Hollywood backlash puts spotlight on ByteDance's sketchy launch of Seedance 2.0.
See full article...
See full article...
Yeah, but have you watched it though? I agree that review bombing is a thing but this was downright unwatchable.You can’t write off something AI generated just because it “reviewed abysmally” as there is an absolutely massive number of people champing at the bit to criticize anything AI related. Let’s wait for the viewer numbers to see if non-reviewer-types actually want to watch it or not.
Is that really much different than soundcloud artists, most gets ignored and the talented end up getting noticed and elevated by other artists or labels? Who is browsing YouTube looking for random feature films?For me the point is not that someone wont be able to create something good using ai. I would bet money on good content creation. For me the question is, do I want to sift though 100, 1000, 1,000,000 times more media content to find something good?
I read an article about some lady that was using chatgpt to create romance novels in about an hour and then self publishing on Amazon. Supposedly already about 5,000 under many different pen names and also is teaching courses in how to do it. Yuck.
Edit typo
If OpenAI is unfortunately still around in 3 years I will be very curious to see how this works out. The model will have spent 3 years training itself with those characters, and they clearly don't have a handle on how to block everything they want.In December, Disney struck a deal with OpenAI, giving Sora access to 200 characters for three years, while investing $1 billion in the technology.
Wouldn't be so sure about that. As I said earlier, that's very quickly become less of an issue.AI still struggles with consistent appearance over numerous shots and video generations. That's thankfully a big hindrance to videos longer than a couple clips.
I've never used genAI so I have no reference for how fast the results are generated after a prompt is submitted. However, IIRC each frame of ents in Jackon's LOTR took 24 hours to render. I imagine genAI to be quite a bit faster.A lot of downvotes going on for comments like yours, which suggests there are a lot of readers here who aren't really keeping up with how damn fast these AI video generations tools are progressing, both in quality and speed (ie energy use).
Like it or not, there is no question that it will not be long before it is indeed possible to turn out at the least a decent short film with this stuff. That's not to belittle the concerns about IP theft and job losses, but sticking our head in the sand and claiming it's not going to happen is naive.
I’m a CG artist. I’ve spent over 30 years honing my craft and passing my knowledge on to others, as a TD.It costs an average of $18,000,000 to make a Hollywood film these days. Films with a large all-star cast and a lot of visual effects often run ten or twenty times that amount.
Making a movie, with the production values and VFX necessary to fully illustrate the writer-director's vision, is completely beyond the reach of the vast majority of independent writer-directors.
Their stories will never be told the way they want them to be. They simply can't afford it.
Unless AI changes that and gives them the tools to explore and share their visions.
It's disruptive, yes. But it's not all downside. I know a lot of people who are very excited about being able to do, in the next 2 to 5 years, things that they've been told for decades are "too risky, too different, simply not worth the studio's funding."
This is similar to my own opinion about generative video tools. They lower the barrier to entry. That means both many people who have no business making a show or movie will be able to and that individuals with a great idea might actually have a chance. I’m kind of looking forward to what sorts of formally gate-kept ideas will be brought outIt costs an average of $18,000,000 to make a Hollywood film these days. Films with a large all-star cast and a lot of visual effects often run ten or twenty times that amount.
Making a movie, with the production values and VFX necessary to fully illustrate the writer-director's vision, is completely beyond the reach of the vast majority of independent writer-directors.
Their stories will never be told the way they want them to be. They simply can't afford it.
Unless AI changes that and gives them the tools to explore and share their visions.
It's disruptive, yes. But it's not all downside. I know a lot of people who are very excited about being able to do, in the next 2 to 5 years, things that they've been told for decades are "too risky, too different, simply not worth the studio's funding."
On YT/FB? Slop has been heavily promoted by YT and Meta, who want to control clicks in any way possible. Funny they invest 300 billion in it, then dont provide filter technologies.More AI cancer. Pure fucking cancer.
it brings out a giant flood of dogshit to the point where scifi magazines have to cut off submissions because they're overwhelmed with AI slopI’m kind of looking forward to what sorts of formally gate-kept ideas will be brought out
Ars editors push the most upvoted comment to 3rd place... What do you mean, criticize MPAA on ARS-condénast? Maximalism rules ok buddy. Time for spiderman 5.AI slop vs notorious copyright maximalists like Disney and other MPAA members. Oh no, how terrible.
