A work of passion, love and Artificial Intelligence
I am an avid consumer of news and I love the dynamics behind politics. After co-founding another startup in Europe and leading it to a successful crowdfunding round I started thinking about something that always bothered me, which is how discussions around the media and politics are always heated and always generically point to some kind of unfairness. Media bias, misinformation, fake news, dishonesty in the news sector is something that every single person that discusses politics and news brings up and it frustrates me because there is a lot of bad news out there but it’s hard to navigate it.
So I started looking around to find tools to help me understand exactly and make sense of the issues with media and coverage. First I discovered a few tools to expose media bias but I rather quickly found them to be of little use for what I wanted: the idea of an artificial intelligence that can detect bias in an article is fascinating but I quickly realized it was solving a problem nobody really has. I don’t need an AI to tell me that Fox News skews right and the NYT skews left, I already know that. I respect services and platform that produce bias charts, they are useful and objective but they don’t tell the story of how much misinformation, bias and narrative pollute the news.
Not bias but coverage
I realized that bias isn’t the metric to judge news by, because what actually matters here is one thing: coverage. We already know which media outlets are conservative and which liberal, we already know HOW conservative and liberal they are and we already know they will treat people, stories and events according to their views. What really matters is the actual coverage. Bias, misinformation and narrative come from how much and for how long news is covered, on its own or compared to other pieces of news.
Bias, misinformation and narrative come from how much and for how long news is covered, on its own or compared to other pieces of news. Ideological bias is just too obvious to be a meaningful indicator…
For example, I already know that a certain media outlet is going to focus on bad news stories that involve Democrats, but how unbalanced is the coverage going to be? How many news stories about a Democrat are published compared with the number of stories about a Republican who is involved in a similar scandal? How many mentions of issues that are dear to a certain side appear on media outlets from the opposite side? How are current events covered compared to events of 5 years ago but with different politicians involved? In other words, how exactly is the media covering the news?
So this takes us to how people consume the news. To this day the vast majority of people get their news from traditional/legacy media outlets (roughly 70% according to Pew Research), be it on on mobile, desktop or in printed form, and all this media share a single point of entry: a main page, which in Web sites is of course the homepage. The home page is where news is given weight: news outlets decide what drives coverage by using pictures, by giving more space to the most important stories and by placing the most important content at the top of the page. So that’s precisely where we’re going to measure coverage…