Presenting: Possible x Artful Intelligence

0:00Starting a business comes with its share of ups and downs, which is why staying true to your vision is essential. a non-negotiable for Romeo and Milka Bregali, Capital One business customers and co-owners of Ra's plant-based restaurant in New York. Romeo and Milka took a leap of faith when starting their own restaurant, gutting an empty space and building it from the ground up. Every pipe, every wall, every detail. But building from scratch came with a heavy financial burden, which is when they turned to their Capital One business card. With the flexibility of the card's no preset spending limit, they were able to spend more and earn more rewards while bringing their vision to life.

0:36Today, Raz's success is proof that with passion and the right support, it's possible to make your dreams a reality. Learn more at CapitalOne.com slash business cards.

0:55Hi, it's your host, Rana here. As we all enjoy the holiday season, the team here at Pioneers of AI wanted to bring you something special. the gift of inspiration. On the show, we think a lot about the impact of new technologies on creative work. And our friends over at the podcast Possible are doing the same with their new three-part mini-series called Artful Intelligence, hosted by Reid Hoffman and Masters of Scale producer alum Chris McLeod. I'm excited to share the first episode here with you all today. So in this episode, you'll hear from an independent filmmaker looking to redefine video art by using AI tools like DALI to compress time, space and reality into surreal, mesmerizing visuals.

1:44Take a listen. And if you're intrigued, explore the full Artful Intelligence miniseries on the possible feed anywhere you listen to podcasts. I'll be back next week with more conversations from the AI frontier. Until then, happy holidays.

2:29So, it all started with a conversation. I was talking with a close friend who is a writer, a musician, and producer. One of the most talented people I know. I was telling him about some of the ways that I've been using AI in my creative work. And he was not only skeptical, but terrified. Not about the Terminator stuff, but about how AI will affect creativity and artistic expression. My friend is not alone. Many artists have a complicated relationship with AI. Some of the greats, like Steven Spielberg, Drake, and Tyler Perry, have raised concerns about the impact the AI is having on the arts. I think the fundamental question many of them are asking is, Are we moving toward a world where humans aren't needed to create anything?

3:23And is that the world that we want to live in? So the last few months, I've been trying to get to the bottom of this question by talking to groundbreaking artists who are fully in on using AI tools and capturing their stories. I found that their often mind-bending work would not be possible without AI. But it also wouldn't be possible without humans. So as we explore their stories in this three-part series, we pose the question, what if? What if the future is still being written? What if we could work toward a world where creativity is not replaced by AI, but amplified by it? So Reid, this is where you come in.

4:09You're an AI creator yourself, but you're also someone who spends a lot of time thinking about how to plan for futures that break humanity's way. So I'd love to play you these stories. From time to time, I want to zoom out to hear your perspective and what we can learn from these creators to ultimately work toward a future where AI is amplifying humans, not replacing them. So, Reid, does that sound okay? Amazingly and wonderfully so. Let's do it. Awesome. Awesome. Welcome to Artful Intelligence. In each episode, we'll share the story of an artist using AI to create what's never been possible before.

4:56Episode 1, Creating the Impossible.

5:05before we begin i have a secret and you can't tell anyone so i built a vocal booth in my apartment a while back and i haven't told anybody but my vocal booth is actually a time machine it's sort of like the police box thing in doctor who except mine is smaller on the inside And honestly, it's pretty cramped in here. To begin our story, we need to go back. Long before the invention of AI. Before the smartphone. And we are calling it iPhone. Before the internet. You've got mail. Or the personal computer. I'm Bill Gates. You're going to see the future. Future. Windows. We're headed to a time before VHS players.

5:58or even the television.

6:04We'll start all the way back in the late 1800s in Paris, France.

6:16At this time, moving pictures were in their infancy. So magicians were the stars of the day. Magicians like Georges Méliès. Melies was fascinated with filmmaking and wanted to incorporate it into his act, so he made his own camera and began capturing footage. One day, Melies was shooting on the streets of Paris when something happened that would change cinema forever. His camera jammed. He fiddled with a hand crank, and it started up again. When he developed the footage and played it back, he saw something extraordinary.

