When in comes to ethics lately, technology is often pointed to as the source of all evils. Bullying, biased or even unfair decisions, surveillance, behavioural manipulation are all enabled, supported or amplified by technology. At the centre of this “tehcnology against ethics” argument is one of the most defining concepts of our time: fake news. But what if technology could be used to solve fake news?
That’s what Lucas Azevedo is trying to explore, as described in his recent article on RTÉ Brainstorm: How to beat fake news with algorithms.
Although it is a term that has seen a very rapid increase in usage, at the highest possible level, fake news is far from new. From political debates to biased media, the need for “fact checking” has been present for several centuries already. What changes in the current world is the scale of the problem – there are too many things to check, and too many channels through which they spread.
What Lucas is trying to do is to train an algorithm to recognise fake news, i.e. to make an automatic fact checker. He uses machine learning approaches which have shown to be very effective recently in tackling many problems, including ones that are directly related to the problem at hand: Reading text to detect the clues that could indicate biased, wrong or false views. Based on those recent advances in AI, Lucas is optimistic:
“With enough data, a good selection of linguistic aspects and a well adjusted machine learning model, a simple and automatic way to the detect fake news might be closer than expected.”
Let’s hope he is right. The scale of the fact checking task is not only growing due to the number of sources of fake news increasing, it is also increasing by its audience. Fact checking is not anymore only needed by journalists. With Lucas’ algorithm, we might all sometimes soon get a little thing in our browser or on our phone telling us how much we can trust what we are reading.