How did David Beckham speak 9 languages and demonstrate that AI can improve video performance by ...

Hello. Today I'd like to talk about the opportunity that AI, artificial intelligence, gives us to create real-time videos speaking in up to 34 languages and to use images of people that some people are branding as deepfakes, but other companies are using to great effect to communicate around the world At low cost, but with high quality, including famous stars like David Beckham. Now I got to thinking about this because I did a shoot for myself with Marcus Ahmed here in Bristol, and when we got there, he had an assistant. There was this moment where everyone was worried whether they should be wearing their masks or not, and of course, one of the issues was that if we had been just he and I, we would perhaps get over that. But if we wanted to an event of some kind or a shoot with multiple models, then we wouldn't be able to do that because with social distancing, it would have just been impossible to all be in the studio together. So there's an issue that's been created by COVID, which I think is being addressed by AI. And I think that because I've spent some time looking to try and solve some problems for myself, which is around What if I want to show the face of some products and some clients, but I can't actually send a film crew to see those people? Now, I'm a big fan of video, as you know, and I'm a fan of having people, but we've all used up our archive photography, and we're all getting out slowly but surely. But with social distancing taking place, it's becoming very difficult to have any group shots, Or any, if you like, press briefings. So in my research, I came up amongst how WPP, the multinational agency originally founded by Sir Martin Sorrell, the the solution that they've been using is a London company called Synthesia. Sounds a little bit like anesthesia, but Synthesia. Now, what WPP have done is they've created some corporate training videos, and they're doing that for their tens of thousands of employees worldwide. Now, what's happening there is that instead of having presenters speaking in individual languages, they've gone to Synthesia and they've created an AI-based presenter. Now, what's interesting about this is that it's not just an AI in terms of a CGI, not a computer graphics, just pure animation. What it is is the use of individuals, real people's faces, and those pictures being uploaded to the Synthesia platform and those then being animated. So I couldn't help but have a look myself, and what one can do is you upload a picture of the person that you'd like to be the spokesperson, and then you upload the text, much like you would subtitles, and the platform then will make the face into an AI avatar. It's amazing the quality of the avatar, both in terms of the facial expressions and also in the lip synchronization. It's not 100% accurate, but it's very close. And considering that most people will be listening to the audio and are less attentive to the lip syncing than they are to the quality of the audio, it's creating an amazing opportunity to have one or many people in effect made into avatars and then to speak. Now that in itself may not sound so exciting, other than what one can then do on this program is you can actually have the avatar speak in up to 34 languages. So you can actually have one person, say the spokesperson of a company, or if it's going to be, for example, in training for different trainers, could be the same person, the same script, but the lip movement and the facial expressions change according to the language that's being spoken. So this seems to be pretty amazing. In the case of WPP, they just had English, Spanish, and Mandarin. Now the WPP Chief Technology Officer, he was covered in Wired just recently, said— his name's Stefan Pretorius— he said, with Synthesia, we have avatars that are diverse and speak your name and your agency and in your language. And the whole package cost $100,000. So this is for an internal training program, but you can only imagine they were able to send out 20 modules of about 5 minutes each To 50,000 employees for what in effect was less than the price of a movie ticket, less than $2 per head. Now, having seen the videos on the website, I can tell you that if you were to watch those, you wouldn't feel as though you had some motionless avatar. It really looks as though someone is, is really speaking to you. Now, of course, it's partly about the technology, but it's also just about the practicalities of it. That now if we are going to try and create videos in the way that I would create before, for example, a video of a CFO, I can't take the whole film crew with me because due to social distancing, people can't necessarily travel. They may not all be allowed to be in the same room, and then that spokesperson may need to wear a mask. So there's a practical issue where the technology is really coming to the fore because of the limitations on our freedom to interact with each other caused by, by the COVID Now, as we know, there are companies, for example, in China, where you are able to try on lipstick using an avatar. So an avatar, a And then you paint the lipstick on the avatar, and you choose the different kinds of skin color, eye color, hair color, and so on. So really, it's becoming this montage where we can go in and create real-looking people or people that look like they're real. And with Synthesia, you're able to animate that. Now, Victor Riparabelli. Who's the CEO and co-founder? This is out of UCL, Stanford, and Cambridge. The the researchers that founded the business in 2017. Victor says, "We're saying let's remove the camera from the equation." Now I have a lot of affection and a lot of friends who are photographers, and my intention is not to try and remove the camera people from this world. But there are many occasions now where communications programs can't take place simply because people with cameras are not allowed to be next to the talent. This is an opportunity then for us to continue to communicate and to circumnavigate the limitations being imposed on us all. Now, last year, David Beckham appeared on a public service announcement talking about malaria, and David Beckham, I think, is not renowned for his linguistic ability, but he was in the in the movie speaking nine languages, including Hindi, Arabic, and Kenya Wanda. So the language of the people in Rwanda. So it made David Beckham, and I've seen the video. Be present for millions of people who otherwise would have been forced to maybe watch a video with subtitles or a very distracting voiceover. Now, the narrated video presentations can also be from slide decks, and I thought this is very interesting because I've been experimenting with loom.com where you can present. The Loom will film a copy of your presentation or your screen and also the camera attached to your computer or your external camera, so that basically you can speak and present and narrate to, for example, a PowerPoint or a screen. Now with Synthesia, you can do the same thing. So