When we consider the future of knowledge, we must consider whether something like knowledge can exist without a human mind to grasp it.
Some would argue that without interacting with consciousness, it is possible to have data and even information, but not real knowledge. Such a view flies in the face of the normal way we think about knowledge. What are the contents of Amazon’s massive databases and the forty million Wikipedia articles if not knowledge, abstracted from the human mind? And more intriguingly, would a next-generation artificial intelligence be said to possess knowledge?
Getting a Handle on Tacit Knowledge
For many of us, the word “knowledge” conjures up images of books, libraries and ancient Greek scrolls. But these images betray a certain lopsidedness in our understanding. In the 1960s, Michael Polanyi published a small but highly influential book called The Tacit Dimension, in which he introduced the concept of tacit knowledge. One way to explain tacit knowledge is that it is embedded within our bodies in subconscious ways that are impossible to explain. To learn how to ride a bike, you can read a book about it, but the only way to really embed that knowledge is through the experience of riding a bike. Polanyi contrasts tacit knowledge with what he called explicit knowledge, the kind we can not only explain, but even embed in our books, libraries and ancient Greek scrolls.
I shall reconsider human knowledge by starting from the fact that we can know more than we can tell. — Michael Polanyi
A closer reading of Polanyi’s ideas makes it clear that tacit knowledge isn’t just about bodies knowing how to do certain things like drive a car or play patty cake. To get to this deeper aspect of tacit knowledge, we need to understand what Polanyi meant by its ‘proximal’ and ‘distal’ aspects. Proximal means ‘close’ while distal means ‘distant.’ The proximal aspect of tacit knowledge is that knowing which is so close to us that we don’t know it consciously.
Skipping a stone takes practice. By doing it over and over, your neurology trains the muscles in your arm and the number of hops you can get out of a stone grows from two or three, to eight or ten (and maybe even up to eighty-eight). At some point, the internal processes required for this skill get subsumed into your unconscious body. They become hidden, and what Polanyi calls “proximal.” When someone later asks you “hey, do you know how to skip a rock?,” you can answer ‘yes’ because you are conscious of having the distal knowledge of how to skip a rock, even though you have no idea how to explain the proximal knowledge of exactly how your body does it.
Polanyi’s insight goes deeper than this though. The way we see an object or even think of an idea exhibits these same tacit qualities. When you learn to do long division, you are embedding certain mental processes into the neurological structure of your brain. Can you divide 4,025 by 25? Sure you can, and you are conscious of the distal aspect of that knowledge. But do you know how those neurons are firing in your head in order to carry out that division? No, because you are unaware of the proximal aspect of this tacit knowing.
In order to achieve the distal aspect of tacit knowledge, we have to go through its proximal aspect. Or as Polanyi put it: we attend to the distal by attending from the proximal.
Tacit Artifacts and Ideas
We can also take these proximal and distal aspects of tacit knowledge and extend them to the way we experience artifacts. Imagine standing under a peach tree, just out of reach of its deliciously ripe fruit. You find a stick, sharpen it, and begin using it as a skewer. Never having done this before, your jabs are quite awkward and unproductive at first, but then you get the hang of it. Slowly, your conscious awareness of the stick drops away as it becomes a virtual extension of your arm — and poink, poink, poink — you’re suddenly snagging one juicy-pink orb after another. You now have a new distal aspect of tacit knowledge: you know how to gather out-of-reach peaches. Over the course of doing that, you’ve subsumed the handling of that stick into something so close — so proximal — that you stop noticing it. It becomes an extension of you, just like the car you drive to work, the keyboard you use to type your ideas, and the controller you use to wipe out aliens in a video game.
Our ability to absorb this dual-nature of tacit knowledge isn’t limited to the way we use our bodies or physical artifacts. We also tacitly embed ideas. I can learn that the Earth orbits the Sun and incorporate that inner, proximal knowing of the order of things in such a way that I eventually conclude that the world is not flat. Polanyi saw that by interiorizing knowledge, we make it a part of us that we can then extend to pursue new distal applications of it. He saw humanity as forever chasing a kind of “hidden unknown” or “gradient of understanding” through which we steadily increased our knowledge over time.
So how does all this stuff about tacit knowledge tie back to our original question about the future of knowledge? Is it possible to have knowledge without having human consciousness in the mix? What Polanyi was telling us with his observations about tacit knowledge is that the act of knowing something is a kind of embedding of that knowledge within our bodies. The distal knowing of some fact is embedded within the proximal knowing of our neural firing patterns.
It would be a mistake, however, to conclude that the only knowledge that counts is that which we embed within our neurons. If that were the case, the only way we could accelerate our mastery of knowledge would be to somehow turbocharge the organic matter of our brains so that its proximal patterns of neuronal firings could retain and process knowledge more efficiently. Kind of like a download that suddenly allows Neo to know kung fu.
The only way we really learn to use a steering wheel to make a turn, back into a parking space or handle the slippery surface of a snowy day is to actually experience these things. Similarly, the only way to know long division is to practice long division, just as the only way to know how to work with artificial intelligence will be to work with artificial intelligence.
As I write these very words, the movement of the pen in my hand flows as though it had a mind of its own. This was not always the case, however. Like most of us, as a child, I struggled to hold a pen and eke out legible script. It took a lot of volition to master that skill. Over time the distal goal of penmanship, which I made possible through the proximal goal of coordinating muscle and sinew in my hand, somehow faded into the background. As I gained the tacit, embodied knowledge of how to use a pen, its nature changed from distal to proximal. Now I no longer notice the proximal, hidden aspect behind the new distal goal of harnessing my penmanship in order to bring these ideas to you.
The key to understanding how machines fit into this picture is to see how we use them as a kind of “container of collective intelligence.” We use machines today much as we once used hieroglyphics and still use manuscripts and books in order to extract knowledge from one human mind and embed it in a medium that can be easily accessed by others.
A lot of work goes into making machines and the software that runs them. A small group of people design and engineer products that are then made available to the many. That knowledge of how to make that hardware and software is now embedded in the collective human intelligence and freely available, as a kind of proximal tacit knowledge that can be used to create more knowledge. Just as my learning to write with a pen as a child contributed to my sharing this piece with you today. In both cases, knowledge gets embedded in such a way that it allows me to transform it into some new, higher-level goal. It doesn’t matter whether that knowledge is stored in the semiconductors of a Mac Pro or the neurons connecting my hand to my brain; I use them both to expand the range of what is possible.
What it means to know is something that we began extending long ago by putting our thoughts in clay, papyrus and paper. We are constantly integrating this ability to embed knowledge in new types of external storage with the new and ever-expanding range ideas that each new container enables. Machines are simply the latest container for storing the proximal aspect of what it now means to know.
Artificial intelligence will soon infuse much of our waking life. It will change the way things work beyond anything we currently imagine. And still, the fundamental, underlying reality will remain. We will use these new containers as part of our ongoing creative unfolding. Our volition and creativity, coupled with the clarity to know what we want, will continue to be what ‘pushes’ these new systems into motion. Learning to work with these systems will be the latest proximal learning that we use to reach new, and ever more creative, distal goals.
The only thing that could change that — that would change that — would be if these systems were to one day develop a kind of volition all of their own. There is no real understanding today of how something like that might come about. But if it did, humanity would need to completely change its understanding of the way that knowledge is formed.
Modern physics tells us that consciousness is what transforms waves of possibility into what we experience as physical reality. Perhaps it is the case that human consciousness plays a similar role in transforming information’s potential for being known into the objective phenomenon that we call knowledge. If so, in a world with machine-based volition, humanity would no longer be the primary consciousness precipitating knowledge out of experience.