Loading summary
Transcript164 lines
- [00:00]
A
Wow. SORA has really thrown all copyright concerns through the window.
- [00:13]
B
All right, welcome to Culture and Code, a podcast about the biggest shifts in culture and tech. I'm Rei Nomoto, a creative entrepreneur and founding partner of iamcore, a global innovation firm based in New York, Tokyo and Singapore.
- [00:25]
A
And my name is Tara Tan. I run Strange Ventures, an early stage fund focused on the future of computing.
- [00:32]
B
So today's topic, AI related, but specifically. And this was your suggestion, and it made me. Gave me a reason to play around with that, with it more. So we're talking about Sora and then the latest version of Sora, which I guess is Sora 2.0 or Sora 2. Why don't you start off by A, why you wanted to talk about this, and B, your early observations.
- [00:55]
A
I mean, I think it's pretty timely. Right. So In a nutshell, OpenAI launched Sora, I want to say, two weeks ago now, and they did it in a pretty. They had a pretty nice growth, hacky guerilla marketing launch, which was through the Sora social app. So this was almost like a TikTok style app, but it was all AI generated content.
- [01:16]
B
Yeah.
- [01:17]
A
And, you know, the way they launched it was that they had invite codes and it was pretty, you know, hard to get and so on. And so it drove a lot of buzz. I think in its first couple of weeks, it got a million. A million downloads, enough to.
- [01:28]
B
Enough to shoot it up to the top of the app chart.
- [01:31]
A
There you go. There you go. So, you know, really great growth, growth hacking there. I would say my initial thoughts were, you know, number one, there was a lot of slop. Right. So one of the fun things about it, you could basically create a digital avatar of yourself and create, you know, sort of things around it. So you had so many videos of St. Altman, you know, speaking in Mandarin or driving a car through New York or something like that. It was just all crazy stuff. But there was a lot of slop. That was my first take.
- [02:03]
B
Wasn't it all slop in a way? Was there anything other than slop?
- [02:07]
A
I would say there's potential for it to be. Not slop.
- [02:11]
B
Potential.
- [02:12]
A
Yes, potential. But in the first couple of days, it was. I think people were just playing with it.
- [02:17]
B
Yeah.
- [02:17]
A
It was clear that the creators were not really there. But you know, what was interesting was that I started seeing it everywhere. I started seeing Sora type content everywhere.
- [02:26]
B
Yeah, yeah.
- [02:27]
A
And some of them were really cute. Right. I would call them. I would say it's brain rot content, you know? Brain rot. Yeah.
- [02:33]
B
Yeah, yeah, yeah, yeah.
- [02:34]
A
So like you would have a dog shaped like a blueberry eating a blueberry. Yeah, like that kind of content. So it's very visually cute, but it's, it's kind of junk food.
- [02:46]
B
So that was your first point. I interrupted you. What was the second point?
- [02:49]
A
So my second point is that, wow, SORA has really thrown all copyright concerns out the window. They've basically said, I don't care. Come at me. Clearly there were people doing like Pikachu content or like clearly copyrighted content. But right now they're like, I don't care. I think that's.
- [03:15]
B
On that note, I have an observation to make. And this sounds a little extreme, but I found it extremely arrogant for OpenAI to do it and some out want to do it, number one. Number two, depending on how you look at it. And because I come from Japan, I found the approach to be very, I'm not going to miss my, my words. But almost racist and white supremacy. What? Let me tell you why. Let me tell you why. So what. Apparently this is what I read. What OpenAI did was that this was an opt in.
- [04:02]
A
Yes.
- [04:03]
B
Situation. Right. So you can opt in to have your content. Oh, I guess you, you opt out to have your content not be used by sora. And then OpenAI informed Disney's and other entertainment companies in the US but they did not, they chose not to inform Japanese entertainment companies. Oh, so this was big news headline in Japan that week because. And there was a statement made by the Japanese foreign minister saying like, why did you choose not to inform Japanese companies when you in. When you clearly informed American companies.
