The story of the Neanderthals used to run something like this: They were cognitively inferior dullards who could barely figure out how to hunt, forget come up with a culture. Until they were relentlessly replaced in a slow-burning genocide and wiped off the face of the earth by more advanced late-comer hominids, Homo sapiens.
That story is wrong. Ongoing research shows they likely ate, bonded, and even made art much like H. sapiens. Many modern humans have a surprising amount of Neanderthal DNA, a “stunning 20%.” These genes are a record of interaction, possibly repeated interactions, across multiple generations. We are the product not of a horrific Darwinian extermination effort lost to the mists of time, but of interbreeding and mingling, not only between H. sapiens and Neanderthals, but likely with other hominid groups like the Denisovans.
The evolution of modern humans is more like the braided headwaters of a river high in the mountains, with small streams flowing in and out, some sinking in the soil, some rejoining to form the beginnings of a river. Once that stream hits the plain, it seems mighty, inevitable. But hundreds of meanders, forks, and tributaries led to this inevitability.
Why am I going into this on a music technology blog?
Because the story of the Neanderthals is instructive when we start thinking about AI. We’re in that headwater moment, and yet we seem to think we’re on the cusp of a battle royale for human existence. (We may be, but that’s climate change and our only enemies are ourselves.)
AI in popular thought and storytelling is almost always confused with Skynet, with some post-Singularity cyberbeing, with some potential general intelligence that would make seemingly arbitrary decisions and destroy humanity. This thinking lurks behind the frequent drumbeat of questions in headlines and interviews: Will robots replace us? Will AI become the next Picasso, Prince, or J.K. Rowling? In our dystopian, comic-book-level imagination, AI is pitted against humanity. Only one force can win.
Just this week, a wave of stories stoked the now familiar panic by claiming that Endel, a mood music app, was the “first ever algorithm to sign a record deal with a major label.” Cherie Hu and other cool heads have already dissected and defused the sensational headlines- it’s a distribution deal, not a “record deal,” for one thing, and it’s the humans who run Endel who inked it. Without a perspective shift, a different AI project will get caught up in the same outrage cycle next week.
What we’re looking at is a thousand braided streams of AI. Some do indeed threaten key values we hold dear and would like society to reflect. Some target innocent people. Some threaten the livelihoods of perfectly good professionals and workers. Some, however, are benign, beautiful, emotionally fulfilling, wildly collaborative. Many AI applications aim not to do the work for humans, but to amplify what humans long to do more of by generating intriguing or random ideas. These ideas can inspire music composition or songwriting (Amadeus Code) or visual imagery (AICAN) or any number of other AI-based assistants. In their expansion of our imaginations, they feel deeply human (because at the core, they are).
These moments, each with its own decision point, are combining and separating, flowing in and out. We have a call to make with each one, and nothing is inevitable--yet.
This braid is creating hybrid humans, something akin to (but perhaps more interesting than) the cyborgs you read about in your favorite cultural studies seminar. Since we started using tools and organizing knowledge socially, we have always been hybrid, as our tools and communities have shaped us and become near animate parts of our minds. AI is an extension of that, a responsive tool that shapes us as we shape it. A knowledge community on a scale unthinkable a generation ago. It will change us utterly, but not make us any less or more human. It will make us otherly human.
We don’t know how we will fold together with it, how it will become a part of us. The braids are just forming, the co-mingling barely underway. Yet where the river will flow has yet to be decided. AI may help us build a society unlike anything we’ve managed so far, more just, more efficient and gentle with our natural resources, more leisurely and creative. It may enable our rapid self-destruction.
We need to pay attention to the small streams, and dream up something far stranger, more wonderful, and more hopeful, if we long for a different river to emerge than the dull science fictions that narrow AI conversations and recycle old paranoia.