AI should be treated as species that could disempower us all, whistleblower warns

Thank you for subscribing!

Josh Engels warned that humans could soon share the planet with systems considerably more intelligent, raising questions about who would ultimately control the future

Artificial intelligence should be viewed as more than a powerful new technology, but rather an increasingly advanced "species" that could disempower us all, a former AI safety researcher has warned.

Josh Engels, who recently left DeepMind’s AI safety team to join non-profit research organisation METR, said the rapid development of artificial intelligence could create “very large risks to humanity”.

He cautioned that if researchers fail to ensure future systems remain aligned with human interests, continued innovation could lead to societal regression.

Speaking to the BBC, Mr Engels argued that society should begin thinking of advanced AI as a potential “second species”, rather than simply another invention.

He warned that humans could soon share the planet with systems considerably more intelligent, raising questions about who would ultimately control the future.

“I think in some ways it's better to think of it not as a technology but as a second species that we are almost giving birth to,” he said.

Mr Engels said the danger did not necessarily depend on AI developing hostility towards humans - but instead could pursue objectives that were fundamentally different from human priorities.

He said researchers currently had limited ability to shape the motivations of advanced AI and warned that its “alien values” could eventually result in humanity being disempowered.

“It's not that these systems will hate humanity,” he said, “it's just that they have alien motivations that don't have humans front and centre.”

The warning comes amid growing concern among researchers and technology executives about the potential consequences of accelerating AI development.

Jacob Coxon, a former senior researcher at Anthropic, has also warned that leading AI companies are taking potentially dangerous risks and that continuing at the current pace could create a significant possibility of human extinction in the near future.

Other figures have estimated that there could be a 10 per cent chance of AI causing the extinction of humanity within the next decade, prompting calls for greater caution from industry leaders, with executives at major AI companies urging a slower approach to development.

The King is due to host leading figures from the artificial intelligence sector at Dumfries House in Ayrshire this week, where he is expected to stress the importance of developing AI strictly for the benefit of humanity.

The Joint Committee on Human Rights is urging the Government to introduce an AI Bill that would impose tougher requirements on systems judged to pose greater risks.

It has also called for certain applications to be prohibited, including AI systems designed to predict whether an individual is likely to commit a crime.

The committee said highly powerful systems, including Artificial General Intelligence and Artificial Superintelligence, could pose “very serious harm” if they became capable of operating beyond effective human control.

The warnings remain defiant in encouraging the benefits of AI, while ensuring increasingly powerful systems remain subject to tangible human oversight.