Creative Singularity
a reflection on the creative potential of Artificial Intelligence
Will there be a ‘creative singularity’?
The imminent arrival of the Technological Singularity is omnipresent, at least in the press, board rooms and TED conferences. You cannot escape it, it’s enthralling and its palpable. It has even began to elbow its way into the last strong hold of humans, the realm of creativity. I am not a Luddite, hardly, I’ve made my living being on the leading edge of tech, however… I’m not buying that there will be a Creative Singularity – ever.
By creative singularity, I mean when Artificial Intelligence can teach itself to become more creative on its own, without any human involvement.
Artificial Intelligence is a tectonic shift in the world of knowledge
This is obvious, it will be able to rip through thousands of years of information to find the best and most logical references. It will suggest and execute combinations of data that we would just not be able to do because of the difference in processing speed between our brain and the latest computer. Herein lies exactly the reason why creativity will remain ours, because AI will always be working from the past, even if it’s a fraction of a second old. Knowledge, by definition, is from the past, It’s data will always be of things that have already happened. Creativity, however, is a nuanced moment of imagination that did not previously exist a second ago.
On the subject of creativity, I have two points to make.
The first is that the higher volume of ideas one pushes out does not necessarily mean there is a singular great idea, or even a good idea in the bunch. Creativity is not like shoveling sand, it’s not a quantitative task in the sense of how much of it did you move? Creativity is not about filling up a roll of trace paper with sketches. I have no doubt that a ‘creative’ AI program can put together 1000 ideas effortlessly. Who is to say which are rubbish, which are good and which is genius?
Which brings me to my second point. Is being random creative? That’s a hard question to answer. I love Tristian Tzara’s Dada poetry and Brian Eno’s algorithmic music video which never repeats, I think they are all magical in their randomness and on the forefront of the creative spectrum. However there can never really be true randomness by us, at some point we make a choice of process. What if someone made every design decision with a roll of the dice (believe it or not, it’s been done)? That would not be quite as satisfying, that’s random for the sake of being random. The idea to do it was good, but the actual execution of the idea was not necessary. How would AI know the difference? The nuance to recognize creativeness in randomness is a far stretch for an AI program.
When to break the rules?
Breaking the rules, the infamous ‘thinking outside the box’, has always been a subjective exploration. We cannot consider an infinite amount of possibilities and by corollary, an infinite amount of ideas. Recognizing when something almost random strikes the right chord, now that’s creative.
Like I said, creativity is a nuanced moment that did not previously exist.
I really liked when you said ”Knowledge, by definition, is from the past, It’s data will always be of things that have already happened. Creativity, however, is a nuanced moment of imagination that did not previously exist a second ago.” How can something be creative if has already been done before. And computers learn by repeating the past to look for mistakes which isn’t what creativity is about.