Riding the AI Bicycle

There are certain phrases that we use without thinking much about their deeper meaning. How often have you said “it’s like riding a bike” without considering that the bicycle wasn’t invented until 1817. One has to wonder what, for example, Mozart might have said to indicate that a thing was easy to remember how to do even after a period of absence. Perhaps he would have said “it’s like playing the piano”, though that invention only dates to the early 1700s. It is not so long ago that saying something was “the greatest thing since sliced bread” implied only that it was the greatest invention of the last few decades, the automatic bread slicer itself only having been invented in 1912.

It is undeniable that the last small handful of years have seen a revolution in artificial intelligence that is astounding in its breadth, depth, and potential impact on how we do almost everything. Given this rapid pace of change, it is likewise not surprising that we, as a species, are still very uncertain how to make sense of the changes this technology is foisting upon us. The variety of opinions regarding how to deal with it runs the gamut from an unbridled embrace of all things AI to a naive hope that “this will all blow over” and fade into obscurity.

There is no crystal ball, not even an AI powered one, that can reveal where artificial intelligence will go in the next year, let alone the next decade or more. There is a strain of thought regarding AI that is based on a false premise, though, that is incredibly tempting. This premise, and the opinions that build upon it, are ultimately unhelpful and the more they are embraced the more we risk, as a species, ignoring the important decisions we must make. This is the premise that, because we humans live in “the real world” and because artificial intelligence is “artificial” and dependent upon technology, that dealing with artificial intelligence is optional.

Frequently you might hear those attracted by this premise suggest things like “touching grass” to indicate that, instead of dealing with AI, one can simply put down their phone, close their laptop, and forget about the existence of AI. While it is certainly a good idea not to stay stuck inside, in front of a screen, for all hours of the day, taking a break from technology is not the same thing as being able to deny its existence and increasing intrusion on our daily lives. When you are done “touching grass”, the reality is that AI will still be there and all the unpleasant, uncomfortable challenges that its advance presents remain.

To understand the foolhardiness of assuming that dealing with AI is optional, because it is dependent upon technology, we need to consider a glass of milk. We humans are mammals and therefore, by definition, the first food we consume as babies is milk. Contained within this milk is the sugar lactose, a sugar that takes its name from lactation, the act of producing milk. As babies we, like all mammals, can digest and derive nutrition from this milk. However unlike nearly all other mammals, humans are unique in that we don’t universally lose the ability to digest lactose when we are weaned from our mother’s milk.

Why are humans different in this regard? Quite simply, at some point in the distant past, ancient humans figured out that we could domesticate and raise other mammals that produce milk for their own young and that we could divert some of this milk into our diets in the form of cheese or a nice cool glass of milk. While farming and animal husbandry are not subjects that we typically include under the banner of “technology” in the modern age, they were the very definition of technology for ancient humans. Even the suggestion to “touch grass” betrays the fact that having grass to touch is a byproduct of agrarian technology.

What’s more, when it happens that a human does lose the ability to digest lactose, increasingly instead of suggesting they ignore the problem because it is related to technology, we have developed even more technology to allow these individuals to continue to participate in the fruits of dairying technology by developing forms of dairy without that core constituent or by providing alternative means for enabling lactose digestion in the form of technologically-produced lactase enzyme supplements. In other words, when it comes to the consumption of dairy, we humans, as a species, have actually evolved to anticipate the existence of agricultural technology.

There is another, more recent and somewhat more uncomfortable way in which humans are in the process of evolving in the anticipation of technology. To understand this example we have to consider the form of the pelvis, the realities of an upright gait, and what it means to rely on human intelligence as an evolutionary advantage. Nearly all mammals give live birth to their young, a process that involves newborn individuals passing through an opening in their mother’s pelvic bone. Evolution has, not surprisingly, selected for a concordance between the size of a newborn’s head (the head often being the largest feature on a newborn) and the opening in the pelvic bone. Unfortunately for humans, this push for concordance butts up against two competing evolutionary pressures. First, walking upright favors a more narrow opening in the pelvic bone. Second, high intelligence favors newborns with a larger head.

The net effect of these competing pressures on human evolution is that childbirth for humans is a much more difficult, even traumatic, process than it is for nearly all other mammals. Still, over millions of years of evolution a careful balance has been achieved by evolution wherein our pelvic bone structure is good, but (as any older person with lower back pain can tell you) not great for walking upright and our babies’ heads are large enough to get by (literally and figuratively). Then along came the practice of modern medicine and a technique known as a Cesarean Section. Humans have only been able to perform C-sections with sufficient skill that there is a high probability of both the mother and child surviving the procedure for the last century or so. As such, it is far too soon to say with any confidence whether the availability of C-sections is altering the balance evolution has struck, but the question even the possibility raises is unsettling. What if the availability of C-sections allowed humans to evolve larger heads to the point that birth without a C-section becomes impossible?

Of course, it is not necessary to idly speculate on the future possibility of a human race that is unable to survive, in the most basic sense, in the absence of technology. This is already the situation we find ourselves in. Our very day-to-day survival is dependent on one of the humanity’s (or possibly their predecessor’s) first technological advances: fire. It turns out that cooking food makes more of the calories and nutrients contained within the food available for processing by the body. The simple practice of cooking food was likely key to kicking off the runaway evolution of more and more intelligent humans, with larger, more complex, more calorie demanding brains. Without fire, and the cooking it enables, survival is essentially impossible¹.

So if, in some hypothetical future, it comes to pass that humans evolve to the point that birth without C-section is impossible, or that controlling technology without AI is impossible, it would not be the first or even second or third time such a technology-dependent transition in human evolution has occurred. Yes, there is still something deeply embedded in the human psyche that romanticizes the Robinson Crusoe or Castaway story of a human dropped into the middle of the wilderness with nothing but their own two hands and human ingenuity, figuring out how to survive in the absence of all technology or modern convenience. Don’t be fooled by this relic of our primitive ancestors carried through generations of evolution. Humans are unique on the Earth as the only species not only to develop technology, but also to depend on and to evolve in concert with it.

This is not to say that we should blindly stumble forward, accepting every AI-based takeover of any element of our day-to-day lives. Human history is littered with technological “advances” that we would have been better off without. Instead of focusing on questions of “AI or not”, we should spend our time thinking about “which AI, and why”. After all, while the expression is that “it’s just like riding a bike”, I’ve found that going long periods without driving a car does not diminish my ability to get right back behind the wheel. Still, I think the world might be a better place if we did more bike riding and less driving.