This morning, I watched an ad on TV for a new Master Vintner at a winery – IBM Watson, a super-computer – which will replace the people that make the wine now. Two days ago, Elon Musk was talking to a conference of US Governors about the dangers of unregulated Artificial Intelligence; which is more commonly called “Machine Learning.”
These made me realize that a new world of automated audio and video editing will soon be upon us. In fact, I’ve seen a variety of tests over the past 18 months that leads me to believe automated editing will be a common reality in two years or less.
Naturally, this new technology – like all technology – will be pitched as a “time-saving device” for today’s overworked editors. And, while that’s true, the Law of Unintended Consequences also says that this new automated editing technology will put lots of us out of work. Probably lots and lots of us. And that’s scary.
The Law of Unintended Consequences says that this new
automated editing technology will put lots of us out of work.
Because much of the editing that most of us do isn’t high-end story-telling requiring a high-degree of craft. It’s illustrating a talking head, creating a text video for social media, or editing highlights of a sporting event. And these tasks are easier to automate than story-telling.
Developers are rushing to create these new automated tools because this new technology is exciting and truly promises to revolutionize editing; and, like the rest of us, developers need to eat. But, as editors, we are on the other side of the coin. If we are not careful, we’ll find ourselves on the outside looking in.
Now, I’m not suggesting we rail against technology and burn our computers. First, because it’s pointless; complaining about technological change is as old as humanity and tech hasn’t slowed down yet. Second, because complaining misses the bigger point: if we know this tectonic shift is coming, then we need to start preparing for it today. The worst time to plan ahead is when you are out of work with a mortgage looming.
In the past, I’ve written that we need to improve our story-telling skills. But, I realized this morning that story-telling is not sufficient, because many of the videos we create are not, strictly-speaking, stories. Instead, they are illustrations. As I thought further about this, I realized that there are five things we can do to guard against becoming obsolete.
The worst time to plan ahead is when you are
out of work with a mortgage looming.
First, it is said that people skills are 50% of the skill set of an editor. That has never been more true that today. What will preserve our jobs going forward is not our story-telling, but our humanity. The one thing people will always do better than machines is build relationships. This means that we need to put a renewed focus on finding, keeping and growing our clients.
Second, take a close look at the finance side of your business. What’s profitable, what’s not profitable? What’s growing, what’s falling into disuse? Everyone is different, but if you don’t know, deep down, where your money is coming from, now is precisely the right time to figure this out. Lean days are coming – start thinking today about what you need to keep and what you need to let go.
Third, stay on top of technology – and not just the professional video space. The software that threatens us is coming from consumer software and mobile devices. Consumers want faster ways to edit their iPhone videos. At a certain level, video is video. If it works on a mobile device, it will fairly soon migrate up into the professional level. We only need to look back to 1999 – 2000 and track what the introduction of DV video did to our industry to see the impact of a “consumer” format on the “professional” industry.
Fourth, don’t be content with your current products and services. Talk with your clients and customers to discover what new products, new services, or new approaches they want. There is nothing scarier than asking a client: “What do you need that I’m not currently providing?” or “What can I do better?” But, out of that conversation will come ideas that you may not have thought about before. Our industry is in a constant state of re-invention. Have the courage to apply some of that reinvention to ourselves.
Fifth, when Apple introduced Final Cut Pro X to howls of dismay from many, the problem wasn’t simply how different FCP X was from FCP 7. The bigger problem was that we, as editors, were defining ourselves by the tools we use, not the results we created. We proudly walked around saying: “I’m an Avid editor,” or “I’m a Final Cut editor.” As those tools evolved, as they must, our view of who we are became threatened.
We can’t do this to ourselves again. Don’t define yourself by your tools. Don’t define yourself by your workflow. Define yourself by your results and the benefits you deliver for your clients. Tools change. Workflows change. We need to create the flexibility in our mind and our skills to change with them, without redefining who we are in the process.
Don’t define yourself by your tools. Don’t define yourself by your workflow. Define yourself by your results and the benefits you deliver for your clients.
I’m not saying the sky is falling. And I’m certainly not saying we are doomed. But I am saying we are about to undergo massive changes in what is already a complex and challenging industry. Consider yourself warned and start to become prepared:
The last fifteen years have been exciting, scary, challenging and fast-paced. The next five years will be worse. Don’t spend your time looking at your feet as you move forward. Look up and make sure the road you are on is the right one.
This is an important discussion, and, if you agree, please share it with your friends. And, as always, I’m interested in your opinions.
19 Responses to I’m Worried About The Future of EditingNewer Comments →
I responded to Larry in an email, and posted a long response on my own blog http://www.philiphodgetts.com/2017/07/thoughts-about-larry-jordans-worries-on-the-future-of-editing/ where I started by saying:
Larry Jordan posted a thoughtful blog about the future of editing in an AI- infested world. I think we should all be aware of what’s happening, and I’ve certainly been trying to do my part there, as recent posts attest. But I’m not sure I’m quite as concerned about editing jobs as Larry is. Assistant editors, perhaps.
