Why do I often prefer stock video over AI videos for my service prototypes?
Why do I often prefer stock video over AI videos for my service prototypes?
Q&A: Prototyping and Service Design
Prototyping Planning and Strategy
Prototyping Planning and Strategy
Theater prototyping
Theater prototyping
Digital Prototyping
Digital Prototyping
Backstage Prototyping
Backstage Prototyping
Presenting Service Design prototypes
Presenting Service Design prototypes
In short: AI videos take way too much time to create, are way too expensive, have a shitload of stupid mistakes, are not that better than stock videos and people forgive them less.
For years I've created tiny videos to pitch and show new service ideas with stock video. In the last yearI've tried several times to do the same with AI videos, but it just sucks. Here's why.
Takes too much time
AI often doesn't get things right the first time. That's okay with texts that are generated fast. But with video it takes each time several minutes to generate a new version. And often even after several rounds you don't get to what you wanted.
Compared to browsing stock videos from Storyblocks or any other tool that integrates it live Screenflow or Descript, AI videos are much slower. In two minutes I can see dozens of possibilities that don't have a shitload of weird errors. Sure they are not what I exactly had in mind, but I can move the voice over fill the gaps of what can't be seen.
The other problems
If it's slower and gets you were you wanted or helps you create something better that would be okay. But AI videos, in my experience have other problems too:
the results are not better: there many weird mistakes, weird faces, that attract the eye more than the story you want to tell.
it's more expensive: if you already use a fool live Descript, stock video is included in your subscription. The same if you use Canva. Otherwise you can have unlimited videos for one month for 21 bucks at storyblocks. That's cheaper than the AI tools you'll need.
people are less forgiving: with AI we often think that we can make it show exactly what we have in our head. Therefor we and the people who are watching the video are way less forgiving. What is shown should be exactly what you pitch, because AI can show anything, right?! A stock made video is clearly a scrap-booking exercise so people know they need to fill in the gaps.
What I prefer
I still prefer to do video stories of service ideas in two ways:
Stock + images + voiceover: a bit like a collage of all what I can use fast.
sketches + voice over : if I need to show something impossible to showcase with stock material then I can use illustrations that I lightly animate with Zoom - ins and zoom - outs
This approach then also pushes me to have a good voice over text as the foundation for the video so that the voice over feels the gaps of what can't be show visually.
Backstage of this article
This article was written and illustrated by hand on a refurbished Remarkable II tablet. The hand written text was converted into typed text using the Connect Service by Remarkable. If you are curious you can download below the original note.