New AI-based tools are gradually becoming available. With Firefly, Adobe has its own AI model at the start, which works for still images and should also help with video editing in the future - Adobe announced this in a big way at NAB in April. In a trade show interview about this, which interestingly has only now been taken up ymcinema.com/2023/08/14/the-inevitable-future-of-ai-filmmaking/?expand_article=1, Michael Cioni, then Senior Director of Global Innovation at Adobe, did not mince his words on the subject of AI. His explicit predictions regarding the impact of AI on the creative industry go far beyond what one would expect in this setting.
Generating instead of filming?
"AI is going to be the most polarizing bloodbath of the last 20 years" (original quote) - the changes will be incredible and it will happen very quickly, said Cioni at NAB. Very much of what is still recorded with cameras today will be generated by AI at the push of a button in the future. News and sports will still be filmed, he said, but in the narrative area, he expects that most of it will not only be generated automatically, but of course will no longer be edited as it was before - "Computational editing is going to start to replace a lot of the creative editing at some level." This is all inevitable and he can only urgently recommend that every filmmaker position himself accordingly now, because those who are creative can continue to be so, just in a different way. Those who don't embrace the new technical possibilities, on the other hand, are in danger of going under.
That was pretty tough talk to hear at the Adobe booth - maybe too tough? As it so happens, Adobe and Michael Cioni went their separate ways shortly thereafter. It's unclear on whose behest, because in a recent YouTube clip titled I Quit My Job, Cioni says something contradictory about this; at one point it also says "I lost my job.
However, Cioni is not alone in making such assumptions, as it's clear that moving image production is about to change - anyone who has followed our news over the last 12 months will know the buzz around generative AI technologies in text and photography, and also that it's only a matter of time before moving images can be created in better quality by way of text prompts. Whether it will actually be possible to simulate absolutely photo-realistic videos, and if so, when that might be, is unclear - what is certain is that tremendous progress has been made in an unexpectedly short period of time.
It is absolutely possible that things will continue at this pace, but it is just as conceivable that AI representations of human motion sequences, for example, will remain stuck in Uncanny Valley for longer. However, there is no doubt in our minds that AI technologies will have a disruptive effect, even for filmmakers. The question is to what extent and at what speed.
