B (14:35)

Don't worry. This is an AI generated video about a group of friends on a night out and they make a couple of pretty bad decisions. We tried to depict the story as realistically as possible. However, we regenerated the ending just because we can don't drink and drive. Real life can't be regenerated, generated.

A (15:01)

So here AI actually becomes part of the message. And I think that works well because you kind of put people like beyond the detective mode of trying to spot whether it's AI or not and just get a sort of acceptance and curiosity about the story. And so in some cases this is a very interesting approach. I don't know how long this will last as of course at some point the novelty will die off and maybe people, people will be like, okay, yeah, so what, it's an AI ad, but in any case, interesting for now.

C (15:43)

I made $420 uploading my homework. I paid my car payment with money from my data. Upload your data and get paid. I got paid 320 for foot pictures. I made $800 for all my Streamo videos. Upload your data, make money. AI companies are paying a ton of money for training data. So upload your content and make it rain, baby.

A (16:29)

So the second option, definitely a one I like personally a lot, is basically embracing the medium and even going harder on it. So making it even more uncanny, embracing the weirdness. I think it's an interesting approach and again, it's something we see a lot happening now, especially with some of the work by P.G. aseturo, who's like, leading the way in this style. And I think it's interesting because it recognizes that, yes, AI is flawed, but we can tell interesting stories with it. And I have another one that's also embracing this in a really cool way.

C (17:11)

Well, that was good.

A (17:13)

Short, but good. Well, there goes another satisfied customer. So, as you can see in both of these films, it's clear that, like, things are a bit weird.

D (18:10)

Work has been weird lately. He's going to fly the coupe and fast. Luckily, AI mode and Google search can help him hatch a plan. It's got to be quick and it's got to be. Yeah, that should do it.

A (18:28)

Phew.

D (18:29)

Things were getting dicey there for a moment, but now everything's gravy. Plenty of quick getaway. Just ask Google.

A (18:39)

