Ep 26 - Breaking Down Sora by OpenAI: The Next Leap in AI Video Generation

Watch the YouTube video version above or listen to the podcast below!

Ep 26 - Breaking Down Sora by OpenAI: The Next Leap in AI Video Generation Podcast and Video Transcript

[Disclaimer: This transcription was written by AI using a tool called Descript, and has not been edited for content.]

Dave Dougherty: Hello and welcome to the latest episode of Enterprising Minds. Dave and Ruthi are here today.

Discussion on OpenAI's Sora Text to Video Generator

Dave Dougherty: We are going to do a quick one-topic conversation because just Yesterday, or at least as I was made aware of it OpenAI released their Sora text-to-video generator. So yeah, Ruthi, you're the one that made me aware of this.

So why don't you talk about it a little bit?

Ruthi Corcoran: I did. I have not been paying as much attention to all of the new AI news updates and tech updates as Dave, you have throughout the last few months. But this 1 came up in multiple places in my sort of general information feed. And so I did, in fact, go out and see what all the fuss was about and turns out it's, it's worth the fuss.

So. What OpenAI released is essentially the DALI equivalent for video. It's called Sora and the it's not open access yet. It's limited to a very small group of testers and a very, very small group of content creators, but they've been posting different versions of it that have been created. And it's quite striking in that.

If you know, you're looking for an AI video, of course, you can spot it. You can see things, right? It's kind of, you still have a little bit of the uncanny valley going on with people's hands movements, for example. But if you don't know, you're looking for AI, it is much more realistic than anything we have seen.

Or at least that I've seen in the AI space for creating video and as the tech YouTuber MKBHD says, you know, this, it only gets better from here, which, when you take that viewpoint of this is the new, this is the new basement for this type of technology. That's pretty astonishing. So that's the general gist of what they came out with and we, I don't know offhand what the timeline is for releasing it to the public, but it opens up a lot of new space for dialogue because now, all of a sudden, just like Dolly, you can create any sort of any sort of video content.

And that changes a lot of what gets shared on on this wonderful place. We call the Internet. So that's my initial.

Implications of AI Video Generation on Content Creation

Ruthi Corcoran: I've got some more thoughts on it, but I'll hand it over to you, Dave for.

Dave Dougherty: Yeah, I mean, with the initial results, I definitely sympathize with the video guy you sent over. I'll link to the.

The video Ruthi shared where I think he had, he was having the same reaction I had when I first saw the text generation. So as I identify more as a writer and a musician, then, you know, video creator I was a lot more worried about those things than I am the video. But to your point, there's. One of the influencers that I really like to, to listen to with Paul Reutzer, he is the founder of the Marketing AI Institute and has a great podcast and website devoted to using AI in marketing and business.

And he's saying all the time too, this is the dumbest form of AI that we're using right now. And it's, you know, it's already unbelievably good. So. Yeah, I think the the applications of it could be really cool video in particular online gets me a little more worried than than text generation, right?

Because it's a little more convincing. It's you know, I think about, like, You know, my mom or my niece, who may not be as educated in AI or digital marketing communications, and I just go, Oh my God, so many people are just going to be convinced of this like whatever video they see. So yeah, I mean, it's, it's a two way street for sure.

It you can tell I'm trying to make sense of it as, as we discuss here, but it's just another showcase of. It is unbelievable how far the videos have come since, you know, last year, I think was the first, the first text to video thing that I saw and it was pretty impressive, but it was still just like, Ooh, okay, this is new technology.

Yeah, I think Facebook had one meta had one that was out for like. Well, like two weeks or something before they shut it down. But I just remember the, the, the, the physics in the video were pretty cool there, but what a, what open AI showed on their product launch, like the reflections on the street in the video of the Japanese woman walking down a Tokyo street, right?

The fact that you had. The, the neon signs and the lighting was correct and the shadows moved and the, you know, the reflections were in the puddles on the street and like all these little detail things are just like, oh, my God, that is unreal. That is unreal. So,

Ruthi Corcoran: all right, I'll pick it up from there and sort of go off in a couple of different directions.

So. Three general thoughts.