I'd do a 'let them fight' jpg, but i don't want a 'Cease and Desist' letter...
It's Ars Technica, not ARS Technica. I swear, we can't even get good trolls these days.Why is this less voted comment pinned at the top, and the blatent ARS revolt againt MPAA and copyright abuse is pegged down in comments?
And if the book writers are the very same people writing the Marvel slop scripts?It seems that with AI, making a movie will soon become like writing a book. We might soon be getting some fantastic movies instead of Marvel slop.
Nothing's pinned, it's just in time order, and that's when I commented.Why is this less voted comment pinned at the top, and the blatent ARS revolt againt MPAA and copyright abuse is pegged down in comments?
When the bubble pops, it won't be the AI boosters left holding the bag.I want them all to lose.
While I agree that GenAI video isn’t there yet, it’s a mistake to look at self-driving and conclude that video AI will be equally slow to develop. The biggest difference between AI self-driving and AI video generation is that, when these AIs inadvertently drive Tom Cruise’s car through a concrete wall, the self-driving AI’s operator is left with a billion-dollar lawsuit, while the video AI’s operator can simply try again. This makes iteration quite a bit easier for the latter case.GenAI video doesn't scale to a full movie. It's fine for augmenting shots, fixing shots, changing things... but the amount of work involved in replacing a full movie, with good performances, with nothing wrong... it's just huge. The overhyped demos that get widely shared don't pass a quick visual inspection, and they aren't going to replace full movies any time soon. It's good as a replacement CGI and for memes.
It's a mistake to think that because it's gotten better, it'll keep getting better at the same pace. Tesla's been promising self-driving cars for a long, long time, and it's always just out of reach.
Now, coding... that's far more ordered, and it has gotten a lot better recently. And images are easier too. Yes, this is getting better, but many videos will need real people saying real things for a long time yet.
If anyone really wants to dig into AI for creative uses, I wrote this book which came out recently: https://aiforcreativeproduction.com
This is just... automation. Once upon a time it was about spinning yarns, now it's about the act of taking pictures. CGI was decried as the death of an art many years ago, since you no longer had to carefully construct miniatures and arrange for crazy stunts to be performed; film acting was looked down on by theatre performers. The Luddites did have some good points, but in the end, in the bigger picture, automation will liberate humanity from hard, tedious work and let us focus on bigger things (as we have seen in the last 200 years... mostly. Consumerism has sort of muddied the waters). The problem is just who has access to and control of such automation, and that is the real point of contention. I'd say that people should read more Marx, but Americans place him in a similar place to Hitler.I’m a CG artist. I’ve spent over 30 years honing my craft and passing my knowledge on to others, as a TD.
When I hear this kind of philosophizing, I roll my eyes and die a little inside, because it almost always comes from a place that seems to have some good intention, yet is paving the way for the destruction of an entire culture.
I have lots of ideas, more than I will ever get to see realized, but I won’t steal other peoples work and call it “standing on their shoulders” to do so … this is not the same as cutting out newspaper clippings or searching google images for textures to use on a matte painting: it’s theft, wholesale and without conscience.
And this is coming from someone who can see a genuine use for AI tools, but remembers how everyone in the 90s made such a big deal of how “the internet is going to improve the world by bringing us closer together.”
Follow the money for the truth.
To be fair, the West did steal lots of Chinese IP before that, such as when a pair of monks smuggled silkworm eggs for the Roman Empire, somewhere around 650AD. Or when Robert Fortune smuggled tea plants out of the country around 1848.After all, it's not like the Chinese government and its government led/compliant companies haven't been stealing the West's IP since 1972...
Ironically you could use LLMs to filter all of that excess out.= so you wouldn't have to do that.For me the point is not that someone wont be able to create something good using ai. I would bet money on good content creation. For me the question is, do I want to sift though 100, 1000, 1,000,000 times more media content to find something good?
Like it or not, there is no question that it will not be long before it is indeed possible to turn out at the least a decent short film with this stuff. That's not to belittle the concerns about IP theft and job losses, but sticking our head in the sand and claiming it's not going to happen is naive.
So is AI a threat or is it not a threat?
Artists are upset that it's threatening jobs.
Also, artists are upset that it is generating "slop" that can never replace human-created content.
If it's just creating "slop" then people won't consume and it's not a threat. Why not let it die on the vine?