- [04:49]
A
Interesting.
- [04:49]
B
Like it was, it was filled with Studio Gibbie like content people because OpenAI didn't bother to inform them to opt out of their content being used. So I find it not only extremely arrogant, but also selectively informing people is very disrespectful and therefore racist.
- [05:18]
A
I had no idea. I thought everyone was under the same blanket rule. But now you're saying, no, it's not.
- [05:24]
B
It's not.
- [05:24]
A
For some reason did not approach Japanese companies. Did they tell American companies to opt out? They did.
- [05:30]
B
They. Apparently that's what they did. They approached, you know, Disney's over the world, but not Japanese companies. Language barrier, you know, whatever the reason may have been. I don't know.
- [05:40]
A
Right, right.
- [05:42]
B
And you know, I'm probably taking it more personally than other people, non Japanese people, but like, because it's not news here. But I noticed it, it's being talked about in Japan. Yeah. And it's being, you know, statements are being made, but it's not getting picked up because it doesn't really affect American companies, but it's affecting Japanese companies. And then they very quickly they say, oh, you know what, like we're not going to use your content unless you let us use it.
- [06:07]
A
Well, yeah, the damage has been done.
- [06:10]
B
Yeah.
- [06:10]
A
And it's interesting. I mean most of the content that was out there were, or the most of the sort of cameos or the reused IP were Japanese characters.
- [06:20]
B
Yes.
- [06:20]
A
So like Pokemon, Pikachu were everywhere I saw, like Dragon Ball Z, like anime type content. So very clearly Japanese ip. It's interesting. Yeah. I do not care.
- [06:31]
B
They chose not. It's not even. They didn't bother to, but they chose not to.
- [06:36]
A
It's very selective. Yeah, that's pretty controversial. Yes. Yes, that's pretty controversial. And it made me think, you know, the US has always been a pretty vocal champion of IP generally, but this move is pretty counter to it. And maybe it's, you know, a stunt, a publicity stunt or a marketing stunt or growth hacking stunt, I don't know. But this is making me think, so China is coming over and looking at Sora and being like, oh wait, these guys don't care about IP laws. Maybe we don't have to care about IP laws because there's no consequences. So, you know, this is going to start off a whole chain of well, if they didn't do it, why should we?
- [07:21]
B
So, you know, you know what's happening with Chinese entertainment companies and this is happening even before Sora, is that Chinese entertainment companies or production companies are taking old kung fu movies and kung fu stars. Right. And making fictional sequels to those movies where the movie actor might be dead or the film director might be dead or gone or, you know, the production company might not exist, but they're just making these movies without getting permission and some of them are getting sued by the family estates of those individuals.
- [07:57]
A
I mean, it's crazy fanfic. This sort of fan fiction type stuff has all has been around for a long time. Right. But it's always existed in text. So people would spin off like fan fiction universes or fan created universes based on the characters and so on. But now that it's visual, it becomes harder to sometimes tell if it's fan made content or studio made content or.
- [08:26]
B
AI fan made AI made content.
- [08:28]
A
Exactly.
- [08:29]
B
Very great blurred line between the two area.
- [08:32]
A
Yeah, exactly. So, yeah, you know, whether we like it or not, I think we're in a new era. I'm pretty sure. The studios here are pretty concerned about it, but, you know, it is this kind of rolling beast that is kind of coming and encroaching on a lot of IP rights in many ways.
- [08:51]
B
Yeah. By the way, I. I didn't mean to chop. What are the other points that you wanted to make?
- [08:55]
A
Those were my major ones. And the last one was. Did you get a headache? Was watching it. And I felt like the frame rate was a little bit off.
- [09:02]
B
Oh.
- [09:03]
A
But like, by like a hair of a second. So it gave me that same type of headache that wearing AR or VR glasses does.
- [09:12]
B
Yeah, yeah, yeah.
- [09:13]
A
So it made me think that the frame rate is slightly off. I'm very sensitive to that. But the frame rate slightly off.