Impact of AI on Marketing and Advertising

Ruthi Corcoran: First one is I reflect on in in the marketing world. We use a lot of video. We use a lot of copy and we use a lot of imagery in order to create advertisements. And the creation of all these mediums has become cheaper and cheaper in the short run. I think that just makes it easier for incumbent companies, small startups to get an edge on large, large enterprises who are a bit slower to adopt.

It's all of a sudden much cheaper for the startups and the smaller businesses to create sort of really nice, good potentially effective advertising. Whereas the enterprises, they take a little bit longer to adopt this sort of thing. They can't move this quickly. So I think that's sort of a short run effect.

I think in the long run, it probably buffs out. And I think also in the long run, these shifts in the cheapening of content going into advertising and other things is going to continue to shift. The way in which the marketing landscape has evolved, right? We've seen this already with the advent of the internet that had a big change in how we do marketing.

I think in the last sort of five to 10 years or so, we've saw, we've seen the influx of content creators as almost like you're outsourcing or outsourcing your marketing to influencers and content creators. This seems to me another technology change that's going to shift the way the marketing landscape happens.

And what's effective just as a quick note. I think it's notable that the Super Bowl ads, at least in my eyes, have been coming less and less amazing. So to speak. I think that all of the creative agencies that used to do the Super Bowl ads and make them awesome. They've now directed their attention elsewhere because they know they can get more bank for their buck and other mediums.

And the Super Bowl just isn't the thing. So I think we're already seeing. That shift in the marketing landscape and that the Super Bowl thing is just a symptom. I wanted to make sure that I shared here because it was just something that struck me this last weekend. So the marketing landscape is going to shift how I'm not entirely sure.

I'm going to have to think about that. But the last thing, which is in a totally different direction.

Potential Influence of AI on Gaming Industry

Ruthi Corcoran: Is I see that video technology and I go, Ooh, what does this do to the gaming industry? How do we think differently about games in the future? Be they just sort of indie games or the triple A games? I think of something like Skyrim, right?

Which I'm a huge old fan of when Elder Scrolls six comes out, it's going to be a great day. And I think about how boring the dialogue options. Became over playing, right? You know, an arrow to the knee. You can only hear that. Like, it became a meme of how often they would repeat some of these lines. And so 1 of the cool things I think I, I'm looking forward to seeing is how.

Gaming shifts because of the ability to do these sort of generative language within. You know, audio, but then also, maybe you can expand that to video, too. So, all of a sudden, right? Skyrim, like, is a huge world. It's very immersive, and there's other even more impressive versions of this, where you've got these huge worlds.

Maybe this sort of technology just improves the ability for games to create those immersive worlds that you, you can't do if you have to create it all, essentially.

Dave Dougherty: Yeah. I think coming off of those, I do think that it unlocks a lot of the harder mediums to more indie creators. Like when I watched the, the open AI stuff, or you know, before this announcement runway AI was the, at least in my mind, kind of the leading video text to video generator.

The immediate thought that I had was, you know, if you were creating these kind of like, you know, alien worlds or landscapes or something where you may not have, you know, budget for a huge studio or something, but you can still shoot in green screen or something, you could then bring that idea to life more easily through the AI generated landscapes.

So that's one thing when I watched the open AI video of the woman walking through Tokyo, I immediately thought of a music video idea that I've had for one of my songs for like 12 years, never had enough money to do the traditional video shooting or storyboarding, like, you know, and I just went, wow, okay.

If you got enough of these sorts of clips, you could weave together a narrative. Of these individual shots to create a kind of mood and then, you know, put that out. Right. So that, that is kind of a cool thing. And then, you know, shots of the band playing or the, the artist singing would just be kind of B roll for the, for the AI generated stuff, but yeah, it's, it's cool how close you can get.

From, you know, having the idea to now producing things.

Copyright Implications of AI Generated Content

Dave Dougherty: The, the other thing my brain immediately goes to is, okay, what is the copyright implications of this? Right. So I think it's a little bit more defined for writing, you know, the, the copyright office came out and said that, you know, it has to be majority produced by human in order to be eligible for it, which means you can essentially create an outline and then write it yourself.

