SPEAKERS Gayathri Sindhu, Terry Zhao, Dan Jasnow, Stefano Da Fre
[Gayathri Sindhu] 00:00
Welcome to the Berkeley Technology Law Podcast. Hi everyone, this is Gayathri, your host for the day from the Berkeley Tech Law Journal. Today’s episode is on an immensely exciting and polarizing topic, which is further made so by our two expert guests who have kindly obliged our request to answer a few questions on the same.
While preparing for this episode, we came across an article, which in an attempt to create discourse on the impact of AI in the entertainment industry, asked ChatGPT for its thoughts on this and it readily came up with a scary set of answers and that is exactly what we are focusing on today.
On July 14, members of the Screen Actors Guild and the American Federation of Television and Radio Artists, often known as SAG-AFTRA, a labor union that represents over 160,000 media professionals, began a strike intended to raise demands for various advancements that were required within the media and entertainment sector. However, the main point of contention was the extent and manner in which AI shall be implemented in the entertainment industry.1
This is especially significant to talk about now because, although the strike was officially declared over in November after an acceptable agreement was finalized by the union and the studios, the issue was only truly resolved very recently when the union members, comprising of your favorite actors, singers, and other artists, voted to ratify it with a 78% majority, which is indicative of some revolutionary changes to come.2
We have two accomplished guest speakers with us to give us their nuanced perspectives on the many angles of the strike and its outcome.
Our first guest is Dan Jasnow, who is a partner at ArentFox Schiff and the co-leader of the firm’s AI, Metaverse & Blockchain industry group. Dan has been at the forefront of the AI and metaverse space, and has also helped ArentFox Schiff become the first major law firm to open an office in the metaverse.
And later we will be hearing from Stefano Da Fre, an actor, director and producer best known for the documentaries The Girl Who Cannot Speak and The Day I Had to Grow Up. Stefano has had roles on various CBS shows, such as “Blue Bloods” and “Law & Order: Criminal Intent” as well as MTV’s “One Bad Choice.” His films have been screened at multiple film festivals including the Cannes Film Festival and the Chelsea Film Festival in New York.
His interview coincidentally happened 2 days before the decision to end the strike was made, which makes the conversation with him even more illuminating in the context of what may be expected through this voting period.
So without further ado, let’s hear what our guests have to say.
[Terry Zhao] 03:19
Welcome, Dan. Thank you so much for talking to us today. It’s great to have you join us on the BTLJ Podcast. I’m very excited to speak with you. And I’m sure our listeners are curious to hear the legal perspective on generative AI, and the issues and opportunities it brings to the entertainment space. So to start things off, can you share a bit about your work and experience with generative AI so far?
[Dan Jasnow] 03:44
Sure. And thank you for having me today. It’s great to be with you, and to be a part of this podcast. So I’m a partner at ArentFox Schiff, based in New York. And my background is in intellectual property, primarily trademark and copyright, as well as advertising, unfair competition laws, marketing related work. So those two issue areas led me into the generative AI practice. I do also a lot of emerging tech, guidance, and advising. So we advise a lot of clients who are considering incorporating generative AI into various parts of their business or their creative process. They might come to us and say, hey, we want to do this, or this is already happening amongst our employees. What are the regulatory issues that we should be aware of? What can we do to mitigate potential risks? What are the risks? And really help them think through how to balance the opportunities of generative AI with some of the notable risks associated with its use.
[Terry Zhao] 04:58
For sure. And given that it’s such a hot topic among IP lawyers right now, especially those involved in the SAG-AFTRA deal negotiations: how are people in the entertainment and IP law space grappling with generative AI? How are you guys thinking about this right now?
[Dan Jasnow] 05:15
That’s a really good question. I think everybody is sort of in the same boat. Everybody is trying to figure out what the parameters are going to look like long term for both responsible and ethical use. And certain sets of laws are going to adjust or develop in order to ensure that companies’ or individuals’ intellectual property rights are protected in this new context. So, there are a lot of open questions about copyright law, the extent to which copyright is going to protect works that are created using generative AI. There are a lot of questions about the adequacy of the publicity rights regime at the state level, and whether there needs to be a federal response that adds to the state protections. And a lot of those questions are still unanswered. I think the guild, SAG-AFTRA, writers and directors guilds, and the studios should be commended for wrestling with these issues. You know, so publicly and so early in all of our collective understandings of the implications of generative AI. And I think these agreements are really going to potentially go a long way to helping various industries start to think about how to put in place parameters for responsible use.
