Monday, May 15, 2017

GUEST BLOG POST: The Coming Human-Machine Partnership In Creativity — Sam Arbesma

We are tool users. While this skill is not unique to humanity (Wikipedia has helpfully documented other cases of “Tooluse by animals”), human civilization is suffused with tool use. We use tools to help us eat, tools to help us build homes and other structures, tools for transportation, even tools to assist us in getting exercise.

Awash in the many tools of our technological society, we are seeing a new type of tool on the horizon, one which is similar in some ways to what has come before and in other ways quite different. These are tools that augment human creativity. We are increasingly, as a society, building tools to help assist human creativity, whether in art, design, or even scientific discovery.

Obviously, we have long had tools in the cognitive realm: tools to help us store and retrieve information (paper!) and tools to help us better perceive the world around us (eyeglasses). And some of these tools have even been used to drive further discovery and creativity, such as when the microscope and telescope were invented. We would not have been able to see the wonders revealed by these tools on our own, but due to them, humanity reached new creative heights.

But these new computational tools for creativity are different in one particular way: they seem to be replicating some of the creative processes within our own minds. Whether the underlying algorithms actually faithfully mimic the processes in the human brain or not, their output is becoming increasingly human-like.

For some of us, this can be a bit distressing. Creative output has long seemed to be one of the few things that was immune to the march of machines and computation. But it’s not. We have software for automatically generating news articles, software for creating music and art whole cloth, algorithms for designing novel optimized shapes, and even computers programmes that tell jokes.

Here is a small sampling of what is increasingly possible:


This realm of computational creativity, or creative AI, is broad and rapidly growing (see CreativeAI.net to stay on top of the newest developments in this space). If the last ten years were about building AI to help analytic workers, the next ten will be about building AI to augment creative workers across all disciplines. This marks a material shift we’ve seen and a thesis we’re actively investigating in at Lux. And whether you like it or not, it’s coming.

So how should we respond to this new era? Delight and wonder, of course. But even more importantly, we need to work together with these new tools: we need to be partners with our machines in the very act of creation.

Creativity has always been a collaborative act. We can now increasingly count machines as members of our team as well.

Sam is Scientist in Residence at Lux Capital and the author of the book Overcomplicated.
Follow him on Twitter.

No comments :

Post a Comment