Artificial intelligence should be considered a new species with alien values that humans may struggle to control, a research scientist has warned. Josh Engels, a former safety team researcher at Google DeepMind, has quit the tech giant to join a non-profit organization focused on the threats posed by AI.
Engels stated in a BBC interview that AI has the potential to present "very large risks to humanity." He described the advancing technology not just as a tool, but as a "second species that we are almost giving birth to," adding that soon Earth will be inhabited by entities far more intelligent than humans.
"I'm worried that if we don't have the ability to ensure they have our best interests in mind, there's no guarantee humans will keep control of the future," Engels said. He elaborated that an artificial species could "disempower humanity" through motivations that are not centered on human well-being. "It's not that these systems will hate humanity. It's just that they have alien motivations that don't have humans front and centre."
Engels's concerns echo those of other researchers. Jacob Coxon, who left Anthropic, another leading AI firm, similarly warned that companies like his and OpenAI are "gambling with our lives." Some industry predictions suggest a 10 percent chance of AI leading to human extinction within the next decade.
In response to these escalating warnings, executives from major AI firms have publicly called for a slowdown in development. This week, King Charles III is scheduled to host an AI summit at Dumfries House in Ayrshire, where he intends to encourage industry leaders to direct the technology's use "for the good of humanity."
A royal source indicated that the discussions aim to foster a broader understanding of the responsibilities and opportunities in AI development, ensuring that technological advancements prioritize human benefit.
Meanwhile, in the UK Parliament, lawmakers are preparing to advocate for a new AI Bill. The Joint Committee on Human Rights plans to propose a regulatory framework that would impose greater obligations on higher-risk AI systems. The committee has suggested outright bans on certain AI applications, such as predictive profiling for criminal risk assessment, drawing parallels to the science fiction film "Minority Report." The proposed legislation would prohibit powerful AI systems like Artificial General Intelligence and Artificial Superintelligence that risk causing widespread harm or evading human control.