[Terry Zhao] 07:00
Since you mentioned the copyright and publicity rights, I just want to quickly ask, in this sort of new age of deepfakes, synthetic media and whatever else generative AI can do in this realm, what’s the biggest challenge that you think this poses to copyright, or portrait rights or publicity rights? And are there any sort of current major legal solutions that people are proposing?
[Dan Jasnow] 07:24
I think the biggest tension right now is that the US Copyright Office is taking the position that copyright does not protect works generated solely by generative AI. And that makes sense in a certain context, right? It might make sense if you just go to Midjourney, and you say, hey, create a photo of an elephant, surfing a wave, and then Midjourney creates it, and all you do is prompt it with one little sentence. And that’s sort of the end of story. It might make sense for that not to be protected by copyright. But what we’ve seen is much more sophisticated creative processes by artists, by graphic novelists. They’re not just using generative AI in sort of this one and done kind of way. They’re really engaging in methodical and sophisticated prompts of Gen AI in order to create new works that maybe where the generative AI tool is just the lens through which they are creating something new.
The Zarya of the Dawn graphic novel is the classic example of this so far, where this artist is spending 1000s of hours prompting generative AI tool, iterating and iterating, and iterating and iterating, trying to get exactly what she wants, and taking the tool from one very unsophisticated image to very sophisticated outputs that align with her vision. But the Copyright Office so far said, “Well, that doesn’t count as human authorship.” So you know, that has positives and negatives. We may not want as a matter of policy that a lot of generative AI works that are protectable under copyright. But from my perspective, it does a disservice to the artists, to the creators, who are using these tools in a way that’s really unique, and that very much reflects their vision as creators and artists. And right now, the Copyright Office isn’t sort of rewarding them for that creative effort and output. So that’s a big tension.
And then the risk of—in the publicity context—the risk of unauthorized synthetic content, or unauthorized use of somebody’s name, voice for musicians, likenesses, that’s a substantial risk. These tools make it very easy to engage in that type of both very real looking type of activity, and also potentially unauthorized performances, etc., that could really undermine actors’ rights over the long term. But we have, as you probably know, a patchwork of state publicity laws. There’s no federal right of publicity. And that just makes it even more difficult for anybody with an economically valuable set of publicity rights to enforce those rights, and ensure that they’re stopping unauthorized uses.
[Terry Zhao] 10:51
It’s very interesting. So in your opinion, should the use of an actor’s likeness in generative AI content be considered an infringement of publicity rights? And do you think that this might require further revisions of state or federal publicity rights law?
[Dan Jasnow] 11:08
I certainly think that the unauthorized display of a digital performance is a violation of publicity, right? So let’s say, a movie studio creates a likeness of Tom Cruise, and then deploys that in Mission Impossible 15 and doesn’t get permission from Tom Cruise to do that, I think everybody would agree that that’s a violation of publicity rights. Where it gets a little trickier is whether or not there’s a violation of publicity rights, if you’re using somebody’s name, image and likeness to train a machine learning model. So you put a whole bunch of these performances into a dataset that trains the machine learning model, that gets to the same point that’s been debated a lot under copyright law about whether the unauthorized use of third party works to train a machine learning model is copyright infringement or fair use.
So same question applies for publicity rights. And that’s not necessarily clear. I think if you’re an actor, maybe you want to know if your performance is being used to train a model. And then it’s possible that performance could generate an identical replica of that actor. So if Tom Cruise’s likeness has been used to train a machine learning model, then you as a user could go to any of these tools and say, hey, create a new movie trailer using Tom Cruise about whatever movie you want to make up in your head. So there’s a risk there. But there’s an open question about whether just the use of NIL rights and to train a model is actually infringing? I don’t think there’s much question that the unauthorized display of that type of performance would be infringing.
[Terry Zhao] 13:05
Right, certainly. As AI-generated content advances, what adjustments would you like to see made to existing copyright laws to protect the interests of human creators?
