Jun 2, 2020 in Research

Artificial Intelligence

Artificial intelligence (AI) is a term coined in early 1956 to denote ‘intelligent’ machines that help people. In a broader and more contemporary sense, AI is a branch of computer science that specializes in computational devices, machines and systems that possess a computerized analogue of a conscience that enables them to perform various tasks, both pre-programmed and relatively autonomous. Here is a perfect definition of AI: “AI is technology that appears to emulate human performance by learning, coming to its own conclusions, understanding complex content, engaging in dialog with people, enhancing human cognitive performance, or replacing humans in executing both routine and non-routine tasks”. Currently, humankind and the global economy are at a stage of technological evolution when many fictitious sci-fi visions of AI are becoming everyday realities. AI is now firmly and almost fully integrated into the people’s daily and economic life. 

The artificial intelligence that exists in the present day includes many examples and covers many areas, such as virtual personal assistance and targeted advertising. For the most part, AI has become such an ordinary part of the daily life that people have stopped noticing it and are now taking it for granted. For example, people actively use virtual personal assistants in the form of Siri or Google Now with voice control. Video games is another example of AI, even though seldom perceived as such. People also drive smart cars, and self-driving cars are now being actively developed and tested by Tesla and Google. AI is also used in purchase prediction on Amazon, etc. In addition, AI is now used in fraud detection, security surveillance and online customer support in banks and other institutions. One more prominent use is AI in smart homes. To top it all, AI-powered websites are now capable of writing news stories, Wordsmith’s Automated Insights being a bring example. 

 

The face of the future AI is also shaped today, in the form of visionaries’ ideas and designs. One of the brightest and most discussed examples of the AI of the future is robots with human-like processing capabilities and a human-like face. Analyzing the current trends and implications for the future, one of the studies made a prognosis for the following decade. They predict that AI of the nearest future will have a wide array of applications. For example, they mention the emergence and spread of machines with improved natural language processing, i.e. machines that use human language as an interface. The applications may include drafting, transcribing and translating texts. Another example of the future popular AI is a virtual personal assistant able to issue reminders, help people organize their finances, schedule their appointments, and locate providers of sought-for services. Also, there will be machine vision able to distinguish objects and activities in visual images. The first AI examples of this type already exist and include machines that assist blind people by providing object descriptions, or car-safety systems that rely on visual identification of road users and road scenes, street-view maps, etc. Ultimately, the scientists envision the emergence of machines that can function without human assistance based on automated analytical model building. The examples will include self-driving cars, fraud protection AI, drug detection AI, etc. 

As any technological artifact, AI has its positive and negative sides. Among the pros of AI is work efficiency. Machines have a potential to substitute and replace manpower, partially or even totally, for machines never complain and are capable of completing the tasks on time and with the highest degree of precision. Another pro is technological growth rate. AI itself is a springboard for adjacent breakthroughs and technological revolutions. In other words, artificial intelligence can stimulate and enable further advancements in technological development. The third advantage is the fact that, unlike all living organisms including humans, AI does not need to rest and can operate continually if having access to the energy source. Ultimately, AI is best suited for risky tasks where humans can potentially get hurt or cannot survive at all. On the other hand, AI does have its cons. The first one is the dependency and over-dependency humans tend to develop. Indeed, people of the modern age rely on technology even in routine daily tasks not to mention production, exploration and other fields. Hence, if suddenly deprived of technology, they may become vulnerable and helpless. The second con is the so-called lack of human feel. The AI can be too objective and logical, thus, potentially inhumane and harmful to people’s subjective values and vision that, to a machine, would seem faulty. Another issue is the threat of AI misuse since any technology can turn into a weapon if in the wrong hands. Last but not the least, there is a reasonable concern regarding job security. It is the darker side of the medal of replacing manpower with machines – people will lose their jobs.  

