In an age where we can’t even agree on what facts are anymore, it can seem like a lost cause to even attempt political fact-checking.

And to be fair, right now things don’t look great.  However – and this is a topic that I have a history with – I think it’s a cause absolutely worth supporting.

Back in the day, I was a cofounder of a startup called fuseGap.  It was a political/educational social app with the goal to make it cool to be informed about politically-relevant facts.  So instead of who can bring the spiciest political opinions…it would be more about who could bring the best facts.

Given that I’m here writing this…yeah it didn’t work out.  We learned a lot, but ultimately it’s really, really hard to get people motivated to learn facts in a golden age of the political hot take.

Lesson learned: people are way more interested in providing and consuming opinions than in actually looking to see if those opinions have any basis in reality.  Of course, few people would admit that they feel this way personally…but I’m pretty sure it’s true.

Anyways, all that doesn’t mean that it’s not important to have the general population informed about political matters – if anything, it’s more important than ever.

Which brings me to a cool startup I’ve been looking into: Full Fact.  They’re a startup based in the UK, and they’re using AI to automate political fact checking.

 

Why does it matter?

Given that most news today is more about ‘engagement’ and basically getting the viewers riled up (as opposed to…reporting on reality) – it seems that fact-checking is unfortunately becoming a relic of the past.

This has been going on for a while now: flashback to the 2012 presidential campaign, when Mitt Romney’s campaign hit us with the “We’re not going to let our campaign be dictated by fact-checkers.”  Fun times.

Things haven’t gotten much better since then.  Depending on who you ask, some might say that the UK and USA are starting to get torn apart by rampant polarization.

When many media outlets are more interested in making you think your fellow citizens are the enemy, vs actually reporting the facts…I’m not sure that will end well.  But hey, at least the audience will be ‘engaged!’

So getting back to Full Fact.  While most politicians are out there throwing hot takes left and right, it’d be nice to have someone checking whether their spiciest statements have any basis in reality.

The problem is, fact-checking is hard.  While also being not exactly the most fun activity, and not the most profitable either.

And with more and more people getting politically outraged and outspoken about all sorts of things, there are more and more impactful hot takes to sift through.  It’s becoming more of a scalability problem…

…which sounds like a great use case for artificial intelligence!

 

Full Fact’s automated fact-checking project: some traction, momentum

Full Fact has some decent momentum, along with some big-name players supporting them in a few different ways.

In 2016 they announced a partnership with Google’s Digital News Initiative, and their automated AI fact-checking approach has been featured on BBC.  They’ve also been covered by TechCrunch, Wired, The Guardian, among others.

More recently, in May 2019 they were co-winners of the Google AI Impact Challenge.

Again, fact-checking is hard.  It’s generally a thankless endeavor, and pretty much no one gets passionate about supporting the fact-checkers.  It’s only when the facts happen to support someone’s opinion that people will generally even acknowledge the effort.

 

Conclusion

I think it’s cool to see people using AI to attack very impactful, very immediate real-world problems.

I do have some longer-term questions about how well Full Fact’s approach can handle the complexities and subtleties of fact-checking, but so far they certainly are off to a good start.

Now if only we could get people to care more about facts instead of opinions…

 

The views expressed on this site are my own and do not represent the views of any current or former employer or client.