This interview was edited for clarity.
Take a job like Microsoft’s Rembrandt project [which generates AI-generated images in the style of Rembrandt]. Do not we’ve got fantastic Rembrandts currently?” There is a point, and that’s that it might help us understand new items from the work of art. If you have a look in Jackson Pollock’s work from a mathematical standpoint, we can observe new items that we missed before. So there’s an intriguing function that AI can play in showing new structures which we might have overlooked in works of art that we take for granted.
One of the challenges of AI now is that lots of the machine learning applications produce code, but we do not know quite how it’s working. The Google DeepDream project is helping us find a way to understand how that happens. So as artwork for us as people is a method of helping us get in the mind of the other individual, maybe art created by AI can help us get inside the workings of the code that is rather mysterious.
Let’s start by breaking down”imagination” What are those and what do these breakdowns mean for the function of AI?
There are lots of creative fields. Is there one that you locate AI and the maximum struggle?
The other interesting story that I think is really important to this whole book is really in the sphere of visual arts and that is Google’s DeepDream. Google requested its visual recognition software to find out exactly what it saw in a random array of pixels, also by dialing the pictures that it had been seeing we heard something about the way the AI was thinking, how it was seeing, and the way it had been programmed.
And where does AI fit into each of those patterns?
The Verge spoke to du Sautoy about different kinds of imagination, AI helping humans become more creative (rather than replacing them), and the creative areas where artificial intelligence battles most.
So I feel this is one of the exciting roles that AI will play going forward. Very humans start repeating patterns of behavior. It helped reawaken his creativity because he had been revealed there were things that he could do using the components that he already had and that he had not realized was possible. I wished to demonstrate that the function an AI can play in imagination is possibly to enhance human imagination, this will be a partnership moving forward, that together we could make things more interesting than when we just worked independently.
There’s exploratory imagination, which will be taking the principles of this game and pushing them into the intense, such as Bach. There’s combinatorial imagination, where you’re taking two thoughts that have nothing to do with one another to see how associations in one can help stimulate new thoughts in another. The next imagination, which is the most mysterious, are such moments that somehow seem to come out of nowhere — those phase changes when suddenly you’re boiling water, and water becomes steam and affects state completely.
What was striking was that the jazz musician’s reaction. He explained,”That which I hear, I understand. That’s my world of music. It is playing just like how I perform, except it’s playing things I’ve never considered doing before with my music audio world.”
What’s the significance of this?
Well, in the area of movie, let’s choose the Netflix algorithm that urges the films we would like. It may divide films in interesting new ways. A few of the groupings we could identify as”it grouped all of the comedies collectively,” but sometimes it had been grouping films in clusters according to people’ expressions of likes and dislikes where we couldn’t know what was the common theme. It recognized a genre of film that we didn’t even have a word for. It is like saying,”there’s another sort of taste inside here and you need to name this” It perhaps see things , subconsciously, we’re expressing and can take our creative outcome but have not made aware. It can assist us articulate what might be there within our artwork.
One of the surprises for me was challenging the word would be. There is so much written word available for AI to learn on. I was surprised that although AI is now quite good at writing short-form literature, it is still not able to really sustain the written word during a lengthy term. It does not have a sense of narrative arc, for instance. I haven’t seen anything that keeps a narrative. I’m really looking to see whether that’s achievable, and that I really don’t see why it can not be, as we do, but it might be very challenging for AI to have the ability to articulate language. Maybe it requires more than just exposure to data, perhaps it requires a period of development as we have been through, and the question is: How long will it require?
Traditionally, it had been believed,”How could AI violate the rules? Isn’t it stuck inside a system since it’s programmed to do work in a particular way? How can it leap outside?” But if an AI is told,”I”m going to break the rules,” this is a rule in itself. You own a meta-code which tells the app to break the code inherent.
So when du Sautoy watched DeepMind’s AlphaGo beat Lee Sedol, he thought there had been a sea change in artificial intelligence that could affect other imaginative realms.
Those creativities Each offer a different challenge. Exploratory creativity seems perfect for a computer because that’s exactly what a computer can do, so a lot more calculations than the usual human mind could ever perform. Combinatorial creativity is interesting, [an AI] could learn patterns and use them to new areas. But I feel the hardest one is your concept of getting something fresh and breaking out of this machine.
“I’ve always compared doing mathematics to enjoying the game of Go,” he says, and Go isn’t assumed to be a game that a computer can play since it requires intuition and creativity.
In the book, you discuss a great deal of creative AI jobs. Which ones were interesting for you?
A lot of people think that creativity is all about expressing what it means to be human, and therefore could AI get anywhere near that? I look at a great deal of artists and reveal that quite a great deal of art has pattern and structure behind it, which is very mathematical in nature. That’s why I believe artistic creativity may be about pattern and algorithm than we give it credit for, and very often the patterns are hidden, and perhaps that is something AI can discover because it seems to be very good at detecting hidden patterns.
That sort of pattern-detection isn’t confined to only the visual arts, right?