There is no doubt that the accelerated development of technology will change entire professional industries. Sergey Markov discusses whether there will be a place for a person in the future labor market
Paradise for tech-pessimists
The debate about how quickly and on what scale robots can replace people in various professional fields has been going on for a long time.
Some believe that this will mainly affect routine operations, which will free people for creativity. Others warn that, firstly, humanity will not need so many creative specialties, and secondly, AI systems are already doing, and often successfully, what has traditionally been considered the prerogative of a person – compose stories, write articles, draw pictures and invent musical works. Still others are sure that our fears about this are exaggerated, and if robots are ever able to replace us everywhere, then this will not happen soon. Who is right, argues Sergei Markov.
About the expert: Sergey Markov — Head of the Department of Experimental Machine Learning Systems, SberDevices Department, Sberbank PJSC.
Substitution of concepts
“Very soon AI will replace humans!” – similar words can often be found in the media, but it is not entirely correct to say this for several reasons. First, it is important to understand the difference between “AI” in the global sense and “AI systems”. By “artificial intelligence” experts mean the direction in science and technology that automates the solution of intellectual problems or, as the inventor and futurist Raymond Kurzweil succinctly put it, “the art of creating machines that perform functions that require intelligence in cases where they are performed by people.” In this sense, about the same unpleasant story happened to AI as to ecology, when the name of science began to be widely used instead of the concept of “environment”.
If we are talking specifically about AI systems, real and hypothetical, then they can all be divided into two varieties:
Such “smart machines” do not yet exist.
“Smart” chatbot
People often confuse these two varieties, which leads to a distortion of the picture of the world. Someone begins to suspect the presence of consciousness in modern virtual interlocutors, someone – to accuse the creators of applied systems that they are engaged in “fake” AI.
In the 1950s, no one doubted that an AI system capable of defeating a world chess champion was “real AI.”
After the victory of Deep Blue over Kasparov, the game of Go was cited as a “watershed”, supposedly beyond the control of “real AI” and so on. Some of the seemingly unattainable tasks for him have long been solved, although people may simply not know about it. Some will say that machines cannot learn (although machine learning has been around since the 1950s), others that they cannot understand human emotions (although systems of so-called sentiment analysis or text sentiment analysis successfully cope with this in during the last decade). Some might argue that machines can’t, say, compose music (although in 2019, for the first time in history, a contract was signed between a major music label, Warner Music, and Endel, a company that created a system for algorithmic music generation).
Art versus routine
It is important to understand that AI systems are created, first of all, as tools for solving various urgent problems: increasing labor productivity, reducing the proportion of procedures that are monotonous or dangerous for people, and, finally, to expand human capabilities.
Today, writers and artists use tools in their work that simply did not exist a century ago. – text editors with automatic spell checking, change history tracking, the ability to copy text, graphics tablets and graphic editors with many features, etc. They saved people from tedious operations, giving them the opportunity to focus on creativity.
Have these tools replaced humans? After all, now the writer and artist began to take less time to create works. Does this mean that our society needs fewer artists and writers now? It is obvious that this is not the case. Moreover, in modern society, the number of people involved in the production of texts and images is much greater than a hundred years ago. This seemingly paradoxical fact is due to the fact that in reality the amount of textual and visual content consumed by society has increased significantly.
In the 1950s, the United States accounted for several dozen computers. All the computing power of the NASA computers used in the American lunar program is inferior to an inexpensive modern smartphone. Does this mean that all computational problems have already been solved, and mathematicians have lost their jobs? No, we are making use of amazing computing resources. The emergence of new “smart” tools opens up new opportunities for humanity.
This is well understood by researchers building AI systems today, and as stated in the Social AI and Extended Intelligence project manifesto of the Human Dynamics research group at MIT, instead of competing with machines, you need to understand that the combination of human + AI will work better, than humans and AI acting alone.
Subscribe to the Trends Telegram channel and stay up to date with current trends and forecasts about the future of technology, economics, education and innovation.