Putting AI in charge of a millisecond war that can ‘occur faster than human decision making’ is ultimate threat: Eric Schmidt warns Congress

Former Google CEO Eric Schmidt tells Congress that the ultimate threat of AI would be a scenario where AI is in charge of a war that occurs faster than human decision making, lasting only a millisecond.

“I can’t figure out how we’re going to build offensive and defensive systems reliable enough to put them in charge of a war that occurs faster than human decision making” — Eric Schmidt

“We need to consider some of life’s bigger philosophical questions like how will machine intelligence affect what it means to have human thought, free will” — Eric Schmidt

Providing expert witness testimony before the House Oversight Subcommittee on Cybersecurity, Information Technology, and Government Innovation on Wednesday, Schmidt told lawmakers that he worried about the following scenario:

“In the future there’s a war. It’s an attack by North Korea on the US. China stops the war between North Korea and the US, and the entire war took one millisecond.

“And the reason I worry about that is that I can’t figure out how we’re going to build offensive and defensive systems reliable enough to put them in charge of a war that occurs faster than human decision making.

“That to me is the ultimate threat of this technology.”

“We need to make choices about whether we care more about freedom or more about efficiency” — Eric Schmidt

AI “will impact all aspects of human life, ranging from reasoning to how we work and to how it will shape the international order” — Eric Schmidt

In his opening remarks, the former Google CEO and close friend of Henry Kissinger testified that generative AI “has the potential to change the world as we know it.”

“Generative AI,” according to Schmidt, “refers to the ability of machines to create new and original content, such as images, music, or text, that is indistinguishable from content created by humans.”

“As generative AI is woven into exciting technology applications, like social media or predictive healthcare analysis, we will need to critically think of creative ways to address potential challenges,” Schmidt submitted in his written testimony.

He then recommended three principles for platforms to follow:

1. Platforms must, at minimum, be able to establish the origin of the content published on their platform
2. We need to know who specifically is on the platform representing each user or organization profile
3. The site needs to publish and be held accountable to its published algorithms for promoting and choosing content

On the topic of anonymity, Schmidt told lawmakers, “True anonymity hidden behind a paywall would allow nation state attacks.”

“AI can be used to generate good things in biology, but also lots of bad viruses” — Eric Schmidt

In his written testimony, Schmidt added:

“The decisions for AI and machine development need to be made by multidisciplinary experts – not just technologists.

“We need to consider some of life’s bigger philosophical questions like how will machine intelligence affect what it means to have human thought, free will.”

Also testifying in Wednesday’s hearing on “Advances in AI: Are We Ready For a Tech Revolution?” were Scott Crowder, VP at IBM Quantum; and Merve Hickock, Senior Research Director at the Center for AI and Digital Policy.