There are times, in today’s reasoned and rational political discourse, where you’ll see some politician say something and find yourself thinking “Is that actually true? I wish there was a way I could easily check if they’re making that up.” Well, Curious George, it turns out that others have not only had the exact same thought, but they’ve done something about it.
The Washington Post has released the prototype of “Truth Teller,” a new app that offers the ability to fact check political speeches in “as close to real time as possible.” The origins of the app came when Steven Ginsberg, the app’s creator and national political director for the Washington Post, found himself at a political rally listening to information he knew was false. “It was one of those small parking lot affairs outside a sports bar and the candidate was there speaking to about 30 people,” Ginsberg said. “For about 45 minutes [she] said a lot of things that I knew to not be true, and nobody else there knew that.”
Ginsberg and the Post took the idea to the Knight Foundation and applied to the Knight News Challenge – which funds “breakthrough ideas in news and information” from a collective pot of $5 million. The team ultimately received $500,000 from the Foundation’s parallel initiative, the Knight Prototype Fund, created to offer similar funding for new ideas to push media and news reporting forward technologically, but with a shorter time cycle than the News Challenge.
Writing about the app on the TruthTeller site, the Post’s Executive Producer for Digital News, Cory Haik, described that the current prototype combines multiple technologies, including audio/video indexing and transcription, as well as approximate string matching and “fuzzy string searching” algorithms to try to recognize content of the speeches. “We are effectively taking in video, converting the audio to text (the rough transcript below the video), matching that text to our database, and then displaying, in real time, what’s true and what’s false,” he explained, adding that the result is that “what you see in the prototype is actual live fact checking — each time the video is played the fact checking starts anew.”
The prototype is available here, pre-loaded with multiple speeches to demonstrate the app in action.
Haik admits that it’s a work in progress. “It needs more technical work and we need more facts,” he wrote. “For instance if you stumble across what you think is a false positive, let us know… we’re tuning the algorithm as we go. It’s a proof of concept, a prototype in the truest sense.” He’s bullish about the potential of the app, however. “Do we think this can be applied to streaming video in the future? Yes. Can this work if someone is holding up a phone to record a politician in the middle of a field in Iowa? Presenting the truth is without dispute one of the most important missions of journalism. So yes, we believe it can.”
- A robot named Heliograf got hundreds of stories published last year
- Bye humans! The Washington Post is using a robot to report on the Rio Olympics
- Reading coverage on the RNC? It might be coming from bots