Warning AI may develop 'will to survive', beyond human control

Leading computer scientist Yoshua Bengio warns that if AI gets out of control, 'humanity could disappear in 10 years'.

Scientists worry about the risk of losing control of AI - Illustration: ChatGPT

On February 7, a leading computer scientist warned about the possibility of humans "losing control" over artificial intelligence (AI) systems in the future when these artificial products develop a "will to survive".

Speaking at the AI Action Summit taking place in Paris, computer scientist Yoshua Bengio called for tighter international regulation and more extensive research into AI safety, which currently accounts for only a small fraction of the major investments being made in the field.

The emergence of the high-performance, low-cost AI model DeepSeek last month has accelerated the race to safety, according to the winner of the prestigious 2018 Turing Award in computer science. Yoshua Bengio warned that if AI gets out of control, “humanity could disappear in 10 years.”

He also called for government leadership in AI management. The call comes as the world's leading AI nation, the United States, has taken many steps to "untie" the development of this technology under new President Donald Trump.

At present, the International AI Safety Report presented by Mr. Yoshua Bengio at the conference also points out well-known risks such as the use of AI to create fake or misleading content on online platforms, while additional risks are constantly emerging, such as biological attacks or cyber attacks.

With contributions from about 100 experts from 30 countries, the United Nations, the European Union and the Organization for Economic Cooperation and Development (OECD), this is the first report on international AI safety, conducted since the AI safety summit in the UK in November 2023.

Source: https://tuoitre.vn/canh-bao-ai-co-the-phat-trien-y-chi-sinh-ton-vuot-kiem-soat-cua-con-nguoi-20250207104202015.htm