Difficulties in understanding
If we take as a starting point the opinion of the Bauman Moscow State Technical University, one of the most authoritative scientific institutions engaged in the development of robotic systems, including for defense purposes, it turns out that there are at least ten (!) Different understandings of the term “robot”. And that's not counting the classic definition by Harry Domine, CEO of Rossum Universal Robots, who proclaimed that robots are technical devices that reproduce human actions. Moreover, they must have systems for receiving, converting energy and information.
To be precise, this term belongs to the Czech writer Karel Czapek, who coined the character Domina for the play "RUR" back in 1920. It is important that initially all robots had to be necessarily smart and anthropomorphic, that is, similar to humans. Webster's English Dictionary very clearly in this regard characterizes a robot as an automatic device resembling a human shape and performing functions normally inherent in a person or a machine. And it is not difficult to find a decent job for such a technique - to replace a soldier on the battlefield or, in extreme cases, to become a personal guard. A typical example of an ideal combat robot is the protagonist of the following video:
This, of course, is a skillfully shot parody that refers us to the modest achievements of Boston Dynamics, whose products so far can only do this:
Or like this:
In general, human-like (or dog-like) robots that are now widespread in the world are still very far from the classical understanding of the Czech term "robot". And the Boston Dynamics products, as it has now become clear, are not particularly needed by the buyers - the equipment for the most part remains in the status of a technology demonstrator.
But back to the problem of identifying robots. After Čapek, such devices were treated as
"Automatic machines, including a reprogrammable control device and other technical means that ensure the performance of certain actions inherent in a person in the course of his labor activity."
A very broad definition! In this way, even a washing machine can be ranked as a robot, not to mention complex industrial manipulators such as KUKA.
So are robots or manipulators? In foreign technical literature, everything is mixed: robots are called
"A reprogrammable multifunctional manipulator designed to move materials, parts, tools or specialized devices through a variety of programmable movements to perform a wide range of tasks."
Not mentioning the rudiments of artificial intelligence, autonomy and self-learning, which are now being spoken about from almost every iron. Much more complicated and, as it seems, closer to the truth, the following definition of the concept of "robot":
"A programmable autonomous machine capable of moving objects along a path with a large number of points."
Moreover, the number and characteristics of these points should be easily and quickly changed by reprogramming; the working cycle of the machine must start and continue depending on external signals without human intervention. This, by the way, is very similar to the robotic autopilot systems of cars, which will be discussed below. The engineers and researchers of the M. V. N. E. Bauman stopped (at least for now) on the following cumbersome definition of a robot:
"A universal reprogramming or self-learning machine, controlled by an operator, or acting automatically, designed to perform a variety of tasks instead of a person, as a rule, in a priori unknown conditions."
Have you read it? It is clear that MSTU rightly decided not to complicate their work and simply mixed robots and industrial manipulators with their strictly "learned" actions, Lego Mindstorms school sets, and artificial intelligence systems, used, for example, in court proceedings in the United States.
There is a simpler, but no less paradoxical definition:
"A robot is a mechanism, system or program that perceives, thinks, acts and communicates."
Again, with the modern development of the Internet of Things, when refrigerators already know how to think in their own way no worse than cell phones, a lot of gadgets are suitable for this concept of a robot. Further study of robotic scholasticism leads us to options such as
"A robot is an artifact that functions autonomously."
Here, even a helium-filled balloon fits the description of a robot. Or like this:
"A robot is a machine (more precisely, an 'automaton') whose behavior looks reasonable."
The helplessness of this formulation is obvious. For each person, the criterion of rationality is different. For one individual, a newfangled crossover, automatically slowing down in front of a child who ran out onto the road, is already the height of rationality, especially if this child of his escaped. And for the second, even the automatic landing of the "Buran" will not create the impression of rationality. It seems that even the classic saying of the American engineer and inventor Joseph Engelberger (1925-2015), who is often called the "father of robotics", is gradually losing its meaning:
"I can't define a robot, but I will definitely recognize it when I see it."
With such a vague term, Engelberg would not recognize modern robots - they simply become indistinguishable from "non-robots."
Who to blame
Actually, due to such confusion about robots in the modern world, it seems that they do not know what to do with them in the future. No, of course, regarding various smart gadgets that simplify our life, everything is clear: here they seriously and not long captured our future. But answer yourself honestly: would you buy yourself a plane ticket that has no pilots? Imagine, an aircraft with several hundred passengers is controlled autonomously for most of the route, and only during takeoff / landing, operators from the ground take on the role of pilots. At present, technology allows it, but public opinion does not allow it. Just as it does not allow the introduction of full-fledged automation of road transport management. And there are conditions for that. Parts of the A9 Berlin - Munich highway were re-equipped several years ago for autonomous cars of the fourth and even fifth level of automation several years ago. That is, on this autobahn, a suitably equipped car can move completely automatically - the driver can just sleep or talk peacefully with fellow travelers. And, by the way, outwardly such a robotic car will differ little from a car in the classical sense. Why don't we implement it? The whole problem is responsibility for the outcome of possible accidents both on the ground and in the air. Think of the noise caused by the fatal accidents of the unmanned Uber and the autonomous Tesla. It would seem that thousands die on the roads every hour around the world, but death from artificial intelligence is perceived especially acutely. At the same time, public opinion does not want to hear that even a partial introduction of unmanned vehicles will save thousands of lives. Society cannot get along with the idea that the notorious "trolley problem" will be solved not by a person, but by an artificial mind.
What is the crux of the problem? Philip Foote, a British philosopher, formulated it back in 1967, much earlier than the advent of drones:
“A heavy, uncontrollable trolley rushes along the rails. On its way there are five people tied to the rails by a mad philosopher. Fortunately, you can switch the switch - and then the trolley will go in a different way, a side track. Unfortunately, there is one person on the siding, also tied to the rails. What are your actions?"
In solving such problems, you can rely on public opinion, as was done in the Russian Cognitive Technologies, when in 2015 they were working on the project of an autonomous KamAZ. The respondents were offered test tasks "What should an unmanned vehicle do?" with several solutions. As a result, moral recommendations were developed for the algorithms of future unmanned vehicles. But there is one catch: only 80 thousand people from Russia took part in the survey, which is only about 0.05% of the country's population. This part of society will decide who will live and who will die?
Taken together, this is precisely why, despite the inevitability of a robotic future, we do not even know approximately what it will be. And largely due to the fact that we have no idea what a robot is!
