DEV Community

Cover image for Programming Emotion and Approximating Human Life
Jose Maria Iriarte
Jose Maria Iriarte

Posted on

Programming Emotion and Approximating Human Life

This article explores the intersection of robotics, artificial intelligence, biology, and psychology to examine the potential for creating artificial beings with human-like traits, including emotions and personalities. It discusses how advances in programming, mathematical modeling, and interdisciplinary science could enable machines to emulate human feelings by demystifying emotions as logical and programmable phenomena. While contemplating the uncanny valley and the philosophical implications of artificially engineered humans, the article highlights the profound distinction between creating human-like beings and replicating the soul, proposing that this elusive essence may remain the defining boundary between humanity and its creations.


Can men become Gods by virtue of their knowledge?

Advances in robotics and artificial intelligence have given birth to mechanical creatures able to perform sophisticated calculations that allow them to traverse obstacle filled environments autonomously with remarkable speed. In the field of cell biology cellular material can now be programmed to spawn and develop into any human part or organ.

Such advances should have us pondering the day we'll be able to create human life by means of combined computer programming and biological engineering. The mere thought may be off-putting, but the birth of semi-synthetic human life is approaching.

Certainly, in robotics we have seen the creation of the shells of beings that look deceptively human. Contemplating these carcasses alone can be an unsettling experience. This is the notion of the uncanny valley: A dip in the spectrum of our perception that produces in us revulsion instead of acceptance, when we observe an anthropomorphic figure that looks very human, but falls short of it by enough.

Surely, those shells can be perfected to look exactly like humans. But computers can never be human... Or can they?

No matter how much they may act like us, talk like us, and look like us, they are not, nor will they ever be us: Computers don't feel. They can't. Emotion is as foreign to artificially intelligent robots, as absolute mathematical perfection to human beings.

Programmed Emotion

But we can program machines to feel, for there is nothing esoteric about feelings and emotions other than our own ignorance and reluctance to engage them scientifically. Feelings and emotional reactions are no less logical in nature than the cognitive act of reading a sentence, and no less programmable than the mechanical act of walking.

Emotions have escaped the rationality dominated scientific community, not because they are any less logical than reason, but rather, quite plausibly, because the scientific community has had an inherent cultural and personality bias against it, and also because the very nature of scientific inquiry requires a degree of concreteness and mesurability often dissociated from common conceptions of what feelings and emotions are.

In Game Theory there already are interesting models that describe and accurately approximate the making of decisions, the framing of problems, and the resolution of conflicts. In Neurobiology similar models exist to determine the measurable impact of emotions in our organisms.

Psychological models for understanding personality and emotions such as those that have developed based on the foundational work of Carl Jung are less conspicuous and lend themselves less to interdisciplinarity, yet they do exist, and are mathematically sound, symmetrical in design, and eerily accurate.

But how could emotions be reduced to a model? After all feelings are subjective and context sensitive, and in a universe of infinitely unique individuals and circumstances, sheer mathematics demands that none of us can be exactly alike.

Modeling Emotion and Personality

Consider this: Anna is an extrovert. As such, she feels energized socializing. For this reason she has a wide group of friends and can often be seen in parties. Being surrounded by people most of the time recharges her, makes her happy; being alone for excessive amounts of time, drains her, and takes a toll on her mood.

What is the relationship between time spent in a crowd and the intensity of her happiness? How much alone-time is too much for her? By means of observation and quantifiable self-assessment we can arrive at such figures - holding all other variables constant.

And although Anna is unique, her preference for extroversion is not. This preference, predominant to varying degrees in roughly half of the human race, carries certain emotional, behavioral substance that is measurable and consistent. Knowing someone is an extrovert, we can predict a lot about their behavior, the effect of particular external stimuli, their internal emotional state, and the organic effects of such behavior and emotions, ceteris paribus.

Ceteris paribus is a large assumption. The field of economics and social science at large is often attacked for abusing the condition to model phenomena. Suffice it to say that there are established personality and emotion psychology models that bundle traits and make predictions - unsettlingly accurate to the untrained eye - based on the shared combinations of these traits across individuals. All humans are unique, but given a large sample and the right body of knowledge, just a cross section of these individuals can reveal a large number of significant behavioral and emotional patterns.

If we also mathematically account for the impact of environmental factors influencing individuals, we can model their behavior, internal emotional states, and organic chemical reactions to those states.

What then but time and toil holds us back from designing equations to model personality, formulas to describe behavior based on those personalities, and algorithms to program emotional states?

What then holds us back from creating computers that feel?

After all, the ordinary notion that says emotion and logic are mutually exclusive is spurious. Individuals acting on emotions do behave logically, clearly not acting solely on sensorially collected facts, but also on their given emotional state. It would not be logical for an individual feeling the biological chemical reactions triggered by an outburst of rage to behave in a calm and collected manner.

But if programs based on mathematical models can be made to produce such outbursts of anger, joy, sadness, fear - if they can account for probabilistic patterns of how we will behave given external and internal stimuli in a fashion congruent with a personality, computers can be made to feel and be imbued with yet another layer of humanity.

What is Human?

When robotics, biology, cognitive science, personality psychology and computer science arrive at a confluence, we will be able to give computers bodies, intelligence and full fledged human personalities and emotions, leaving outside the realm of our creation only the soul itself, whose existence, subject to faith alone, can never be accounted for by empiricism, being neither material nor verifiable by our senses, and irreplicable by means of science.

As time passes, our knowledge evolves and interdisciplinary practice becomes the norm, humans engineered by humans may come to life. They will be everything we are, except they will be created artificially, by means of our sheer knowledge and understanding of mathematics, science and the human condition. They will also, however, not have souls; at least not deliberately, at least not by design. If there ever is an uncanny valley in them, this missing piece might just be the only thing to account for it.

Top comments (0)