
Joshua Benguio, often called the “Eye Godfather,” warns that fierce competition between open ryes and China’s deep -shak can impair AI’s safety. (Image: Borealis AI)
Artificial intelligence has evolved rapidly, but the risks of the risk are the same. Yo -A Benguio, often called the “Contemporary AI Godfather”, has issued a strong warning on the growth of AI racing among global high -tech giant. His concerns may not only make a bidding for companies to go ahead, as well as technology advances. The important focus of his warning is Openai, a company behind Chatgpt, which is facing the increase in competition with China’s deep seek. Bengio believes that this fierce rivalry may encourage AI companies to prioritize speed over safety, which can lead to potential dangerous results.
AI racing brings safety risks
Bengio recently talked about AI races in an interview with the Guardian and expressed concern about the pressure to maintain the dominance of companies like Openai. He explained, “That would mean a closer race. This is not a good thing from the viewpoint of AI’s safety.” His warning has been reported in the report that DeepSeek, a Chinese emerging company, has made a major progress in AI development and is directly challenging US companies such as Openai, Google Deepmind, and Anthropic.
According to Benguio, when companies feel much ahead of their competitors, there is plenty of safety precautions. However, if your rivals are closely matched, companies may be in a hurry to innovate at the expense of AI’s safety. “Then they may not pay much attention,” he warned.
The possibility of AI’s threat is increasing
Bengio has led the first international AI safety report released in the UK 2023 AI Safety Summit. This report, edited by 96 experts, emphasizes the increase in AI -related risks, including the possibility of using cyber security attacks and biological development.
One of the surprising discoveries was that AI could create very detailed instructions for creating biological threats. It is knowledge that even experts are having a hard time editing. It is unknown whether non -experts can misuse such information, but the report emphasized that these risks must be taken seriously.
The next global AI Summit is scheduled for February in Paris, so the discussion on AI’s future is not the end.