The Frankenstein complex and artificial intelligence

In the summer of 1816, a group of friends met to spend the season at a farm near Lake Geneva in Switzerland.

They were very rainy days, so to liven up the boredom of the confinement, a horror story contest is offered to visitors.

There were literary professionals in the group, so the competition was difficult to guess.

However, the story that terrified them all - let's assume during a stormy night - did not come from the established writers.

Its author: a young woman barely 18 years old remembered today as Mary Shelley;

The story of her was titled 'Frankenstein;

or, The Modern Prometheus'.

The cinema put together its own folklore by manipulating Shelley's original idea.

The popular version revolves around a crazed scientist (Victor Frankenstein) who brings to life a clumsy and evil monster, with an ending engulfed in flames.

It is the monster that is identified today with the name Frankenstein.

Shelley's original account is much richer and more tragic, of course.

There will be the young science student Victor Frankenstein obsessing with reanimating dead tissue, there will be 'the Creature' as a result.

Despite its torrid birth and escape, the Creature will be by nature good and noble;

on his own he will learn to read, speak and philosophize, he will aspire to love humanity, but when he gets closer that love will not be reciprocated.

The Creature will be feared, hated and despised for its appearance.

She will be denied any right to her pursuit of happiness and now, filled with spite, what she will seek is revenge.

The monster was everyone's doing.

What remains is a hunt to the ends of the Earth.

There is an element in common between the novel and the popular version: the danger of creation out of control.

Isaac Asimov in his stories explored the fear of such a thing happening in the case of his robots and named it "the Frankenstein complex."

That fear was a commercial threat to the manufacturers of these artificial beings.

The answer was the so-called "Three Laws of Robotics": ethical rules implanted in robots to guarantee the safety of humans.

This is not alien to us;

For today's airline industry, business critically depends on security.

In today's reality we see elements of the science fiction of the past although with less drama.

We have autonomous vehicles, drones, industrial and home robots that incorporate more and more -very limited- artificial intelligence techniques.

These models themselves are also products: applications that process information in human language, music or images.

Creations that interact with us and that begin to show ethical problems.

Each failure or misbehavior that these advances manifest -call it ChatGPT, for example- feeds the Frankenstein complex.

The designers of these avant-garde techniques must resort to ethical guarantees if they want them to be accepted by society.

Ethical behavior and the biases that are induced are the children of ethics and the biases of designers and of society as a whole.

Artificial intelligence models have no experience of their own;

they train with samples of the real world in which we have been leaving the sublime but also the monstrous of us.

Where is it appropriate for designers to put inhibitory brakes?

At the entrance, the exit or both ends?

Today combinations are feverishly tested in the laboratory like Victor Frankenstein: ChatGPT writes poems on request in favor of certain irrevindicable figures, but refuses in other cases.

Like the novel, together we can create a useful tool or a monster.

We can write a great horror story or a virtuous one.

There is the mutual challenge that awaits us, like those restless people in the stormy summer of 1816.

Sandra Pitta is a Pharmacist, Biotechnologist.

CONICET researcher.

Damián Gulich is a Physicist, Full Professor at the University of the City of Buenos Aires and Adjunct at UNLP, CONICET researcher