Danielle Van Dalen
The cracks are starting to show on the fancy computer algorithms that we rely so heavily upon these days.
The recent news that Facebook was involved in sharing the personal information of quiz participants and their friends with Cambridge Analytica has illustrated the creeping influence algorithms have had on our lives.
After the story broke, the New York Times reported that “the hashtag #DeleteFacebook appeared more than 10,000 times within a two-hour period,” showing that people are starting to recognise that our increasing reliance on algorithms is not all positive.
We need to remember that technology cannot be a substitute for relationships.
Powerful tools
Algorithms are powerful tools for creating efficient and targeted processes used in our day-to-day lives. They do a lot of awesome things.
Companies like Netflix and Uber harness them to ensure that the right information gets to the right people, like which movies to advertise for whom, or which driver to connect with which passenger.
Unfortunately, as the Cambridge Analytica example suggests, there are also risks to the widespread use of algorithms and the personal data they so often rely upon; and privacy breaches are just one example. In many less-obvious areas of our lives, the algorithms are showing their fair share of bugs too.
Healthcare options
In 2016, the Arkansas Department of Human Services began to use algorithms to determine healthcare options and supports for different people. The algorithm led, however, to significant changes and reductions in many people’s medical care – one person even claimed, “they were hospitalised because their care was cut.”
A court eventually discovered that the software vendor had “mistakenly used a version of the software that didn’t account for diabetes issues.”
Bad predictions
While not quite as life-or-death, companies like Shell have also been using algorithms to fill job roles but are now finding that these algorithms are “bad at predicting rare events, such as when employees would excel at a task they haven’t encountered before.”
So, while we have seen that the potential of technologies like this and get understandably excited about the future possibilities and efficient systems we might create, these examples show that completely relying on these processes is not always the answer.
Sometimes they get it wrong.
It is easy to quickly trust the outputs of a computer, but as these examples show, we simply cannot afford to be so naïve.
The response
So how do we respond?
While quitting Facebook seems like a good start for many – even futurist and technological entrepreneur Elon Musk deleted his company’s accounts – becoming a technological hermit does not seem helpful, or even possible, anymore.
Besides, the efficiency-related benefits of algorithms are hard to give up.
Instead, we need to be aware of their limitations and wary of their potential unintended consequences.
We need to remember that algorithms often fail to see the nuance that only human interaction can detect.
From social media to medical care or potential employment, let us introduce some more healthy suspicion when it comes algorithms.
Danielle van Dalen is a Researcher at the Maxim Institute based in Auckland.
