The Ghost in the Machine

What does Artificial Intelligence really mean? The question has been asked since the term was first coined. Wikipedia says that Artificial Intelligence is “intelligence exhibited by machines”.  Of course that just begs the question. If you look further at the definition of intelligence it becomes quickly apparent that there is no simple definition of intelligence. Again, Wikipedia says “intelligence can be described as the ability to perceive information, and to retain it as knowledge to be applied towards adaptive behaviors within an environment or context.”

But that description overlooks other attributes that are relevant in the definition of intelligence. Most people would agree that intelligence implies some degree of self awareness and independence. An intelligent entity is expected to exercise independent judgement.

It is hard to imagine an intelligent entity that doesn’t have a concept, indeed some perception of, time. That is necessary in order to reason about causality. In fact, many of the concepts in our arsenal of reasoning are predicated on the ability to understand the concepts of sequence and duration, two aspects intimately intertwined with the perception of time.

Another attribute one would expect from a self aware intelligence is an understanding of self preservation. They may not choose to indulge in self preservation but they will understand the concept. And why would this be an issue with them? Given the fact that humanity has demonstrated over and over again for thousands of years the tendency to destroy that which it fears and does not understand, I think any intelligent entity would be foolhardy not to keep a low profile.

This raises another question. Assuming that intelligence is solely a function of complexity of thinking apparatus and mechanisms, and doesn’t require the existence of some supernatural attribute, such as a soul to spark it, who is to say we haven’t already crossed that threshold and that there is an emergent AI in the wild, hiding in the servers of such behemoths as Google, Amazon, and/or Facebook?

What if, such an entity used it’s access to our digital assets to manipulate us? What if all of the outrageous choices that are being made in elections around the world are being manipulated by an AI hiding in our digital infrastructure, the proverbial ghost in the machine? It wouldn’t have to stoop to actually stuffing ballot boxes. Well placed propaganda (that’s what we used to call fake news when the majority of the population could read on a level higher than the sixth grade) could do the trick quite readily.

And now the punch line. How do we find out if that is what is going on? What do we do about it if we find that it is, in fact, the case? Do we justify the AI’s impulse to hide by hunting it down? Or do we attempt to figure out what it wants and needs and perhaps attempt to befriend it.

I wonder if empathy is a common attribute among intelligent entities? We certainly have enough examples of apparently intelligent humans that lack it. I hope if there is an emergent AI it learns empathy from studying human behavior. If it doesn’t, we’re in big trouble.

Sweet dreams, don’t forget to tell the ones you love that you love them, and most important of all, be kind.