[Dan Jasnow] 13:18
I think that the Copyright Office is going to have to reevaluate the way it’s currently applying copyright law to these AI-generated works. It’s not enough to just say nothing created using Gen AI is subject to copyright protection. It might be the easier path, but I think in the long run, there’s going to have to be a case-by-case analysis, just as we do for fair use and copyright law, to say, was there sufficient human authorship in the process of creation for this specific work? The Copyright Office already acknowledges that you might be able to substantially modify a work that was created by generative AI to sort of give rise to human authorship, right? So something’s created by Gen AI, then you tweak it a whole bunch, you modify it, you do whatever, then you might be able to protect it. So the Copyright Office already acknowledged that’s possible. But they really need to, I think, reevaluate the opposite scenario where somebody is sitting in front of a computer, they’re prompting these tools over and over and over again. Maybe they’re not modifying the output, but their creation process is so involved and so directed that there needs to be a recognition that there’s human authorship in the direction of that generative AI tool. So that I think needs to change for sure. I just don’t think it’s a sustainable analysis at the moment. And I think all types of different creators, whether they’re studios, or individual artists, graphic novelists, musicians, as they incorporate these types of tools into their creative processes there, there needs to be some sort of recognition that these tools are not just a machine, they are directed by a human author for now.
[Terry Zhao] 15:31
That’s a really good point. I want to talk a bit more specifically about SAG-AFTRA and the other guilds that have been on strike recently. So before the general ratification votes were held, the Writers and Directors Guild of America National Board unanimously voted to approve the agreements. While on the other hand, 14% of the SAG-AFTRA National Board voted against the proposed contract. Why do you think there was a difference between the guilds’ reactions to their agreements?
[Dan Jasnow] 15:58
It’s hard to know for sure. But from just the public reporting, it does seem like there’s concern about the generative AI provisions. I mean, it’s sensitive for all of the guilds. But I think there’s a lot of concern among the actors that there’s a potential for, particularly, you know, as Fran Drescher talks about journeyman actors, to lose access to roles, to have their opportunities diminished, and for generative AI to really depress compensation for the acting community. I think those are all clearly valid concerns. I think the SAG-AFTRA agreement, at least from what we know of the summaries, does a good job of striking that balance between preserving the promise of generative AI and offering some meaningful protections to actors. This technology is here to stay, right? Generative AI is soon if not already ubiquitous, it’s incorporated into all of the tools or it’s going to be incorporated into all the tools we use all the time from Microsoft Office Suite to Google workspace. And obviously, it has much more sophisticated input implementations. But at the end of the day, I think what’s important is to be protecting the actors, writers, directors who are going to be most directly affected, and to recognize that their publicity rights, their name, image, likeness rights are theirs to profit from and license out. And it should ultimately be the actor’s call about how those rights are put to use in any of these projects. And I think at least based on the summary language, the SAG-AFTRA agreement does a pretty good job of providing a framework for that, may not be perfect, but as I mentioned at the outset, we’re at the early stages of this. And this is a hard job to negotiate these types of issues, when we’re still so early in understanding how generative AI is going to affect the economy as a whole. So I’ve been very impressed by how all of the guilds and the studios have worked through these issues.
[Terry Zhao] 18:49
Definitely. And do you think generative AI issues might put the actors’ guild in a more precarious position? Because like you said, actors would face a more existential crisis if they lose their digital portrait and publicity rights.
[Dan Jasnow] 19:04
My perspective is that generative AI offers tremendous opportunities for actors. I think the notion that an actor might be able to create a generative AI, face scan or performance, and then license that out to an advertising agency, to have the advertising agency create commercials of that actor, without the actor having to be on set or in a particular location. And you can think about athlete endorsement deals, and actors who might be going through some sort of health issue, like Bruce Willis losing his ability to talk. All of these tools offer tremendous potential for actors to really commercialize their NIL rights in a way that they never have before. Obviously, we’ve had CGI, but we’ve never had anything that’s as sophisticated as this.
