Wednesday 17 March 2021

What Technology will Predict About Us in the Future - Felicia Yap

 

I predict that computers will soon know us better than we do ourselves. They will be able to predict our futures, how we will live – and perhaps even when we will die. More controversially, computers will also be able to predict a person’s likelihood of becoming a murderer (how this information will be used is an altogether different matter). 

I also predict that humanity will be divided into two kinds of people: people who will tell computers what to do, and people who will be told by computers what to do. A large proportion of the human population will fall in the second category, perhaps more than 85 percent. In the future, the digital divide will not be between the haves and have-nots. It will be about the control of technology: those who control or those who are controlled.

Here are my predictions of what computers will predict: 

Our Likelihood of Becoming Murderers

Law enforcement units are increasingly able to predict when and where crimes will occur. Scientists at Carnegie Mellon have developed CrimeScan, a crime-predicting software based on the premise that violent crimes tend to happen in geographic clusters and which takes into account day-of-the-week and seasonal trends. At least fourteen police forces in the United Kingdom have deployed software (such as PredPol) which identifies ‘hotspot’ areas where crimes are more likely to happen. Some forces are also involved in the National Data Analytics Solution project which combines machine-learning and information held by the police (such as conviction histories) to work out risk scores for individuals and predict their likelihood of committing crimes. 

The interesting question is who, not just when or where. Our data trails can already pinpoint our profiles with increasing accuracy. Cambridge University scientists have shown that Facebook Likes can ‘automatically and accurately predict a range of highly sensitive personal attributes including: sexual orientation, ethnicity, religious and political views, personality traits, intelligence, happiness, use of addictive substances [and] parental separation’. 

Will computers predict if we are likely to become murderers? My novel Future Perfect is partly about a secret service software named CriminalX that scans people’s backgrounds and data trails (such as their Google histories) to prevent future homicides and terrorist attacks. Research trends are already heading in this direction. Scientists at the University of Texas have shown that murderers are more likely to have lower IQ and had suffered greater exposure to violence, while researchers at the University of Pennsylvania have found that convicted murderers are more likely to have offended at a young age. Researchers from Harrisburg University have apparently built software that can predict ‘if someone is a criminal based solely on a picture of their face’ with ‘80 percent accuracy and with no racial bias’, although their research have yet to be published by a reputable journal. In 2016, researchers from Shanghai Jiao Tong University suggested that criminals have upper lip curvatures that are 23.4% larger than non-criminals and a slightly-narrower distance (5.6%) between the inner corners of their eyes, only to be refuted by researchers at Google (some of these projects have sparked public backlashes and much hand-wringing about the inherent biases of the algorithms used). 

How Our Days Will Unfold 

I predict that computers will be able to tell us how our days will look like, what we will be doing next. This is because humans are predictable creatures of habit. Computers will soon gather so much data about us (including our daily schedules and consumption preferences), they will be able to predict our short-term prospects with ease. In Future Perfect, an app called iPredict provides forecasts about what will happen to a person over the next two days – and a large swathe of humanity is, predictably, addicted to the app.

Our Health (and Our Likelihood of Dying)

I predict the widespread use of digital-phenotyping gadgets, such as mirrors that can read facial features, expressions and even emotions to infer a person’s health status. Researchers are already using the bounteous data on YouTube, Facebook, Twitter, and Instagram to create algorithms that detect autism, HIV, obesity, schizophrenia, Parkinson’s and Alzheimer’s. Our mobile phones, Fitbits and sleep-tracking rings will amass so many health metrics about us (such as steps taken, respiratory rates, body temperature and sleep quality), they will be able to predict our life expectancy, perhaps even our likelihood of dying on a particular day.

The computer scientist Alan Kay wrote that ‘the best way to predict the future is to invent it’. Yet some think the best way to predict the future is to prevent it. Having set down these predictions, I doubt whether some of these are truly desirable. After writing a book set in the future, I have realised that pre-cognition may not necessarily make a person happier. Will knowing when and how we are going to die improve our lives? The main problem with information is knowing what to do with it, to use it in a way that improves our physical and mental well-being. Knowledge can paralyse more than liberate. As the world hurtles inexorably in the direction of predictability (and as we get increasingly deluged by data and smothered by gadgetry), we may realise that unpredictability is what makes us human, what makes life worth living. In a world where computers will know us better than we do, the essence of our humanity may lie in not wanting to know. 

Future Perfect by Felicia Yap (Published by Headline Publishing Group) Out Now

What if today was your last day... A bomb has exploded during a fashion show, killing a beautiful model on the catwalk. The murderer is still at large... and he may strike again. Yet this is the least of Police Commissioner Christian Verger's worries. His fiancee Viola has left him. He has to keep his tumultuous past a secret. To make things worse, his voice assistant Alexa is 99.74% sure he will die tomorrow. Moving from snowy 1980s Montana to chic 1990s Manhattan to a drone-filled 2030s Britain, Future Perfect is an electrifying race to solve a murder before it's too late. Yet it is also a love story, a riveting portrait of a couple torn apart by secrets, grief and guilt. A twisted tale of how the past can haunt a person's future and be used to predict if he will die... or kill.

No comments: