US "Foreign Policy" magazine: Artificial intelligence will exacerbate global inequality

Article in the US "Foreign Policy" magazine on May 29, original title: Artificial Intelligence will exacerbate global inequality The race for artificial intelligence (AI) is intensifying, and its risks have reached unprecedented levels. Major companies, including Alibaba, Google and Microsoft, are using powerful computing capabilities to promote AI progress. In addition, hundreds of private businesses and nonprofit organizations are launching software and programs in an effort to gain a foothold in this emerging market. Concerned people predict that these achievements will disrupt the way we work, play, trade, create wealth and manage things.

Will robots replace humans?

In an atmosphere of optimism, fear is also growing. More and more technology giants and computer scientists have expressed concerns about complex algorithms, and they believe that in the near future, super omnics will quickly replace humans. According to a 2022 survey, nearly 50% of AI experts believe that these technologies may lead to at least a 10% risk of human destruction.

Concerns about the risks of machine intelligence have been around for a long time. Since 1872, British writer Samuel Butler has predicted in his book "Nowhere" that autonomous robots will eventually replace humans. In 1942, the Russian-American science fiction master Isaac Asimov proposed the famous "Three Laws of Robotics": robots must not harm humans; robots must obey human orders unless these orders conflict with the first law; robots must Human existence must be protected so long as this protection does not conflict with the First or Second Law.

The current level of popularity and confusion of AI is unprecedented, which can be confirmed by history. Large language models and the powerful computing power behind them are no longer limited to laboratories, but are now used in the real lives of hundreds of millions of people. However, some staunch supporters of AI believe that if left unchecked, AI will soon bring about fatal disasters. The theoretical threat that was once far away has now become more real. Recently, more than 31,000 people have signed a petition calling for a suspension of the training of powerful AI models because they are "the biggest risk facing society and mankind today."

Potential hazards are concentrated in 4 aspects

Before this anxiety began to spread, governments, businesses and universities in North America and Western Europe had already begun discussions about artificial intelligence and its potential harms. The focus is mainly on four aspects: first, super-intelligent machines may quickly replace humans, posing an existential threat; second, AI will aggravate the unemployment situation; third, the way AI imitates and transmits text, voice and video will It will intensify the spread of false and misinformation; fourth, AI may be used to develop doomsday technologies such as biochemical or network viruses, which will bring devastating blows.

As awareness of AI risks increases, the standards used to buffer risks are also gradually improving. Most of these standards are voluntary, such as a large number of agreements and guidelines on design responsibility and self-regulation. Maximizing human interests, improving design security, and optimizing algorithms are all common AI development principles. Algorithmic transparency, responsibility and fairness in applications, privacy and data protection, human supervision and control, and compliance are all development goals of AI. Technology companies are actively participating in pushing for the establishment of AI governance agencies and better institutional systems, as voluntary self-regulation gradually loses focus.

However, the scope of attention paid to the security system is far from enough. Currently, discussions about AI and strategies to mitigate its harms are focused primarily on the Western world. The vast majority of current government and industry standards originate from EU, US or OECD member countries.

The European Union is about to launch a new AI bill, focusing on high-risk AI applications and systems. The reason why the Western world pays so much attention to AI is that there are a large number of concentrated AI companies, investors and research institutions in various regions. In Latin America, Southern Africa, South Asia, and Southeast Asia, although AI is also developing rapidly, its needs and concerns have not appeared in AI-related discussions. In other words, discussions about the impact and regulation of AI are largely focused on a small minority of the population. In poor and underdeveloped countries, attention and funding are far from adequate. The global population has reached 6.7 billion.

The harm is even worse in low-income areas

Due to the lack of AI regulation, this situation poses significant risks, with the global South facing particularly severe negative impacts. The deeper negative impact of AI is its violation of equal opportunities. Allowing AI to develop freely will widen the social, economic and digital divide between countries, promote the concentration of corporate power, deepen technological monopoly, and make the West's already riddled democratic system worse.

Although the harm of AI is global, its impact on various regions is not the same. Its harm is particularly serious in low- and middle-income countries and regions with weak regulatory systems and institutions. On the one hand, algorithms and data developed in rich countries and their subsequent application in developing countries can exacerbate prejudice and discrimination caused by weak awareness and lack of diversity. Due to inadequate wages and lack of labor protection, workers with lower incomes and lower skill levels are more likely to face the risk of being replaced by artificial intelligence. It is true that the development of AI in the global South will bring huge benefits, but without proper regulation, ethical controls and the strengthening of public awareness, we will not be able to limit its negative impacts.

The above is the detailed content of US "Foreign Policy" magazine: Artificial intelligence will exacerbate global inequality. For more information, please follow other related articles on the PHP Chinese website!