AI-Generated Video is About to Change Everything
For the past year we’ve been saying that AI is your digital superpower. Now it’s about to get a whole lot more powerful. AI-generated video is going to move us from an era where what we could produce was limited by our resources and production capabilities, to one where the only limiting factor is our own ideas.
Sora: OpenAI’s image generator
This week, OpenAI (the makers of ChatGPT, DALL-E, and others) released a series of demonstrations of their newest AI tool called Sora, which will generate high quality video from any prompt you can dream up. The demos are mind-blowing, and when we all get the ability to create any videos we want with only a simple prompt, the possibilities for the volume and types of content we can produce will be nearly endless.
Until recently, creative production — in all of its forms — had been the limitation. We didn’t have the time, knowledge, or resources to produce the things that we know will help us to achieve our goals.
Enter: AI
First, it came for written content, with ChatGPT, Bard (Gemini), and others. Our ability to produce drafts of web copy, blog posts, press releases, grant applications, and all other copy-based content exploded.
Then it came for images with DALL-E, Midjourney, and others. That made it possible to create any image we could imagine for backgrounds, decks, social posts, and blog content.
Now it’s coming for the most resource-intensive medium of all: Video.
So far, based on the demo provided by OpenAI, we know that the videos are:
- Super sharp
- High quality
- Built from a combination of your own prompts + the full knowledge base of its Large Language Model (GPT-4 in this case)
- Whatever you can imagine
It’s easy to imagine that these AI-generated videos will be used for (just to name a few):
- Social content
- B-roll for longer videos
- Explainer videos
- News and updates with voiceovers
- Mockups, or drafts for full production videos
At this point we all know that video content fairly consistently outperforms all other forms of content, whether that’s on TikTok, YouTube, Instagram, or even how we explain our businesses on our websites. But many of us don’t have the resources it takes to consistently produce the ideas that are in our heads.
Welcome to the Ideas Era
Imagine for a moment a marketer from the 80s being introduced to tools like Canva, Google Docs, CapCut, or Photoshop. The limitations that they faced would be blown apart, and they would see a world where just about anything could be created with a fraction of the effort they were used to.
As we now know, the best work came from people who were able to use tools like Canva in creative, original ways. The same is about to be true for AI.
I’ll give you some examples, just to show how ideas + AI have been the real unlock, rather than just AI on its own:
- Heinz used the image generator, DALL-E to show that their brand is so well loved, that even AI bots thought of Heinz first when asked to draw ketchup
- A Food-Tech VC blew up on Linkedin when he had ChatGPT write a rap about the future of meat
- A Redditor was planning a trip to Japan, and had ChatGPT decide every detail of his visit, down to where he’d shop and what he’d eat
- Farfetch had AI tools turn their email marketing drafts into a variety of versions, when they tested and increased their open rate by 7%
- ClickUp used a tool called SurferSEO to grow its blog traffic 85% not by writing the posts for it, but by having it identify keywords and optimization opportunities
All of those were generated by AI tools, but the ideas came from a human.
That’s why the best thing we can all be doing is learning the capabilities of these things, and starting to build the muscle in our minds that can think about what’s possible, and how we can use these tools creatively.
Of course, there will be limitations
Even the creators of these tools see that they won’t be a magic wand to solve all of our content problems. Here’s a quote from OpenAI about Sora:
“It may struggle with accurately simulating the physics of a complex scene, and may not understand specific instances of cause and effect. For example, a person might take a bite out of a cookie, but afterward, the cookie may not have a bite mark.
The model may also confuse spatial details of a prompt, for example, mixing up left and right, and may struggle with precise descriptions of events that take place over time, like following a specific camera trajectory.”
As we think about what will be possible with video-generating tools, here are a few things that we probably won’t be able to do (right away):
- Specific videos about your product or property
- Include logos, text, or conversation
Of course there’s also the concern about AI-generated video being used for misinformation, and given the power of video, the impact could be dramatic (especially with several important elections coming up around the world).
There is some work being done to combat that effect, starting with Meta’s effort to tag all AI-generated content and disclose it to their users. It’s imperfect, but it’s a start. The better solution is for everyone to get educated about how content can be created so that we can think critically about what we’re seeing. But as anyone who’s had to take a tech support call from a less-than-digitally-savvy relative can tell you, we’re a long way off from that.
So What?
The first thing we can all be doing is exploring the tools. I find that, for many of us, creative ideas are limited by what we believe is possible and available to us. When we start to explore the possibilities of AI-generated content, new ideas, and new inspiration will start to open up.
Next, try them out for yourself. The best way to learn about something so fundamentally different from what we’re used to is to simply try them out. I strongly suggest checking out an image generator and try to push yourself to consider new and different ideas.
Then, recognize that the world is about to become overwhelmed by AI-generated video content. When resource limitations are no longer a factor, it only makes sense that we’d see a massive increase in the amount of content being published.
As with any major trend, there will be counter-trends, and in this case we’re predicting that many people will eventually gravitate towards the types of content that can’t be mass-generated by AI. That will likely look like raw videos, thoughtful pieces with human imperfections, and content that tells real, first-person stories that can’t be artificially created.
It seems that the old adage is true: The more things change, the more they stay the same; and in this case, that’s a desire for human connection.