7:03The film showed a horse-drawn omnibus carrying workers up the street. Suddenly, the omnibus had been transformed into a hearse. He realized that when he was fiddling with the hand crank, the camera stopped shooting. Then when he fixed the problem and the camera started up again, the omnibus was gone and the hearse had entered the frame. Melies had accidentally created the first ever special effect.

7:37This opened up a whole new world of what was possible in filmmaking. Melies would go on to invent the split-screen effect, the stop-motion effect, dissolves and fades, and even color film by hand-painting every single frame. These special effects enabled him to create films that transported his audience on fantastical adventures to haunted castles, the center of the earth, down into hell and up to the moon. There's one particular effect that Milliers invented, which laid the groundwork for our story today. The time lapse. Nearly a century later. Wait, I need to set a course for the mid-90s. Okay, let's go.

8:27In 1995, another French filmmaker would push the technique even further by asking, what if you could compress both time and space? Guy Rowland from Montreal created an effect that would become known as the hyperlapse in his experimental short film called Pacer. No dialogue, no actors, just footage of the streets of Montreal. Each scene only lasts a few seconds. You see time-lapses of repeated objects like windows on a building or gravestones. Unlike a traditional time-lapse, there's constant movement. In each frame, where one window or gravestone used to be, there's now another. The result is disorienting and surreal, but almost impossible to stop watching.

9:22So that's a hyperlapse. But what if you could compress time, space, and reality? That would require another groundbreaking filmmaker. They might as well be French too. Okay, so my name is Malik Lombion. Meet Malik Lombion. Malik is a video content creator from Montreal. He and a friend invented a style of filmmaking that would have never been possible without AI. They call it... This work, I call it Hyperlapse XAI. Hyperlapse XAI. What is Hyperlapse XAI, you ask? ChatGPT, describe a Hyperlapse XAI video. A Hyperlapse X AI video is a type of video that combines the traditional hyperlapse technique with AI-generated imagery.

10:11You might take a hyperlapse of a cityscape and then use Dal-E to transform each frame into a different artistic style, or to replace elements like the sky, buildings, or lighting with imaginative alternatives. These videos are mesmerizing. Images change before your eyes as the camera pans around an object. Malik made one particular Hyperlapse XAI video with a 360 camera that went viral in 2022. When other video creators from all over the internet saw what was possible to make with AI, the style exploded. And I had to hear the story of how he made it. Hey, can you hear me? Hi, Malik. How are you?

10:53I'm doing great. I'm doing great. How are you? Through his story, we'll ask the question, what if? What if you, yes you listener, could create something with the help of AI that's never been possible before? You might think it's as easy as typing in the right prompt. But Malik told me there's so much more to it than that. Most people, when they see AI, they just see the really surface level work that people do, which is you go on the software, you type a prompt, you have a beautiful image. Yeah, but this is the most basic way of using it. But I've seen some incredible artists create some artwork that was just not possible before.

11:32And Malik is absolutely one of these artists. So strap in. As we traverse the timeline of Malik's story, we'll discover ways to use AI to create what's never been possible before. Part one, to create the impossible with AI. The right humans are required.

12:03Setting a trajectory for 2022.

12:09A new tool was released to a handful of waitlisters, including Malik Lambian, that would change the landscape of creativity forever. Dali 2. When Dali 2 came out and I saw what it was capable of doing, I was like, wow, you can literally create some beautiful artwork, some beautiful images with just a sentence, just some text. So I was like, yo, if this is what we can do right now, this is the worst it's ever going to be. So can you imagine what is going to be in like two years, three years? I made it a mission to stay updated in what AI can do. Do you remember what the first thing you typed in was by any chance?

12:50Do you remember like the first prompt you put in? It was probably something basic, like a picture of a cat, like the classic. Yeah, totally. I think that was it, yeah. But Malik instinctively knew there was so much more to be explored than cat pics. If he wanted to get the most out of this brand new tool, he would need an expert. Yeah, his name is Manuel Sensili. He is a creator from Guadeloupe, and we both live in Montreal. We knew that we wanted to collaborate together on something. We just didn't know what. I knew that he was kind of an expert in AI. So I was like, okay, let's put our expertise together.

