Engineering Artificial Relationships
We’re approaching a new digital age. Tools like Google Now, Siri, Cortana, and most recently Facebook M are becoming more and more common. The age of artificial intelligence is dawning and companies are scrambling to capitalize on machine learning and the ability to autonomously meet the demands of humans. As these tools become more ubiquitous in every day life a strange phenomena begins to occur. We are eager for these robots to be human-esque in their responses and behavior. We want to see how life-like and “real” each interaction feels. And yet, skepticism, doubt, and mistrust pervade. Why does this negative and deep-seated distrust occur and what causes these feelings to perpetuate? What is the root issue and can it be overcome?
As humans we are driven by several powerful forces, our logic, our intelligence, and our emotions. Logically we look for solutions to problems, our intelligence empowers us to make choices and decisions based on knowledge and memory, and everything we do is colored by our emotions. This intimate mixture is what makes us, as a species, unique. This blend is what forms the foundation of our id, our psyche, and our self-worth. We desire to make better choices, to gain more knowledge, and to establish relationships with others; this exists at the core of our being.
Logically we look for solutions to problems, our intelligence empowers us to make choices and decisions based on knowledge and memory, and everything we do is colored by our emotions.
Now, as we turn our attention to the problem of artificial intelligence and the desire to create a machine to respond as a human would we are faced with a glaring problem. Intelligence and knowledge can be accomplished rather easily thanks to the information age. Acquiring facts is not difficult. Logic is inherently machine-oriented processes and therefore can also be easily attained and implemented. The part where our quest for “AI perfection” fails comes from the third aspect of the human triad: emotions.
We want these machines to be human. We want to interact with them as we would interact with other humans. This is perfectly acceptable when dealing with logic and intelligence, and as a race we have no problem communicating with machines, interacting with them and carrying on a discussion. But ultimately, subconsciously, we want to form a relationship with them. Again, relationships form the core of human connection and naturally any interaction where we are attempting to mimic human behavior will also require a deeper relationship bond.
Herein lies the problem, we recognize emotions cannot be “faked” by a machine. We are skeptical of fake emotions and fake relationships. We are highly tuned to identify when another human is not being genuine with their emotions or feelings. When a false emotion is spotted or felt in a conversation the entire relationship is put in jeopardy. Why? Because as humans we immediately raise our guard in self-defense when we detect this insincerity. We are more alert, more observant, more cautious as a result. And if the situation is not resolved then the relationship dynamic changes fundamentally. And the single most important aspect of the relationship is destroyed: trust.
And now we arrive at the heart of the problem with artificial intelligence and machine learning, we can’t engineer an emotional relationship with a robot and thus no inherent trust is established.
In order for AI to become what has been portrayed in the movies and science fiction novels everywhere (man’s new best friend) we have to find a way to overcome the emotional aspects of building a relationship. And secondly, we have to establish some basis of trust. Once we are able to build a semblance of trust in these autonomous figures and overcome the emotional dynamic of a relationship we will be ready to enter this next digital age with confidence.
