ImproveTheNews.org is a website, created by Prof. Max Tegmark at MIT, that continuously ingests content from all (or at least most) news outlets around the globe and serves them up to readers based on simple selections they make on one or multiple sliders (follow Max Tegmark on Twitter). The two main ones are:
- Political Stance: Left — Right
- Establishment Stance: Critical — Pro
Under “more preferences” you will find a number of additional ways of shaping your daily news stories, but let us focus on the two main sliders at the top right. These two sliders basically let us view the world through the eyes of the “other guys,” the ones that voted differently from us. The ones that only focus on scandals from one side of the political spectrum versus the others that only focus on the opposite side‘s missteps. Go and play with these sliders and you will instantly know what I mean.
Yes, there are plenty of aggregators where you can simply provide a list of sources that are to your liking and read those on a daily basis. However, the “magic of machine learning” evaluates each individual article for its place on the political spectrum and serves it up accordingly. This provides readers with the nuance needed to understand the political “color” of what they are reading in every instance. In short, a news outlet can be in different places of the same side of the political spectrum, depending on author, topic, ownership, and lots of other factors. Machine learning will pick up on this and maybe surprise readers once or twice a day/week/month.
ImproveTheNews.org is a very interesting example of how we can easily use machine learning to reverse the effects of divisive media content, by letting readers “take a peak” of what’s going on at the “other side” of the spectrum. And as it could be our neighbor saying “Hello,” it is definitely worth taking a look and working on our understanding of why other people might see things differently. As importantly, exploring the site might inspire us to figure out more ways of leveraging machine learning for good.