cover of episode The Role of AI in Hollywood

The Role of AI in Hollywood

2024/11/29
logo of podcast WSJ Tech News Briefing

WSJ Tech News Briefing

People
J
Joseph Gordon-Levitt
P
Prem Akkaraju
Topics
Joseph Gordon-Levitt认为,AI模型的训练依赖于大量数据,而这些数据中包含了演员的表演,这些数据未经许可和补偿就被使用,这严重威胁到演员的生计。他主张,根据数据尊严的原则,人类应该对其生成的数据拥有所有权、同意权、署名权和补偿权。他认为好莱坞可以作为AI技术潜在问题的预警信号,并以音乐行业的发展为例,说明技术发展与法律法规的调整需要时间,但最终会走向公平合理的解决方案。他强调,演员与观众之间的深层联系是无法被AI完全复制的。 Prem Akkaraju则表示,AI模型的训练离不开数据,数据如果没有AI则价值远不如现在。Stability AI致力于保护知识产权,并与艺术家合作,为其数据付费。他承认硅谷长期以来一直未对数据使用者进行补偿,并表示Stability AI与其他AI公司不同,它来自电影行业和科技行业,具有高度的创造力,这使其在AI行业中独树一帜。他认为AI的责任贯穿整个技术栈,Stability AI致力于承担其责任。他介绍了Stability AI在数据获取方面的做法,一部分来自公开的公共领域数据,一部分来自有版权内容的许可库,并正在与体育、游戏和电影工作室进行合作,以获得数据的许可。他认为,鉴于AI技术的革命性,我们需要重新思考知识产权的含义,并以电影制作为例,说明电影制作中,导演、摄影机和演员之间的互动是不可或缺的,而一些后期制作工作更容易受到AI的影响。 他认为,演员与观众之间的深层联系是无法被AI完全复制的。

Deep Dive

Shownotes Transcript

Translations:
中文

Say this is your financial life. Over time, things can get more complex. With a personalized plan, Merrill can help you navigate it all. Learn more at ml.com slash bullish. Merrill, a Bank of America company. What would you like the power to do? Investing involves risk. Merrill Lynch, Pierce, Fenner & Smith, Inc., registered broker-dealer, registered investment advisor, member SIPC, a wholly owned subsidiary of Bank of America Corp.

Welcome to Tech News Briefing. It's Friday, November 29th. I'm Julie Chang for The Wall Street Journal.

Can Hollywood work with artificial intelligence? Stability AI CEO Prem Akaraju and actor Joseph Gordon-Levitt went head-to-head last month at WSJ Tech Live discussing just that. Image generation startup Stability AI is best known for its generative AI model Stable Diffusion, an open-source tool that companies can use for free. Its investors include ex-Facebook president Sean Parker.

and Prem Akaraju is no stranger to the entertainment industry. He was the CEO of Weta Digital, an Academy Award-winning film production and visual effects company behind films like Avatar and Avengers: Endgame.

Joseph Gordon-Levitt, known for films such as 500 Days of Summer and Inception, says that performances by him and other actors are being used by large language models to train new AI technology without permission or compensation. They were joined by WSJ senior personal tech columnist Joanna Stern. Here are highlights of their conversation. Joseph, I know you have a lot of thoughts about how these models are made, how they're trained.

Yeah, well, that's exactly right. The first thing to say is for anybody who maybe doesn't know that much about how the tech works, and I'm no engineer, but these models, they can't do anything without a ton of data to quote unquote train the models.

So that it can then spit out these new outputs. And the sort of sleight of hand of calling something artificial intelligence is it kind of makes you ignore the fact that these things were created by humans. Because who made all that training data that went into the AI models? Well, humans did.

And so this is something that concerns me a lot as an actor who works in show business because, well, frankly, you know, my livelihood and the livelihood of the people I have worked with for my entire career are all being threatened. But I would also say Hollywood could serve as a canary in the coal mine. And you've written some really strong editorial pieces about how you think that can be solved.