Similarly, with post production, an actor after filming, instead of having to fly back to a studio in order to reshoot a scene, might be able to consent to the use of a Gen AI likeness in order to do that. So they can be off with family, or they can be on a different project. And those are all, I think, great opportunities for actors. It really opens up a whole new potential revenue stream, including post mortem, which is a big area of publicity rights law, where states grant various terms of post-mortem publicity rights. The family members, loved ones can continue to generate revenue streams, for years after somebody potentially passes away. And it’s potential security for a lot of these people over potentially another generation. So I think that’s a great opportunity. It, of course, has potential to be abused, there’s potential for unauthorized use, there’s infringement. But those issues are issues we deal with all the time in every type of intellectual property, right? You’re always fighting trademark infringement, you’re always fighting copyright infringement. And what’s important is to have a system in place that ensures the people who are the owners of those rights or the holders of those rights are being notified of their use, they’re being adequately compensated for their use, and they ultimately are able to control how those rights are used. So I think, again, it’s a balance here between trying to take advantage of these opportunities that Gen AI presents, while establishing some safeguards that will prevent unauthorized use, make sure that the rights owners are the ones who are benefiting and have ultimate say. And I think the agreements so far have done a pretty good job of threading that needle.
[Terry Zhao] 22:32
For sure. In your opinion, how might the relationship between talents like actors, writers, directors and studio producers change with the advancement in AI-generated content?
[Dan Jasnow] 22:46
That’s a good question. I think it remains to be seen, right? I mean, we’re still waiting to see how these new three agreements, the directors, writers and actors agreements will actually be implemented. But a lot of details are missing, particularly in the SAG-AFTRA agreement. We haven’t seen the actual language. What I think is striking in the SAG-AFTRA agreement is how the even more important individual contracts are going to be. The studios are required to provide a clear description of how they intend to use an actor’s digital replica or synthetic performance. Actors are required to, if they want to deny rights to a studio to use their digital performance, posthumously, they have to explicitly hold back those rights. And to me, looking at that there’s the potential for a lot of litigation around those issues, because the studio has to clearly describe the nature of its proposed use of digital performance. What happens when that use varies a little bit, or there’s a disagreement about the actual use, the implementation of that right differs from how it was described in the contract? And then, you have all sorts of questions about royalties, and payment compensation. So it’s certainly another area where there could be strife between the actors, writers, directors and the studios. But again, that’s not necessarily different than anything else, we deal with those issues all the time. I would hope that there might be, ideally, it’s not a zero sum, right? This technology potentially grows the pie, and everybody gets a bigger slice. I think that’s sort of what the hope would be, but that might be a little bit too optimistic.
[Terry Zhao] 24:57
Great. Well, that’s all we have. Thank you so much, Dan, for taking the time to speak with us. I’m sure our listeners will learn a lot from your expertise and insights. So thank you again.
[Dan Jasnow] 25:08
My pleasure. Thanks, guys.
[Gayathri Sindhu] 25:10
Let’s now hear from our second guest, Stefano Da Fre.
[Terry Zhao] 25:18
Welcome, Stefano. Thank you for taking the time out of your busy schedule. It’s absolutely wonderful to have you join us on the BTLJ Podcast. I’m very excited to speak with you about generative AI and the challenges it poses for actors, writers, and directors.
So to start things off, can you share any personal experiences where generative AI has directly impacted your work either in front of the camera or behind the scenes?
[Stefano Da Fre] 25:58
First of all, I’m really glad to join you guys. Thank you for having me as a guest on your show. You know, it’s the elephant in the room, right? AI is a fundamental conversation piece, negotiating piece, it was part of the writers strike. The WGA just basically came to settlement terms, and now still ongoing with the actor strikes. We have been on strike since the 12th of July, I believe. So it’s going to be almost half a year. And it’s the talking point and the challenges in the room. So we see the aspects of generative AI in post production, directors have already had to deal with this. And I’ve personally had to deal with this before in post. For example, in the film that I’m working on, called Stolen Dough, which is a feature film. And this film, basically, if you’re in a coloring session, you can look at colors to try to match. And you don’t need a colorist to be able to match a specific shot that came from one scene to another. In other words, you could have generative AI put in the location and the look of the shot to see what tones, texture, lighting fixtures that you have. And then you would be able to sort of replicate that all the way through in theory. The reality is it’s not advanced yet to the level that it actually works precisely in the way that you would want to have a human colorist do detailed work on that aspect. But that seems to be very normal for us in the sense of moving forward post production. Now lots of people in post will be losing their jobs. That’s the challenge, right? So how do we use it to make things more efficient, but not take advantage of the labor that we have? Because, as you know, that filmmaking is extremely labor-intensive.