In fact, concern regarding the AI’s ability to make ‘half of the world’ unemployed within only thirty years has been snowballing in the media reports of the recent years. Some scholars argue AI is first and foremost a threat to the economy, and not necessarily a benefit. In fact, it is not the matter of a distant future. China – one of the world’s fast-growing economies – already demonstrates the first signs of the looming AI-related crisis. There, robots already crowd people out of their positions as electronics plant workers. Moreover, in the bigger picture of the hypothesized future, the displaced workforce is expected to lose qualification, thus, become economically worthless in addition to being jobless. Let alone China, the world’s hegemon USA is also experiencing troubles associated with technologization and automation. According to some scholars, the United States have been in the non-apparent economic crisis for the last three decades. In particular, it was noted that even though the Americans “drive GDP with increasing productivity”, the employment peak was back in 1980 and the trend for wages has been declining since then. Some associate this decline with automation. Interestingly, some jobs in the labor market appear to be more vulnerable to automation compared to others. For example, the well-paid jobs of doctors and low-paid ones of health aides and landscapers are thought to be least affected by the AI revolution. In contrast, CEO’s jobs can be automated up to twenty percent, and file clerk’s up to eighty percent. Telemarketers, cooks and legal secretaries are also put at risk. 

Along with scholars’ speculations, there are studies – both ongoing and completed – focusing on the already existing effects of robotization and automation. One of such studies is a comprehensive study of the economic effects of industrial robots conducted by Georg Graetz (Uppsala University) and Guy Michaels (London School of Economics) and issued in 2015. The study embraces as many as seventeen countries, including the USA, within the period between 1993 and 2007. The findings are as follows. The average density of robots was found to rise by more than one hundred and fifty percent along with the fifty-percent decrease in their prices within the time frame of 1990-2005 (with six countries demonstrating a dramatic price fall of as much as eighty percent). The fastest increase in the quantity of robots per million hours worked occurred in the transportation equipment (8.1), chemical (3.3) and metal (1.7) industries. According to the data, robots did not affect the hours worked by high-skilled employees, whereas the machines tended to crowd out middle-skilled and, especially, low-skilled employees. The latter fact proves that concerns of the scholars regarding the threats AI poses for these categories of workers are quite reasonable. Also, the researchers found that the exploitation of robotic power led to an increase in wages, total factory productivity, as well as labor productivity and value added from labor. However, they also estimated that the increase in robot density results in a decline of economic returns. The dependency was directly proportional:  the larger the increases in robot density, the smaller the productivity. The latter finding suggests that there exist certain congestion effects from increased application of robots. Nevertheless, the overall increase in the use of industrial robots proved to raise the annual growth of GDP (by 0.37 percentage points) and labor productivity (by 0.36 points).

In addition to retrospective studies of the effect of automation and AI on the world economies and economics, there also exist studies focusing on the future. Inter alia, a study by Analysis Group, Inc. sponsored by Facebook assumes the AI’s economic impact over the next decade will manifest in a boost to global economic output within the range of $1.49 trillion to $2.95 trillion. They also mention the possibility of the upper bound rising as high as $5.89 trillion. Unlike the pessimists whose viewpoints regarding AI’s impact on employment have been mentioned above, the Analysis Group is more optimistic. They believe that the modern labor market is flexible enough to adapt to new technologies introduced into it and incorporate these technologies successfully and efficiently, at the same time, producing new jobs in newly emerging areas. They call it the labor market’s “historical resilience”.

In conclusion, AI stands for a ‘smart’ machine that is or will be able to assist, complement, or even substitute humans in their basic daily activities, as well as on a larger socio-economic scale, for example, in production industry. Aside from its evident benefits in making people’s life easier and helping humanity in making breakthrough discoveries and previously unheard-of operations, AI also has its downsides and pitfalls. The main concern is associated with AI’s impact on the labor market, both nowadays and in the future. Scholars disagree on how AI advancement will affect the economics of employment. While some worry that AI will cause out-of-proportion unemployment, others find empirical proof that automation is economically viable. Either way, the AI evolution is unstoppable, and the AI revolution is approaching.

logotype

Related essays