Wonder-lust.
And, of course, we should now all be thanking LinkedIn, from the deepest and sincerest cockles of our nary-a-drop-of-irony hearts, for giving us a brand new option when posting thoughts on the site.
It’s a button featuring three diminutive greyed-back words: “Rewrite with AI.”
Talk about a trigger for the insecure; you can almost hear a visceral click of the mental firing pin sliding into position. Does my written shit perfume the digital air so badly that it needs a machine to improve it?
Don’t answer, don’t want to know.
Still, you have to appreciate how this implied commentary on the state of human literary skills concentrates the mind. In part, by crystallizing a question that’s been pecking at my decidedly non-technological head for the past few weeks.
It’s this: what if Turing, the man who kicked off the AI madness, had it wrong? About his eponymously named test which tells us that computers will have attained true “artificial” intelligence when a remote human researcher, communicating with two others—one AI, another a second researcher—cannot tell the wetware from the machine.
It’s really a standard based on mimicry and deception, which is why it also made a terrific movie about the man himself. But is it the most accurate measure of the moment when AI demonstrably breaks into the most stratospheric realms of cognition?
I wonder if there’s something else, a sharper line of demarcation or a brilliant beacon; a more sensitive signal that will let us know we’ve met, or, maybe better, built our match.
All expressed in exactly those two magical words: I wonder. Or, put in Turing-like terms, “the point a machine reveals it has a sense of curiosity.”
This is where the conversation has to remain philosophical, not technological. Because I’m quite sure the Effective Accelerationists—the e/acc cheer squad for unconstrained AI development—would sing the large language modeling song in a way that ends with the lyric, “if we can prompt to achieve a similar trip, who gives a binary flip?”
I do, and for this reason: AI isn’t the only computational system out there that dazzles with its fluid and seemingly effortless capabilities. The other one is called the human subconscious and, as I get older, I grow increasingly respectful of its prowess.
Both, when you think about it, rely on the existing universe of knowledge and recorded experience.
Both, it seems, are capable of bias and error.
Both respond to instruction.
But only one has a singular advantage—the ability to wander, to muse, to veer off the prompted path and connect two or three or a half dozen previously unconnected dots.
And maybe that’s the real value that links a fruitful AI and human future. It will bring us the ability to source, iterate, even interpolate the whole damned internet in time pacing at quantum speed.
We’ll bring the curiosity. And the wonder.