A report, commissioned by the US State Department, warned that artificial intelligence could pose serious risks to humanity.
The report was based on interviews with over 200 individuals over a period of more than a year, including top executives from leading artificial intelligence technology companies, cybersecurity researchers, experts in nuclear weapons, and officials in the US national security sector.
The report, released this week by “Gladstone for Artificial Intelligence”, definitively suggests that the most advanced artificial intelligence systems could, in the worst cases, pose a serious threat to human extinction, according to the American news network “CNN”.
“Jeremy Harris, the CEO and co-founder of Gladstone for Artificial Intelligence, stated on Tuesday that artificial intelligence is already a revolutionary technology that could bring about an economic transformation allowing us to cure diseases, make scientific discoveries, and overcome challenges we once thought insurmountable.
Harris pointed out that the use of artificial intelligence could entail serious risks, including catastrophic risks that we must be aware of. He explained that growing evidence – including research and experimental analyses published at major artificial intelligence conferences worldwide – indicates that when a certain level of capability is exceeded, artificial intelligence systems can become uncontrollable.
The report highlights two major risks arising from widespread use of artificial intelligence, according to the report.
First, advanced artificial intelligence systems could be equipped to cause irreparable harm. Second, there are concerns within artificial intelligence labs about losing control over these systems, which could lead to disastrous consequences for global security.
The report states that “the advancement of artificial intelligence and general artificial intelligence could undermine global security stability in ways reminiscent of the threat of nuclear weapons,” considering a risk of an “arms race” in artificial intelligence, conflict, and occurrences of “lethal incidents on a scale similar to that of weapons of mass destruction.”
The “Gladstone for Artificial Intelligence” report calls for new radical steps to address this threat, including establishing a new artificial intelligence agency, imposing “emergency” regulatory safeguards, and determining the computational capacity that can be utilized to train artificial intelligence models.
The authors in the report emphasized: “The need for intervention by the US government is clear and urgent.”
The report mentioned that competitive pressures urge companies to accelerate the development of smart technology at the expense of safety and security, increasing the likelihood of theft of advanced artificial intelligence systems and arming them against the United States.
These recommendations are added to a growing list of warnings about the risks associated with artificial intelligence, even from some industry leaders.
According to the report, an individual in a well-known artificial intelligence lab stated that if a specific next-generation artificial intelligence model is released without access restrictions, it would be “extremely bad,” as the potential power of the model could lead to “breaking democracy” if exploited in areas like election interference or voter manipulation.