13:33And so we just decided to set one day that I would come over to his place and try and fail and make and break some videos and be creative with AI. So Malik captured some stunning video of the city of Montreal and went to Manuel's place to see what they could come up with. He took the footage he had captured and created a hyperlapse. We basically spent like 24 hours, no sleep, scene by scene. I'm editing a scene, the rotation, the zoom in, the transition. And I'm sure that this scene is pretty much picture locked. Then he and Manuel brought in a new collaborator, this time non-human. They started experimenting with Dali, but Dali was for images only.

14:18To work with video, they would have to modify each and every individual frame. So at 24 frames per second, that's 24 images for every second of video that he wants to bring into DALI. I take these images and I upload them one by one into DALI. In DALI 2, one of the incredible features was that you could select a small area of an image that you uploaded, then regenerate only that portion of the image. I select my area that I want to modify, and I type the prompt, and I export that. Sometimes the immediate result that you get is not that good, and I had to really iterate. Sometimes working on one second of video could take me like two hours.

15:09So yeah, that was a long, tedious process. Footage of objects like church steeples or the banner on an ice cream shop transformed into new surreal images in every frame as the camera panned around them. Through Malik and Manuel's tedious experimentation, they created a brand new style of video editing, Hyperlapse X AI. It became a tool set that Malik would utilize to show what was possible when AI tools amplified human creativity.

15:43But first, Reid, I've heard you talk about how humans should always be a part of the equation with AI. In the case of creating hyperlapse XAI videos with Dali, it required two very specific humans, a video expert and an AI expert. But let's zoom out for a moment. On your podcast, Possible, you have guests from many various fields discussing how to impact the future. Leaders in fields like science, political strategists, technologists. I'm always amazed at the caliber of people you have on the show, all working to make the future a little better. When you're looking for collaborators to shape AI advancement, whether a company founder or a podcast guest, what kind of criteria do you consider so that the right humans are brought into the equation?

16:36Well, obviously, you know, kind of there isn't like a one size fits all because it's, you know, different human context, different goals, different audiences, different creation moments. Like, you know, if you're talking about the deep tech of AI, then it has to be this kind of technologist or this kind of humanist or both. If it's the question about how we experience things and how we think about societies, then it's a different person. And it's not like, you know, I always bring in Sue or I always bring in Fred. You know, it's kind of like it depends on the circumstance. You know, and by the way, it doesn't have to be like, I think it could be perfectly entertaining to create a, in some circumstances, a humanist discourse that's entirely like chatbots talking to each other.

17:23I think it's totally doable. So, it doesn't like, you don't have to say like, well, there always has to be a human here. It just you have to have kind of human beings and humanity and humanism in mind in what you're doing. One of the things I did is I did this podcast with with Ali Miller, where we had a set of custom GPTs that were listening to our discussion. And then we were kind of calling them out for commentary and questions and different things. And so you had a whole panoply of agents, you know, participating. And that was great. And Ali and I were the humans in the loop in this case. And I think the thing to remember pretty centrally in all of this is that, you know, you want people who are imagining how to get to really good futures with human beings, with technology, with AI.

18:17But it doesn't necessarily mean you need promoters only or advocates only. It's actually very useful to have critics, warriors, you know, people who are uncertain in the dialogue, too, because, look, we are it's a navigation. Navigation isn't just, hey, this is the path to getting there. But also, here's how you avoid the potholes and landmines and cliffs and other things that could be part of the journey. So it's a range of folks. You know, I'll just go to kind of a personal conversation that I've been having recently with Yuval Harari, who's a spectacularly intelligent and humanist thinker, you know, gave him my new book, Super Agency, that'll be coming out in January.

19:05And we were talking about it. One of the questions is, can human society deal with the speed that we're getting to? And it's unique. We haven't had this much speed within technological revolution in human history. On the other hand, you know, each technological revolution has been much faster than the previous one, whether it's the printing press or the car or electricity or the Internet or the mobile phone, you know, etc. You know, he's more hopeful than I am that you could persuade people to slow down. And he's basically like, look, I hope you're right, but I think you're wrong for the following reasons.

19:42You know, then I'm the critic because I'm like, well, no, I don't think you can do that. I don't think it works. You know, maybe it's a good thing if it works. And let's talk about like your ideas to make it happen. And I'll give you my two cents on those ideas. And that dialogue gets you a much more productive kind of like, okay, here's how we're steering into the future. And so we ended up, he ended up blurring the book. So that's an example of a conversation that wasn't on a microphone, wasn't at the possible podcast or anything else, but was a conversation that was an example of a very good dialogue between humanist optimist in AI and a humanist, skeptic's probably not the right word, but deep concern with AI.

