When watching an interview on TV, I’ve always wondered how much of what an interviewee says actually made it on-air. Or even more important, how much of their intent is being conveyed due to editing in post-production.
Editing a video in post-production is normal and typically honest. To fit within time constraints, for example. However, there are times when nefarious networks and editors will clip and/or re-order the interview to push an agenda or dictate the narrative.
A potential example of this, and one of the clips inspiring this app, is Elon Musk’s 60 minute interview where Musk pointed out “a very misleading edit“, calling for 60 minutes to release the full transcript.
What could have Musk done to prevent – or at least give viewers the information at their disposal to determine if and when edits were made? What if 60 minutes actually did release the full video. How could Musk “sign off” on the authenticity of the video, and more important, how could viewers confirm this claim?
The Idea
As far as I could tell, the problem of protecting a video clip, and being able to verify its authenticity was an unsolved problem.
What if Elon had a timer in frame during the interview? That would give viewers information they have never had, information on when and how long edits are. That would solve the “hey they re-ordered what I said” and “wait, they cut off my entire second sentence”.
What if there was a way that some sort of signature could be displayed, that only Elon had the “key” to reproduce AND it would be really cool if there was an easy way to share that key AFTER the interview aired?
“I know” popped into my head. I could use a Pseudorandom number Generator (PRNG) to create a reproducible signature from a key that only the interviewee knew. “I could also display a timer along side the data”. Hmmmm. Nerdy…. novel…. I like it!
I’ve been wanting to learn NativeScript, especially since they support deep Angular integration. This would be a fairly trivial implementation and something good to evaluate NativeScript with.
The Solution
ClipGuard. An app that displays a timer synchronized with data points created by a PRNG:
All Elon has to do now, is get the app in frame (put on table, mount, hold etc) and start the sequence. While the interview is taking place, every second that the counter ticks, a new PRNG datapoint is generated. The value of the datapoint is computed based off a “seed” (a number between 1 and approx. 2 billion) that was chosen by Elon before starting the sequence. When the interview concludes, he hits “Stop” and is given a QR code that includes the “key” to reproduce the sequence. He saves the QR code someplace safe.
After the interview airs, he shares the QR code on Twitter. Viewers can scan the QR code inside the ClipGuard app to reproduce the sequence at any point in time – authenticating what they are watching on the video.
Try It Out
Admittedly this app is only useful for a handful of people (like if you are ever stuck giving a deposition). However, if you want to try it out, the app store links are below. If you have any suggestions I’d love to hear ’em. Head over to the post on Hacker News or open an issue on github.
Ryan Pendergast 