- [09:20]
B
I did not pick up on that. I mean, the clips are pretty short and each scene in, you know, is pretty short. That may be why I watch a lot of.
- [09:28]
A
Lot of short content. I watch a lot of short content.
- [09:30]
B
I see, I see.
- [09:30]
A
But like, I.
- [09:31]
B
You don't get that.
- [09:32]
A
No. It tingled the part of my brain that triggers when I wear AR or VR glasses.
- [09:37]
B
Maybe. Maybe it's psychological just because.
- [09:39]
A
No, no, no, no. It was pretty. It was pretty clear. Yeah. I think this. I think the frame rate is like a hair off. The frame rate is interesting.
- [09:44]
B
Wow, that's great.
- [09:45]
A
This happens all. All the time. Right. You know, when the first films came out in the world, Lumiere Brothers.
- [09:51]
B
Yeah.
- [09:52]
A
And people watched the first film, which was a shot of a train coming into a train station.
- [09:57]
B
Yeah, yeah, yeah, yeah.
- [09:58]
A
People threw up people in the cinema. You remember the story.
- [10:01]
B
Right.
- [10:02]
A
And it's because the frame rate is. Is off, but I feel like it's a similar. It was a similar sort of tingling in my brain, and I tried it a couple of times. I went back, watched other content, came back to it, and it triggered the same part of my brain. So. I know, I know they were evolving is all I'm saying. Yeah.
- [10:17]
B
Interesting. Well, my take is I did find the approach to be very clever, and they did it quite effectively in terms of generating a buzz. So instead of, you know, releasing Sora purely as a tool to make content, they made it so that it's social media, social network. Right. And I want to like that approach because it was effective in generating a buzz. And on top of that, they restricted it so that, you know, it was harder to get in and it was hard. Like, I had to ask you for my access code. Yeah. And that approach was quite clever. Very Silicon Valley, and obviously, like, we talked about, effective enough that it made the app go number one. What I did notice, I played around with it. And by the way, like, the way that I came across that content was outside of Sora. Like, some of my friends are posting content that was generated and some of them, like, I knew immediately, oh, this is not him or her. Like, I knew immediately it wasn't them, but it was still entertaining. But one of them, I had to watch it three times to realize that it wasn't them. It was real enough. It was a person and a dog, and both of those subjects were real enough that I was like, oh, you know, this is, this is a funny clip. And this individual that I know happens to be a funny guy. So it was very him. And then the clip was realistic enough that it didn't feel like a piece of slop content. But then I watched it the third time as I, okay, this is off. This is not him. And then realized that it was generated by Sora. And then, yeah, like I browsed through bunch of clips on Sora and it'll be interesting, like how they evolve this platform, whether to leave it as is and just make it. It reminded me of giphy. So it's, you know, just purely entertainment content. Purely, like brain candy content. And nothing more than that, nothing less than that. And there's probably big enough market for them to keep at it. You know, obviously. Am I the target audience? Am I, you know, interested in that content? No, like, I just found it. So, you know, so. So swap that.
- [12:52]
A
Some of it's really creative, though. Like the.
- [12:54]
B
Yeah, I don't disagree with that.
- [12:55]
A
I love. I actually like the ASMR videos where they're like, oh, cutting a piece of crystal, man. Like, those are, I mean, those are kind of like interesting and I feel like they exist probably better intertwined within your other, you know, social media platforms like TikTok or Instagram or whatever, versus just pure AI, you know? Pure AI.
- [13:18]
B
Yeah. Whether Sora stays AI only content or not is TBD. Well, it's probably. It is going to be, but like, what's going to be interesting and potentially complicated moving forward is that this content being part of, you know, Instagrams and TikToks and, and X's and, and linkedins and people don't, don't know the difference between what's real versus what's not.
- [13:42]
A
Yeah, yeah.
- [13:43]
B
And it's already happening. It's already happening.
- [13:44]
A
It's already happening.