Larry shared his post with me, asking for feedback, and having written a fairly comprehensive response, I decided to share it here as well. While I mostly address the areas where AI/Machine Learning might be used, and why pervasive automated editing is probably way further in the future than Larry’s concern would indicate, none of that negates Larry’s excellent advice on managing your career.
There’s more detail in the blog post.
I think some form of editing automation is reasonably close.
I’m thinking of how today certain kinds of “ordinary” news stories (sports primarily, I believe) can be “written” by a computer program that understands the semantics of this well enough to do a basic news story about “the game.”
I’m thinking about how facial recognition has become commonplace.
I’m thinking about how many informational videos are really edited around what is said and how voice recognition and automated transcription are now widely used.
I can see these all coming together in a way that will not necessarily replace a human editor, but for a large number of ordinary projects, it will deliver a rough cut automatically, along with a list of probable second choices. This machine-created rough cut will often contain hilarious errors at first, but it will easily prove itself to be an efficient starting point for the “real edit.”
Clearly, the more creative, the more novel the story to be told, regardless of subject, the less useful such a tool will be. But it’s not difficult to see a basic tool growing more supple over time.
A related kind of development — dealing with intake and metadata has been mentioned — could be a combination of media management and voice recognition for the editor. “Show me all of the clips of Joe like this one” could be a pretty useful bit of automation. “Show me all the B-roll and let me name each clip verbally” — no typing. And many more things like this.
The automation of video editing won’t be a monolithic thing but there are plenty of different on-ramps we will be seeing in the near future.
Those are great points to think about, Larry.
Even beyond automated editing, we face many other changes ahead: 360/VR video, the changes of television going from cable to streaming services, content overload/competition from video websites like YouTube and Vimeo, outsourcing, formats changing, and even just the tools themselves (NLE’s, etc.) constantly updating and adding features.
It’s never been more important to be flexible and innovative in your approach to your career.
Your statement really resounded with me when you said, “The bigger problem was that we, as editors, were defining ourselves by the tools we use, not the results we created. We proudly walked around saying: ‘I’m an Avid editor,’ or ‘I’m a Final Cut editor.'”
In my fast-paced, change-heavy career, I have worked on Avid, Final Cut Pro, back to Avid, back to Final Cut Pro, back to Avid, then Adobe Premiere, and now, I find myself using several of the Adobe Creative Cloud applications, beyond just editing. And don’t even get me started on how many times I’ve switched from PC to Mac to PC, haha.
Thanks for the insight, Larry.
Such a great and necessary discussion! I think Larry’s key word in all of this was “humanity”. Storytelling at its best contains elements that move the soul. AI is moving fast indeed and I think the best way of keeping our work most relevant as technology advances is to utilize our skills in the area farthest from what computers can learn and mimic: focus on ways of storytelling which evokes feeling and empathy.
What about us wannabees or dip my toe in the water types? The available options are frighteningly endless. And while I realize “One size fits all” ain’t so, what’s a wanna be to do?
Stop worrying. Too many people spend time trying to find the “best” solution. Just pick anything and start. The best way to learn how to tell stories is to tell them.
As you gain experience, you’ll learn what you like to do and what you are good at. That will help you decide the best tools to use to tell the stories you want to tell.
All major software offers free trials. Use them. Experiment. Play. Discover.
See if this is something you want to do with a career. At which point, you’ll be able to ask MUCH better questions.
There is a simple, elegant, honest response to the threat of machines taking over our world: Be HUMAN. “To be free is to be capable of thinking one’s own thoughts – not the thoughts merely of the body, or of society, but thoughts generated by one’s deepest, most original, most essential and spiritual self, one’s individuality.” -Rudolf Steiner 1904. People tell original stories all the time. Machines can’t. One way for Editors, Producers, Directors, Actors, Designers -People! – to keep fresh is to play Movie Games. http://www.moviegames.ca
Loved your post and the topic a lot!
Don’t you think that the risk is coming more from humans than from machines ?
– The producers who are pushing for something to finish in 1/3 of the proper time/budget
– Directors who just want to deliver something and get it done with to get the cheque
– The managers who just want “content” no matter how good/bad/useful/useless is that content
and the most important factor:
– Viewers who do not care about quality in anything anymore ? the “yeah whatever”-generation
I see the industry getting affected by many things at the moment, but I think that A.I. editing will be a small result of the situation of the industry more than a threat to it
maybe a bit too provocative, but Vimeos Staff Pick of the Year 2014 “Watchtower of Turkey” https://vimeo.com/108018156
… could it been edited by some algorithm?
a) edit to the beat of music? a clear yes.
b) analyse footage for possible whip-pan transitions? … yes.
c) edit rule #1 ‘in the flow’/keep main movement? … easy.
d) edit rule #2, wide to close/close to wide? … yep.
e) keep two clips within same colour scheme? yessir.
f) set accents, ramp lengths, set pauses, climax? why not?
