Love it or hate it, AI video is here. And yes, it can be terrible. “Slopitalist garbage,” as one Reddit user so eloquently put it (we see you, vq2k).
But before you blame the tool, remember: humans have been making sloppy content long before AI existed. The problem was never the technology. It was the thinking behind it.
One of the sloppiest videos ever?
Pepsi officially pulled the ad less than 24 hours after it debuted in 2017, it’s no longer available on their official channels. Missing the mark on so many levels, it was widely panned for trivialising real-world social justice movements like Black Lives Matter. It was (rightly) criticised for being “sloppy” and tone-deaf.
So where does AI video actually add value?
When used by people who know what they’re doing, AI video doesn’t just save time, it unlocks capabilities that weren’t financially or logistically possible before.
If you’re a marketer sitting on the fence about AI video, here’s a run-down of what you need to know:
Intelligent video editing
AI can automatically cut, sequence, and pace footage based on storytelling best practices. Turning hours of editorial grunt work into minutes. The quality doesn’t take any hit. But the AI still needs someone who knows what story they’re telling in the first place. The edit is only as good as the brief behind it.
Multi-platform content repurposing
One video, automatically reformatted for TikTok, YouTube, Instagram, and LinkedIn. Each version optimised for that platform’s specs, audience behaviour, and aspect ratio. No manual re-editing. Yes please!
For in-house marketers juggling a dozen channels, this one’s a game-changer. Maximum reach, minimum effort. Just make sure a social media specialist has signed off on the messaging and distribution plan, because the same words don’t always land the same way on every platform.
Real-time collaboration tools
AI-powered creative tools let teams brainstorm, iterate, and produce together in real-time, no matter where they are. Faster feedback loops, more diverse creative input, fewer meetings arguing about which cut to use.
AI-powered personalisation
Videos can now adapt to individual viewers in real-time – different dialogue, pacing, even different endings depending on who’s watching. Done well, this makes your content hyper-personalised, more engaging, more relevant, and significantly more likely to convert. Done without a clear strategy behind it, you’re just creating multiple versions of the same mediocre content.
Voice cloning and synthetic narration
Want to create videos in French, Spanish, and Mandarin? AI can replicate a voice and translate with remarkable accuracy – so you can produce multilingual content instantly, with consistent voicing, at a fraction of traditional localisation costs.
The alternative? Relying on Google Translate or paying for translation support, then working with a spokesperson who doesn’t speak a word of French. Which is how you end up pulling an Emily in Paris. Say what you need to say in your language, and let AI handle the rest.
Text-to-video
With a well-crafted text prompt, video specialists using AI can now generate cinematic visuals without expensive cameras, studios, or crew. I can’t stress this enough – this needs to be in the hands of someone who understands storytelling. Because in the hands of someone who doesn’t, this is exactly where AI slop comes from.
AI avatars and digital presenters
AI-generated presenters now look genuinely indistinguishable from real people. Your brand can have a consistent, always-available digital spokesperson for training videos, company updates, product explainers, at a fraction of traditional production costs. There are lots of considerations around authenticity when it comes to specific content, so this isn’t going to suit every brief. And remember: the script that makes that spokesperson sound like a human worth listening to? That’s still a writer’s job.
So about that AI slop you keep seeing
That video made by someone in their living room, the one you’re using as proof that “AI videos suck”? That’s not an AI problem. That’s a skills problem.
A brand strategist who understands your objective. A prompt engineer who can translate that into something the model can work with. A video expert who can look at the output and tell you whether it’s technically polished but emotionally hollow – that combination produces something completely different. And yes, you’ll most likely need an expert to finalise the job.
AI video right now is a bit like early television. Nobody looked at a grainy black-and-white broadcast in 1950 and said “television is a terrible medium.” They watched it get better. And it did, beyond anything anyone predicted.
AI video has evolved more in the last 12 months than TV did in its first decade. The brands getting ahead right now aren’t waiting for it to be perfect. They’re experimenting, learning what works and building a competitive advantage while everyone else is arguing about whether it counts as “real” creativity.
And human-made content isn’t going anywhere either. Authentic, filmed, human storytelling still carries emotional weight that no avatar or generated visual has fully replicated. The smartest brands won’t choose between the two. They’ll know when to use which.
The future of video isn’t AI replacing humans. It’s humans who know how to use AI, are experts in what they do, and can’t be replaced by humans who don’t.
Written by Sian Evans, Head of Video at Definition