- [13:45]
B
It's already happening.
- [13:46]
A
You know, there was a, there was a post. I think I screenshotted it and Somebody on Reddit was like, this post had like 700,000 likes on Facebook.
- [13:56]
B
Okay.
- [13:57]
A
And it was clearly AI.
- [14:01]
B
Was it a clip generated from Sora?
- [14:03]
A
Yeah, it was like a little clip about some story about a kid. I can't remember what it was exactly, but this Redditor was like, we're cooked because it's like 700,000 likes on Facebook and people couldn't tell it was fake. But I think that that's pretty common generally and it's going to become even more prevalent. I mean, my thought is that this version of Sora, the social media app, is probably already tamed down. They could have gone really, really radical. So like the minute.
- [14:32]
B
What do you mean by that?
- [14:33]
A
For example? Like the minute you record a video, it starts multiplying it on its own. You can actually do that.
- [14:43]
B
Yeah, yeah.
- [14:45]
A
So it could actually be an entirely self generating social media feed.
- [14:50]
B
Yeah, right.
- [14:51]
A
Entirely, like without even human input, you.
- [14:54]
B
Know, so you could, you know. So your point about the, the hype sort of coming down a bit. What I did notice was the rating in the app store of this app is pretty bad.
- [15:06]
A
Oh yeah, interesting.
- [15:07]
B
It's like barely three out of five. And I mean usually, you know, most of the apps that are that popular.
- [15:15]
A
Right.
- [15:15]
B
Are four or above, right? Usually, yeah. And then when the, the rating is three or lower, I think it's going to struggle. Like if it stays the way it is, I think that's indicating. I mean, yeah, it create, it's created a lot of buzz, at least in the first two weeks or so, whether you will stay like, you know, ChatGPT managed to stay at the top of the, the app chart.
- [15:39]
A
Yeah, I think it's very experimental for them.
- [15:41]
B
Yeah, I agree with that.
- [15:42]
A
I don't, I don't know if it's like, oh, they're going to put a lot of eggs into it.
- [15:45]
B
Yeah, I mean it's a toy at this point. Right.
- [15:48]
A
And obviously Meta also launched Vibes that same, same week. So we're all, we're all in there. I mean, I in general love, I love the democratization of creative work. You know, a long time ago I said 25 years ago I was very much into film and I was making film and that process.
- [16:09]
B
You were like shooting. You were shooting.
- [16:10]
A
I was shooting, I was writing, I was editing, I was directing. Like, I was really big into film and that learning curve was actually quite high to use the tools and to go out and to shoot everything and to like edit it. It was a lot of work, hours and hours and Hours isn't almost like, you know, it was. It was. It was so hard. But now you can kind of create these amazing universes.
- [16:34]
B
Yeah.
- [16:35]
A
Through these tools.
- [16:36]
B
So I think with almost no knowledge, with almost no skill, I always argue.
- [16:43]
A
That there is skill. It's different type of skill, in a sense.
- [16:47]
B
Same one on that.
- [16:49]
A
So, for example, you still need to know, like, the people who get really, really good results.
- [16:55]
B
Yeah.
- [16:55]
A
Can prompt it a certain way. So they know what they're looking for. Right. They know, like that type of angle. Maybe they know a reference from a specific style or aesthetic. And you kind of use that, you know, and that. And that sort of knowledge informs the. The prompt itself.
- [17:13]
B
Yeah.
- [17:13]
A
So I would say that there is still skill. It's just less manual.
- [17:20]
B
Yeah.
- [17:21]
A
You know, and I think that when the true creatives get their hands on it versus the, you know, the tech. The tech bubble of Silicon Valley, I think that's where we'll actually see the real craft and magic come in. Yeah.
