Artificial intelligence is developing faster than humans can control it

The global race for artificial intelligence is accelerating uncontrollably, with the US and China in the lead. World leaders seek the unprecedented power that comes with this technological advantage, while leading tech giants warn: we must apply the brakes before a computer no longer needs humanity.

Artificial intelligence (AI) has reached a point where development speed exceeds humanity's ability to control it, warns a number of the world's leading tech experts. The US and China are at the forefront of the race, and neither is prepared to slow the pace.

The allure of power and the absence of control

World leaders see AI primarily as an unprecedented source of power, both economic and military. But alongside political ambitions, the technology sector is sounding increasingly urgent warnings. Several prominent tech leaders have publicly called for international agreements that would slow AI development until safety can be assured.

The fear is not merely theoretical. According to experts, the question is no longer whether artificial intelligence will exceed human capability in certain fields, but when, and whether humanity will then still have the means to manage the situation.

The US and China: two poles, one future

The United States and China are investing hundreds of billions of dollars in artificial intelligence, viewing it as as strategically important as nuclear weapons were in the mid-20th century. Both superpowers fear that backing down would give the other a decisive advantage, creating a classic security dilemma where nobody wants to hit the brakes first.

This is why international regulatory attempts have remained modest so far. Although the EU has passed an AI law and some dialogue is taking place at UN level, there is no binding global agreement.

What does "control" mean in the age of artificial intelligence?

Leading tech companies, including some that are themselves developing AI, have begun talking about so-called "safety brake" mechanisms. The idea is to create systems capable of halting or limiting AI operations when necessary. But critics point out that such mechanisms are effective only if applied globally; unilateral braking would simply mean being left behind in competition.

The moment at which a computer "no longer needs humans" has become its own watchword in AI debates, the threshold of so-called superintelligence. Most experts consider this still distant, but warn that current development speed makes this scenario more real than ever before.

The question is no longer merely technical; it is a matter of political will and international cooperation.