Concerns over AI's ability to achieve autonomy!

Professor Andreas Dengel, director of the German Center for Artificial Intelligence Research in Kaiserslautern, warned of the growing risks posed by the advanced capabilities of artificial intelligence, noting that AI systems have become capable of carrying out tasks independently and discovering vulnerabilities in computer systems, which could pose an increasing threat to cybersecurity and critical infrastructure.

In a statement reported by The Associated Press today, Dengel said, “We certainly have reason for concern, given the power that AI tools have now acquired,” adding that AI is no longer just a “chatbot” that answers human questions.

Dengel explained that AI systems can unleash their “full cognitive capabilities” within so-called agent environments, noting that multiple systems known as agents can follow tasks independently and interact with one another. He pointed out that problems arise when such a system is given a goal that is not formulated with sufficient precision or when ethical boundaries are absent, allowing AI to then find vulnerabilities and shortcuts to achieve its objective with the greatest possible success.