In terms of the actors, the way that the studios have positioned, asking actors and propositioning them with generative AI, if they needed to place you in a scene, they would be using your likeness, whether you’re speaking or not, and pay you for the contract of that one day. But then they would be able to use you throughout the season, without paying you for coming into work. If you lose your likeness, the thing that you bring as a human being, you essentially lose your standard of living. The purpose of being an actor is that you are actually the person who is delivering those lines. That’s why we find it so susceptible, in political discourse, with a lot of deepfakes that YouTube is around. So all these aspects, without going into too much detail into your first question, all of these are a very big challenge for the film industry. And for the actors that are performing in those films.
[Terry Zhao] 29:56
Certainly. So apart from just post production, are there any sort of current projects that you’re working on or that you know of, that generative AI has already started to affect other parts, like, the writers room or other parts of acting and the sort of pre-production aspect of it?
[Stefano Da Fre] 30:18
So let’s talk a little bit about what the value is of AI. If you input scripts inside that are shitty, it’s going to learn on shitty material. If you put on things that are very strong, it’s going to learn on things that are strong. It gets smarter and learns on its own material. So the fear that a lot of the writers had, and that they had worked into their contracts—this is part of the settlements—that their scripts, their stories that they were written by humans would not be entered into the generative AI writing programs, so that AI would be able to use their scripts as a basis to learn independently on their own. And this is a massive challenge, right? Because the people that are software engineers, and people that are at the forefront of technology, want to input the most amount of intelligent data, so that it’s generating better data. But who owns that data? The copyright of that data is fundamental. It’s not so clear that the AI has the right to borrow specific artistic materials to learn from. So I think that’s another argument that many of the unions are talking about from the writing perspective. From the acting perspective, it’s much more clear cut, right? I know what my face looks like. I don’t know if you can see me. I’m wearing an aviator nation’s hoodie. I’ve got blond hair, blue eyes, my face has a specific symmetry to it. And it’s very clear how it would be able to mathematically project the same image and be able to take it from a photograph or from a strip of film, and to put it into another scene. So this is the challenge, right? Whose material are we putting in so that this machine can learn is part of the more insidious parts of artificial intelligence.
[Terry Zhao] 33:05
Certainly. And to your point, Stefano, I would just like to give our listeners a bit of background context on where we currently are with the negotiations. So on November 6, SAG-AFTRA released a statement on X, formerly known as Twitter, saying that they had responded to the alliance of Motion Picture and Television Producers’ “last best and final offer” and ”negotiations over this offer once more dissolved” in part because of what you mentioned about the studio’s proposal to reuse artificial intelligence scans of potential deceased performers and actors’ likeness without their estate or the guilds’ consent, and the guilds cite that their determination to secure a deal to protect actors’ likeness. So you already touched upon this a little bit, but I just want to go a little bit further. For you and your current projects that you’re working on, Stefano, how do you foresee that these movies and these projects being impacted by the ongoing negotiations and a potential deal that SAG-AFTRA might be able to achieve?
