Does Ai Stunt Creativity?
My article last month highlighted not only the increasing quality but also the increasing cost of the latest video making Ai, and it painted a somewhat bleak future for the amateur Ai video maker.
On the face of it, it would seem that amateur enthusiasts will not be able to access the superiority of software such as Veo3 (and others of a similar standard and expense), but instead have to satisfy themselves by using the lesser offerings, perhaps stifling or reducing their creativity.
On reflection though, I suppose in many ways that has always been the dilemma for amateurs.
It was never likely that a small group of film makers would ever be able to achieve the quality of the big studios blockbusters. Instead they would make the best of what was realistically available to them; average scripts, cheap special effects, or simpler locations, and employ their creativity to make up for the lack of dollars.
Thinking about the videos I have created, its apparent that they had two characteristics; firstly showcase the capabilities of Ai at that time, and secondly push the boundaries by combining a range of different Ai utilities and making them deliver things that might not have been originally intended. It is this second aspect that I find most enjoyable and satisfying.
My first ever Ai video was Rip Van Wrinkly, made in January 2021, and for that I combined two very different programs (FaceApp and EbSynth) to both add age and de-age myself in a real video. FaceApp was originally intended to create enhanced static images for social media vanity and EbSynth animated paintings. They weren’t designed to work together.
More recently HeyGen was originally offered as a multi language text to video talking-head avatar creator for corporate videos, but users found a way to upload their own videos and create multi-language lip-synced dubbing. It was this technique that I used in ‘An Editors Struggle’ last year.
And Googles NotebookLM was created to analyse large documents, reducing reading time for complex reports or making plain English sense of legal jargon. It also had the option to create an audio discussion (deep-dive) between two Ai generated voices. I uploaded my own fictitious document, and used this ‘deep-dive’ feature in NotebookLM to create the interesting (and real) dialogue between two Ai generated voices who discussed between them the possibility that they were not real.
So in an odd way, rather than reducing or stifling amateur creativity, the release of Googles Veo3 is perhaps most likely to inspire even more creativity.
Veo3 has simply raised the bench mark for the capability and quality of Ai films, so we can clearly see what is possible with the latest Ai today. Now it’s up to us amateur film makers to search out even more creative ways to stretch toward that new standard by using and adapting tools that are comfortably within our budget.