Sora has just been announced. Even though AI video tools have been on the rise for a long time, like ChatGPT when it was released, this new paradigm is certainly a little nerve-wracking. But there are a lot of reasons why this is a good thing, and why just being nervous about how it’s all going to change is an unhelpful approach.
Here’s the TLDR;
- Work steadily being augmented by tech is always how careers & skillsets have changed
- Letting a huge global paradigm shift signal the end of your creative identity is a well known and lazy existential trap
- It’s still early days and there’s an immense amount of time to leverage it ahead of us before a ‘video producer’ is no longer a thing
- Getting clear on what you’re ultimate goal is will help you ride the change of ‘how’ you do it
So to help myself understand and get a little more grounded from the hype, here are some thoughts to put this in context for a Video Creator in 2024. I’ll cover:
- Putting this, and other developments in context
- Leveraging the wave
- A look at my current skillset and how it’s been affected by AI
- Waking up to just how much this is going to expand & improve our work
- Getting clear on what it’s all ultimately for
- We’re always on a tech wave
- Leveraging the wave
- How is this going to be good?
- The AI paradigms
- How text & image models are impacting my core skillset
- Researching
- Learning Professional Tools faster
- Troubleshooting issues quicker
- Coding my own extensions, scripts, and expressions
- Dummy data & images
- Where I think this is going over the next 3 years
- Video creators will start to become Directors
- Why do we create content?
We’re always on a tech wave
At every point in history, the world is changing. Since at least my grandparents, there have always been huge seismic changes across a generation. But when you’re going through it and the wave is shifting under your feet — it’s easy to think that your generations changes are the greatest.
Now, I am no historian, but to help me appreciate this — I found it useful to superficially (and selectively) summarize the major shifts that my immediate generations have faced (toggle open to read).
- World Wars
- The Great Depression
- Technological Advancements
- Cold War and Decolonization
- Civil Rights Movement
- Digital Revolution
- Rise of the Internet and Social Media
- Globalization
- Climate Change
- Rise of AI
Of course this is all speculation… 😉
- Major AI Improvements
- Technological Integration and Human Enhancement
- Space Exploration and Industry
- Quantum Computing
- Unimaginable discoveries and developments brought about by the development of AI
And what do you notice about every generation? Technology. Technology is always shifting and growing. I know this is probably the most obvious statement you've read this month… but I think it's a kind of spiritual mantra that one needs to remind oneself of during times like these.
I've always loved technology. I always get excited about technology developing, until these seismic moments like the release of Sora or ChatGPT. Then suddenly, the radical shifting of the landscape becomes a bit scary and uncomfortable. I feel uneasy in my stomach, I’ll be honest.
And with this, questions arise like:
- "Can I keep up with this?"
- "Will this undo what I've been working on for so long?"
- "Will the business that I'm working for still be here in five years?"
- "What will my children do for work?" etc etc 😅
But… I have to remember. We have to remember
Technology always disrupts, democratizes, and we all develop.
It's a tried and true example, but imagine you're a writer before the advent of the printing press. You have your methods of production that are well established and give you a great income and reach, and then suddenly the printing press is invented. Suddenly you realize all of the people who don't have the workflows that you have in order to disseminate your knowledge are going to have the same advantage as you, and therefore the playing field will be leveled and your edge will have gone.
Orrr….. the Renaissance will happen. And you’ll enjoy that and your work will change.
In other words, huge paradigm shifts are uncomfortable but are always happening and bring good new things that you can’t imagine. That continual change is part of the human experience if you live in society.
What is also part of the human experience is that we really want stability and/or to know what is going to happen. Right now, you don't know that, I don't know that, and I certainly find it really uncomfortable. But… I’m trying to remind myself that the Renaissance might be about to happen — and that I’ll enjoy that, and my work will change.
But… this is big-picture stuff. It’s not practical. How is my work going to shift soon?
Well, let’s get practical then — how do we start shifting with the landscape and truly leverage this? I don’t ultimately know, but if I think just about the next 3 years — I have some good ideas.
Leveraging the wave
I used to be a composer, and when I started writing music, computers had the shittiest virtual instruments — I wrote pieces on terrible violins that sounded like duck quacks being scratched on vinyl. To actually realize my music, I had to learn how to read and write music, write it out on sheet music, find great musicians, I learned how to conduct orchestras, I paid musicians, I learned how to record music well, and produce it. In other words, to be creative I had to manage a fairly complex pipeline.
But at the same time, during all of this, technology for composers was radically evolving. Towards the end of my composing career (which didn't last ages. I'm only 34 as of writing this), I could realize entire orchestral scores on my laptop without reading or writing notation, or recording anything. Companies like Spitfire Audio developed and took creative tools and technology to new heights.
This evolution was totally natural to me, but for some composers it probably caused them to feel threatened. Their skill sets had built up years before mine around the ability to manage that complex pipeline from idea to recording studio, and leveraging all of the business decisions and infrastructure that a full-scale orchestral recording required.
