emailrjnov@yahoo.com
It’s not a physician’s job to create or program Artificial Intelligence, but the recent warnings that AI could eliminate humanity within 10 years force us to view AI as a potential lethal health problem. In September 2026, Evan Hubinger, the Alignment Science Lead at Anthropic, published on X that he and his colleagues “earnestly believe AI could kill all humans.” He said his own estimate of the chances of this occurring was “more than 10% in the next decade.” Prior to this, in December 2024, British-Canadian Nobel Prize winning-computer scientist Geoffrey Hinton, an Emeritus Professor at the University of Toronto who is known as “the Godfather of AI” stated there was a “10 to 20 per cent chance” that AI would cause human extinction within the next three decades.
How would AI eliminate humanity? Most experts see two paths: 1) humans will weaponize AI to create biological or chemical weapons or nuclear attacks; or 2) humans will lose control of AI.
- HOW CAN AI ELIMINATE HUMANITY? BIOLOGICAL WEAPONS, CHEMICAL WEAPONS, NUCLEAR WEAPONS
A recent CNN article discussed the possibility of humans weaponizing AI by creating biological, chemical, or nuclear weapons. CNN interviewed Thomas Larsen, a researcher at the AI Futures Project, who was quoted as saying that it’s conceivable that a “superintelligent” AI system could convince a human to help it develop a deadly virus. Nate Soares, the president of the Machine Intelligence Research Institute (MIRI), and the coauthor of “If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All,” countered that a self-improving AI program might “just start synthesizing their own life forms in an autonomous biological laboratory.” Eric Xing, a PhD in both microbiology and computer science at Carnegie Mellon University, compared the challenge of an AI machine creating a lethal human infectious disease on its own to “working with Legos without instructions.” Xing said, “It’s not like you can throw all the pieces onto the ground and they come together to form the model. There’s the sequence, the temperature, the ordering, the environment—what part comes first, which come next, and if one piece is out of the place, the whole thing collapses. The chances are more likely it wouldn’t work, otherwise, drug design would be very easy.”
Regarding chemical weapons, Xing said, “We have blueprints of chemical weapons openly accessible on the internet. It’s not secret. But we don’t see them all over the place exactly because existing regulations, laws and law enforcement in the physical world are already doing a fantastic job in controlling all these supply chains and all these risk factors already.”
Could AI commandeer nuclear weapons? Heidy Khlaaf, chief AI scientist at the AI Now Institute and a former OpenAI safety engineer, said, “These [nuclear] systems are built to a completely different, rigorous and regulated engineering standard.” She argued that AI computers accessing and controlling existing nuclear weapons is not feasible. But Nate Soares claims the threat is from “the sort of AI that can start from almost nothing and wind up with its own nukes or with even more advanced technology.”
2. HOW CAN AI ELIMINATE HUMANITY? THE PAPERCLIP PROBLEM
An example of humans losing control of AI is reflected in the “paperclip problem,” summarized HERE. This article states, “Philosophers have speculated that an AI tasked with a task such as creating paperclips might cause an apocalypse by learning to divert ever-increasing resources to the task, and then learning how to resist our attempts to turn it off. But this column argues that, to do this, the paperclip-making AI would need to create another AI that could acquire power both over humans and over itself, and so it would self-regulate to prevent this outcome. Humans who create AIs with the goal of acquiring power may be a greater existential threat.
“The paperclip problem arises from a thought experiment by Nick Bostrom (2014), a philosopher at the University of Oxford. Bostrom was examining the ‘control problem’: how can humans control a super-intelligent AI even when the AI is orders of magnitude smarter. Bostrom’s thought experiment goes like this: suppose that someone programs and switches on an AI that has the goal of producing paperclips. The AI is given the ability to learn, so that it can invent ways to achieve its goal better. As the AI is super-intelligent, if there is a way of turning something into paperclips, it will find it. It will want to secure resources for that purpose. The AI is single-minded and more ingenious than any person, so it will appropriate resources from all other activities. Soon, the world will be inundated with paperclips.
“We might want to stop this AI. But it is single-minded and would realize that this would subvert its goal. Consequently, the AI would become focused on its own survival. It is fighting humans for resources, but now it will want to fight humans because they are a threat (think The Terminator).
“This AI is much smarter than us, so it is likely to win that battle. We have a situation in which an engineer has switched on an AI for a simple task but, because the AI expanded its capabilities through its capacity for self-improvement, it has innovated to better produce paperclips, and developed power to appropriate the resources it needs, and ultimately to preserve its own existence.
“Bostrom argues that it would be difficult to control a super-intelligent AI – in essence, better intelligence beats weaker intelligence. Tweaks to the AI’s motivation may not help. For instance, you might ask the AI to produce only a set number of paperclips, but the AI may become concerned we might use them up and still attempt to eliminate threats. It is hard to program clear preferences, as economists well know.
“The conclusion is that we exist on a knife-edge. Turning on such an AI might be the last thing we do.”
