’t help. So when Jensen Huang got up on stage to demonstrate an artificial intelligence (AI) assisted application from NVIDIA called GauGAN I had a near religious experience. Just like the computerized spell checker this thing could change my life. More importantly, it could save us all from the coming wave of IP theft problems as the different companies move to protect creators from those who use their pictures, from the web, without permission or license.

Let me explain.


What GauGAN does is it takes colors translates them into objects, and then places the objects so the result looks like a photograph. In other words to create a landscape all you have to do is draw a blob of one color to create a forest, a blob of another color to create a lake, a blob of another color to create clouds, tell the computer what you want and you end up with what looks like a picture of a beautiful valley that only existed in your—and the AI’s—minds.

If you want to draw people, a future version of this, would just require you to define what the person you wanted to draw looked like, connect that person to a symbol you could draw, and the computer—when building the picture—will create the people you wanted in the picture you were creating and it all will look like they were actually in that imaginary landscape.

It is very much like if you drew a rough abstract drawing and the computer translated that into a photo realistic picture. You could draw a photorealistic full screen landscape in 5 minutes and likely still have time to grab coffee where typically that kind of detail would require a graphics artist an hour to create. This would be a massive time saver.

Saving Litigation Cost

However, those of us that post on the web or write, often need pictures of things to make our points so we pull pictures off the web or from licensing sites to populate our posts. But we don’t really need the specific picture we pulled, we just need an illustration and there is always the chance that the one we choose could result in a takedown notice.

But if we can create the picture we want from lines, colors, and stick figures, we can create the pictures we need without borrowing from anyone else. We can get a far better picture for our needs, because we generally know what we want, and avoid any related problems.

I expect we’ll also be able to train the AI with other pictures so we can do product reviews. Take pictures of the product and have the AI not only auto-edit the pictures but help us place them in compelling places and interesting poses. For instance, I’d like a picture of a laptop I’m doing a review on—if it is good—to look like a superhero jumping off a tall building, and—if I don’t like it—burning in hell. I think it would make the points better than most anything I could find.

Wrapping Up

Much like spell checkers changed the lives of those of us that can’t spell, GauGAN will change the lives of those of us that can’t draw. On top of dramatically improving our ability to play Pictionary, it would allow us to better create graphics for our posts and presentations, more fun things to post for social media, and allow those of us that are color and artistically challenged to better explore our visual imaginations without making those that see our work laugh excessively or lose their collective lunches.

Put more succinctly, GauGAN is just short of a miracle that can take those of us who are challenged artistically and turn us into impressive artists. There isn’t a single thing I’m waiting for more eagerly than this miracle of technology from NVIDIA. Much like someone that was blind, if given sight, might scream “I can see” I’m looking forward to the day I can scream “I can draw” and my wife just stops screaming at my drawings.

Check it out, if you are like me, this could change your life.

Comments are closed.

Scroll to Top