- [17:36]
B
Yeah. I mean, at the end of the day, I think I may have said this in another episode, but at the end of the day, the more interesting stuff will come from people who will. They don't have to be like pros and experts, but the people who are willing to spend more time with it and to spend more, to experiment more with it, to push it. And those people will bubble to the top. They will be not even 10%, 5%, 2% of the population, or 2% of the people who are posting on these things. So I think right now there's a lot of slopes. Yeah. Give it another week or two. Maybe there'll be interesting things that are coming out of it. But if it stays this way, there's just more junk to fill our screens.
- [18:24]
A
I mean, one of my favorite. I always go back to this art project, one of my favorite art projects in AI that I absolutely love. I have no idea who's doing this is probably. I think the person is still stealth or not stealth or anonymous.
- [18:39]
B
Okay.
- [18:39]
A
But one of my very, very favorite projects, I don't know if I've shown you this, called the Visual Dome. And this person just created this entire universe of these sort of civilization. And they live in different districts. And within each district, there are, like, histories. The residents look different. There's entirely new rules that they, you know, they sort of live in. And it's just absolutely stunning. I wish I can show you. So, yeah, five districts, different bloodlines, different population. This person mapped out an entire universe of like, you know, what, what these, what These people look like the residents. Their behavior like, look at this.
- [19:27]
B
Oh, wow.
- [19:28]
A
So beautiful.
- [19:29]
B
A world of diversity. Talk about diversity.
- [19:32]
A
I know. And this person, I think, created this probably it's gotta be a year ago, year and a half ago. I've been following their universe for a while, but I'm like, this is pure craft. Like, this is pure craft. So, like, you know, stuff like this really excites me. And I think that. I think we're going to see, like, creativity on a different scale. Like, look at this.
- [19:56]
B
So just to describe this, what you're showing, it's these realistically rendered.
- [20:01]
A
Yeah.
- [20:01]
B
Fictional worlds. Some of them look like people, but some of them, like these, you know, look like aliens from another planet or something. Somewhat future futuristic. Somewhat retro as well. Yeah.
- [20:14]
A
You know, I mean, like the. Just the creations, like, these are high craft.
- [20:19]
B
I would imagine that this person is a photographer by trade.
- [20:24]
A
I'm sure, like, somehow really get to the creative field. Like just the knowledge about, obviously, depth of field. Lighting.
- [20:32]
B
Lighting is very consistent.
- [20:33]
A
Lighting is very consistent. Even the color palette.
- [20:36]
B
Yeah.
- [20:37]
A
Art direction is. It's like supreme. It's amazing. But. Yeah. So stuff like that kind of makes me think, as we evolve beyond AI slop, what are we going to get? I can't wait to unlock this sort of type of creativity. Right. And this is absolutely stunning. So one of my favorite projects.
- [20:59]
B
So this is. What is it called?
- [21:00]
A
The Visual Dome.
- [21:02]
B
Visual Dome, yeah, The Visual Dome, yeah. We'll put a. Let's put a. Yeah, yeah. So I think, like, these types of creations and the people who are, you know, no doubt that this person spent hours and hours crafting each image. And like you said, the way each image is rendered, there's a certain. There's certain lighting, certain depth of field, certain colors, palette. There's certain taste and tone to it. So he or she who's doing this is aware of those nuances and putting into the prompts to make sure that there's consistency in the. The output. And I would bet this person is. Has photography in their background one way or another.
- [21:43]
A
Yeah. Or. Or film, you know, like. Or art direction. I don't know. But, you know, for me, this is a future of cinema in many ways. This is the future of a studio. I mean, it could be even a collective of people, a team doing this. I don't know. But for me, this is.
- [21:58]
B
Yeah. I mean, you know, what. What used to take 300 people or even thousand people to create. It could be done with 10 people or even maybe less.
- [22:07]
A
Yeah. Or you could have a thousand person studio, but create something that, you know, a universe that's so complex and intricate, you know, that you could go several layers down, which is amazing.
- [22:22]
B
So it would be interesting just not even five years from now, but like a year from now. And you know, the obvious industry to get hit by this is, is Hollywood. And I think it's already, you know, we're already seeing some impact of that. But then, you know, there's always something else that comes out out of it.