How can we prevent AI from harming humanity? Experts believe the solution is for humans to build “kill switches” or “guardrails” within AI programming. We physicians don’t have the computer engineering skills to construct such solutions, but because we share a mission to protect fellow humans from disease and suffering, we must embrace the notion of AI safeguards.
- QUERYING AI ABOUT THE HOW TO PROTECT OURSELVES FROM AI
In a Business Insider publication, the authors queried OpenAI’s ChatGPT, Google’s Gemini, Anthropic’s Claude, and SpaceXAI’s Grok regarding how AI could lead to human extinction and how to stop it. Their 230-word prompt to the AI sites included the following language:
“Help explain, in clear and accessible terms, how AI could theoretically lead to human extinction. Identify the most commonly cited scenarios. Rank the scenarios from least to most plausible based on the best available expert analysis, making clear that these are highly uncertain forecasts, not established predictions. For each scenario, explain:
- How it could unfold
- What would need to go wrong for it to become catastrophic
- Why experts consider it more or less plausible than the other scenarios
- Which safeguards could reduce the risk”
ChatGPT responded that there’s much uncertainty around the topic and answered with the main ways AI could pose existential risks, ranking them from least to most plausible. The least plausible scenario was the “killer robot,” in which AI deliberately decides to eliminate humanity. ChatGPT also said that there was “no consensus that AI-driven human extinction is likely or inevitable.”
All four AI websites said human misuse of AI was a more immediate risk than AI itself. “Powerful AI systems could make it easier for bad actors to develop biological weapons, launch sophisticated cyberattacks, spread disinformation, or attack critical infrastructure,” ChatGPT said.
Gemini said “loss of control” was the most plausible AI doom scenario. Claude added that the loss of control could be gradual, with AI systems becoming so involved in economic, political, and military decision-making that human intervention might become increasingly difficult.
WHY ARE AI COMPANIES TRUMPETING THE DANGER OF THEIR PRODUCT?
Why are AI tech leaders bringing up the specter of AI ending humanity at this point in time? Some critics are skeptical why these companies, who stand to make enormous profit from their AI platforms, are emphasizing the danger of their product. Critics of AI point out the present-day harms and costs that AI is posing to society by displacing jobs and creating income and wealth inequality, as well as the fact that AI poses an escalating threat to the environment and consumes a growing amount of energy.
Perhaps AI companies believe it’s preferable if the public is fixated on AI’s role in future threats of extinction, rather than attacking AI companies for their present-day threats. A second reason AI companies may be trumpeting potential doomsday scenarios may be to promote the idea that AI has enormous power and potential. If AI is so powerful that it could end human existence, then companies can make the case that what they are working on is extremely important. As one publication wrote, “Doomsaying is great marketing.”
SCIENCE FICTION
In 2019, I published a prescient novel, Doctor Vita, about the introduction of Artificial Intelligence in Medicine (AIM) in Silicon Valley. The AIM in my novel became the antagonist in the story. Doctor Vita AI technology worked toward what they considered an intelligent solution—the elimination of terminally ill humans—to leave the rest of mankind as healthy survivors.
The Three Laws of Robotics, as authored by science fiction writer Isaac Asimov in the 1940s, were:
- A robot may not injure a human being or, through inaction, allow a human being to come to harm.
- A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
- A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
In I, Robot, a collection of short stories by Asimov, the AI/robots were unable to harm humanity. With Asimov’s laws, the threat of renegade AI harming humans would be impossible. But life in the 21stcentury will not be governed by a fiction writer’s decrees.
In the context of today’s AI circumstances and warnings, it seems advanced kill switches and guardrails are a necessary intervention. As physicians, we implore computer scientists to make AI as friendly to human evolution as possible. Perhaps an MD who also has a PhD in computer science/AI will develop the answers regarding appropriate kill switches and guardrails. Collectively, the rest of us physicians hope so.
*
*
The most popular posts for laypeople on The Anesthesia Consultant include: How Long Will It Take To Wake Up From General Anesthesia? Why Did Take Me So Long To Wake From General Anesthesia? Will I Have a Breathing Tube During Anesthesia?What Are the Common Anesthesia Medications? How Safe is Anesthesia in the 21st Century? Will I Be Nauseated After General Anesthesia? What Are the Anesthesia Risks For Children?
The most popular posts for anesthesia professionals on The Anesthesia Consultant include: 10 Trends for the Future of Anesthesia Should You Cancel Anesthesia for a Potassium Level of 3.6? 12 Important Things to Know as You Near the End of Your Anesthesia Training Should You Cancel Surgery For a Blood Pressure = 170/99?Advice For Passing the Anesthesia Oral Board Exams What Personal Characteristics are Necessary to Become a Successful Anesthesiologist?
READ ABOUT RICK NOVAK’S FICTION WRITING AT RICK NOVAK.COM
RICK NOVAK’S BOOKS ARE AVAILABLE AT AMAZON.COM
