Creators of infamous 'AI death calculator' issue warning about fake sites stealing user data
Technology
Copycat apps have flooded the market
(Web Desk) - Creators of the 'AI death calculator,' which claims to predict when a person will die with 78 percent accuracy, have issued a warning to the public about a new threat.
Researchers at the Technical University of Denmark have discovered copycat apps have flooded the market since they unveiled their technology, called life2vec, last year - but the team noted their AI would not be made publicly available online.
The team uncovered four, fake 'death-foretelling' AIs touting 'counterfeit services': AI Doom Calculator, Intelligent Death AI, Death Predictor, and the ambiguously named Telecharger.
'We are aware of social media accounts and at least one fraudulent website claiming to be associated with the life2vec model,' the researchers alerted morbidly curious internet users via their life2vec AI's official homepage.
'We are not affiliated with these or any other entities that claim to use our technology,' they noted.
The Life2vec team, who first revealed their powerful AI last December in the journal Nature Computational Science, hopes that the public will 'be careful of' dangerous copy-cat scammers that 'have nothing to do with us and our work.'
Despite a public clamoring for the chance to test their own future against live2vec's predictions, researchers has had to keep the operational AI under wraps, to protect the personal information of the people whose data were used to train the system.
That is because personal information on over six million real people, including income, profession, place of residence, injuries, and pregnancy history, was fed into the algorithm, as lead researcher Sune Lehmann told DailyMail.com.
And, as such, Lehmann said, live2vec is not available for the general public — or companies — to use.
'We are actively working on ways to share some of the results more openly, but this requires further research to be done in a way that can guarantee the privacy of the people in the study,' said Lehmann, professor of networks and complex systems.
Even when the model is finally available to the public, Danish privacy laws would make it illegal to use life2vec to make decisions about individuals - like writing insurance policies or making hiring decisions.
'The model opens up important positive and negative perspectives to discuss and address politically,' as Lehmann told Newswise.
'Similar technologies for predicting life events and human behavior are already used today inside tech companies that, for example, track our behavior on social networks, profile us extremely accurately, and use these profiles to predict our behavior and influence us.
'This discussion needs to be part of the democratic conversation so that we consider where technology is taking us and whether this is a development we want.'
This potential for corporate or governmental abuse is of particular concern to the researchers behind live2vec, because of their systems above-average accuracy.
The model, trained on data from 2008 to 2016, proved capable of correctly predicting who had died by 2020 more than three-quarters of the time.
The tech does so by analyzing the user's life story - their text prompts.
Lehmann and his team assigned different tokens to each piece of information, and these pieces of data were all mapped out in relation to each other.
Categories in people's life stories run the whole range of human experiences: a forearm fracture is represented as S52; working in a tobacco shop is coded as IND4726, income is represented by 100 different digital tokens; and 'postpartum hemorrhage' is O72.
Many of these relationships are intuitive, like profession and income - certain jobs make more money.
But what life2vec does is map the huge constellation of factors that make up an individual's life, allowing someone to ask it to make a prediction based on millions of other people and many many factors.
It can also make predictions about people's personality.
The test asks respondents to rate 10 items based on how much they agree, items such as 'The first thing that I always do in a new place is to make friends,' or I rarely express my opinions in group meetings.'
It's important to note, Lehmann said, that the data were all from Denmark, so these predictions may not hold true for people living in other places - besides the fact that most people probably don't actually want to know when they will die.