So in this video, I would say this is like the, the. The last option and. And it's something we've been seeing a lot more recently. The Allen ad that I showed you was a similar technique. Technique where basically you circumvent the uncanny by working with animals, working in different, less realistic styles. This helps you tell a story without people being overly focused on, like, you know, the humans are not real discussions. And I think it's one of those things, especially for brands. It's very interesting because you can basically create a very distinctive art style and really own that. And there is also a sense of like, a bit of like, texture and humanism that goes back into it by showing, like, by using things like stop motion techniques. So I think it's an interesting one. In any case, generative AI shouldn't replace videography or photography in my opinion. It's a new medium. This is a film for Gucci and there is this scene that uses generative AI and it's just something that would be almost like impossible to do with VFX even. And that's why I love it. Like, it's. It's something that feels right and natural in this story and it's the right technique at the right moment. And I think that will be something that we'll see more and more. There's even like artists that are imitating a video, which I think is hilarious. So, yeah, the future is definitely hybrid and I think that's where a lot of the most interesting stuff will happen in that latent Space in using real video, mixing it and so on. There's a couple of techniques I want to show you quickly that will help us. One, there is of course video editing that is becoming a lot smarter where basically you can add or remove elements, you can tweak things in a way that is very interesting and was very much impossible before if you weren't like a heavy VFX artist. So I think this will give a lot of creators new possibilities. There's also motion style transfer, which is again something interesting where you can take existing movement from whatever, from sports, from dance, from anything and you can style transfer it, like putting it in a new world basically. And again, this is interesting for brands because you can remix things in your own aesthetic. And then of course we will also have more creative control. Things like detailed annotations where you can basically write on the screen exactly what motion you want to happen will allow us to tell more intricate stories and be less repetitive, reliant on the model to fill in these things. And also very interesting is it allows us to close the gap between idea and execution. I think that's something that has worked for me a lot when I was trying to pitch new ideas. You know, if you're working for brands, it's sometimes hard to convince them early on because you don't have all the materials yet to showcase what it could look like. And now thanks to Generative AI, you can basically go from sketch to a mock ad in a couple of minutes, which is of course great to, to bring an ID to life, not to, to do as the final product because I said, as I said like one shoting, this is not per se the right way, but at least to bring ideas to life, it works. And look, the real power is still human creativity. This is a recent ad for McDonald's where everything was handcrafted, hand designed, hand animated and filmed. And yes, I mean it's beautifully made, but the reality is that this is like only possible for a few brands in the world to make this type of ad. And I hope we keep doing it, but at the same time for a lot of other brands, smaller brands, but as you've seen also Google and Coca Cola, it's a reality that they won't always be able to make this type of thing. Thanks to Generative AI, you can make your CFO happier too. No, it's not cheap, fast, good. It's cheaper, faster and good if you invest enough time in it. So how do you make your brand future proof? How do we prepare it for the Generative Era. There's three rules I want to discuss with you and give you some examples and then we're done for today. So the first rule, and this is just the first rule of branding and brand building in general, is you have to be unmistakably you. And how can you be unmistakably you is you have to be very iconic, very recognizable. Heinz did a really cool campaign where they asked Dall E, which is a bit outdated by now, but to basically draw ketchup and what, what they got was their product in different versions. But it really shows that if you're iconic enough, even AI will prompt you. And I think that will also be an interesting case where we as brands have to show up in these models visually, verbally, of course. But the real interesting thing about being unmistakably you is that if we look at the data, the most recognizable assets is not your logo. I'm very sorry. It's actually characters, it's jingles and it's sonic devices, which means these are living, talking, singing assets. And I think that's interesting because that's exactly what generative AI unlocks. Designing these characters, bringing them to life is a lot easier with generative AI. It used to be something for only a few that have big budgets. Rule number two is be highly entertaining. It's not enough to have memorable assets. You need people to watch them, to grab the attention, to do cool stuff with them, and only then will you be remembered. Of course, and this is a real problem because, and we see that actually showmanship, the thing, you know, the skill where you bring entertaining creative ads, is dropping in the last decades. And salesmanship, or what we also see in Les Binette and Peter Field's idea of the long and short of it, sales activation has been on the rise. You know, digital performance marketing, conversion oriented rational marketing has been very much on the rise. And we've abandoned showmanship. But if we look at the data, showmanship, entertainment creativity is actually what creates more lasting brand growth. We see it on all the metrics. If you do that properly, you get more bang for your buck. And let's face it, it's a very hard environment out there. The same slice of reach you bought 20 years ago now holds less attention in someone's head. And it cost you 220% more than a decade, which is kind of crazy if you think about it. So of course, attention is the new gold. And the way to grab attention is to really stand out. And I really like this framework which is called like, fuck what cool. Where Basically, the first stage of anything you create must be shock and grab attention. And this is also again, what slop does really well. But then you also have to build a layer of intrigue, pull people in, make them more