Will AI create the next social divide?

When AI becomes mainstream

Imagine a world where AI is ubiquitous. We have cars that drive themselves and you to places, where your refrigerator is smart enough to place orders of local food for you knowing what you like, and so on. It does sound like an utopian world for the consumer.

Consumer benefits are typically the popular news but some of the business benefits can be more far reaching. Let us consider an AI application in the area of driver-less automobiles. When driver-less automobiles become mainstream, there is an additional benefit for the society namely, The movement of traffic on the roads will be smoother.

Understanding traffic patterns

If you have driven on a crowded freeway but everyone is going at a good speed and suddenly there is a slowdown to a crawl and then you move along slowly and then the traffic start to pickup and then in a few minutes everyone is going full speed. You have just experienced what is called a Phantom Jam.

The scientific explanation for a phantom jam is that when a stream of cars are going steadily on a highway, even if one car decides to slow down slightly, the car behind them has to slowdown as well and typically over-compensating to maintain a safe distance between the two cars thus the slowdown spreads more and more downstream, while intensifying thus causing a phantom jam.

Now let us take a more extensive example of this phenomenon. If there is an accident say a mile ahead and all the lanes are blocked, now the first set of cars that are stopped is not visible a mile down the road. Now as each set of drivers notice the rapid slowdown and are slamming on the brakes the slowdown rapidly propagates down the freeway. The speed of its propagation tends to be quite high while the intensity of slowdown is dropping as one observes the phenomenon on say an overpass over the freeway.

The first set of drivers who see the accident are slamming on their brakes to bring their vehicles to halt while the ones behind them seeing the red light also start breaking even earlier because they see the rapid break lights and realize something is wrong. Thus the information is propagating backwards that there is a slowdown at the speed of (brake) light causing more drivers to start the slowdown process even before they realize the car ahead of them slowing down. Thus they don’t have to slam on their breaks to slowdown and can gradually decelerate based on the feedback that is propagating downstream.

How would AI solve this?

Now if automobiles where self-driven, instead of looking at each car as an individual unit that is only reacting to traffic and obstacles around it, the entire columns of cars can be made into a single homogeneous organism that is fed by traffic situation information ahead of them allowing them to react as a collective thus allowing them to slowdown together automatically. Further, we can evolve AI algorithms that go beyond a single car and manage the entire organism can start re-routing a group of them through another exit on the expressway. This kind of “smoothening” of the flow of traffic can be akin to laminar flows in fluid dynamics. There is quite a bit of deep research in the area of fluid dynamics that can be applied to construct algorithms to manage traffic patterns when they can be controlled enmass through real-time traffic inputs.

The net result of being able to look at, shape and manage entire traffic stream on freeways allows us to improve the traffic flows on various roads. Thus AI applied at solving transportation as a problem and not just for driving cars can help improve commutation of people and goods on roads thereby allowing us to extract more value from existing road infrastructure rather than building expensive new infrastructure to support the economic growth. This improvement in business outcomes will naturally translate as a business benefit to the end customer both as time and cost benefits. Thus society as a whole get a new advantage that was not available before AI driven automobiles existed.

Adopting AI and its impact

This social benefit is going to be available where driver-less automobiles can be easily adopted. What happens to countries where there are roadblocks to driver-less automobiles? There are clearly

That will come in the way of adopting driverless automobiles. With such roadblocks obviously some countries will be ahead on the AI adoption race than the others. These countries automatically get the societal advantage that AI provides.

Emergence of a new “third world” ?

What happens to the countries that don’t adopt AI based solutions? The driverless automobile benefits to society is a single example. AI is permeating so many industries and even very critical ones such as medicine, doctors, finance, education etc., What happens to countries and societies that refuse the adoption? Do they become the next generation’s “third world”? Would their cultural / regulatory challenges be in the way of their growth as a modern society? Will a similar AI adoption change the economics of a country?

If we look at history we see a few examples in the past in terms of adoption / availability of Internet and the benefits that countries such as Africa have seen from Internet penetration, AI adoption or the lack of it will have an impact on each country. Whether they become the next “third world” only time will tell.