When headlines proclaimed that artificial intelligence (AI) can be used to create a “death calculator” that predicts the day you will die, it sounded like something from a terrifying science-fictional story. The reaction showed how readily people believe that AI has magical fortune-telling powers.
The reality was not as far-fetched. The paper that spawned the fracas, in the journal Nature Computational Science, did involve using AI to predict death, but it was not very precise. Using both economic and health data on thousands of people in Denmark, an AI-based system was able to predict with about 78 percent accuracy which people would die within the next four years.
The algorithms used to create actuarial tables already do this kind of statistical forecasting, but the new system, called life2vec, is more accurate and works in a completely different way. The lead author on the paper, University of Copenhagen complexity science professor Sune Lehmann, said life2vec predicts life events much the way ChatGPT predicts words.
The findings matter not because they might create a scarily accurate “death calculator,” but because of how the forecasts could be used. Such algorithms could be used for ill — to discriminate or deny people healthcare or insurance. Or they could be used for good, by highlighting factors that affect lifespan and helping people live longer. Or they might improve lifespan calculations, which some people use to plan their retirements.
It was “wild to see how the results were misrepresented,” Lehmann said. “People said this AI can predict the second you will die with incredible accuracy.”
This is because people do not understand the technology yet, and as science fiction legend Arthur C. Clarke has said, any sufficiently advanced technology is indistinguishable from magic.
At the same time, hospitals are incorporating AI to do all sorts of jobs. Will doctors and hospital administrators put too much faith in the decisions or forecasts of AI because it is fast and sounds confident? Can the medical system use AI responsibly if people have unrealistic or magical ideas about what it can do?
Lehmann said his work in this area is aimed at testing the powers of prediction for all kinds of life events, including job changes, income changes and moving.
He is looking for a more coherent scientific understanding of the way algorithms can predict complex phenomena, he said.
Often algorithms’ workings are treated as a mysterious black box. The researchers did not choose death out of any morbid preoccupation but because it is something that is precisely measured and recorded.
In groups of young people, the question is too easy — you would be mostly correct if you predicted that nobody dies over the next four years. Predicting death within one year is not too hard — you would just have to know who was sickest. The further out you go, the harder the future is to predict, until you get far enough ahead that almost everyone will have died.
At this stage, then, AI is not likely to surprise anyone about their life expectancy. If you are healthy and not extremely old, it would predict you would live more than four years. It cannot foresee that you would get in a freak accident or predict whether you will die in 10, 15 or 20 years, said Andrew Beam, a professor of biomedical informatics at Harvard Medical School.
Beam said there is a risk that AI could prompt humans to be misled by authority bias: “If you think someone is smarter than you or has access to information that you don’t have, there’s a real tendency to turn off critical thinking and believe anything that comes out — whether it’s a person or an AI.”
ChatGPT is good at synthesizing information, but it is not very selective and could fold in bad studies and flawed information. “So, if you’re in an area where the science is unsettled or the human knowledge is just not there yet, ChatGPT is going to be just as bad if not worse than a person,” he said.
Predicting a healthy person’s long-off death is just science fiction, he said: “We need to be careful when we’re asking it to do things that are still clearly sci-fi.”
Sometimes fiction can provide a reality check by reminding us that our actions influence the future — even in cases of life and death. Consider what happened in the classic Charles Dickens story A Christmas Carol. The Ghost of Christmas Future gave Ebenezer Scrooge a terrifying preview of loneliness, grief and death. Scrooge then asks a smart, critical question: “Are these the shadows of the things that Will be, or are they shadows of things that May be, only?”
If the reporters trying to scare people with life2vac had asked that question, they would have gotten the same answer Scrooge did from the ghost: Of course our actions can change the future. A forecast does not seal our fate in stone.
This new system reinforces what other studies have shown — that income and job type can affect the length of your life. Being poor and having a job where others have power over you is correlated with premature death. It is something Dickens recognized long ago. Maybe AI can turn this general observation into poignant real-life scenarios that could motivate modern-day Scrooges to address the inequalities that shorten so many lives.
F.D. Flam is a Bloomberg Opinion columnist covering science. She is host of the “Follow the Science” podcast. This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Trying to force a partnership between Taiwan Semiconductor Manufacturing Co (TSMC) and Intel Corp would be a wildly complex ordeal. Already, the reported request from the Trump administration for TSMC to take a controlling stake in Intel’s US factories is facing valid questions about feasibility from all sides. Washington would likely not support a foreign company operating Intel’s domestic factories, Reuters reported — just look at how that is going over in the steel sector. Meanwhile, many in Taiwan are concerned about the company being forced to transfer its bleeding-edge tech capabilities and give up its strategic advantage. This is especially
US President Donald Trump’s second administration has gotten off to a fast start with a blizzard of initiatives focused on domestic commitments made during his campaign. His tariff-based approach to re-ordering global trade in a manner more favorable to the United States appears to be in its infancy, but the significant scale and scope are undeniable. That said, while China looms largest on the list of national security challenges, to date we have heard little from the administration, bar the 10 percent tariffs directed at China, on specific priorities vis-a-vis China. The Congressional hearings for President Trump’s cabinet have, so far,
US political scientist Francis Fukuyama, during an interview with the UK’s Times Radio, reacted to US President Donald Trump’s overturning of decades of US foreign policy by saying that “the chance for serious instability is very great.” That is something of an understatement. Fukuyama said that Trump’s apparent moves to expand US territory and that he “seems to be actively siding with” authoritarian states is concerning, not just for Europe, but also for Taiwan. He said that “if I were China I would see this as a golden opportunity” to annex Taiwan, and that every European country needs to think
For years, the use of insecure smart home appliances and other Internet-connected devices has resulted in personal data leaks. Many smart devices require users’ location, contact details or access to cameras and microphones to set up, which expose people’s personal information, but are unnecessary to use the product. As a result, data breaches and security incidents continue to emerge worldwide through smartphone apps, smart speakers, TVs, air fryers and robot vacuums. Last week, another major data breach was added to the list: Mars Hydro, a Chinese company that makes Internet of Things (IoT) devices such as LED grow lights and the
RSS