All things remaining unchanged, I’m going to die on Saturday, Nov 6, 2049. So the Death Clock told me. The Death Clock uses your age, your BMI, your gender, your Mode (e.g., whether you are optimistic, pessimistic, sadistic, or “normal”), and your smoking status to calculate the date of your demise. More or less, it's the same data that actuarists use to estimate your health insurance. The problem is your BMI could change, and the Death Clock ignores variables like family history, environment, race, economic status, career choices stress, alcohol, sleep patterns and nutrition. Given the incompleteness of this test, hopefully I’ll die far older.
In a tractate of the Talmud, Ethics of the Fathers, Rabbi Eliezer tells his students “Repent one day before your death.” They question him: "How can you know when you’re going to die?" He tells them: "Since you never know, repent each day."
In January, 2018, Stanford scientists developed a deep learning model that tells your death to the minute, five years before you die. If you're seriously ill, that is.
How does it work?
The Stanford team used a deep learning approach where they fed their computers data on two million patients from Stanford Hospital and Lucile Packard Children’s hospital who had died between 1995 to 2014. The data came from the Electronic Health Record (EHR) containing details of the diseases, medical histories and the severity of diseases of these patients.
The computers were trained to answer the following algorithm:
Given a patient and a date, predict the mortality of that patient within 12 months from that date, using EHR data of that patient from the prior year.
When the algorithm was tested on the records of another 40,000 patients, it was correct in predicting when the people would die in three to twelve month within 90 per cent of the cases.
Ken Jung, a Stanford Medicine research scientist and co-author of the study, told Gizmodo:
“All too often, advanced illness turns to a medical crisis, and patients end up in the ICU. There, events can attain a momentum of their own, resulting in increasingly aggressive interventions that often do not serve patients and their families well. One of the goals of the palliative care team is to engage in conversations with patients so that they can think through and articulate their preferences before they are in a crisis.”
As Dr. Mukherje in the New York Times observed, cats can sometimes detect death better than doctors. Too many patients who need palliative care are turned away, or people who are mistakenly thought to be at the end of their lives are referred for palliative care. This AI helps doctors correctly inform patients when their time is up.
What does this AI give you?
Studies show that more than 80 percent of individuals want to die at home peacefully with partners, spouses, and their children holding their hands. In fact, when 92-year-old Barbara Bush recently died, NPR reported that a high interest term was “end-of-life” comfort care.. People wanted to know how she could predict her death to make her own arrangments, and how they could, too.
As the Stanford researchers noted, only six out of every eight people know where they're ready to die. Doctors get it wrong in 20 percent of the cases. Doctors are overly optimistic that their patients will live. They try to push life to and beyond the end with invasive, painful procedures. Some clinicians don’t refer patients because they believe in positive thinking and don’t want their charges to give up. Other times, stress of their jobs causes professionals to overlook those who are terminally, or critically, ill.
“We want to make sure the sickest patients and their families get a chance to talk about what they want to happen before they become critically ill and they end up in the ICU,” Jung says.
That’s the function of this deep-learning model.
The “death tool” helps your family prepare to survive after you’re gone. If you have underage children and are a single parent, it gives you time for final arrangements. A parent? To stuff in time with your children (or anyone you deem valuable). Generally, patients use it to prepare for their deaths and to tie up loose ends.
Others go the Queen Latifah’s path in Last Holiday emptying their savings on a luxury holiday in Europe before they die. Or riding motorcycles on the Great Wall of China, as Carter Chambers and billionaire Edward Cole in the movie, Bucket List.
In his book, Barking Up The Wrong Tree, Eric Barker writes that SEALS leverage their grit by dwelling on their deaths.
What could be a stronger motivator than this AI machine that tells you your death to the minute?
Over the years, several people who have ‘died’ or had near death life experiences told me how it felt like. Few accounts contained the alleged inclusions of soul slipping through tunnel, of meeting a Jesus-like being, or of reunion with loved ones who had died.
All described the actual moment of death as intensely pleasurable, although the journey to that moment was harrowing and painful. For most people and their families the greatest agony came from not knowing when they or their loved ones would die.
Stanford's AI system has a 90 percent predicted outcome tested on 40,000 patients. And that’s far more conclusive than the prognosis of the Death Clock.