Anxiety about the threat from artificial intelligence tends to focus on a distant point in the future, when the whole world will be “captured” by the so-called universal artificial intelligence. This is the moment when machines will be able to reason and perform operations on human or even superhuman levels in most activities, including even critical thinking and creativity. However, there are already reasons to worry about the seizure of the world by artificial intelligence, writes The Washington Post.
Some analysts have focused on the potential danger to human civilization from artificial intelligence in the future. Such warnings have already sounded from entrepreneurs Bill Gates and Elon Musk. In fact, you should be afraid – not of what the AI might become, but of what it already is.
Almost all progress in the field of artificial intelligence today has little to do with the so-called universal artificial intelligence. Now progress has focused on narrow tasks. The capabilities of AI do not yet provide for the implementation of the most dire predictions, but the consequences can be disastrous right now.
AI is driving out workers. A study by analysts Jonathan Hazell and Pasqual Restrepo found that companies that increase their use of AI by 1 percent also reduce the number of living workers by about 1 percent.
And, of course, AI provides new monitoring technologies used by corporations and governments, often in a negative way. It is also used in the US justice system to make bail decisions, and sometimes even for sentencing. Artificial intelligence often “distorts” public discourse on social media, interfering with the functioning of modern democracies.
The impact of AI on the labor market may be the most critical. For example, the US economy once provided good jobs, paying decent wages and guaranteeing job security and career opportunities for workers of different backgrounds and skills. From the end of World War II to the mid-1970s, the United States saw not only significant growth in employment, but also skyrocketing wages.
This growth stopped long before the development of AI. Since the 1980s, average wages have fallen into relative stagnation. People who did not receive higher education began to earn less.
During this period, automation and jobs, as well as the transfer of companies to other countries, caused a decline. But now AI is accelerating these negative trends. Many cost-cutting employers will seize every opportunity to eliminate jobs through new technology and automation.
Some economists believe that the fear of automation and displacement of AI workers is an exaggeration. They argue that as work becomes more automated with AI, the resulting productivity gains will fuel demand for labor in other parts of the economy, and sometimes even in the same companies that are implementing AI-driven automation.
If AI technologies were truly productive enough for the tasks they perform today, this argument would carry some weight. Modern AI technologies are not only far from universal artificial intelligence, they often fail at tasks that should have already become the norm, such as facial recognition or speech understanding. This means a double harm to the labor market, since AI technology is crowding out living workers and does not create an increase in labor demand, which should provide significant productivity gains.
Other uses of AI are likely to enhance the growth of corporate power over the workforce. AI enables better monitoring of workers.
The use of AI in government decision-making, particularly in the criminal justice system, is a matter of concern. Evidence suggests that algorithms often inherit and sometimes reinforce existing biases and inequities.
AI then wreaks havoc on democratic discourse and politics. This is due not only to the algorithms of disinformation on social media, but also to the expansion of the ability of companies and governments to control and manipulate the behavior of millions of people, which is fundamentally inconsistent with true democracy.
Each new technology creates critical decision challenges that determine who will win and who will lose, and whether the benefits justify the harm.
This statement sounds true for AI, and not only because of the consequences of its influence throughout society and its influence on areas that should remain far from being regulated by technology, such as the justice system, for example.
There are many different ways in which you can predict the future of AI: will the AI start to actively work to supplant living workers and spy on people, or will it will focus on expanding human capabilities, creating new perspectives for workers?
These choices require public and government scrutiny to prevent the abuse of technology and regulate its impact on the economy and democracy. If the choice remains in the hands of the loudest advocates of AI, it can have negative consequences for society.