- [22:42]
A
Yeah, there's always something that it has to evolve. I mean it's, you know, it's getting surreal. I think. You know, we talked about K Pop Demon Hunters and I think I saw an AI video probably a couple of weeks ago where people were doing a behind the scenes of a live action shoot of K Pop Demon Hunters and it looked so real for a minute I was like, whoa, this looks amazing. I'm going to try to see it. Yeah.
- [23:09]
B
And was it AI generated as well?
- [23:12]
A
That gotta be because it's not.
- [23:13]
B
Oh, right, right, right, right.
- [23:14]
A
It's not a.
- [23:15]
B
It looked like the film K Pop Demon Hunters.
- [23:18]
A
Yes.
- [23:18]
B
But it was from a different perspective.
- [23:20]
A
It looked like it was taken behind the scenes of a live action shoot. Okay, I think I found it. Okay, should I play it?
- [23:29]
B
Yeah, yeah, play, play.
- [23:30]
A
Okay. Fascinating.
- [23:31]
B
While K Pop Demon hunters live action.
- [23:33]
A
BTs like Behind the scenes of a live action shoot.
- [23:36]
B
Yeah, yeah.
- [23:37]
A
I mean you can kind of see his AI but at first glance I think people are like, wait, what?
- [23:43]
B
Yeah, yeah, yeah, yeah.
- [23:44]
A
Really fun.
- [23:45]
B
Oh, that's pretty good.
- [23:46]
A
It's pretty good.
- [23:47]
B
Even the character.
- [23:48]
A
Yeah, yeah, yeah, yeah, yeah.
- [23:50]
B
So pretty good. Pretty good.
- [23:51]
A
Yeah. So I would say, you know, it's, it's getting interesting. So you know, obviously these are fan created content, but yeah, I think the universe is expanding. Yeah, it's definitely expanding. I mean the universe of content creation and movie making is drastically changing.
- [24:07]
B
Drastically changing. Yeah. What's our takeaway? Yeah, I think the bar for creativity is lower and higher at the same time. Yeah. And what I mean by that is that it's become easier than ever for anyone to create. Content is catch all phrase, but I'll be specific image or video based narrative content for anybody to produce. You know, five years ago you had to either have some degree, you had to have some skill, you had to have some experience to be able to produce even like 10 seconds of content. And for, you know, creators, that's a profession that didn't exist 10 years ago. But you know, iPhones and smartphones have made millions of people to be able to produce it. Now hundreds of millions of people can produce professional looking content. So that bar has gotten infinitely lower than it used to be. But at the same time, because anybody can create, you know, like the one that, that you were showing the, the visual dome. For anybody to do something that's not just better, good and better, but different. Because I think like for something to stand out, it's not just about being better, but you have to be different and distinct. And that bar has gotten that much higher because anybody can do something that looks like, like K Pop Demon Hunters or, you know, Studio Ghibli movies or Japanese anime or Hollywood movie, what have you. Anybody can create it. But for anybody to create something different and distinct is that much more difficult. So therefore the creative bar is higher.
- [26:00]
A
Yeah, I guess from an investor's point of view, I think the demand for entertainment generally is growing exponentially. So a way we can, you know, sort of talk about it is that the traditional ways of making content is not scaling as fast as the demand for it. So there's space, there's space. I think there's space for traditional made content and AI made content to coexist. Yeah, coexist, because the pie is getting bigger. We're spending a lot more time on our phones, we're spending a lot more time consuming content on our phones versus say in the cinemas or movie theaters or all of that. The question is, when are we going to see this inflection of the professionalization of AI content? And I think we're going to see that industry emerge very soon in the next five years. So the industrialization. Yeah, or sooner. But a true industrialization of this new type of studio, which I think will happen.
- [26:57]
B
Yeah. Cool. Well, that's a good place to end. Good topic. Fun topic.
- [27:03]
A
All right, see you next week.
- [27:04]
B
Next week. Yeah, bye.