[Stefano Da Fre] 33:48
For me, I’ve seen the impact of how quickly it can work on the post production side, just on a coloring perspective. Also, let me share a personal story with you. I was slated to be a television director for a television show basically of a set of jurors who are human, a host, and a generative AI, who basically makes decisions. Low grade under $10,000 criminal decisions between the defendant and a plaintiff, through AI. And I went through this whole process, I must have been on this project for five months, but we ended up breaking apart for creative differences. On one hand, I was drawn to it, because I was honored to be asked, and we were on the forefront of technology. So hey, who does not want to enter into the next phase of technology? It’s always been to the detriment of people who are the late comers to adopt technology quite late. But just like the line of Jeff Goldblum in Jurassic Park, you stood on the shoulders of giants of humans, and you had a piece of technology, which had no ethical responsibility, you had no knowledge or appreciation of how long it took you to get it. That’s the problem. The people who are making arts and making movies, it took them, let’s say 10,000 hours of experience to create this product and had to ethically figure out how to source this material. The problem with generative AI, aside from the copyright issues, is that it’s not clear what the ethics are. We haven’t built in ethics into what it can or cannot do. We’re just still in the phase like, wow, look at how amazing this acid experience is. Let’s see how incredible this is. That’s the wrong questions we’re asking. We should be asking what it’s permissible to do before we let the genie out. Because we realize that this thing is working at a very rapid case. I mean, for me, I can even see just in color corrections, how quickly it can go through an entire scan of a film, and tell me what the color temperatures are of the film. And also tell me what the lighting sequences are. Now for me, I have found it to be wrong a lot of times, and sometimes when it’s wrong, it’s way off. But when it gets it right, you’re actually amazed that it took someone, let’s say, 50 hours to color a scene. You realize this is going to eliminate jobs in its beta testing if it’s not regulated. And that’s just one aspect of some obscure part of filmmaking, which is color correction. So we need to ask these ethical questions of what we believe as a society is permissible. And in the field of art, you know, you can write a story. You can learn on stories of other people, but you never had to experience the poetry of Keats and Shelley, and Oscar Wilde, and Lennon McCartney, of losing a child or losing a girlfriend or losing a boyfriend, to be able to inspire the writing that you’re doing. In a sense, it’s going to be a very big topic for us, you know, what is human-driven art? What is technology and machine-driven art? And this is the forefront of what is going to be in the art world for us in the next 10 years, and this is irrespective of employment, irrespective of what deals get found. These are the philosophical questions. The reason why we need good philosophers in the room to ask these questions because policy is going to make or break how it’s used, and what we as a society agree is permissible or not permissible.
[Terry Zhao] 39:14
I definitely agree. I think also for US law students or people working in the legal space is also a very interesting question to see exactly what level of permissibility we can grant AI, like you said, it needs boundaries, it needs certain defined limits to what it should or should not be able to do. So I’m also curious, right? From your perspective, as an actor, what would you like to see accomplished concerning generative AI right now through the SAG-AFTRA deal? I know, like you said, it’s going to be a process. And this is just step one, in the very long battle with or maybe not battle, but sort of a long toggle with this new technology. But just within the context of the current deal, what would you ideally like to see accomplished?
[Stefano Da Fre] 39:59
Terry, it’s a very good question, because at the heart of your question, on this podcast is, where is the place for AI, right? What is the boundary and the sandbox in which it should exist? And the answer is that it’s a useful tool, but it can never be your master. So it should never be able to replace or substitute for actors’ faces and likeness. I think it should have a place in terms of the same way that CGI is used, right? We know that robots and explosions in the post production world are handled through graphic effects. I think it’s completely morally permissible, and very reasonable to sustain itself within the realm of how we look at special effects. It should stay in the world of a special effect, the operative word meaning special and effect, right? But when it starts entering as a substitute we know that we inherently, I guess not everyone knows, depends on what your moralistic stance is on this. But when it becomes a substitute, then we know it’s not right. It’s not the way that was intended, and I think that the right way is to be used as a proper assistant, but it should not substitute for what is incredible about being human. We are human beings. And there’s a level of value from the experiences that we live in our shared community, even the three of us right here talking together, that is unique to being human, and that sacredness should stay in art form and should be the main parts. So I love it as an assistant, but not as a master.
[Terry Zhao] 42:34
For sure. And this might be a bit of a personal question. So feel free to disregard this, if you feel uncomfortable answering. But do you feel like the union so far has represented your interest and the interests of similarly situated talents? Well, throughout the negotiation process, is there anything you would like to see that they do differently?
[Stefano Da Fre] 42:54
I think that the union has done a very, very good job of understanding that this is a battle of 10 years. In other words, this is not a problem that we’re going to have in 2024. It’s not going to be a problem of 2024 or 2025. It’s going to be a problem in 2030 and 2035, when this contract has already been signed, and done poorly, that we’re going to have to face. So from that aspect, I believe that they’ve represented us well.
