Heroes in Business - Victor Riparbelli, Co Founder Synthesia, raised over 66Million AI video positive side of deep fakes

Episode Date: January 1, 2023

Victor Riparbelli, Co Founder Synthesia, raised over 66Million AI video positive side of deep fakes is interviewed by David Cogan host of the Heroes Show and founder of Eliances entrepreneur community.  

Welcome back to Alliances Heroes, where heroes in business align. To be part of our super community and find out more about Alliances, visit www.alliances.com. All right, that's right. We're back. And please continue to provide that feedback to us. Comments, you know where to go. You listen to it on the radio, 50 other outlets, keep the feedback coming. And by the way, again, too, we've been getting so

much feedback regarding when I interviewed Richard Branson, that we may end up bringing him back with the more questions that you've been asking that you want to go ahead and ask him. So make sure that you go to alliances.com. That's E-L-I-A-N-C-E-S.com. Why? Because it's the only place where entrepreneurs align. And we have with us today, Victor Rippobelli. He is the co-founder and CEO of, are you ready for this? Synthesia. You can reach him at synthesia.io. Of course, again, we'll have it on our website. So you can go to alliances.com to also listen to past episodes. You're going to want to go to it. And I'm going to tell you why. And he's going to tell you why also too, because it's pretty cool. And I haven't seen anything else

like it. So Victor, welcome to the show. You got a kick-ass website. I love it. And I love what you're doing. Thank you, David. I really appreciate you inviting me here. All right. So how is Synthesia making deep fakes the hot new corporate training tool? Because it's really cool. In fact, Victor, maybe take us a step back and tell our audience, those that may not know it or use it, what it is. Yeah, sure. So Synthesia is the world's largest AI video platform. That probably sounds a little bit funky. But essentially what we enable our customers to do is to create video content of presenters talking head-style content, but without the need for any physical recording process.

You don't need cameras, don't need microphones, don't need actors, don't need studios. The way it works is that you go to our website, you select one of our AI avatars, or you create yourself with four minutes of footage. And then you create new videos by simply typing out the script. We have this sort of PowerPoint-style editor around it, so you can add in your screen recording,

your background images, text, your logo, whatever you might want to add in there. You click generate, maybe a few minutes, and then we give you a video, which is a video file like anything else. But you've made it without any cameras, microphones. That's kind of fundamentally how it works. The reason this is super interesting, and I think the reason people are using us is really because video is just taking over. We're recording video right now. If you look at the internet as a whole, it's very, very obvious that people, they want to watch and listen to that content. They don't want to read it. And everybody sort of knows this. I don't think that's a big secret any longer.

Video is a better medium almost no matter what you do. But the problem is that video is a thousand times more difficult to create than, you know, writing a blog post or creating a PowerPoint deck. So we're sort of gapping that by offering this AI video tool where if you want to share something, instead of doing it by a text or by slides, you can now create these AI videos with the same amount of effort. And it looks and feels really cool. Today, we're very focused on talking head style content. I'm sure we're going to get into that later but for us what we're really excited about is in the next five to ten years we want to enable anyone on the planet to be able to make a Hollywood film without the need for anything else than a synthetic login at their imagination. That sounds crazy but if you look at a lot of stuff that's going on in the AI space this will will happen sooner than you think. And it's going to be

really cool. It is so cool because you're able to pick, you're able to pick from a huge selection, male, female, different ethnicities in regards to who you want to be able to have whatever you're submitting speak. And it's, I mean, it's something people really, to really understand it, you got to see it. So after this interview, you can go to Sympathia.com. That's S-Y-N-T-H-E-S-I-A dot I-O to see it. It's so easy to use. We use it.