Well, so and this is these ideas are not mine, I will say, but I've found it really inspiring to hear what some technologists and economists write about. There's a movement that gets called data dignity, which is the basic principle that if a person generates some data, the person ought to have an ownership of that data.

And when I say generate data, I mean that could mean you act in a movie, or it could mean you wrote something, or it could mean you took a picture, or it could mean that a camera...

took a picture of you or it could mean that you hit a like button on a social media platform or any number of things that happens all day long human beings are now generating data all the time and the way that we currently have it set up is all those human beings don't have any ownership of that data and it feels to me like with this new revolutionary technology coming

We maybe need to change that. And that we should be paid for that data. Yeah, the idea is that a human who generates some data should have the right to consent, the right to attribution, and the right to compensation if their data is used to make money. Prem, do you agree with Joe here? Is that something your company could support? I like getting paid for what I do.

I think everybody else should too. I think that's actually a much better, healthier ecosystem and economy, as I think your point was alluding to. The fact of the matter is, is that it's absolutely true that AI could not exist without data. There's no doubt about it. And data would be nowhere near as valuable without AI.

So, what I think is, and I've made my entire career off creating IP or protecting IP, and I plan on doing that in the AI industry. And I think that's what Sean and I, and then many people might know that James Cameron invested and joined the board as well. And I can tell you we're all three very like-minded in that exact fashion. I can guarantee you James Cameron wants to get paid when he works. And I think that we... But that's exceptional. Yeah.

Not to blow smoke, but that's exceptional in the tech industry. The general philosophy in Silicon Valley for the last number of decades has been, no, we have the total right to just take all this data and we, meaning the biggest companies in Silicon Valley, just make these unprecedentedly lucrative companies.

by not compensating the humans whose data we're using. I take that as a huge compliment, and I appreciate it. It is true. I mean, I think that that's why, yeah. I mean, the first, there's no other AI company that looks like us.

that came from the film industry, that came from technology, and came from a highly creative place. So the three top people at the company are those things. And so I think that that's why we're important to this industry. That's why it was the time was right for Sean and I to do this. And this was the right platform for us to do it. There's a reason that art and science go together.

And I think that that has to be that case from the actual inception. And that means from, there's multiple layers in AI. You know, we're part of one layer, right? Like we're part of the data set and we're part of the training foundational models and then some of the applied, but then we hand it off to a lot of people. It's about responsibility throughout the entire stack. And I know that we're going to be responsible for this one. And certainly that's what we're bringing to the table to actually do exactly that. Because I think it's hugely important. I think it's very dangerous not to do that.

Coming up, what does it mean to own intellectual property? And what precedent has the music industry set for film and TV? That's after the break. You want a straightforward path to your goals. But at Merrill, we know things may get in the way.

Or new opportunities can put you at a crossroads. With the bull at your back, you get a personalized plan and a clear path forward. Go to ml.com slash bullish to learn more. Merrill, a Bank of America company. What would you like the power to do? Investing involves risk. Merrill Lynch, Pierce, Fenner & Smith Incorporated. Registered broker-dealer. Registered investment advisor. Member SIPC. A wholly owned subsidiary of Bank of America Corp.

So let's drill in, though, a little bit on that, because that would mean your training data, anything that's in there, you'd have to pay to the artists that...

made that content, whether it be music, whether it be audio, images, all the things that Joseph did a great job listing. That's possible? Well, not only possible, we did it. So from day one, Sable Audio was licensed. And from the image side, there's two main areas that we get data from. Number one is truly open source public domain data and images. And the other is licensing. So one in licensing libraries of content that are copyrighted.

which we do and we support. And what we don't, in taking that even further, is that since I've taken over as CEO, I've started multiple conversations across sports, across gaming, across the film studios themselves to actually work with them on licensing data. Just to take this concept a little further, with the technology being as revolutionary as I think it is,

we maybe need to rethink some ideas of what it means to own IP, intellectual property. And I'll use the example you used because it's closest to my life. But again, I think this applies in so many industries throughout our whole economy. You mentioned licensing video or movies from movie studios.