interested, build curiosity. And finally you have to have a resolution and payoff. And as I said before, the art is to do this within the brand, the universe that you are working for. And if you do that, then I think generative AI really allows you to become a show man or woman again and again. We can see this with new generative AI techniques such as Runways Act 2 or video to video, where basically your performance in front of just a normal webcam can become something really entertaining and interesting. And this gives tools to creators. And the last rule, which is, I think one that is often forgotten in this new space, is we have to be playable. We've seen an evolution of user generated content. Where first we had like, okay, people were blogging, there was fan fiction, there was people writing stuff, then there was video. This is still where we see a lot of things happening, especially on TikTok and Instagram. And then there's 3D things like Roblox, but also Web 3.0 and generative AI, of course, where basically it's no longer just like a 2D or like a streaming version, but it's really almost building worlds where fans can play in. And this remixing or this like playing around with content is happening today. We see it, of course, in Sora, we see it with Harry Potter raves, which is the brands currently are fighting this. And of course this is because legally ip, I get it, you've built long and hard on this IP and then suddenly people are just messing with it. It's a problem, but at the same time it's also a reality. Fans want to play with it, want to engage with it, don't just want to be a passive observer. And so I think this is a bit of an interesting playbook where you see a lot of the traditional strong IP brands like Studio Ghibli, they first built these beautiful art pieces, films that people love, then they build theme parks. Actually Netflix is doing this. Disney obviously is doing this. There's a lot of strong IP brands that do this to further expand their IP and monetize it. But the last part, remixing the IP by fans, giving them the sandbox to play around with, it is not happening from the brands. And what happens then is, for example, what you get with Ghibli is the Ghiblification where basically people are just starting to generate Ghibli images with things like OpenAI ChatGPT, because they love living in that world. But it's, of course, it's not on the terms of Ghibli, it's on the terms of defense. And we'll see more of that blurring of the lines between the original and the derivative, you know, ushering in an age of remix. And I think we'll have to rethink what IP means. And the reality is, I think, for example, for Studio Ghibli, imagine if they had created a sandbox where people would have been able to generate this. I think that would have been a smarter play for them. And we see this with some interesting brands like Doodles, who created a very iconic illustration style and now allows fans to not only create their own avatars, but also create their own art with their own custom proprietary model. So again, they created the sandbox and they invited their fans in to play. And we also think about this at Alan, my coach, Alan Anna calls this the distributed brand era, where basically you have less control over your brand and you let fans play around with it and you embrace that. You embrace that fact. But of course, to do that, you need strong codes and as I mentioned before, strong, living, breathing characters sound and so on. This was an interesting campaign we did where we invited our audience in. They created over 27,000 healthy wishes postcards. Send them to their family and their loved ones. So it was a great example of like a simple campaign where you give people the tools to create remix your own codes. And it worked really well. So basically, generative AI allows you to make your brand sandbox. This is a Marcel mascot dot com. This is actually where Marcel lives, where I create all sorts of apps where you can hug with Marcel, you can draw with Marcel, you can do all sorts of stuff. I would say go spend some of my credits if you're up for it. It's kind of fun. It's a kind of fun environment. And it's exactly what I mean with like, brands being a sandbox and even it's great because I'm the smallest possible brand you can be. It's a personal brand and I can still do this. This is something that five years ago would have required huge budget engineers, lots of different people, and now that's not the case. So I think if you're an interesting brand, an entertainment brand, you will stand out. And if you stand out and people memorize you, then they will want to play in your world. And that's something I think that we see now in gaming where things like this are happening, where the Simpsons and Fortnite do a collaboration and people just want to play as Homer Simpson because they, they remember that it's iconic. So Genai not only makes your brand a sandbox, it also makes you conversational. Actually this is the recent launch of Microsoft's new Micro Assistant, which is not the only one. We'll see. You know, it's an obvious play. When you have LLMs, you build a face for it, but people will get used to interacting with these living, breathing characters. You might say like it's the rebirth of Clippy, but in a more natural way. This is another example of that where you have Tolands. And so that will be very interesting to see how characters will not only just be like, you know, moving around talking, but actually being able to go in conversation with you. And the last point I want to make here is that to be playable you also need to be somewhat immersive and real time or interactive. And this is again very interesting to see where things like real time video are coming, where you can basically not just watch a video but actually engage with it and see it change in real time. This really changes the dynamic of content creation. It makes you part of it and that's, I think something very interesting. And of course we have things like Google's Genie 3 or videos, which is basically allowing you to build any possible world, so called world models. So yeah, a lot of interesting stuff happening. And to recap, I think that's important, whatever you want to do in the future is that you are unmistakably you. So that means you have the right assets to stand out, that you use those assets to be highly entertaining, to really grab the attention in a fun way and then to invite your audience in, to play in that sandbox, engage with them, go in conversation, be gamified in a way. So my final question to you, are we cooked or are we cooking? Thanks for the attention.