EVERY day, algorithms make decisions that affect your life, whether it is social media firms determining the news you read or mapping apps choosing the route you will take. We are vaguely aware of this, but few people give it much thought.
It is time that changed. As machine learning advances, we are conceding more control to the algorithms, with potentially dire consequences. This is a particular problem in the public sector, where cash-strapped authorities are turning to algorithms they don鈥檛 fully understand in an effort to cut costs. In the UK, one police force is now investigating half as many reported assaults and public order offences on the say-so of a computer (see 鈥�A UK police force is dropping tricky cases on advice of an algorithm鈥�).
Advertisement
Of course, algorithms can be useful, like a neural network that can identify rare genetic disorders that a human doctor might miss (see 鈥�AI can identify rare genetic disorders by the shape of someone鈥檚 face鈥�). But even these can be misused: this particular algorithm studies faces to look for telltale signs of the disorders, meaning anyone with your mugshot could learn your medical history.
It is impossible for us to stop the march of this technology. For that reason, it is essential that we all understand it.