I guess I wasn’t phased by these developments because my career was just getting started. So all of these improvements to my tools just served to enhance and improve my skill set. Even if I could see that it was leveling the playing field, I enjoyed the opportunities it afforded me. I was in a sense riding that technological wave.
Today as a video producer, I'm in a different position. My livelihood is my work. And I've grown a skill set that is marketable. As the landscape around that skill set shifts, it's nerve-wracking.
But crucially, it's only nerve-wracking in the moment I consider it. In other words, our thought processes only ever happen at a particular moment in time, they don't happen across years or months.
In this one moment, this sliver of time, the worried part of my mind wonders if things will shift so quickly under my feet that the rug is pulled from underneath me and I'm suddenly left with a worthless skill set.
But in reality, if you're conscientious, things develop slow enough for you to ride the wave and leverage the developments. And that, that is what we need to do.
How is this going to be good?
Whenever you're scared, whenever you're facing conflict, whenever there's a disruption or a change, or something is not going the way you want it to, it's always really useful to ask yourself the question, "How is this going to be good?”
As you know, as humans, many of us have a negativity bias. In other words, it's often quicker and easier for our minds to jump to the negative connotation or conclusion or worry. It takes imagination, emotional regulation and maturity to slow down and actually ask the more productive question that creates a better future. “How is this going to be good?”
Let’s start with the things that are already happening in my skillset and then move to where I think this could go in the next 3 years. The goal of me writing all of these out is to get you excited if you haven't thought about these already. And to get you in the mindset of realising that all that is happening here is we are becoming more powerful and therefore more able to express our ideas and the way in which they are expressed — without being bogged down with the highly technical or incidental aspects of our work.
The AI paradigms
Let's break open the top level paradigms that are being affected by AI that are relevant to a video producers skillset.
- Text (GPT4)
- Image (DALL.E, Midjourney)
- Audio (Stableaudio)
- Video (Sora)
Obviously, there are more, like 3D generative tools, but these top ones encapsulate the main elements that actually go into a video:
- Text for writing and planning
- Images for use within your production, design, mood board, or storyboard
- Audio for music and sound effects
- and of course Video
Right now, the models that are actually good and widely available are text & image. Audio and video models are very close, but the good ones (like Sora) are not yet widely available.
So, before I look at where I think this is going, let’s have a look at how Text & Image are currently impacting my skillset.
How text & image models are impacting my core skillset
While text & image models are being used by developers to create a big wave of new AI/Automated video creation tools — these are still very early, and mostly focused on non-professionals.
I have little need for these just now, but I imagine they're going to get good very fast. So I’m going to be tracking them closely.
Researching
This is obviously not specific to a video creator, but it's still highly relevant. The research phase, for me, has always been lengthy. And now, for almost all of us, it's being radically improved by LLMs. If I have an idea, I always chat first with a model and then refine it, even sometimes moving to scripting stage within one conversation.
I also am beginning to use image gen within my look development. Look dev is now a dance between finding images that are right, and learning to describe images that are right. I find the process of learning to describe the visual style improves my understanding of it and helps my intentionality in production.
Learning Professional Tools faster
After Effects took me a couple of years to master. I learned the essentials of Cinema 4D and 3D rendering in about five months — on the side of a full time startup job while raising two young kids. I find that impressive even if you don’t 😉.
ChatGPT was my tutor — and it massively improved my learning — helping me grok and parse out the core principles and techniques needed without having to read the insanely huge documentation libraries or go down YouTube or Course rabbit holes.
It helped me follow the 80/20 principle all the way through my initial learning journey.
Troubleshooting issues quicker
There is a non-trivial technical aspect to being a video producer, especially when you work in 3D. The field is just still too young for all of the complexity to be abstracted away. We're still dealing with the raw constituent parts of what makes up a great render, and it can get really fiddly.
It's a bit like coding; because you're working with the raw parts of the thing you're building, rather than abstractions that are on rails. You're more likely to hit bugs and issues.
I started out my video career in 2019 and back then we were still in the age of Google search. I absolutely hate trawling through forums. My brain just starts switching off after two threads, and very often I’d just end up on a dead end.
But now that doesn’t happen. I just ask ChatBots and they help me 80% of the time get what I need.
Coding my own extensions, scripts, and expressions
Software like After Effects or Blender allows you to write in JavaScript or Python to develop your own extensions, making the software better and more specific to your needs. That's great if you know how to code, but it's useless if you don't.
Back in the day, the only people that really did this were big studios who had their own in-house developers. Now, however, with ChatGPT, I am regularly writing extensions to build out my own personal workflows. It's absolutely game-changing. And this is only going to get better.
Dummy data & images
My role at Glide requires me to make tons of educational and marketing videos about the software. Glide is heavily data-driven, so there's an immense amount of dummy data to fill out and make believable. Since ChatGPT, DALL.E, and MidJourney have come out, I have realized just how much time I used to spend collecting and creating fake data — just to make a video shine. Now, I just generate it, and can get on with making the actual video.
I don't think this is exclusive though to software videos. Many times as video producers we need to find a way to fill out background data that isn't the main focus but needs to look realistic, like images on the wall, letters on the table, or book titles on the shelves.