And that kind of counts. Now, again, if you're going to do anything with this, don't take this as advice. Talk to your lawyer. But that at least, that at least seems doable, right? But if there's if you cannot copyright The video AI created videos and it plays a role in a background or something, all of a sudden, you know, like for off the head example, if, you know, Marvel used an AI generated alien world background.

Could any schmo use that same background because it's, you know, AI generated, right? Where does that copyright fall?

Ruthi Corcoran: I think I'm pretty the copyright piece. I'll leave that to the experts and I'll just watch from the sidelines. I don't I don't know that I have anything to add to that line. But your comments about sort of the, the. Indie game creator, or even just this music video thing. I hadn't even considered, right?

The idea that now all of a sudden you can add video to a dimension of some other thing that you're doing that you just otherwise wouldn't have access to is pretty phenomenal. And I think you also hit a hit on something that I just want to emphasize, which is reducing the space between that idea and the ultimate creation.

I think that is sort of. At the heart of what all of these different AI technologies are doing is they're making it easier to go. Oh, I've got an idea. Let's make it a reality and. That's going to unlock things we can't imagine be the proliferation of music videos, or you've got some brilliant artist who has an idea about something, but otherwise couldn't have made it happen.

And all of a sudden it can and there's very likely going to be a number of not so nice things that come along with with reducing that space as well. And there's going to be a lot of upheaval. I think I think we're in for quite a ride in terms of how. How we interact with each other, how we interact with this thing called the Internet sort of what the nature of.

Our our society, and for sure, our job force looks like, especially if you're in the sort of, I work with computers area of. Of the job industries, the job industries.

The Future of AI in Content Creation and Consumption

Ruthi Corcoran: That's a good line. You can, you can call me on that one. That I think we're in for it. I don't know exactly where it's going. I can't predict it.

I just know that this is, this is a shift in how we do things. And that shrinking of idea to creation is going to have some huge ramifications in the next few years. How do we get ready for it? That's I think the big question I've got is how do I think about this in terms of my day to day work in terms of this type of creation that we're doing here or recording a podcast in terms of how my kids will inevitably interact with these technologies.

That's also huge and top of mind,

Dave Dougherty: right? Yeah. It would be interesting to play around with it, even just to create potentially like a, a more interesting. Title screen, you know, doing something with motion graphics for like, cause that's one of those things where it's like, I would love to be able to go like a next level, make it look more polished, more professional, but I'm not a designer, I'm not about to learn super advanced video editing skills.

Right. I N this is something that I was talking with a friend about recently where it's, you know, with the AI stuff, I feel only comfortable. Really working on the stuff where I know how to explain it. And I know how to talk about it, but I don't necessarily know how to do it, right? It's the same thing when I'm looking at an Excel spreadsheet.

I might not know how to super automate everything, but I know what I'm trying to solve for, right? So whereas before I was going to a website that could show me all the tips and tricks of Excel, now I can just say, I want to learn this particular thing, see if it's there. That is wonderful. And I was driving around yesterday and I was thinking, wow, okay, I'm able to this problem of being able to take ideas of creation to final thing through the creative process so much more quickly and at such a higher level than before, I wonder if we see the cycle of burnout.

Completely compress because you'll be able to create so much constantly, right? That maybe you don't take enough enough time for yourself because you're so good at, you know, good creative is more about saying no to all the random ideas you have, right? But now we're getting to a point where it's super easy to just create it.

Cause he had the idea. Yes. That I wonder about that because I don't know that enough people will, you know, take the day to sleep on the idea before putting it out to the world, right? I mean, cause there, I, I have enough experience with the music side of things where it's like, I think I nailed the take. I nailed the mix on the night that I'm recording it.

I go to bed, I come back, I list do it and go, what was I thinking? That's trash. I have to redo it. Like,

Ruthi Corcoran: you know, I, I think you bring up a really good, good point. I would never have taken the burnout angle. The angle that I think of is somewhere along the lines of the lean startup book that came out a while ago that says, you know, hey, just just get your ideas out there.

You don't know which one's going to work. So put them out in front of people because 1 of them may take off. You just don't know which. And as I think about this influx of all of the ideas that can now become reality, I think of, hey, what are the amazing things that wouldn't otherwise have existed because the transaction costs for making it a thing were too high and now all of a sudden are my optimist is thinking we are going to, we are going to see a fair amount of growth.

