Inside the AI Conversation on BNN Bloomberg: Interview with Brask's CEO, Maria Chmir

In a recent interview with BNN Bloomberg, the CEO and founder of Brask Maria Chmir engaged in a thought-provoking discussion with hosts Amber Kanwar and Paige Ellis. They explored the world of AI technology, deepfaking, and the ethical considerations that surround it.

You can watch the full interview here.

AMBER: Maria Chmir is the CEO and founder of Brask. Maria, thank you so much, I know it was so much effort to do that. And for all the concern about getting deepfaked without your knowledge, this was actually quite a process–I mean, I had to sit and speak for twenty minutes! So did Paige. Is it not actually that easy to pull off a deepfake?

MARIA: Hi, Amber! I’m really grateful to be here with you today. The process is really smooth. If you want to, as you said, deepfake yourself, you just need twenty minutes of video with your emotions, your identity, and that’s enough for our neural network to replace your appearance and voice.

AMBER: You know there’s a little bit of a taboo around it, concerns about how these images can be used. So it’s interesting that you’ve built an overt business around this, offering this as a service. Tell me about who’s using it, who’s coming to you to get this done?

MARIA: We are working with both sides of the market. I mean, not only production companies and studios, but we are also trying to provide our service for talent managers and actors. Not only A-list actors, celebrities, but also their understudies. Because, as you know, right now it’s impossible to replace the charisma of a human. That’s why we need to work with understudies, and this is a win-win story for both sides.

PAIGE: This has been a very contentious issue during these strikes, the actors strike in particular. There’ve been concerns raised about how AI will be deployed and potentially harm the actors. So, how do you respond to those concerns that have been voiced?

MARIA: Of course, being the CEO of a global AI company, I’m looking at new tech positively. Our team considers it a tool to supplement actors, but not to replace them in any way. Brask Doubles is a way for actors to find new commercial opportunities. A-list actors, as I said, can participate in several projects at once. Meanwhile, their stunt doubles’ vast contribution, necessary for the algorithms to work, gets more orders. For studios and independent production we are providing the service to optimize their costs. This is a really great opportunity for the entire market, and we hope that it’ll be the new norm for the market.

AMBER: Well, but talk about what rules you adhere to. What are some principles that you govern yourself by to make sure that it’s not misused or you don’t run afoul of, you know, licensing issues, copyright issues, likeness issues?

MARIA: I totally agree with you that this technology could be dangerous not only for the market, but for the users. When I talk about generative AI, I like to use the metaphor of uranium. At its core, it’s just an element. And generative AI, at its core, can give you the pure energy for new ways of making fascinating creative content. At the same time, it could be a weapon of misunderstanding. So, when I talked about normalization, I meant that we hope that celebrities and talent managers and the union will help us normalize this technology. And I think that after this normalization happens, we will utilize the technology for good reasons, and the next step will be regulation. Because of course we need to figure out together the rules on how we can use this technology for good.

AMBER: And do you imprint the clips with some sort of a digital signature that would allow people who know how to figure these things out realize that “it’s not really me”?

MARIA: I think that labeling is one of the best ways we can recognize that we are currently in contact with generative content. Yes, we can use it, of course. And also I hope that platforms, big players like Meta, or Snapchat, or other social media platforms, will help us understand which content was made with the help of GenAI. This is also very important–not labeling, but recognizing technologies.

PAIGE: I’m curious if you’ve had conversations with the estates of dead celebrities. You know, the potential for an Elizabeth Taylor film to come out, brand new, never seen before. And if you have had those conversations, what types of concerns they’ve brought to you?

MARIA: Yes, we had these conversations with relatives. I mean we’re talking about the rights of famous people who passed away, so we need to ask their relatives if they’re willing to work in such a way. And right now we are making really great documentary projects, documentary films which will include this kind of deepfakes. But I think that this project will be very useful for the entire market because we will see the great historical figures again, on the screen, in a new format. This is very fascinating to imagine.

Curious to learn more about the topic? Follow these links to delve into Brask Doubles and approach to
‍ethics in generative AI.

Beyond its pioneering work on digital doubles, Brask is leading the way in AI-driven content localization.
‍Rask AI is a tool that lets you effortlessly translate videos into 130+ languages while keeping the tone of the original voice intact.

As AI technology is evolving, Brask remains at the forefront of responsible tech innovation. Stay tuned for more exciting developments from the team!