'Extremely accurate' AI death calculator predicts when you'll die

If there weren’t enough reasons to be fearful of the artificial intelligence boom, scientists have found that it can be used to predict when you’ll die.

Just like with the introduction of any new technology, artificial intelligence (AI) has sparked a slew of fears.

We might look back and laugh just like we did at those who were fearful of the radio, television and the internet.

And while not everyone wants to know exactly how long they have to live, a newly developed AI death calculator, known as 'life2vec', seems to be able to predict when a person will die with surprising accuracy.

You would be pleased to know we haven’t strayed so far into sci-fi that a computer program can simply look at you and give you a date of death - but it is still slightly alarming regardless.

“We use the technology behind ChatGPT (something called transformer models) to analyze human lives by representing each person as the sequence of events that happens in their life,” Sune Lehmann, lead author of the December 2023 study told the New York Post.

“Using sequence of life-events to predict human lives”

Getty Stock Image

Lehmann, who is a professor of network and complex systems from the Technical University of Denmark, explained how the algorithm they'd developed works.

It factors in the typical things such as income, profession, residence and health history — to determine life expectancy with 78 percent correctness.

So essentially with enough information, the model can get determine what your personality is like - as well as when you are likely to die.

An impressive piece of tech to put it bluntly.

Lehmann’s group used the algorithm and predicted the life expectancy of 6 million Danish people who varied in sex and age to discover which of the subjects would likely live for at least four years beyond January 1, 2016.

Getty Stock Image

Using the public information of its subjects and assigning specific 'digital tokens' to each piece of data, life2vec correctly predicted who had died by 2020 more than three quarters of the time.

Thanks to the sheer scale of the data collected, it had allowed them to 'construct sequence-level representations' of their 'individual human life trajectories'.

The report adds: "We can observe how individual lives evolve in a space of diverse event types (information about a heart attack is mixed with salary increases or information about moving from an urban to a rural area).”

Lenham also stressed that none of the participants had been told their predictions, calling it 'very irresponsible' to do so.

It's certainly eerie stuff but less so when you consider the facts that with enough information and time, you could probably do the same thing - though it would admittedly make for some morbid conversation.