There is a new artificial intelligence model that has a macabre and disturbing ability: predicting death.
Dubbed as a new “Death Bot,” it was developed by Danish scientists at the Technical University of Denmark, and its abilities published in scientific journals, according to a report by U.K. outlet The Independent.
The A.I. is similar to ChatGPT in that language embedded with data is used as the input to create a predictive styled outcome.
Input data included details such as health history, education, profession and income.
According to one of the developers, Sune Lehmann, the new A.I. program named life2vec, analyzed “human life as a long sequence of events similar to how a sentence in a language consists of a series of words.”
Sune said the fundamental question was, “To what extent can we predict events in your future based on conditions and events in your past?”
The data set was immense: the scientists used de-identified personal data of 6 million people residing in Denmark to feed into the Death Bot’s algorithm.
To see if life2vec could find patterns and correlate life events and choices into a predictive model, the data set included an age range of 35 to 60.
Data was drawn from the years of 2008 to 2016.
Should this technology be made available to the public?
The scientists then asked life2vec to predict which people would have died between the years of 2016 and 2020, based on the information given to it. The researchers already knew the real number of decedents, around half of the individuals in the data set.
Shockingly, the “Death Bot” accurately predicted the deaths of 79 percent of the included individuals.
These results make life2vec 11 per cent more accurate than any other current A.I. model attempting to correlate life events and death.
According to Science magazine, some of the factors that correlated to a higher risk of premature death were mostly intuitive: having a low income, having a mental health diagnosis, and being male.
Events such as accidents and heart attacks expectedly distorted the predictive outcome of the algorithm.
Beyond the grim prediction of untimely death, life2vec was also able to provide predictions of lifetime earnings and even personality aspects, such as being extroverted or introverted.
Lehmann said he hopes a good use of this new technology would be for improving health by identifying disease risk.
The developer also noted, however, that this also raises the ugly prospect of life insurance companies using life2vec as a means of assessing policy eligibility.
“Clearly, our model should not be used by an insurance company, because the whole idea of insurance is that, by sharing the lack of knowledge of who is going to be the unlucky person struck by some incident, or death, or losing your backpack, we can kind of share this burden,” Lehmann told New Scientist.
“We stress that our work is an exploration of what is possible but should only be used in real-world applications under regulations that protect the rights of individuals,” Lehmann said.
A glance at social media showed that Lehmann’s words were not creating a ton of confidence.
I think I’ve seen this film before.. and I didn’t like the ending
— JuJuBee (@bjuju5487) December 22, 2023
Ping @katherinebassil Let’s discuss the ethical concerns related to AI and the Life2vec model. Is it worth exploring the risk analysis and figuring out what it would entail? Let’s make sure society is fully informed and prepared to tackle any challenges that come one way. 🔍 👀 https://t.co/jQdk2ltSAy
— ninawickman (@wickman_) December 18, 2023
One X user cited Revelation 13:13-14 to question the wisdom of putting any stock into this technology.
That user quoted the verse as: “This second beast performed great miracles; it made fire come down out of heaven to earth in the sight of everyone.
“And it deceived all the people living on earth by means of the miracles which it was allowed to perform in the presence of the first beast. The beast told them to build an image in honor of the beast that had been wounded by the sword and yet lived.”
