Hacker News new | ask | show | jobs
by golergka 3051 days ago
Can't we expect in a few years to dynamically change what actors have been cast as what charavters with a flip of a button? Deep fakes already look pretty convincing.

Also, insert some scenes, remove others, change the cut a little bit, change color grading, music mix – all these changes look like something modern AI could learn to do.

1 comments

Absolutely not. None of this is possible or even desirable. We don't want to watch optimized video. We want to watch stories that move us, created by geniuses like Coppola, or whoever you think is good.
What if a story could be modified dynamically in order to maximize the chance that it moves you? Or to deepen its impact? I figured that’s what the GP post was getting at.

Just more tools for geniuses to use to move us and move us more deeply.

It makes a lot of sense, and though what I say may sound a bit ridiculous, I don't think most people are capable of making such decisions. They'd rather have some mindless entertainment than moving stuff (we already know this from what's shared on social media, and how TV news and reality shows thrive).
Why would you think that what we want, or, to be more exact, what we think we want, matters? The only thing that matters is what we end up watching more and not closing.

A lot of people say that they want long, challenging reading but they end up reading Buzzfeed articles. Don't ask people what they want to do - watch what they actually do.

The question is probably a bit different - if there is a place for a new genere that is not a linear, sequential narrative model.
There is, they're called video games.