Artificial intelligence is not what it seems.
Today, artificial intelligence (AI) is a faithful human assistant. But many science fiction writers, scientists, and the general public are wondering: if it becomes smarter than a person tomorrow, what then?
Actually, nothing. There is a snag in the word “smarter.” So far, modern computers are fundamentally no different from their ancestors on punched cards, where “zero” is the absence of a signal, “one” is the presence of a signal. The computer is not able to solve several tasks in parallel – only sequentially, through the interrupt system.
As a result, no matter what tasks are put before the computer – the pattern recognition system or the game of chess. Anyway, it all comes down to a mechanical search of options and the selection of the optimal one. Most importantly, it all comes down to an optimal option in terms of the algorithm and the objective function set by the user. In general, you cannot compare the intelligence of a person and machine. These are fundamentally different processes.
By the way, the machine computing power is greatly exaggerated. Sure, a person cannot compete with a machine in calculating. What absorbs the engine resource mostly? Any gamer will tell you – video processing. Honestly, a person hasn’t any problems with this. The processing and analysis of video information by humans still exceed the capabilities of the machine. Moreover, if you take into account that both hearing, olfactory, tactile information, and coordination of movements are processed in parallel online – there are no doubts.
Pattern recognition is a difficult task for a machine. Even if we are talking about, for example, an ordinary glass. A computer will confuse a glass with a pot or simply with a truncated cone. The reason is that the connection of the image with the object functional purpose behind the image is an extremely difficult task for the machine. So, how can we compare apples with oranges?
Here we should discuss an important point. Speaking about AI, the average person intuitively uses not “intelligence”, but “brains”. Honestly, these are very different things. Intelligence is the ability to optionally solve tasks. What about brains? There is still no clear definition. In order to define artificial intelligence term, computer science utilizes “strong artificial intelligence” term. Thus, it removes the term “mind” from consideration.
Indeed, it is believed that the sign of the mind is consciousness (another vague term, most often implying the presence of abstract thinking), self-awareness (as a reflection), self-determination, ability to learn and self-study, rational behaviour, and intuition.
Actually, all this is quite applicable, for example, to a domestic cat: of course, we do not know how she realizes herself (and whether she is reflexive), because of the lack of a speech centre. However, even deaf people are well aware of themselves without using this centre. Even better is the case with collective insects: ants, termites, bees. There are a difficult language (bees do not have a brain, only ganglia, but they can tell where the best flower stalks are located), a complex hierarchy, ideally adapted to the system environmental conditions. If we consider a separate individual as a kind of “transistor”, then the collective mind of the hive, of course, takes place. However, it is not human.
I would say that the main feature of the human mind is the will for irrational actions. For example, the conquest of space as the most vivid example. Attempt to go beyond the habitat – what could be more unreasonable? Or search for “brothers in mind”. It is both unreasonable and dangerous.
Now, let’s come back to artificial intelligence. This is a tool and a useful stuck. Strong artificial intelligence (brain) is still very hypothetical. Some are inclined to see the possibility of its occurrence with the advent of quantum computers, but this is not all that simple. Will it be the mind of a conditional cat, a conditional anthill (the embodiment of rationality), a conditional person or a conditional something? No answer.
There are still lovers to speculate about the upcoming synthesis of the human mind and artificial intelligence.
So far no biological synthesis is expected. It is possible only in the case of creating AI on a biological basis, that is, not on the simulation of neural networks. It is based on the cultivation of neural networks on DNA with simultaneous programming. I’ve faced this fiction.
It is connected with biological synthesis.
However, social synthesis has already happened. Computers compose music and play games, are engaged in recruiting (staff recruitment) and analyzing our tomograms with the issuance of diagnoses and treatment recommendations, predicting consumer behaviour on the Internet, up to creating point-sensitive contextual advertising for you personally. That is a routine.
Therefore, the currently developing AI is in some sense a human competitor in the labour market. In the automotive industry (a classic example: a robotic car assembly line has displaced millions of people from the labour market), in the medical sphere, recruitment clerks; it replaces staff psychologists. However, I do not see the future for AI right now. Thus, when we will face a strong artificial intelligence desiring to make something strange, then it will be a hot topic for discussion.