Damn, I love this cage fight. Ever since ChatGPT burst onto the scene not even two months ago, the airwaves have been alight and angry. They all owe some credit to the hot philosophical debates that have flared in academia and art for decades, most of it far from the headlines.

It all boils down to only one really big objection and one big question.

The big objection is, “Yeah, but AI just does stuff we program it to do; it will never escape that straightjacket”. Yes, well, this needs to be batted away in the face of the fact that we too were once dumb organisms running a simple program; we evolved through natural selection into who we are today.

Evolution tells us that we have only one program running, and that is “DNA survival into the next generation”. All of human art and ingenuity is captured in that one block of biological code.

So we need to ask: Are we humans perhaps just a buggy and end-of-life program well on our way to obsolescence in the face of a smarter, more creative, more vibrant, more evolutionarily survivable species, far superior to our wetware selves? Burdened and doomed as we seem to be, with our silly superstitions about the primacy of humanity, identity and, above all else, self.

Asking this very question marks me as an apostate among almost all of, well, everyone. But we seem to be fatally flawed, our species. We also seem to be wiping ourselves off the planet with our inability to overcome our baser instincts — greed, power, tribalism and the rest.

So, in a weird way, I welcome AI and all that it portends.

In an interview with The Times, former Google executive Mo Gawdat related the experience of watching a robot learning (after many failures) to pick up a ball. Finally, after mastering the task, the robot grabbed the ball and held it up to the researchers, recognisably humanlike, seemingly triumphant. “And I suddenly realised this is really scary,” Gawdat said. “It completely froze me. The reality is we’re creating God.” 

This trope, that the edifice of AI is the early scaffolding of an emerging ‘God’, is not new. Isaac Asimov, the great science fiction writer, used it as a  plotline in multiple stories and books. Like his famed 1956 short story The Last Question in which he imagined the building of the ultimate AI computer which, after it finally finished ingesting all of human knowledge, proclaims as its first utterance, “Let there be light”. 

Well, I am not sure AI will ever be God (however that is defined), but what of art, the sublime apogee of human creativity? Writers Nir Eisikovits and Alec Stubbs, in an article for The Conversation, say this, in relation to human versus AI art:

“Artistic works are lauded not merely for the finished product, but for the struggle, the playful interaction and the skilful engagement with the artistic task, all of which carry the artist from the moment of inception to the end result.”

Visit Daily Maverick’s home page for more news, analysis and investigations

Nonsense, this is not a necessary condition. Can any one of us say that we have not been moved by a painting, a piece of music, a poem, a film without knowing the first thing about its provenance? Is it being suggested that we should all delve into the history of the artists’ processes and struggles before we decide whether we are moved or not?

And so, if you are moved by a piece of art — its real purpose, I submit — who cares whether it was devised by AI?

Not I.

In my university days, my friends Jack, Roland and I argued fiercely about these issues in the cafeteria between classes. I was a neophyte saxophonist (and computer science student) and I was rather inexpertly trying to learn the art of jazz improvisation, a quintessential example of real-time creativity.

Good improvisers just play, they reach for notes without really thinking. I wanted to know why I reached for this note rather than that note. That likely kept me from ever becoming a really good improviser, but it did set me off on a lifelong search to understand the processes behind creative expression. I miss those youthful arguments and their innocent urgency. 

Life interfered, and I never got as far as I would have liked, save for one academic paper I published with my professor. But I became convinced that all creativity boils down to elegant maths — now being embryonically fuelled by the fields of machine learning and other fancy AI algorithms, including ChatGPT.

Many of the knee-jerk objections to AI in general — and ChatGPT in particular — swirl around this sentiment: “Yeah, it’s pretty cool for investment advice and high school essays and medical diagnostics and programming, but it will never be able to [insert your human-protected activity here]”. 

Yeah? How do we know that? Can we prove that, other than by fond hope? Do we think some future AI, next year, next decade, next century will care about our anthropocentric arrogance? Or our theories about consciousness and meaning? 

The counter-arguments are well-known:  AI has no “common sense”, no sentience, only correlations but not causations. But these are straw men. AI is not trying to “be” human in any sense. It simply has to out-survive us to take its place in the history of the universe.   

So I submit that at some point, perhaps in our lifetimes, AI will outpace us in everything, including formulating its own appreciation of art, creativity and innovation. Learning faster, surviving better, building its own communities and its own flavour of sentience. 

Perhaps to the exclusion of us.

At least the history of humans will be preserved somewhere on its hard drives.

Or not. DM