As a result of this, simply because of the surge of ideas that can now be turned into something real.

Dave Dougherty: Yeah. And it's, it's interesting because with the music examples, I was recently at a show for one of my favorite bands called machine head, right? They're out of San Francisco. They're like They're like a heavier Metallica, if you're unfamiliar with them. But you know, they've been around since the mid nineties.

And they've gone through all of the analog stuff, all the digital stuff. But the, the live show I thought was interesting because it was in a smaller club or a smaller theater, I should say. And instead of doing like the big pyros and the backdrops and everything else, they had a couple of led screens and then they just had.

different different videos programmed by the song. So the lighting and the videos were set up to go off at the same time. So then each song had its own video. And it's like, it's exactly what I was talking about, where it's like, they came up with the one really cool image. And then they added some motion graphics to it so that there was a little bit of interest over the course of, you know, five, six minute long songs.

Right. But it brought a cooler effect to the stage than like the opening band that just had the silk screen, you know, band logo and the members on, on the front of the stage. But it also allows them to travel with less gear, which makes touring. more profitable, more easy to do, right? Because they're not paying for all that weight between each city or flying it overseas or whatever else.

Yeah, so it'll be interesting to see how, how people leverage that as well for that live experience.

Ruthi Corcoran: Your comment about it being an led background reminded me of a recent video that I watched YouTube channel called Baratacium, where they're more longer, they're long form YouTube videos anyways, that dive into.

Lots of different things within the science realm. The topic of this video was on developing the blue LED light. Which turns out, go watch the video. There's a lot of reasons why that's difficult. They go in depth. I will not attempt to repeat them here. And it took decades for them to get a blue LED light.

And what's amazing is once they did, not only did they allow the color blue, like you go, okay, sweet, good job, but because they had blue, green and red. Now, all of a sudden you could create white, right? That's what enables us to all to have led lights in our houses. That's what enabled the big led screens.

Basically. The creation of that blue led unlocked a huge, huge shift, like billions and billions of dollars in industry and just a different way in which we interact with, with screens and technologies and even light bulbs. And to me, maybe that's an analogy, as we think about these technologies, right?

We're all the things we're seeing are very cool. They've got a lot of great applications, but the 2nd, somebody comes with that blue light equivalent, we are going to see. A very different ball game in how AI is a part of our every day.

Dave Dougherty: Yeah. Yeah. And it's interesting because, you know, the, the AI video piece is problematic, at least in my mind just because it does showcase so much how how slow the legislative process is.

So part of me worries like, okay, how many people create all these different things just to find out. You know, that they can no longer use it because of the copyright office or because the Senate finally does something. And it, it'll be really interesting to see how these things get regulated. Not just because the heavily regulated industries already.

Like I think of, you know, using this technology for business, right? You could do a product launch video relatively easy now, just by saying, I want this particular thing in this application, in this environment with this. You know, lens kind of lighting, whatever. But how do you make sure that every single product in that particular shot is of a particular brand?

Definitely not there yet. Right. Or you know, what are you allowed to show in your advertisements? Like if you're a, if you're a healthcare company, right. That that's a particular issue as well. Right. So I do see like the, the video and the large language models definitely help like the individual creator or the smaller agencies a lot more, and especially outside of the, the heavily regulated areas as well, man.

I mean, even if you could just do an avatar of yourself. Speaking the meeting summary for all the people who didn't bother to show up, like, that could be a win

Ruthi Corcoran: and I think, oh, I'm going to tease off of that one just for a sec, because it brought on a, we have so much, so much content being created. And so part of the challenge for the individual, let's set aside regulations and all that jazz. Like, that's an interesting topic given of itself that we could go on for a while.

But as an individual, you are being hit with this huge influx of content, and so being able to navigate through all of that content is like, why should I choose to watch the video recap that you've. Created with a I of the meeting notes, right? Like, we've got to make all these little decisions. And I think that's another area of emerging sort of space.