Yes, I completely ignore the talent of Leonardo Dalessandri, but I’m pretty sure, with THAT kind of footage, a bot could create a movie, which impresses a lot of people.
What are the criteria of the TURING TEST? 😉
If the trend follows the same pattern as previous upheavals, the disruption will come first and fastest in the “low end” of our business: “event” videos, simple news clips, cleaned-up lecture-with-powerpoint type things. The high-end compositing and FX functions will still hold out for a while longer because those kinds of jobs still require a bit more craft and artistry.
It doesn’t take fancy A.I. to disrupt the editing business. It could just take a paradigm shift in how the work is assigned. I made a prediction some time back that with suitable bandwidth, many editing jobs would start becoming outsourced to third-world countries with much lower human capital costs and no regulations. It’s possible now that a guy in Bangor could be competing for the same editing gig with someone in Bangalore.
You know, you can ride the wave in to shore, or you can let it tumble you thru the coral. To stay ahead of this trend, you may have to change the nature of what you do.
It’s not all bad, though. There are positive trends too.
One area that I find very exciting and full of promise involves projection mapping in real-time. There is a lot of fascinating stuff being done in this area already, though it’s a relatively small community (for now) and I can imagine the field growing even more in the coming decade, with applications in advertising and marketing displays as well as artistic performance. This kind of graphics work requires a lot of capital investment in high-grunt hardware, and a lot of finnicky data wrangling across multiple disciplines like 3-d modeling, animation, compositing, etc. It’s something AI tools can help with but not something that they can do on their own. It’s something I would point a new grad to as an area that will see growth and career possibilities.
The other thing I see that’s encouraging is the continued boom in content creation for all the streaming video applications. More new shows are being made and tried for this market than ever before, and it’s an area where you’ll still need high-level skills and artistic craft, and there will be people willing to pay a living wage for it.
Looking back at the desktop printing and DV video “disruptions”, the first things they took over were the ephemeral, low-end work that can be done by new, relatively unskilled people with a small(er) capital investment. These were jobs that paid a low margin to begin with, like “weekend warrior” wedding and event guys… and democratizing the “means of production” thru automation just makes it easier for everyday non-pros to get the results they want for free, or at a price they’re willing to pay. That may not be a price that makes for a sustainable career for one of us pros, and that’s just how it is world-wide: ask an automobile assembly worker.
The next step in this revolution, as Larry laid out, points to the need for you to “step up your Game”, if you wanna keep playing it. Or find a new, different, better “game”.
This may be overly simplistic here, but I’ve been working in this field long enough to have seen several big changes in it, or to have had my bosses tell me about what they lived through.
There was the transition from feature films to a world that included television. I wasn’t around for that but I’ve been told how people looked up and saw that the sky was falling then.
Then there was a transition to digital audio, from analog. There was more debris coming from the sky then. And it is certainly true that the number of apprentices (we call them Post PAs now) dropped to nearly zero, and crew sizes shrunk, but it enabled people to work from all over the world, in smaller editing spaces. And this increased the number of projects being created for those who were willing to make the transition.
Then the same thing happened on the picture post side. First videotape and now files, but the transition to digital editing caused more pieces of the sky to tall. Crew sizes shrunk, of course. It was now possible to do simple VFX, audio, color and more without going to large post houses. And, today, I see more people editing, not fewer. Wedding and event videos were a much smaller piece of the market years ago, but because now most people can make a wedding video themselves. Of course, they don’t because they know they would suck at it. So they hire people, at various price ranges, to do it.
And John Landgraf talks about too much Peak TV. But there are people who are editing them, people who are doing sound work on them, people who finish them, people who distribute them in the gazillion formats that are necessary for delivery. Some of these are the same people who had no idea how to do those things ten years ago, but many of them are people who had the flexibility to change, and the lack of ego to realize that they needed to.
That is why, Larry, I completely agree that one of the most important factors a good filmmaker needs to have today is flexibility. Editors who refuse to learn how to do split screens to improve performance (what Alan Bell calls Performance Enhanced Visuals) are called EX-EDITORS.
Finally (wow, I HAVE been going on for a while. I should restart my blog), the move to AI and machine learning will create tons of opportunities for filmmakers and post people to change. I believe that there will be new jobs, as well as new tools for our present filmmakers. Will it be easy to replace someone who churns out cookie cutter, uninspiring wedding or event videos? You bet. So they’ll need to up their artistic game and utilize the time that they will save on getting to the uninspiring cut, in order to give them time to get to the better one.
Or, perhaps, learn to be a carpenter or plumber. I can’t see those jobs getting replaced with AI.