So in my career, I've performed in many movies. And I've had deals with those movie studios pay me to act in the movie. And then they own the IP and I don't own any of it. And now they're likely going to go license a lot of that material to AI companies that will then...

be used to train AI models that can now produce outputs that theoretically replace what I do and what many of my fellow brothers and sisters in the film industry do. Now, if I had known when I made those agreements, "Oh yeah, I'm signing this agreement with you, film studio, and you'll have full ownership of the movie, and you'll then go use that to replace me and I'll be out of a job forever,"

I probably wouldn't have done that deal. So it seems to me like the whole idea of intellectual property, all these deals, they really, if we're being fair, should be renegotiated in light of this new technology. So we've seen this movie before. It's called The Music Business. So this happened in music. Didn't turn out well for musicians. Well, no, it turned out great now. So it took a while, but if you look at...

So technology always is going to be ahead. That's sort of almost the definition of technology is creating these novel products and services that the world's never seen before. And certainly I'm dedicated to doing that myself. So you look in the music business, is that those rights, those contracts with the artists, with the labels were outdated. They didn't contemplate a world of streaming.

Now, it took a very long time, but now, even if you put a 10-second, whatever, 15-second clip of a song on TikTok or any of these platforms, those artists are getting paid.

So that's totally changed. That's a solved problem now. There's no problem anymore. Now, it took a very long time to get there. I think that we're moving at light speed in the visual media world compared to music, but it's still not done. There's still work to be done. Are the artists getting paid or are the music studios getting paid? Well, that would be an issue with the label and the artist or the artist and the studio. So when I say also that I'm talking to...

early stage conversation with independent film studios. There's a lot of data there that has, in no offense, nothing to do with an actor's performance. So there's so many elements to a film. And when I think about what's not going to change, I think on a film set, there's going to be a director, there's going to be a camera, and there's going to be an actor in front of that camera. And I think that physicality of film production is super important to the creative process. Actually,

Don't know how you do that without that. And I think that kind of action and reaction and the direction is a human process. I think I call that the visible layer. Then there's this whole invisible layer, which is rotoscoping and pain and camera match and plate reconstruction. I could go on and on and bore you guys.

But that part of it is really can be largely affected significantly more than an actor's performance. I think it's just better to put an actor in front of a camera and do it than, you know, kind of mimicking those types of actions in some other avatar that maybe not an avatar of the movie, but that, you know, doesn't exist. I think that deep connection between an audience and an actor is extremely human.

That was Stability AI CEO Prem Akaraju and actor Joseph Gordon-Levitt speaking with WSJ Senior Personal Tech Columnist Joanna Stern at WSJ Tech Live last month. Subscribers can watch the full conversation on WSJ.com. We'll also link it in our show notes.

And that's it for Tech News Briefing. Today's show was produced by me, Julie Chang. Additional support this week from Belle Lin. Jessica Fenton and Michael LaValle wrote our theme music. Our supervising producer is Catherine Millsap. Our development producer is Aisha Al-Muslim. Scott Salloway and Chris Zinsley are the deputy editors. And Falana Patterson is The Wall Street Journal's head of news audio.

And a quick programming note, we won't have TNB Tech Minute this afternoon, but tomorrow we'll have a new episode of our series Bold Names right here in the TNB feed. And we'll be back on Monday with a new episode of Tech News Briefing. Thanks for listening. Say this is your financial life. Over time, things can get more complex with a personalized plan.

Merrill can help you navigate it all. Learn more at ml.com slash bullish. Merrill, a Bank of America company. What would you like the power to do? Investing involves risk. Merrill Lynch, Pierce, Fenner & Smith, Inc., registered broker-dealer, registered investment advisor, member SIPC, a wholly owned subsidiary of Bank of America Corp.