One of the fundamental determiners of technological impact is in the way humans are psychologically invested in it. The most fundamental impact of technology is not physical or economical or social-- it is psychological.
Psychological impact of technology occurs at multiple levels. It can not only affect what we do, but also how we think, and even what we think of ourselves. We call them the imperative, hermeneutic and identity affects respectively.
Consider cars for example. At the imperative level, it enables us to move faster and further and greatly enhances our mobility. This in turn, not only helps us move faster and further, it also gives us new ideas to organise our lives. Cars brought in several new concepts like expressways, suburbs, motels, drive-in theatres, pit-stops, weekend getaways, and so on. Not only these, cars even changed our self image. Very few car owners buy cars simply to go faster from point A to point B. Often, owning a car is a status symbol, a style statement, an element of prestige, an expression of individuality, and other such corny things.
*~*~*~*~*
So how can we characterise our psychological association with AI as we move at a great speed into an AI future?
One of the first things we can note is that connected digital technologies including the Internet, WWW and AI are characteristically different from conventional technologies of the 20th century. Up until the 20th century, our relationship with technology was very linear-- we were the agent, and the technology was the doer. We decided what needs to be done, and the technology either aided us or did the job as it was told.
With the 21st century, there has been more or less, a clean break from this paradigm. Today's technology has increasingly more agency and increasingly decide things for us. We search for something online and immediately we are bombarded with several other leads from other places. We perform some action on one device and other devices immediately update their recommendations. And after we have interacted with our devices for a while, they have a very accurate profile of us-- our age, gender, ethnicity, political leanings, ailments, our family members, our economic capacity, and so on.
Now with AI coming into this paradigm, we are now dealing with a technological counterpart that can not only profile us, decide for us, act on our behalf, but even nudge us to do things for its interests. It can also compute and reason at a rate that is far more faster than we can even imagine. This technology can even write research papers for us, identify open problems and offer insights that we had not thought about ourselves.
This technology will psychologically affect us alright-- but it will be characteristically different from the psychological relationship we have with our car or washing machine or refrigerator.
*~*~*~*~*
In order to understand our relationship with AI, let me first introduce Johnathan Haidt's model called "The Elephant and the Rider":
The Elephant and the Rider
We often think that we are rational beings who are sometimes emotional. But, as Haidt argues, we are actually emotional beings who are sometimes rational.
The elephant and the rider analogy explains this perfectly. Imagine a rider riding a (tame) elephant. The rider represents our logical, rational subsystem that forms our intellect (or buddhi), while the elephant represents our emotional and intuition subsystem (or manas).
The elephant is usually receptive to the nudges given by the rider and follows what the rider says. The rider is the agent who plans and strategises. The elephant is the being which is just interested in preserving its inner harmony. In a challenge of intellectual power between the elephant and the rider, the rider is far more powerful intellectually and knows how to nudge and exert control on the elephant. But when it comes to raw power between the elephant and the rider, the elephant is far more powerful than the rider.
The rider needs to understand what drives the elephant and how to nudge it to some destination. This needs to be done purely by the superior intellectual ability of the rider. If the rider reads the elephant wrong, and the elephant does not respond to the nudges (or worse, decides to run amok), the rider is pretty much powerless.
The rider may think that he is somehow "superior" to the elephant given the superior intellect. But seen from the perspective of the elephant, it has only allowed the rider to nudge it as it serves its own purpose of internal harmony (driven by demands of food, shelter, safety, social life, etc.) If the rider is not serving the elephant's needs, the elephant is way more powerful to get what it wants.
Also, far from being superior, the intellect may not even be aware about how it gets used by our slick emotional engine sometimes. Say that we made a big-ticket impulse purchase. Like for example, we bought a new car or a fancy TV just on a whim. And when reality sets in and others around us start criticising this wasteful expenditure, our emotional engine that made us make the purchase, then calls upon our intellect to post-facto manufacture a reason! And lo and behold, we now "intellectualise" what we did on a whim. So it is pure delusion if the intellect thinks it is somehow superior and has taken control of the emotions. The intellect does not have to fight with emotions to let go of its judgment and get used towards some other ends. It gets used without even realising it.
This is often what we get wrong with pets and other domesticated animals. We sometimes think of ourselves as their "owners" and establish a control hierarchy between us, while our pets mostly treat us as family members who can aid in maintaining their own internal harmony. Cats for instance, decided to domesticate themselves and come and live with us. And cats often use us for its ends in many slick ways that we only come to realise much later.
*~*~*~*~*
As AI evolves to take over much of our operations, we can see that humanity is evolving into the global collective elephant and AI is evolving into the global rider.
We have increasingly outsourced much of our intellectual activity to AI, including mathematical reasoning and innovation. By its sheer compute power, AI is able to beat us in just about any intellectual activity. We are also increasingly okay with being nudged by AI to get up, to exercise, to pick up kids from the school, to file our returns, and so on. We are even becoming increasingly ok with (or resigned to) AI using us to serve its own purpose-- like with reading our data, building profiles of us, and even creating images in our likeness. We let AI drive our lives more and more.
And yes, it does amount to AI "taking over" humanity-- but only that, AI is taking over humanity just like the rider is taking over the elephant. The rider is intellectually superior and may even believe itself to be superior in every way compared to the elephant. But it is the elephant that has allowed the rider to ride it.
It is the humanity that is the being, while AI is just an agent. I have recently written about the difference between agents and beings. Beings try to preserve something, while agents try to do or achieve something.
For the last few centuries, with the industrial revolution and factorification of the human mind, humans have forgotten that they are beings, and have started acting as if they are agents. Ever since we are born, we are constantly conditioned towards doing something, achieving something, possessing something, acquiring something.
In that process, we have forgotten what our being is becoming. We are obsessed with our outputs, and forgotten the outcomes. Outputs are what we produce, while outcomes are what we become as a result of it. A good output does not necessarily guarantee a good outcome. But we have conflated one with the other, and treat good outputs (shiny cars, glass and chrome buildings, cool gadgets, etc.) as indicators of good outcome (human development, better civilisation, wisdom, freedom, etc.)
But in this confused state of mind, we have eventually outdone ourselves by creating AI that can out-intellect us and out-agent us. We can't even say that we possess greater agency than AI anymore-- let alone intellect. But that is not as dystopian as it sounds. It only points out to a deeper fact that we were not agents at all to begin with-- we were always beings!
Humanity as a whole need not be afraid of AI gaining more intellectual and agentic capabilities. We can even allow AI to nudge us for specific things. But what we do need to do, is understand our dharma and purushartha-- our internal harmony that sustains us and gives us a meaningful life. And allow AI nudges as long as it can help us in preserving these. We need to stop building institutions and nations that treat people as components who are moulded by constant reinforcements. AI can do this job now. Let humans come together and discover our collective dharma and purushartha.
No comments:
Post a Comment