20:36Part two. To create the impossible with AI, plan on failure in a very public way. Firing up time travel circuits, moving forward just a few months. As a successful content creator on Instagram, Malik was constantly receiving DMs. I was receiving a lot of DMs, like more DMs that I could handle. And some of them were just spam. One day, reluctantly scrolling through his inbox. I received a message from a woman, Olivia. She sent me a DM saying that she works for Insta360 and she would like to have my email. Insta360 is a company that makes cameras that shoot in all 360 degrees by placing two cameras on either side of the device.

21:30Malik was intrigued, but he figured it was too good to be true. At first, I thought it was fake, but I was like, okay, let me do my due diligence a little bit. So I went through her profile and I Googled her and I found out that she actually worked for Insta360. So I gave her my email. She actually sent me an email saying that they saw my work, they loved it, and we give you some gear and you make a video with their next camera. It was a camera that was not even out yet. They basically sent me the camera and I was like, wow, they just sent you some gear for free? Like what? Are you crazy? Awesome.

22:10Malik unboxed the camera, looked down at the small silver device with a lens on each side, and realized he had no idea what he was doing. I never shot any 360 footage before. I knew I wasn't going to get it right. It's going to be crap. I just knew it. Malik had said yes to this incredible opportunity to make content for a brand that he loved, but having zero experience working in 360 video. It felt like a recipe for failure. Luckily, failure is something that Malik plans on. Let's jump back a few years to the moment when Malik learned that failure can actually be a resource. He had just quit his stable job to try for a career in video.

23:21Once the bills piled up, he knew he had to find a better way. He saw other content creators making a living on Instagram. But he only had about a thousand followers. If he could grow his social following, it could lead to better opportunities. Time for some public experimentation. One day I decided, hey, I'm just going to post every single day for like a month and try and fail and learn from my mistakes. A lot of different types of content. I tried some fitness, some funny videos, some trends, some kind of music videos. And each of the ideas pretty much flopped. When I scrolled back through Malik's feed to see these horrific failures, I was honestly shocked.

24:07There is some incredible work there, with almost no engagement. But this didn't deter Malik. He kept trying different types of videos to see if anything stuck. After probably like three weeks of getting almost no views and no likes, I started to see a pattern. I was still trying and failing, but I was like, hey, maybe I should do more of this because last time I tried this, I actually had more engagement on my post. Malik realized that failure, even in such a public way, could be a resource, perhaps even essential. I saw that the hyperlapses were the type of videos that were working. Malik had discovered his niche.

24:54At first it was subtle, but his hyperlapse videos were beginning to move the needle. Until finally, the day came when it all paid off, one of his hyperlapse videos went viral. I had millions of views for the very first time. Malik would begin to utilize failure as an essential part of his creative process. So back in the present, Malik is staring down at his 360 camera with no idea what he's doing. So he planned on failure. So I had two days in mind. One day to just try and fail with a camera and one day to practice with what I learned the day before. Day one, experimentation. The camera came with a three-meter selfie stick attachment.

25:48And for us Americans, I googled it, and three meters is about 10 feet tall. The longer it was, the more range you had to actually film everything around you. How do you capture 360 footage on a 10-foot pole? How about a backpack? The first thing I tried was put it in between my backpack and my back. When you walk, you actually take some steps, so you go up and down. I tried to be as steady as possible, but this was not the most optimal way because it was really shaking. Okay. How about a bicycle? Then I tried on my backpack riding a bicycle and that kind of worked. It wasn't the most stable way.

26:29Okay, a little better. Any other modes of transportation? Then I tried with a car. In a car is the most optimal way. So I knew after a few trials and errors what I had to do. I just like love the image of somebody's on the street and you got this thing sticking out of your backpack, riding your bicycle. Like what would somebody have seen if they were like on the sidewalk as you go by? What would they have witnessed? They would have witnessed something weird, obviously. Like I definitely got a few looks. And sometimes even some people came to me and asked me what I was doing. So I was explaining to them.

27:08And some people knew what I was doing. And I told them, yeah, this is the next one. Yeah, I'm privileged enough to have the next one. You don't know about that one, but I do, you know. So that was nice.