Human of the Gaps
Gaps in AI capabilities are no more evidence of human exceptionalism than are gaps in science evidence of God
It is said that when the French mathematician Pierre-Simon Laplace presented his model of the nebular system to Napoleon Bonaparte, the latter asked “And where is God in your model?” To which Laplace is said to have replied, “Your Highness, I have no need of that hypothesis.”
The story is almost certainly apocryphal, but it conveys an important point: where we are ignorant we are inclined to invoke a higher power, a Deus-ex-Machina. Where we are certain, we have no need for such subterfuge.
The tendency to invoke God and other permutations of the supernatural when an observed phenomenon is shrouded in mystery is a tradition as old as humanity itself. The Ancients saw their deities in everything from pestilence to earthquakes. Isaac Newton wrote with concrete precision when discussing the motions of the celestial bodies and the mechanics of optics. He had deduced how those things worked and had no need for nebulous language. But when describing how an acorn becomes a great oak he fell back, like the rest of us mortals, on the majesty and mystery of the good Lord.
This habit is so practiced, so well-established that it’s earned a sobriquet of sorts: God of the Gaps. Fundamentally, we dislike not knowing. We need explanations. False ones, we seem to think, are better than none at all. More generally, we need explanatory models to draw order out of a disordered cosmos. It is a cognitive reflex and one that’s served us well in some respects.
Still, some take solace in the knowledge that the God of the Gaps is a well understood logical fallacy. In this light, it is either a relic of a bygone era or the purview of the scientifically illiterate . The rest of us know better.
But do we?
This same logical fallacy appears again and again in a slightly different context: the question of our humanity.
What is it that makes us uniquely human? What distinguishes us from other animals? And perhaps most pertinent of all, what sets us apart from Artificial Intelligence? These questions touch the egoist nerve within us all. In so doing they trigger a defensive riposte.
We humans are special. We dream. We write poetry. We imagine. We love and grieve. We use tools and have syntactical language. The boldest among us claim that there is a spirit, an immaterial, free-willed, instigator of thought and action. (Never mind that this same conception was used to justify the genocide and enslavement of “soulless savages” across the ages.)
Within each and every crevice in AI capabilities we squeeze in our humanity. The problem is, these little gaps don’t last very long.
Once upon a time it was said that human level artificial intelligence will have been demonstrated if a machine could beat a Grand Master at chess. Today, the dinkiest chess program can trounce the world’s number one. The goal post had to be moved. And then moved again. Then again, and again. We like to say only humans can do this or that. Only humans write poetry and philosophize, dream and imagine, love and grieve. In reality, these are all tentative goalposts, many of which have already been passed.
It’s crucial to recognize one simple fact: we are material. Humans are matter. We are stuff. There is no magic in us, there is no transcendent, supernatural, divine anything or other.
This is it. This is us.
And in being stuff, the things we do are manifestation of this stuff in action. Our hardware mediates everything we are, from the banal to the sublime.
When asked how a machine could ever do a thing that so far only we could do, the answer is always: well how does a human do it?
Picture a Venn Diagram representing the overlap of what a machine and a human can do. That Venn Diagram is not static. What humans can do is essentially biologically limited. It is a finite set of physical and cognitive capacities - speaking in sub-evolutionary timescales. However, what machines can do is dynamic. In fact, it’s dynamic in only one direction: it grows. It follows that those gaps in machine capability are likely temporary. It follows too that those who point to those gaps as proof of the irreproducible prowess of the human mind are destined to disappointment.
The Human of the Gaps is a logical fallacy as counterproductive as it’s theistic counterpart. The domain of uniquely human capacities is an ever shrinking one. In hanging our hopes there we are destined for disappointment. Charles Alfred Coulson said of the God of the Gaps, “Either God is in the whole of Nature, with no gaps, or He's not there at all.” Coulson was a devout Christian but he was closer to the truth than he realised.