He highlights AlphaGo’s approach to the game Go as his example. AlphaGo, the first AI to defeat a Go grandmaster, applied a never-seen-before move that no human would have made. Indeed. Most humans saw the move as a mistake when the AI made it, failing to recognize anything beyond the two choices that the grandmaster expected the program to make. But the algorithm didn’t care about 4,000 years of Go history. It was programmed to win. It did.
Rubin is right in that the AlphaGo win is a teachable moment for human beings. Sometimes, we carry too much emotional, intellectual, and historical baggage around with us to be truly creative. Ergo, divergent thinking is still king when it comes to creativity.
Divergent thinking is also where the proliferation of AI ceases to interest me. Don’t get me wrong. I still pay attention, especially when my colleagues point it out.
Hugh Behm-Steinberg, who teaches writing at the California College of the Arts, recently did so when he mentioned: “One of the dark pleasures of teaching uncanonical work is reading the AI hallucinations my students think I won’t notice.”
To be clear, Behm-Steinberg allows AI assistance if his students include their prompts with the work. He says it’s better than forcing them to sneak it into their assignments and then failing them when he spots what he calls AI hallucinations (something nonsensical, akin to those crazy hand defects that litter some graphics).
I don’t know. After seeing the first official music video made with OpenAI’s Sora on LinkedIn, I still struggle to condone its broader applications that attempt to supplant human creativity. The video is largely unoriginal, with horrible camera angles and bad morphing effects that cause some people motion sickness — AI hallucinations that we can see rather than read. It’s a fail, propped up only by the crutch of AI infancy.
So, what is the status of AI creativity? There isn’t any. I mean, using AI editors as a prompt to improve sentence structure is one thing, along with applying a photo effect that saves some tedious pixel tweaking or creating elements that can inform a component of a bigger project. Those are suitable solutions. This continued pursuit of trying to make it capture a human’s imagination, on the other hand, is faulty by design.
At its core, the true strength of art in all its forms is one human’s mastery over some medium so they may share their unique perspective of the world with others. These perspectives — a lifetime of experience and knowledge and, sometimes, the purposeful subtraction of said experience and knowledge — is more unique than a human iris. And this is why AI, programmed to mix and match other people’s work, will never truly obtain human creativity — even if it is constructed to be born and live like a human being. Because, even if it were built to be born, then it would still only represent a single point in an infinite ocean of stars.
No. More likely, AI merely represents a reverse renaissance or a great reduction in creativity. As humans allow machines to copy processes, techniques, and rules, they may become even lazier in the pursuit of original thinking. And it will be only then that AI may succeed in simulating something superior, not because it’s creative but only because we will cease to be.
Ho hum. I liked it better when programmers focused on teaching AI to wash the dishes and mop the floors so that we could have more time to be creative. Instead, this trend to program AI to be faux creative will only give us more time to wash the dishes and mop the floors. And we’ll all be too dumbed down to even know the difference. Good night and good luck.
