On this we should agree: AI is so last year. And so this year. And so on and so forth as we march, Tiktok dance, or leave fingernail scratches on the rocky cliffsides of a progressively less predictable future.

Progress, of course, being wholly subjective.

That’s because the real question is whether any of this bodes well for what we wetware circuit bundles like to think of as humanity. No question, the signs, portents, and augurs are decidedly undecided.

Reading the bravado from the 80 senior industry executives Ad Age asked to assess the year ahead, it’s pretty clear most think you can’t sit at the cool kids table unless you’re all-in on AI. For those breathing rarified corporate air, the convolutions required for creative “leaders” to justify the embrace are purely a survival strategy.

Reading a grim WaPo report titled, “AI is learning from images of child sexual abuse,” you realize that the legal issues relating to Large Language Model training sources and provenance are neither trivial nor easy to sweep under the insect-infested rug. 

Not unless we’re willing to accept the social impact of making it perfectly legal to train AI on dark web child exploitation and distribute the highly modified images for fun and profit.

Eeeek! Or, maybe better, e/acc, also known as Effective Accelerationists, a Silicon Valley tribe now pushing for unrestricted AI development. If I have this right: no regulatory safety nets, no policy guardrails, no legal IP protections, not even a measured pause for “just in case we’re wrong” consideration.

Of late, I’ve been thinking that the CTO I met last year was one of them. His confidently espoused view was that as the cost of “compute” (computational capacity) drops off a cliff, the price of swapping in a fully functional AI for a lesser-speed human will wind up around $1,000 per install in the next 9 years. 

When compared with your average healthcare professional, knowledge worker, lawyer, clock puncher, phone answerer or, okay, advertising creative who charges an order of magnitude greater per month, it’s the technological Godfather’s unrefusable bargain.

You don’t need to see around corners to see how entire species of gainful and productive chances to work in things that make a difference will be scythed down.

But e/acc says not to worry: as my CTO future tour guide gleefully pronounced, this will result in a time of unparalleled plenty—with humans liberated from the need to do anything in particular in a world where the distance between problem and solution is reduced to quantum mechanics.

The side-benefit: we’re going to need fewer humans.

Now, there’s no doubt that a reduction from a planet-overburdening, climate inflicting, 9-billion people is a desirable outcome. But when you consider that we’re at an inflection point predicted since the dawn of automation without anything in the way of meaningful “how to’s,” you have to worry. Especially if the timeframe is so radically foreshortened.

You want to see where all this leads, punch over to https://effectiveacceleration.tech and the startling web headline, “Effective Acceleration means accepting the future.” Makes me think Grimes, aka Claire Boucher, aka, Elon Musk’s former, had a point after DJing a set for an e/acc cocktail-party-cum-evangelism session. She called it “a visit to enemy territory.”

Like I say, fasten up. The pilot has illuminated the light.