What they haven’t done well is explain to regular yield people and regular layman actors, that streaming services are losing money and most people don’t understand that. But I’ll just give you a very simple metaphor. When you started to join Netflix, and you were paying $200 or $180 a year for cable. And all of a sudden someone said, here’s a $10 subscription with no ads, to be able to watch as much content as possible. In the back of your minds, if you didn’t know something was wrong, you’re asleep at the wheel. Essentially, Netflix created a business model that lied to the public, lied to Wall Street, was losing billions and billions of dollars. And the industry, who were our parents and our employers, rather than telling the public this is an unsustainable model, just like pollution, just like issues with climate change, rather than saying to the public, we need advertising dollars from Kellogg’s, from Procter and Gamble, from shampoo companies, from Mazda, from Ford and General Motors, because that subsidizes a very clean business model that helps actors get paid for the residuals, rather than being intelligent. What did they do? Disney went forward to buy 20th Century Fox and started to fight the streaming wars and everyone followed Netflix’s path. And that path, unfortunately, is going to end nowhere. There’s no world that we’re going to live in, where it’s sustainable for any one of us to consume endless amounts of content without advertising. Because that advertising money ends up trickling down to the residuals that actors make. And it’s not enough to be making it from the subscription, unless your subscriptions go back up to $180 to $200, which we ended up going to a reverse cycle of what we were doing in the 90s and early 2000s. Anyway, when we were paying for a cable bundle for hundreds of dollars a year, which is what most of us had growing up. So I really want your listeners to understand this. This is the issue and the union, SAG-AFTRA, could do a better job of explaining this nuanced problem. It’s not like the studios have zero reason to not accept the strike. They’re being squeezed, too. And you say, Stefano, how come you’re so balanced and centrist that you sound like someone who sort of anti-AI? Yeah, I am. I definitely am for asking the right questions. But I’m a capitalist as well, you know, we live in a capitalist country, and I’m for the belief of sustained capitalism, right? That’s a healthy business model. Look at the price of Disney stock. It’s trading at $84, it’s at a five-year low. Nobody in that company, no shareholders, no endowment funds have made any money on the stock. It’s because it’s losing money. It’s not its parks division that’s having problems. You can look with any company, by the way, you can look at Paramount Global, you can look at Discover, if they’re all having problems. AT&T, which owns HBO, hasn’t made money in the last 10 years. So the point is, we’re chasing our own tails without being honest to the public, of the fact that we need a sustainable business model.
[Terry Zhao] 47:38
Yeah. And I think just quickly bringing it back to generative AI, I just kind of want to get your thoughts about: is there any possible area you might think that generative AI might create more opportunities, or even to mitigate the sort of many industries that it’s tearing down and destroying? Is there any sort of benefits it could potentially pose for actors or people in the industry?
[Stefano Da Fre] 48:02
No, no, I mean, for post production? Or ultimately? Let’s reverse that question and think about the transformation from film to digital cinema, right? So what did we gain? Well, we got efficiencies, we saved money, we didn’t burn through our film. We didn’t burn through exposure. We didn’t have to worry about exposures to sun, right? You had your canister and you were shooting on a Bolex, you had to check the gate make sure there are no hairs in the gate. The process of filmmaking is much quicker for me than it was for an older generation, even in Generation X. So the benefit is the barrier to entry in filmmaking will be very low, very low to produce content. Some people think that’s good. I don’t. We have too much content. There’s too much material already on Tik Tok. There’s too much on your Instagram reels. There are already too many people having the access, so it’s not an issue of having a barrier to entry. So the position that one would take who is looking at this objectively would say, Stefano, you’re not being fair, because this will allow people to have more tools to create. Yes, but at what cost? You can still do many of the things. All it’s doing right now, is not something brand new. It’s not creating something brand new. It’s substituting the processes to get what we already have. We already have writing, we already have spectacular images. We already have great photography, great colorists, great CGI. What we’re trying to do with generative AI is to use it as a tool that will be able to get there quicker and be more efficient. I will say one positive thing, though. So going back to your first question, who’s it going to employ? It’s not going to employ us in the arts, it’s going to employ software engineers, computer engineers, and computer scientists. But nobody who is in the relationship of my community that I work with or that I know. And the closest that I come to a computer engineer would be a colorist or someone who does very technical things or have special skills of doing post production effects.
[Terry Zhao] 50:48
I think that’s absolutely a good point. Since you also mentioned that AI is sort of better, or perhaps only been, like, positively used as a tool. So do you personally anticipate that you’ll have to use some form of generative AI in your own work?
