What’s more important - a robot or a dog?

The classical laws of robotics do not answer this question. We’ll try to remedy this situation.

Isaac Asimov, a renowned twentieth century science-fiction writer, formulated three laws describing how smart robots should behave around humans. Today, instead of robots, we are surrounded by artificial intelligence. The laws that were designed for androids turned out to be very useful in understanding how AI should behave. However, during the last century, taking care of animals wasn’t that big a deal and mankind couldn’t quite decide how AI should behave, for example, with cats and dogs. Now the time has come and we just might have the solution.

An inspiring story of a Kaspersky employee

Kaspersky expert Oleg Zaitsev decided to create a "Smart Home" on his own. From an engineering point of view, it turned out to be a masterpiece, but it is neither accessible nor affordable to the average buyer. Artificial Intelligence controls hundreds of parameters inside and outside Oleg's cottage, chooses an optimal mode of operation for almost everything in the house, knows what to do in unusual circumstances and is able to maintain its viability in extreme situations. It seems that AI is truly running the house. This, however, is not quite the case. Oleg has a smartphone which operates the AI according to the bidding of the real home owner.

The “smart home” created by Oleg Zaitsev has another unusual feature – besides looking after humans, it also takes care of their pets. Oleg has plenty of animals at his home: fish, turtles, a dog and two cats. AI maintains the desired water level and temperature in the fish tanks, and monitors the conditions of the pumps, filters and lighting. And if it notices the dog or the cats strolling around in the yard, it will make sure not to hurt any of the animals when closing the garage door.

But Oleg’s engineering genius stumbled upon an unsolved ethical problem. What protocol should AI follow if it started to run out of resources – should it give the last of the heat to the turtles in the aquarium or to the person indoors? Oleg hopes that in such difficult circumstances AI will make an ethical decision. But what would happen if entire cities were run by AI? Would we have the time to deal with every such situation? Obviously, not. So, let us take a look the three laws of Isaac Asimov and try to apply them to our pets.

Kaspersky engineer

When I was in the process of creating my “Smart Home”, keeping in mind that I have several pets of my own, I tried to create the same level of comfort and safety for the animals, as one would for people.

Oleg Zaitsev,

Kaspersky expert:

First law

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

Our comment

It turns out that it is forbidden to cause harm to human beings. But what about animals? It would be good to prevent robots from harming all living creatures. However, since a significant part of humanity likes to eat certain animals, we could not, for example, use AI to fish. Therefore, we would need to make a distinction between the concept of "person/human" and "living creature".

! Our version of the law:

A robot cannot harm a living being without a human’s permission.

Second law

A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law

Our comment

Let’s say the robot is playing with a dog. Should it play along with the dog and submit to its will? Besides humans, can a robot actually take orders, for example, from chimpanzees? Definitely not. Being aware of the possible consequences, a person consciously empowers the machine. Animals, on the other hand, are not ready to take responsibility for their actions, therefore, it is impossible to give them power or control over robots. All algorithms and protocols of interaction between animals and robots should clearly be regulated by humans. Therefore, the second law of robotics is left almost unchanged.

! Our version:

A robot must only obey orders that are given to them by humans.

Third law

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Our comment

When a robot faces a tough choice it must sacrifice itself, not a human being. But can a robot sacrifice the life of an animal? Let’s say an eagle attacks a drone; can AI use force to protect it? The problem is that the term “animal” is very broad. We simply cannot avoid the fact that we would be dividing living beings into those that cannot be harmed without the permission of a person, and those which AI can sacrifice for the sake of self-preservation. This is a matter of vigorous public debate. However, to us it is quite clear what the Third Law should be.