AI Revelation? Google Paper predicts that AI could soon match human intelligence and “destroy humans forever”

A new research paper from Google DeepMind warns that artificial general intelligence (AGI) is a form of human-grade AI that may appear as early as 2030 and may pose a threat to “permanently destroy humanity.” this study Emphasizes the need for AI developers to take precautions to reduce these potential hazards.

Study warns of serious risks of AGI

Co-author Shane Legg, DeepMind co-founder, does not elaborate on how AGI leads to extinction, but outlines the types of damage that can be caused. It identifies four categories of risks: abuse, misalignment, error and structural risks. Among them, the abuse of AI systems (intentional use of artificial intelligence caused harm by individuals or groups) is a key issue.
“Given the huge potential impact of AGI, we hope it could also cause a potential risk of serious injury,” the study noted. “The existence risk of permanent destruction of humans is a clear example of serious injury.”
It further states: “Whether between these ends of the scope, the question of whether a given injury is serious is not a problem for Google DeepMind; rather than the authority of society, guided by its collective risk tolerance and concept of injury.”

DeepMind paper requires social participation

The paper calls for wider industry and social participation to determine which risks should be considered serious and how to deal with the dangers.

The study also highlights the importance of collaboration between developers, governments and international institutions to establish guidelines and oversight mechanisms before AGI reaches maturity.

CEO demands global oversight

In February, Demis Hassabis, CEO of DeepMind, said that Agi Systems, which are as smart as humans and even smarter as they are, may start to appear in the next five to ten years. He called for a new international structure to monitor and guide AGI development. “I would argue for a CERN of AGI, which refers to an international study focused on the cutting-edge AGI development to be as safe as possible,” Mr Hassabis said.

“You also have to pair it with institutes like IAEA to monitor and deal with unsafe projects. Finally, some kind of super enterprise involves many countries around the world, and they enter the way you want to use and deploy these systems. So, some kind of UN umbrella, which is suitable for this purpose, is one for that system, he adds that.

What is Agi?

AGI refers to artificial intelligence that goes beyond performing specific tasks. Unlike traditional AI designed for limited applications, AGI will be able to perform general reasoning, learn and solve problems in various fields, similar to human intelligence. Such a system will be able to apply knowledge in different fields without programming to a specific task.