Artificial Intelligence in Humans and Machines

"If the caudate nucleus plays an active role at inhibiting brain activity in the hippocampus, then people should be cautious when engaged in activities that stimulate the caudate nucleus, especially as people are getting older. Instead, healthy older adults should engage in activities that stimulate the hippocampus, such as mental mapping," says Dr. Bohbot”.

We need to change the environment in order to stay alive.

This is not what you think. I’m not talking about the climate change.

Humans are hacked by the reinforcement learning. There is a Frankenstein’s monster growing inside each of us. It’s artificially intelligent but not alive. People who believe that they can perpetuate themselves with their artificial copies will be disappointed. They won’t stay alive.

To stay alive (and to preserve ourselves as individuals) we need real intelligence. Fortunately, it was discovered after decades of dedicated work of generations of scientists. Their groundbreaking achievements culminated in the Nobel prize awarded in 2014 for the GPS system of the brain. They continue and expand their quest for the real intelligence, the implicit reasoning of the brain.

All living creatures evolve with two separate devices: GPS and autopilot. Only engineering implementation and technical capabilities differ but the purpose and the practice of usage remain intact for all except the civilized man.

Animals in the wild use the GPS device to extract structure out of the world around them, to set goals and make plans to achieve them, as well as to control and guide the execution of plans by the autopilot device.

The switch between the two devices and their integration can only be triggered by the environment but not intentionally or forcefully applied.

Animal pups and human kids alike use the GPS device to explore and structure the world around them and the autopilot device to satisfy their regular needs and wants. They operate both devices separately and mostly independently from each other.

During adolescence the integration of the two devices should take place as the brain matures. Teens need ambiguous environment to make the integration successful. Teens with underdeveloped GPS devices instead take risks and experiment with substances.

The GPS device in animals and humans is based on the implicit reasoning but not on human conscious logical reasoning. It applies statistical calculations and Bayesian reasoning to infer the structure from the uncertainty of the world and make predictions how that structure will evolve.

In mammals, including humans, the GPS device is built around the hippocampus and the autopilot device - around caudate nucleus.

The reinforcement learning was invented by humans during domestication of animals and other humans. It hijacks the autopilot device by cutting its connection with GPS device.The subject succumbs to external control instead of applying the internal one.

Addictions are the product of an uncontrolled autopilot. The reinforcement learning creates a window for addictions because it overrides the control by a person’s own GPS device but can’t substitute it by a total external control, though totalitarian regimes tried to do so more than once.

The milder but, probably, more hazardous way of manipulation is to exploit addictions of people instead of trying to control them directly. It is hard to name an area of human activity now which don’t exploit addictions. They all make the caudate nucleus grow at the expense of the hippocampus.

Aerobic physical exercises and 3D platformer video games show some promise as tools to make the hippocampus grow. Dr. Veronique Bohbot from McGill University is investigating the relationship of brain’s GPS and autopilot devices for decades. She suggests that people have to make mental mapping a habit to train their hippocampuses.

The problem remains that if the autopilot is responsible for habits than any activity to train the hippocampus will start training the caudate nucleus as soon as the activity becomes habitual.

Have we learned with reinforcement to behave like humans or is humanness an intellectual level beyond a certain threshold that enables human and humane behavior and can be secured only by the lifelong cognitive development? In the former case we are humans artificially created like machines. In the latter we have the naturally evolved human intelligence. What difference does it make?

Critters in the wild have no other choice but to navigate by themselves through the white waters of their short, unpredictable and dangerous lives. Civilization gave humans the choice to either navigate by themselves or just to go with the flow. The reinforcement mechanism, the artificial intelligence enabled civilized (or domesticated?) humans to build from themselves a very efficient and complex machine. The vast majority of its parts in fact had no choice but to go with the flow. Family, education, work, social life are all developing habits, expect and favor habitual behavior run by the autopilot device. The entire life became a habit for the vast majority of people. What for do they need the GPS device?

They can't operate properly without it, that it's for. We are not designed to be on autopilot all the time. The problem of lost self control manifests itself in the vast array of psychological disorders which have no obvious cue for psychologists who were trained ( with reinforcement learning, naturally) only how to replace bad habits with good habits and in most cases unsuccessfully or unsustainably.

Multibillion dollar industries of psychoanalysis, life coaching, mindfulness and alike are addressing milder conditions with attempts of reprogramming the autopilot instead of putting the GPS on.

So far, so good, but we, humans, have created such a sophisticated civilization that the autopilot is not enough any more to navigate its complexity. The information explosion is far from over. The mobile broadband internet is yet to become accessible and affordable for all. However, the number of autopilot reprogramming attempts increased exponentially. Broadband information flow requires to navigate it the fully operational GPS device that uses broad cognitive maps.

Maybe, we can take just one step back and return into the industrial age? It’s highly unlikely, especially after the emergence of artificial intelligence in machines that is taking place right now. What for should we need so much humans with artificial intelligence after machines will be able to perform all their jobs?

All living creatures use their internal GPS devices to navigate with Bayesian reasoning. Hebbian learning is the machine learning in a sense that it is used by all living creatures to automate patterns of behavior particularly suitable for the stable and well structured world. The GPS device powers the natural intelligence that is required to extract structure from the stochastic world and establish the limits within which the automatic behavior is applicable. The natural intelligence is the only difference between alive critters and the dead matter.

Only natural intelligence converts the variational free energy into information hence it resists the increase of the Shannon entropy and the Bolzman entropy through it. Life uses the natural intelligence to avoid decay. Reason is the only reason for life.

Artificial Intelligence has been existing already for a very long time but in humans, not in machines. It was developed by humans to successfully function in the artificial environment of civilization. Soon machines will learn how to use it. Humans in the meantime need to learn how to use natural intelligence anew because the artificial environment became so complex that the artificial intelligence is not enough to navigate it successfully.

Photo by Saksham Choudhary from Pexels

Curious about life and intelligence

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store