It's, I use it. And it's, I mean, you can't get any more easier click submit you're done I mean you you've done phenomenal with that so I'm so excited to see what the future is gonna be with it well tell us more about this now you made David Beckham speak nine languages to stop the spread of malaria made Snoop Dogg appear on Australian TV ads without having to film him, and how you personalized Lionel Messi video campaign generating more than 4 million videos. Like, how is that possible? Yeah, so a lot of those campaigns that go a few

years back, and I think, you know, what we really wanted to do back then was we sort of built the initial version of the technology that we now offer to everyone on our website. You could even create a free video. And we really want to showcase all the amazing things this technology is going to enable. Because one thing is you can create videos easier, like you're probably doing today, David. But with AI, you could actually do a lot of things you can't do with a normal video the first thing we did was uh this back in 2018 we worked with david beckham one of the world's most famous soccer players and we recorded this video of him where he's of course speaking in english and then we used the ai technology to actually replace that speech with him speaking nine different languages um reanimating the facial expressions and his mouth,

and it looks like he's actually, you know, talking these different languages. And I mean, I'll let people be the judge here, but I think done in a way where it's basically impossible to tell that it's not a real video that you're watching, right? The key thing that I find

very interesting about this is that, you know, visual effects is not a new thing. Hollywood has been able to create digital humans for a very long time. What AI really changes is, first of all, it's a lot easier to use, as evident on our website today. But the second thing is that, for some reason, AI just sort of closes that uncanny valley gap to some extent.

AI can just produce these photorealistic videos in a much more believable manner than even Hollywood can do. And that's what we show with the Bearcam campaign. It's what we show with the Lionel Messi campaign, where for those who haven't tried it out, you can go to a website, you put in your name and your friend's name. And it will give you this sort of selfie video of Messi inviting your friend to watch the soccer game with you. Of course, that would be completely impossible. the soccer game with you. Of course, that would be completely impossible.

Messi was not going to sit down and create 6 million videos, but using the power of AI, we actually enabled anyone to create these personalized messages, which is super cool. I think a peek into the future of what we'll be able to do with these technologies. It is still really early, and I think the next couple of years will be absolutely mind-blowing.

Absolutely. You know what else is mind-blowing, Victor? Those that are watching, listening to our interview together, because you're watching, listening to me, David Kogan, host of the Alliances Hero Show. Make sure that you go to alliances.com. That's E-L-I-A-N-C-E-S.com,

the only place where entrepreneurs align. We have with us Victor Repropelli, co-founder and CEO of Synthesia. You can reach him at S-Y-N-T-H-E-S-I-A dot I-O. Victor, how in the world did you come up with this idea, though? That's a great question. So my background is I'm Danish. I grew up in Copenhagen, you know, other side of the world, from the U.S.

And I grew up as a huge up in Copenhagen, you know, other side of the world from the US. And I grew up as a huge sci-fi nerd, loved computers, figured out in my late teens that I could actually turn this into a hobby or sort of a job. Built e-commerce stores for local businesses and sort of that started my entrepreneurial career, I guess. Moved into the Danish startup ecosystem, was part of building some cool companies. A lot of this was very like business software-ish type of products, which is also very fun. But after doing that for some years, I just felt like I really wanted to start my own company, but I wanted to combine it with this very deep interest that I have in science fiction and the future and all the amazing things we've got to be able to do with technology in the future. So to make a very long story short, I actually moved from Copenhagen to London, because London is in Europe, the center of

AI research, and there's so many exciting things going on. And after sort of being in London for nine months, working as a consultant on different projects, a lot about virtual reality and those type of technologies, I met this guy called Matthias Niesner. He was a professor at Stanford at the time. And he was actually like the first person or the first researcher in the world who proved this sort of deepfake. Obviously, it comes with a slightly negative connotation. But he is, you could say, the first one to actually make that technology work. It's back in 2016. So it's a while back.

He got invited to the Jimmy Kimmel show. We were showcasing this tech. It was really cool. It was very fun. And when I saw the technology for the first time, it was very basic. And obviously, you know, it didn't really work that well. But I was just blown away. And I felt like I'd seen magic in front of me.

And I'm, as a person, someone who loves creative things. I love films. I love movies. I produce music in my free time. And I just saw that as, like, this is going to change the world. Like, this is going to be a game changer in how we produce media content in the future. At that point, it was hard to, you know, map out exactly what that's going to look like and how it would look today.

