people complaining about AI features have clearly never wasted hours editing video or lost time and money discovering a technical flaw in a rush shot three days ago. For actual workflows, these tools are lifesavers
Because those people see the phrase "AI features" and the first thought is those sloppy generative AI stuff where things shift.
Where as marketing at all these corporation is trying to genericize "AI features" into anything using an algorithm.
"Content aware fill", something we've had for over a decade is now "AI object removal"
I’m not sure I agree? For each of your examples there are algorithmic approaches and neural network approaches. Companies have certainly been loose and wild with how they market these, but there remain distinct approaches and implementations for each. Very generally speaking, the neural network based approaches (aka “generative AI”) perform better but with much worse degenerative cases and a higher baseline rate of unwanted side effects (that are normally not immediately visible but tend to cause issues down the line).
My bigger concern is that these neural network based solutions have taken the place of the former rather than supplemented them. Many tools no longer provide the algorithmic/kernel-based approach at all, and have marketed the “AI” (née ML) alternative as a strict superset/upgrade, despite its potential drawbacks.
(Interestingly while the inference-based implementations generally have higher latency (or infinitely worse, cloud and pay-as-you-go requirements), for some computationally difficult kernels the inference-based approach is actually faster!
Editing is a craft. You have to watch everything, otherwise you don't know what you have.
A machine organising stringouts and selects can work for interviews, but not for action. But even then it is only parsing your media for semantic intent. It misses the way things are said, which often imparts a different meaning.
You can use AI features for editing. But it is unlikely you will be making anything very intetesting.
I work professionally in commercials. You literally wrote the very solid counter argument into your own comment.
For interviews, this saves hundreds of hours. Giant amounts of social content are now interview-like, including podcasts.
Your argument makes little sense to me. For narrative editing, this is a minor help. For interview-like content it cuts work by 30-70% maybe. its probably a gamechanger.
So you are satisfied with work where you have not watched all of the material? Fine if that works for you, but I would not think I'd have done a good job.
If something has compiled some great selects for me, I only have material based on what is said. But I don't know how it is said - which to me is everything. Yes, this goes for podcasts too. What I make just won't be as good as if I'd have taken the time to go through everything.
I even notice this with things like NBA highlights. It's clear the editor (or bot) never watched the game. There were moments in it that are interesting and significant. Even small incidents that are actually clutch to the outcome of the match. But it is clear they've quickly compiled all of the goals and bashed out something mediocre which does not tell the story.
I don’t care whether I’m satisfied - I care whether the client is. And does this get me faster to it? Probably yes.
In a narrative film or a project of my own I’d care. In a project for hire under a tight deadline?
Why would I? The delivery is the delivery. That’s how commercials go, at least. Clients dont ask - have you seen every bit of footage - they ask - is it ready for deadline.
It makes things more convenient for you. I'm not arguing against that. And, yes, you may have a satisfied client. But the work will not be as interesting, because you have not comprehended all of the ingested material. That is my only point. I'm not sure why this is so upsetting?
> You can use AI features for editing. But it is unlikely you will be making anything very intetesting.
What? It bears no impact on if what you're making is interesting or not, couldn't matter less. People been creating amazing things with nothing, and absolute trash with everything, and also vice-versa, seems to be all up to the person's taste and skill, and less to do with the actual tools they use.
"Taste and skill" is my exact point. If you are not watching everything you have captured, then what you make will most likely not be as interesting than if you had.
I think your mistake is to assume editing is like painting, where you can just make something brilliant with a few colours and a canvas. But editing is much more analogous to writing a book. If you have read extensively on Ancient Rome and spent time comprehending the subject, you will create something far more interesting than essentially remixing a few primer books and articles that have suggested to you by an LLM.
People have indeed "been creating amazing things with nothing" in the expressive arts, but that approach falls short when the value comes from communicating depth from narrative information.
Where as marketing at all these corporation is trying to genericize "AI features" into anything using an algorithm. "Content aware fill", something we've had for over a decade is now "AI object removal"
"Noise suppression" is "AI voice extraction"
Motion unblur is now "AI motion unblur".