The inventor of a controversial suicide pod is making sure his device keeps up with the times by augmenting it with AI tech — which, we regret to inform you, is not merely some sort of dark joke.
“One of the parts to the device which hadn’t been finished, but is now finished, is the artificial intelligence,” the inventor, Philip Nitschke, told the Daily Mail in a new interview.
Named the Sarco pod after the ancient sarcophagus, the euthanasia chamber, first built in 2019, has been championed by the pro-assisted dying organization The Last Resort. In 2024, it was used to facilitate the suicide of a 64-year-old woman in Switzerland. The 3D-printed pod is activated when the person seeking to take their own life presses a button, filling the sealed, futuristic-looking coffin with nitrogen that causes the user to lose consciousness and “peacefully” pass away within a few minutes.
To date, the woman’s death in Switzerland is the only case of the Sarco pod seeing real-world action. Soon after she died, Swiss authorities showed up at the sylvan cabin where the pod was located and arrested the late Florian Willet, then co-president of Last Resort, who was supervising her death, on suspicion of aiding and abetting a suicide. He was ultimately released two months later.
Assisted dying is technically legal in Switzerland, but only if the person seeking suicide is deemed to have the mental capacity to make the decision, and only if they carry out the suicide themselves, rather than a third-party.
That last bit is why the patient presses the button to activate the chamber, a workaround that stands on legally shaky ground as it is (hence Willet’s arrest). Even more contestable is determining whether the patient is capable of making their mortal decision — which, of course, is where AI enters the picture.
As Nitschke was designing a “Double Dutch” version of the Sarco pod that would allow couples to die together, he stumbled on the idea of using AI to administer a psychiatric “test” to determine their mental capacity. If they pass the AI’s judgment, it activates the “power to switch on the Sarco.”
“That part wasn’t working when we first used the device,” Nitschke told the Daily Mail, referring to the 64-year-old woman’s death.
“Traditionally, that’s done by talking to a psychiatrist for five minutes, and we did that,” he explained. “She had a rather traditional assessment of mental capacity through a Dutch psychiatrist.”
“But with the new Double Dutch, we’ll have the software incorporated,” Nitschke continued, “so you’ll have to do your little test online with an avatar, and if you pass that test, then the avatar tells you you’ve got mental capacity.”
Passing the test will power on the Sarco for the next 24 hours, during which a person or couple can climb in and press the button to end it all. If they miss the day-long window, they’ll have to take the AI-administered test all over again.
How this will actually work in practice is as questionable as the promises of AI itself. AI models are prone to hallucinating, are alarmingly sycophantic, and have frequently failed when deployed in medical scenarios. Entrusting the tech to give the go-ahead to someone killing themselves — especially as AI chatbots are under the microscope for seemingly encouraging many adults and teenagers to take their own lives in cases of so-called AI psychosis — sounds like an ethical disaster waiting to happen.
More on AI: A Man Bought Meta’s AI Glasses, and Ended Up Wandering the Desert Searching for Aliens to Abduct Him