But I felt like there was something powerful enough in it that I wanted to dedicate a part of my life to try and bring this technology to life, give it to people, help them be more creative. So after a very long time, I managed to convince this professor to build the company with me. I was 25 at the time, and I didn't really have a track record or a CV that made me the perfect person to found this company. I got another professor involved, Lourdes Agapito, and I got my friend at the time, Steph Gerald, involved as well. And we all just gathered around this idea of we want to make it easy for people to create video content. But the way we think about that is not by, we don't want to make smaller, more affordable cameras.

We don't want to just create a slightly better video editing app that runs on your phone or something like that. That's how most people thought about making video easier five years ago. We wanna build technology that eventually can replace the physical recording process completely, right? That just means no cameras, no microphone, no studios and all of that stuff. And that was five years ago, certainly a crazy idea.

It maybe still is a crazy idea today, but I think it's much easier to sort of track the path forward of how we're gonna eventually be able to create Hollywoodllywood films on laptops and i always when people say you know that sounds crazy i tell them that it may sound crazy but really this is just a continuation of what we've seen in media production over the last many many years right you take something like photoshop that has enabled people to create photos just from their computer they don't need to go out and shoot a real picture anymore. If you take music production, we can synthesize instruments and effects.

I can command a 65-piece orchestra from my bedroom because we can synthesize sounds, effects, and things like that. With typing and text, we type on keyboards, not on pen and paper anymore. And to us, speech and video just felt like the next natural thing. And that AI would be the thing that could enable this to to happen so we kind of went out with that idea got turned down by i don't know how many investors i think 70 80 something like that nobody believed in it until we sent a cold email to mark cuban who um responded after five minutes that he was very aware of all this tech that we were developing and he loved it and he was very interested in learning more. Then we did 14 hours of emails with Mark Cuban and after those 14 hours at 4 a.m. in the morning here in London, he said, you know what, I'll do the million dollars given that everything checks

out and Mark Cuban became the first investor and that really was the start of Synthesia. That's incredible. We've had other sharks on our show that I've interviewed but that's phenomenal story. And you didn't give up on that. So now things are rocking and rolling. You know, you continue to build, but you still have people that are concerned, concerned about what you're doing. And there will always be critics, but there's a number of critics that are saying, what about the deep fakes that are out there, the fraud, people doing the fake news, political scandals, and the list just goes on, that will want to use your system to be able to create that because it's

becoming more and more realistic. What do you have to say to them about it? I think it's a very valid concern right i think it would be deeply naive to think that bad actors would not pick up this type of technology bad actors have picked up every technology that has ever came out they use smartphones they use cars they use telephones they use email and they'll of course also use deep fake technology to do bad things right um i think it's a very you know it's a very deep topic but i'll think you know the short answer for me is there's two things i care about there's one which is

ensuring that our platform at synthesia is not used by bad actors that one is is pretty easy actually you know we have to submit an avatar for creation you need to give explicit consent you have to go through you know a whole kind process. You can't just upload images or videos of anyone, right? That's sort of a KYC style process. We also do very heavy content moderation, which means that we have to take an active stance on what kind of content you can create with Synthesia. Obviously things like hate speech, we don't allow,

but we also take a stance on what we call the gray zone categories. So even things like gambling, for example, you know, unless you have your own avatar, we don't allow that. So we do kind of build systems that ensure that people can't use our tech for that. So that part of it is fairly easy, actually. But the other part, which I think is more interesting,

is that there's, of course, going to be open source technology that's already out there that anyone can take and use. There's going to be other companies. Maybe they'll have less kind of ethical standards than we do so what can we as a company do to help the world prepare for this technology right and i think there's three things that are very important here the first one is education right you know people need to understand that in the future it will be easier for people to fake speech and video content in a highly believable manner and there's many ways you can educate people there's formal education i think

that's great but actually what i think and i might be slightly biased there is that the best way for people to understand these technologies right is to interact with them to use them i'm very proud of how many people create free videos with synthesia every day that's one more person who knows that this is now a possibility and will be a little bit more skeptical when they see things online. And there's, of course, many other companies who build deep fake tech in various other formats, B2C and things like that, right? So I think that's very important. The second part of it is technology. So this could be a very deep topic. What I'm very, very, very interested in, and we're working on this with Adobe, is what's called media provenance.

