Anthropic co-founder Jack Clark calls for a 'kill switch' for artificial intelligence

Jack Clark, one of the co-founders of the US artificial intelligence company Anthropic, gave an interview to the BBC regarding the recent discussions about artificial intelligence.

A way to disable AI software entirely if it becomes too dangerous is something society "might want to make rules about in the future," Clark said.

Stating that many companies, including Anthropic, have various "off button" methods, Clark said that details about "emergency off button" requirements and control should be part of the "broader policy discussion" on artificial intelligence.

BBC: "Should you require companies to have an emergency shutdown button? Should this emergency shutdown button be under the control of a third party?" "I think this is the kind of issue that society would want to know about and ultimately want to make rules about," Clark said. he replied.

When asked what the percentage probability is that artificial intelligence will destroy all of humanity, Clark said: "I don't think these statistics will be of any use. However, it is a bad idea for artificial intelligence to be a completely uncontrolled industry." he said.

"We're gambling, taking tremendous risks. And the point is we have to change the course of this industry," Clark said. shared his opinion.

Jacob Coxon, who has been working as a researcher at OpenAI and Anthropic for the last three years, resigned from Anthropic, arguing that neither company acted responsibly enough regarding artificial intelligence.

Claiming that companies are in an uncontrolled race for self-improving artificial intelligence systems, Coxon argued that these systems could turn into superhuman systems that can "hack everything" in the near future.

Coxon also claimed that some industry workers who developed artificial intelligence models seriously believed that these systems could "kill humanity" within 10 years.