So that's just some of the ways that AI is impacting my work today. I'm excited about all of them. It’s only been 11 months since GPT4 and I’ve totally integrated it into my workflows.
Where I think this is going over the next 3 years
And I think Sora and audio models will be just the same, at least for the next three years. They will come into our skillsets and make us way more creative and productive. But, of course, the democratization will massively change the landscape of what is created.
- We will see infinitely more video clips that look amazing being created, just like we see infinite beautiful images being generated today. But a video producer is not a ‘video clip creator’. A video producer has to pull together many, many things for a particular purpose and then make that collection of things cohesive and compelling over time.
- The proliferation of automated AI video tools will enable all of us (amateurs and video producers) to produce a higher volume of good or functional videos, covering areas like customer stories, tutorials, promos, reviews, and social media posts. Consequently, the incidence of subpar videos will decrease, effectively elevating the standard for what constitutes a 'good' video.
And so the difference between good and great videos will become more important for us, forcing us to be the ones to truly integrate and ride the wave of video tech, and get infinitely better as creators. The below are just my best guesses, and they will also impact amateur video creation, but as a video producer today, we are well placed to truly leverage this as it happens.
- Video gen: We’ll be able to generate any b-roll we need
- AI & image gen will start radically impacting design. Design is hard and I can’t wait to make this process easier.
- LLMs will get more sophisticated and actually become really beneficial in the nuanced early stages of the creative process: brainstorming, messaging, scripting, storyline, researching etc.
- Audio gen will become actually good so we start using it for underscore & SFX.
- AI editing is in its infancy today, with only simple cuts and extracts. I think in the next three years, we’ll be able to ask an AI to recut a video just like we ask an LLM to rewrite some text.
- Video enhancement: we’ll start to see videos not only be upscaled but also enhanced, so that if your footage is not shot or lit well, you can use AI to enhance it — just like AI Headshot tools today.
- Render times: Right now, by today’s standards I have a powerhouse of a PC that cost me $4.5k. It still takes a whole day to render a relatively short 4K 3D clip at 30fps. That is a HUGE drain on my creative output and I can’t wait for that to improve.
- AI/Automated video tools will get so much better that we can start using them as virtual production tools to manage huge pipelines of video campaigns etc. Today we’re seeing the early stages of that in tools like Riverside and Shuffll.
Video creators will start to become Directors
By traditional standards, I really do not think of myself as a director. The term "director" today refers to a very high-level skillset. But I think that over the next three years and beyond, an individual video producer/creator like me may start to think and market themselves more as a director.
At each stage that technology and my skillset improves, what I notice is that my perspective on my work gets higher level. I start thinking more and more about the video and message that I want to create, rather than the way that I'm creating it.
If we look at all the improvements I outlined above and extrapolate out, we can see that what is happening is that more and more we are being empowered freed up to think more creatively about what we want to create rather than how it is made.
It’s really helpful to remember, that truly great videos are basically
1) great ideas
2) produced with a great vibe
And it takes a lot of skill and nuance to get that right. Most people, and software won’t be able to do that for a long time.
Why do we create content?
If I look at our current workflow at Glide and really ask myself the question "Am I happy with the volume of content that we create?" and I ignore any current context about our current resources/bandwidth. My answer has to be unequivocally no.
I'm totally unsatisfied with it. The number of ideas and concepts around what we're trying to do at Glide is virtually infinite, and we're just one company. There is so much more we could be communicating and delivering to our customers and prospective customers. There is so much to say about software development, and we are just scratching the surface.
Before I get to the point here, let me explain a separate paradigm shift that's happened to me in my workflows recently.
Today, I’m writing more and more using OpenAI’s Whisper Model (using Superwhisper). It transcribes everything perfectly, and I then do minor tweaks and reorganizations with the keyboard.
The impact of dictating has utterly transformed my approach to writing. Because I know I can write so much more easily I feel free and able to express myself more and I'm more and more excited about sharing my ideas. There are MORE ideas coming, because of this freedom, and they are getting better through creating more of them.
Video, will be no different — except it will be more inspiring, creative, expressive, powerful, and impactful.
Video is the communication paradigm today and if we are scared about the volume of video speeding up or the democratization of it, we are no different than the people who were scared about the masses gaining access to words with the printing press.
The printing press was invented over 500 years ago. Everyone today produces words. Is a professional ‘Writer’ still a thing today? Of course.
Why? Because good writers:
- Have great ideas
- Communicate them well
Video is no different.
At the core of all this, for me, is a question which has a clear answer. Why do I make video? For some, it will have a different answer, but for me, it's clear.
I create video because:
- I love expressing ideas
- And I love expressing them well.
However the way I do it changes.
One day, when AI is so good that all AI generation is indistinguishable from those made by humans, there is of course a question as to what makes a human’s creativity valuable or distinguishable.
And at that point, I just think we’ll have much better questions.
Until then, as we see this curve get more and more intense, I’m going to keep reminding myself of why I do this, not how. I will also deeply embrace and watch how my humanity interacts and becomes more valuable as it works alongside this growing artificial intelligence that we are creating.