How I Learned to Stop Worrying
And Love Automated Video Creation
The modern day version of “If a tree falls in a forest and no one is around to hear it, does it make a sound?” centers around computers auto-generating content. A few years ago there was a hue and cry about whether robo-news stories (mostly created with Automated Insights technology) would put journalists out of business. Today there’s a similar uproar over new video tools that automate some — or even all — of the creation process. Could this spell the end of video creators as we know them?
Much of the debate started when the head of TRONC (no, it’s not an energy drink, it’s a media company) said he’d be using these tools to create 750,000 videos a year. Then the New York Times published an article examining the claim, while highlighting two of the new companies that bring these fevered robot dreams into reality: Wibbitz and Wochit.
I heard about the debate from two separate groups. I run VidCon’s industry track, and there was a healthy debate over whether this was a sign of the apocalypse, or simply the death-metal thrashing of old-media dinosaurs. But I also work closely with one of the offending platforms, Wochit, who pulls creators down the path towards quickly creating video — without descending into complete automation like Wibbitz. Wochit CEO, Dror Ginzberg, espoused on the fallacy of robo-generated video in a recent piece on LinkedIn, which is worth a read for background as well.
The real question from both sides was whether computers can create video that’s as compelling as those made by real people. And the subtext from both parties — although of a different degree — was that computer-generated videos would be no better than the flotsam and jetsam of today’s distended tsunami of digital content.
I think they are both wrong. There’s nothing inherently bad about a fully computer-generated video as long as we’re honest about that video’s purpose — and clear on what automation can and cannot do.
Back when I was a print journalist we had two entry points into a story — the headline and the lead. We would attempt to suck people into the story with the second paragraph (the nut graph), and wrote in an inverted pyramid style that dominated journalism for years, but made it impossible to start anywhere but at the top.
Today every story of significance has multiple entry points. The goal of getting the customer (reader, viewer, viewser, whatever) to engage with your story is not a linear process with just 1–2 touchpoints. It’s a multi-faceted pastiche of different editorial elements that lets the customer choose her own adventure into any story of interest.
It might start with a tweet. Or an AI-generated note from a messenger bot. Or an email newsletter. Or a Facebook post. Or a YouTube video. Or something shared directly with you. Or perhaps even a short automated video that piques your interest.
But if we do our job as journalists those mental palate cleansers should lead to more nibbling — and perhaps even an engorgement. The next course might be an explainer video, or a longer medium post, or a story in a more traditional media outlet. It could continue from there into a discussion of the issues raised — or the emotions engendered — on Tumblr, Facebook, Twitter, WeChat or many other places where communities congregate.
Tools like Wibbitz — which do auto content generation — provide simplistic signaling. Wochit offers a broader canvas to the creator, albeit one purposefully foreshortened when compared to complex video storytelling tools like Adobe Premier and Apple’s Final Cut Pro. But just as robo-writing really only found success with a sliver of stories, completely automated videos will have a similarly narrow focus. It’s OK for simply rehashing facts — including sports and market news — but less so for more nuanced formats.
But whether you’re producing a completely automated story, or giving neophytes the tools to rapidly produce compelling video, automation serves a great purpose. Quick entry points into broader stories that lead to different ways to explore, extend and emote on that topic can only be net-positive in the news and trending topic world. By reducing — or eliminating — a dependence on the video priesthood, all content creators can now develop entertaining and informative videos that both stand-alone and provide a jumping-off point for further exploration.
Crap is crap. But if you have a tool that makes it easier to entice someone into a story of interest then I see that as a good thing. Because in the end the worst thing that can happen to a story is that it falls into the internet, and makes no sound.