Amodei thinks ‘AI-driven tyranny’ may be coming – Healthcare Blog
Author: Mike Magee
Last month, Anthropic CEO Dario Amodei surprised society leaders around the world by doing the same thing again — warning them that control over human security and governance was approaching the extinction of artificial intelligence.
Amodei’s opening paragraph in his article titled “The Adolescence of Technology” immediately grabs the reader’s attention. He wrote: “There is a scene in the movie version of Carl Sagan's book touch The protagonist is an astronomer who has detected the first radio signal from an alien civilization and is considering taking on the role of a human representative to meet the aliens. The international team who interviewed her asked: “If you could ask [the aliens] Just a question, what could it be? Her response: “I would ask them, how do you do it? How do you evolve, how do you survive this technological adolescence without destroying yourself?” “
Now, I should be clear. I'm already very nervous. As a historian of medicine, I am preparing a major lecture this spring on the birth of immunology, a field I have been studying. What am I looking for? When exploring the frontiers of scientific advancement, I always find the same thing missing – historical context. In most cases, facts and figures abound, but their impact on complex webs of relationships over the years is often overlooked.
Amodei is trying to provide this context in real time. Real-time news includes headlines like this one from The New York Times: “ICE Already Knows Who the Protesters Are,” powered by facial recognition technology powered by artificial intelligence. But Amoday's concerns are more fundamental. The challenge he faces is the rate of change in generating artificial intelligence, which he makes clear is staggering. As he said, “Because AI is now writing the majority of the code at Anthropic, it has greatly accelerated our progress in building the next generation of AI systems. This feedback loop is growing month by month, and we may only be 1-2 years away from the point where the current generation of AI will autonomously build the next generation of AI.”
Clearly, Amodei has come up with a pretty reasonable modern vulnerability, given Musk's recent attempts at DOGE. He said, “As the CEO of an AI company, it's a bit embarrassing to say this, but I think the next level of risk is actually the AI companies themselves. AI companies control large data centers, train cutting-edge models, have the most expertise in how to use those models, and in some cases touch and potentially impact tens or hundreds of millions of users every day.”
Without directly naming the Musk-controlled Grok and
During an ICE crime last month, a legal observer in Portland, Maine, was filming ICE agents when she was approached by an agent who had just filmed her car and was now filming her face. When asked why he did that, the ICE agent responded: “Because we have a nice little database and now you're considered a domestic terrorist. So, enjoy this.”
The events of the past month and the senseless murder of two innocent American citizens made Amodei’s final warning prescient. “Current authoritarian regimes are limited by the need for humans to carry out their orders, and there are often limits to how inhumane humans can be,” he said. “But an AI-powered authoritarian regime would have no such limitations.”
Mike Magee, MD, is a medical historian and regular contributor to THCB. He is the author of “Code Blue: Inside the American Medical-Industrial Complex.” (Grove/2020)