So this is the idea that every time you create a piece of content onto these, with your camera, it's registered in a global database somewhere. Who created it? When was it created? How is it created? All this sort of metadata about the content

that someone who's consuming that content on, say, YouTube can then actually see. That's going to add this layer of metadata about the content that someone who's consuming that content on say YouTube can then actually see. That's going to add this layer of transparency and it's going to make it much easier for people to understand where content came from and how it was created. And then the last part of it is of course regulatory. That's something people are talking about today.

Should we regulate this? How should we regulate it? I actually just came back last week from spending a lot of time in Brussels, in Europe with the European Parliament. I spending a lot of time in Brussels, in Europe with the European Parliament. I spent a lot of time with the UK government. I'm really just trying to figure out, like, how do we actually safeguard these technologies, right? And I think, you know, what it comes back to is that the crimes you can do with this technology is not new.

People have been using email to fraud people for a long time. People use phones to defraud people. Like, it's the same thing. The technology changes, but the crimes are the same, right? And we need to make sure that our justice system is prepared for what happens if you create non-consensual adult material

of someone, for example. Clearly that should be a crime that we punish really harshly, right? So I think those are the three vectors that we think about. You know, that said, I think every technology that's powerful is always met with a lot of fear before it just becomes integrated into society. That has been true for everything from the internet to cars to, I mean, any technology

you can think of, right? Like people initially fear it and then it becomes a part of the world. And I'm a technological optimist and I truly do believe that these technologies will be by far net, net, net positive for the world. It'll free creativity and allow people to communicate much better than they could before. Are the avatars that are on your site that people can select, are those real people?

Yeah, so today those are real people. They're real actors and they basically get paid for every video that you generate with them. Got it. Interesting. Interesting. Well, what you've created is absolutely phenomenal. We've got time for really one more question here, and this is a very important one too,

is what advice do you have for parents to educate and help maybe their, their children or children are out there that want to, that have a dream of right, creating, building their own company like you have and having done and created something that's not done before, right? This is a whole new field too. So, so that's the one part of the question. Those are things they can do. And the second part is, is where do you see this, you know, into the future five to 10 years of what you've created? Yeah. So I think the most important thing you can teach your kids,

I mean, there's many things, but if I had to pick a few, I would say first one is resilience and actively seeking rejection. That is, I think, a superpower. If you're okay being rejected and you're okay being challenged then you could do a lot and i think i don't know if it's a generational thing but i feel like that's something that is lacking maybe a little bit you know my generation and the generations that come after us and i'm 31 that's very important then i think there's many ways to be a great entrepreneur um you know the style of entrepreneurship that i'm doing here that we're doing is like okay let's do something really crazy and let's see if it works if that's the kind of thing that you're interested in curiosity is the key here and i think i think it's something

that's probably hard to force but stimulating kids curiosities by explaining things when they ask stupid questions and don't shut them down keep explaining things to them i think that's that that will naturally mean that you know they will ask more questions they'll be even more interested and it's okay to ask questions and you know really understand something in depth i would say that's one and then i think the other one which is maybe you know when you're a little bit older but that's something i think about a lot is that being very aware of actively taking a stance that's uh that that's the opposite of what the mainstream thinks this doesn't mean you need to be a conspiracy theorist but always questioning you know this social norm for example or this viewpoint we're seeing in the media is that actually correct um and think of those things

from first principles I think that's a good way to stimulate your own curiosity it's just like always asking questions of like is this actually true or real or is there like what's the different sides of this argument I always try and when people put a position or a narrative in front of me to make sure that I can explain sort of both sides of that narrative because then I have more or less full information I can make up my own mind and I think if you force yourself to always question things your mind will wander and hopefully you'll come up with some perspectives that um that that you maybe want if you just sort of blindly follow what other people tell you excellent well that's right question things well i gotta tell you this is phenomenal victor you make video easy for anyone utilizing ai that's a hero reach out to

victor rip about ripa belly co-founder and ceo synthasia you can reach him at synthasia.io this has been david cogan with the Hero Show. Make sure that you go to alliances.com to check out past interviews. That's E-L-I-A-N-C-S.com, the only place where entrepreneurs align.