what you can actually do is have your avatar embedded into, for example, an infographic or a a slide presentation that you're giving, say an infomercial. Great for training, for example. So, um, a lady from Malaria No More said that the Synthesia exceeded their expectations because they were able to harness the power of the immersive AI video to break new ground and reach millions more people globally because of the ability to do it linguistically. Now Interestingly, they're also working with Reuters to create an automated presenter-led video report. And so in this case, it's using match data to auto-generate a news bulletin, which auto-creates the text for the avatar, for the AI-created avatar. In other words, from end to end, the— that once the news has come in from the pitch, then It's all AI. Now, why bother? Because viewers record and retain 95% of a video message compared to 10% when reading text. This is the result of a survey done. I'll put a link in the notes. But so video, as we know, which is why it's a safety announcement on the plane, is a video, not a book. People remember 95% of a video's message. Now, what's interesting is that Synthesia did a test with a platform called RealEyes, which uses computer vision to track viewers' attention when they're watching things and their emotions when they're watching a video. And this for me was the interesting part because I do a lot of work across the world, you know, in different languages. With my offices of EastWest Public Relations in Singapore and China, as well as India and, and the UK, that people connect with the AI-dubbed content more in their own language than they did if it was dubbed. In other words, using an AI, an avatar speaking in a local language, got a, a better recall rate than a real person with a dubbed-over voice. Okay, and the score according to Synthesia was 175% increase. In other words, it nearly doubles the impact. So they also went on to say that the difference was even more amongst young and old, that the young people had absolutely no illusions at all about what they were watching. And they bought and were convinced almost entirely by the avatars, whereas the old people, people in my generation who watched videos and then started— or TV and then videos and then subtitled— that have been used to having one person speaking in a different voice being out of sync were a little bit more cynical. But what Synthesia's Uh, research shows is that video with this AI-dubbed content can increase engagement up to 1,200%, which is pretty massive. And I guess if you think about David Beckham speaking in the language of the people of Rwanda, that's a pretty, pretty astonishing, uh, event. And no wonder people found that memorable. Now, not everyone wants to make animated movies, uh, all the time. So I was also looking, uh, because I'd done the film shoot and a photo shoot, and I was thinking about people making clothes and so on. I found there's a company called Rosebud, rosebud.ai. Now, Rosebud AI makes the kind of glossy images that are used, for example, in e-commerce, or for example, on Shopify. So what's interesting here is that you could then use Rosebud AI, and I've again had a look at this platform, and basically you can pick from either uploading your own photograph if you've got a particular character or person you'd like to use as the base model, or you can use one of their collections of some 25,000 modeling photos of people that never existed. It's pretty amazing if you think 25,000 unique faces. Um, and then what you can do on the rosebud.ai is that you can swap different elements. And what's amazing about this is that the, the changes are seamless. So it used to be that the joke on those police photo fit photographs was they would sort of stick a piece of paper of the eyes and a piece of paper of the mouth and of the chin and the ears and the nose and so on. And make almost like a child's drawing of the different elements. But with the rosebud.ai online platform, one can change, for example, the hair, and it, and it's changing the light on the forehead. If you change the eyes, it's changing the position of the nose, and so on. They've recently launched a service that enables you to put clothes onto the mannequin. So one opportunity then is that if you are, for example, a young designer and you've got either the, the CGI, maybe you design the clothes yourself online on your computer, you can then use their mannequin service to literally dress the virtual mannequin with your virtual dresses. Now Imagine the impact for product launches, but also for production. I mean, you could then only make product to order. You really wouldn't have any stock. You'd just be designing, not having any real models come into a studio to try things on. And if you got people to buy them, then you could make them. I mean, that's an amazing reduction in manufacturing lead times. And also from an environmental point of view means we're not making things that really people don't need. I have to say, these are amazing looking mannequins, and we've used one of the rosebud.ai images for our East West Public Relations virtual assistant that we have. We call her Val, and you, to be honest, you couldn't tell the difference. You know, they have a self-service app with, um, no photo shoot needed, where you can just drag and drop their images, and that's $19.95 a month. You can only imagine how much money you'd save from photo shoots. Virtual fashion models with full faces and full body poses, you can have a campaign at $250. Now, you know, that's less probably than a cab fare and lunch in New York, or in Singapore you wouldn't, you wouldn't get a crew out for that sort of money. They also have a what they call a white, a white glove service for $1,000 for a campaign, and it only takes a week to do a turnaround where you can also ask them to help you with your demographic targeting. And I'll put the YouTube link in the notes because it's pretty amazing when you look at what's created, because considering most people are watching product launches and PR events and and sort of product displays online on maybe a mobile phone, and maybe with or without broadband, these, these, um, AI videos by Synthesis and also by, um, by Rosebud really probably serve your purpose. Now, Alicia Lee, who's the CEO of Rosebud, says that the company can help small brands with limited resources produce more powerful portfolios of images. And I think that's absolutely true. You can have algorithms now to help you make your portfolio instead of the expense and the delay of models. And there was a company called CAA that signed Lil Miguella for a campaign. And what's amazing about Lil Miguella is that she is an Instagram influencer with 2 million followers, but she's actually an AI-created avatar. I think you get the idea. AI is here and it's very powerful, and it's something that we can all use for good. So with that, I'd like to thank you for listening and to wish you good health, a profitable business, and that you keep on communicating. Online, and maybe with some AI.