this post was submitted on 04 Mar 2024
25 points (82.1% liked)

Futurology

1809 readers
174 users here now

founded 1 year ago
MODERATORS
 

A combination of millions of people with production skills in California, and a huge domestic market able to spend billions has fuelled one of America's biggest exports - TV & Movies. They're not the biggest exports in dollar amounts, petroleum brings in almost 50 times more, but they give America something else apart from money - soft power via an outsized place in global cultural consciousness.

What happens when that sharply recedes? Soft power isn't as easily counted as the size of aircraft carriers or the number of missiles, but its effects are real.

you are viewing a single comment's thread
view the rest of the comments
[–] pimento64@sopuli.xyz 23 points 8 months ago (1 children)

People have been able to easily make their own TV and Movies since consumer-grade videotape cameras became affordable. When was the last time you willingly sat down and watched amateur direct-to-video media that wasn't porn?

[–] Lugh 0 points 8 months ago* (last edited 8 months ago) (3 children)

That's missing the point. Yes, you could have made a version of Game of Thrones with a smartphone & your friends in costumes made from thrift store clothes, but there's a reason everyone wants the one HBO spent a billion dollars to make. Now (or very soon) that advantage will disappear.

With generative video, such as Sora is demonstrating, everyone will be able to make the same standards as Hollywood. Its financial and skills clout won't count for anything, or confer any advantages any more.

[–] andrew_bidlaw@sh.itjust.works 8 points 8 months ago* (last edited 8 months ago)

Problem I see here is not even in tech. Movies' language is ridiculously controlled and intentional, up to the point you can tell who is the director or what influenced them. It comes in completely opposite direction to what LLMs are as a tool.

It's demonstrated by endless remakes\sequels not holding any water, and these done by real people yet lacking many things that made originals stand out. LLMs can help you here and there, probably, but it's not revolutionary to the industry. Even endless uninspired sitcoms can hardly be imitated with generative models. Especially with consistency.

Honestly, I'd suggest you to go ahead and try it yourself. Even a campy movie for friends and family or a simple musical clip requires a lot of effort. And even if you have all scenes at hand (by generating them?), editing can take days and decide if it's good or a flop. That's why some directors only work with familiar editors who they can trust with their movie. And there are many professions like that in the industry. You lose nothing and would probably get a fun hobby.

One thing: you can use your phone, but set focus to locking once you hit the rec button. Camera changing focus is the most annoying shit that can kill a good shot.

[–] dream_weasel@sh.itjust.works 6 points 8 months ago (1 children)

I don't know how you figure an independent studio or individual will be able to keep pace with the likes of Hollywood or HBO. The premise is ridiculous. There's more to it than good cameras, good microphones, and some minimal computer processing. There's the story writing, script and dialogue, sound and sound engineering, sometimes CGI, the location and set that is immersive and complementary, and a slew of other stuff covered in the credits to contend with. Unless you think everyone on the planet is going to prefer blair witch style movies your future isn't close by, it's literally never coming.

Now you might say that AI can solve this problem, but unless you've got lots of people with the vision and ability to write unreasonably explicit prompts, that future is also way, way, WAY off and still improbable: consider how many pictures of people with hands and teeth there are and see how bad AI generations are. You can't give 30 Marvel movies as training data and expect the next Ironman to emerge from your fanfic write up.

[–] Lugh 0 points 8 months ago (1 children)

, sound and sound engineering, sometimes CGI, the location

If you look at Sora now, it seems obvious within a couple of years it will be able to produce TV and movies, indistinguishable from today's human produced ones, entirely with AI. I'm sure good human storytelling will still command attention. However shows like "Game of Thrones" that HBO spent over a billion to make, will be able to be produced by AI alone.

[–] dream_weasel@sh.itjust.works 2 points 8 months ago (1 children)

I haven't looked at sora but I hear good things. Even still though, as someone who works in AI, the outcome you get is only as good as the data you put in for training and there just isn't enough yet. Even if there were, people are super sensitive to plot holes and inconsistencies. Though, you might be able to get Michael Bay to retire ;).

[–] wahming@monyet.cc 1 points 8 months ago

people are super sensitive to plot holes and inconsistencies

I wish. They seem to be everywhere

[–] CanadaPlus@lemmy.sdf.org 3 points 8 months ago

We know so little about how Sora works I'm reluctant to assume it will get better. It could be that it fundamentally isn't easy to force it to render a specific person twice, for example. It definitely has a limitation with panning back to the same stuff.

Otherwise, yeah. I'm not sure why this is controversial. I guess most people here think only Americans can direct or edit? Like, statistically, there's bound to be a Speilberg born in a slum somewhere.