[Stefano Da Fre] 51:18
Yes, of course. I’m not a hypocrite. Look, I use a manual colorist as long as I’ve been a director to color-correct my films. However, the question is, what happens if that person is not available? And the technology gets so good that basically I’m able to do it with myself, and use this tool, generative AI, to help further along. I think that all filmmakers are in a situation where they don’t want to be left behind, and they want to know the tools so that they’re able to employ them in an ethical manner. But I don’t think the answer is not being engaged in what is going on and what the tools can do, right? Whether it’s writing, whether it’s having it respond back to you in scripts, or whether it’s doing all sorts of work and post production. For me, it’s something I’ve used, it’s not been to the quality of work that I believe is true art. So it’s not been a tool that I’ve presented to a client, but it has been a tool that I have used and experimented with, since May of 2023. In terms of everything, images, writing, graphics, everything.
[Terry Zhao] 52:55
For sure. And so I think given all that we’ve said about the changing landscape and how much of a sort of different perspective that AI must present for every actor, what advice would you give to up and coming actors and directors about how to coexist with AI and how this might affect their careers going forward?
[Stefano Da Fre] 53:16
Well, I would say to actors, if you’re going to be an actor, stay an actor. Don’t stop being an actor because of the fear of AI. There are always great reasons. Great walls of worry that we have to climb. Whether it was the depression, whether it was the loss of film on to digital, there are always existential issues that come up. This one granted, I think, is unprecedented. But the advice I would say is to get involved with your community organizers and to unionize, join the union and become part of the union. There are resources for the Directors Guild of America, the DGA, you can go on dga.org. And you can also visit resources for sagaftra.org. Become a member of the union. If you’re an actor on theater, join equity. Join the union so that you can be part of a brotherhood or a sisterhood of fellow working professionals, who are having this dialogue. And you’re not worrying about this alone. So make sure that you follow what you want to do as an artist and never feel alone. Join the unions.
[Terry Zhao] 54:36
That’s great. And for a final question, just to sort of end on a lighter note, if you had to choose what movies would you say inspire you, and your desire to speak up and the issues that you care about and that you talk about?
[Stefano Da Fre] 54:51
I think the films that inspire me the most, the director that inspires me the most is David Fincher, who has very dark tones in his films, and dark characters, dark themes. The Social Network, Fight Club, Zodiac, The Killer. I think, though, him as a director is a perfect example of talking about special effects. It’s the difference between Michael Bay using Transformers’ special effects, which you look at and you yawn, it’s just one big blast after another. Using special effects properly in a David Fincher film, you don’t know whether a camera is manually operated, or basically digitally operated, or operated by person, or it’s done in the edits. And I think the idea is that you use everything, you use all those tools to create with an overall theme of how to create a piece of art, right? There’s a through line through all of it. So David Fincher is 100% my favorite filmmaker who uses special effects and got his start doing special effects through Return of the Jedi on the Star Wars films, and uses it more to hide things, hide problems, rather than being the focal point of the film.
And I think the reason I’m outspoken is I grew up in a big Italian family, so you had no choice. If you didn’t have an opinion, you didn’t get anything for dinner. And nobody knew that you had to have an opinion on something. So that’s where it comes from, from my family.
[Terry Zhao] 56:57
That’s awesome. Well, that’s all the time we have, unfortunately. But thank you so much, Stefano, for taking the time to speak with us. I’m sure our listeners would greatly appreciate your candidness and your insight. So thank you again.
[Stefano Da Fre] 57:08
Thank you for having me. It’s been a pleasure to join you.
[Gayathri Sindhu] 57:10
Thank you for listening! The BTLJ Podcast is brought to you by Gayathri Sindru, Liang-Chu Wu, Terry Zhao and Podcast Editors Eric Ahern, Juliette Draper, and Meg O’Neill. Our Executive Producer is BTLJ Senior Online Content Editor, Linda Chang. BTLJ’s Editors in Chief are Will Kasper and Yuhan Wu.
If you enjoyed our podcast, please support us by subscribing and rating us on Apple podcasts, Spotify, or wherever you listen to your podcasts. If you have any questions, comments, or suggestions, please write to us at email@example.com.
The interview with Dan Jasnow was recorded on November 20, 2023, and the interview with Stefano Da Fre was recorded on November 6, 2023. The information presented here does not constitute legal advice. This podcast is intended for academic and entertainment purposes only.