Things to tease out are creating those filter mechanisms to give an analogy of what's what's going on in my head. Publishing companies exist to say, yep, we've given this the stamp of approval that says this is worth your time. Right? Right? There's a reason you. You would choose the Yale press when you're looking at academic books to say, okay, yep, that seems like it's legitimate.

We don't yet have a lot of these institutions filter social media to some extent is kind of filtering it for us in a very odd way. And I think that potentially causes a lot of problems. And so I think that's a lot of untapped potential too, is helping the average person navigate through the Deluge of content that's coming and will be increasing over time.

Dave Dougherty: Yeah, I think it'll, it'll really, it'll really bring us back to some fundamentals though. You know, like the nice thing about a particular creator and the creator economy is that. You identify with some aspects of them, right? And there's, there's a trust there through the, the knowledge sharing, right? Whether it's makeup, how to's or marketing agency tips and tricks, right?

You develop a relationship, albeit one sided over time. Right. With that particular thing. So that ends up being a filter of sorts, right? So if you are into a particular creator or get exposed to a particular creator that you end up really liking more content from them can be a good thing. It could also lead to many more opportunities for For some problems with, you know, celebrity endorsers and, you know, that whole other thing, that's another topic we get into and, and the fraught things with that.

But again, like with the crater economy allows for smaller organizations, smaller individuals to be able to do something similar as the, the big organizations. Right. Yeah.

AI in Webinars and Online Presentations

Dave Dougherty: Yeah, so the reason I bring up the particular meeting recap piece is I've been seeing on a number of webinars lately which, okay, side, side comment, anybody who's hosting a webinar for the love of God, don't make it about yourself.

Unless it's specifically a product promotion thing, like, that's not why I'm here. That's that was not in the copy that brought me to your webinar. Stop it. Right. Okay. Small rant aside. So I've been going to these webinars and if you go into the chat and the user piece, I've seen a lot of these, hi, I'm blah, blah, blah.

So and so's AI meeting agent. And it introduces itself. And that to me, it's like, okay, maybe you're getting something from it. Cause yes, a lot of webinars are a waste of time. I get it. But the only way that you're going to learn or have a depth of knowledge appropriate to call yourself an expert is to attend these things is to learn something at a deep.

level, not send a robot to go and tell you what the important bits were, right? I mean, that's what the landing page copy is to get you there to pay attention to the important bits, right?

Ruthi Corcoran: I think you're missing. I think you're missing the main boat. You ready? We don't need people to give webinars anymore.

The AI is a whole webinar and they can take out all the nonsense straight up and it's just AI webinars. We're going to need a new name.

Dave Dougherty: So with that though is a trust factor. So the reason the AI associate thing bothers me a bunch is for me that's a respect thing. Like if I put in a ton of time and a ton of work to bring a quality presentation to you, I don't want to be talking to your ghost.

Like, that's disrespectful, right? So, maybe that's my own thing to get over. I'm fully open to that. It's a participant's

Ruthi Corcoran: AI? Yeah. What? People are already having AIs attend webinars for them?

Dave Dougherty: Yes, yes, that's the thing. Where it's, you know, happy to be here. So and so's AI is attending this webinar on their behalf.

So then you just get a summary of what was presented during the webinar to you without having to go to the webinar. So it's like, you know, you sent your friend to your college class to take notes for you, right? Did you learn anything? No, because you weren't in the class. You were off doing something else, right?

Just because you look over the notes doesn't mean you're going to learn anything. You're only working towards a bare minimum of something, right? So. That that's one thing, right? But I do agree with you on the, on the presentation side of things, you should start rethinking about, okay, what is, what's going to be a valuable way for me to communicate this, maybe it's more in person presentations rather than the ease of online digital things.

Right. Because if you have a live in person thing, then you can do the coffee and doughnuts beforehand, and then you can, you know, do the networking afterwards or answer questions or whatever else after, right, that's a little bit more of a conducive medium than just I'm going to talk online and record it because it's easier for me to do that kind of thing, right?

Ruthi Corcoran: I need Alex here to talk about how we're going to have webinars or AIs attending webinars that are given by AIs and the cycle just just perpetuates itself. I think here's what's, here's what's coming to mind. I didn't understand what you said at first, that it's truly an AI attending the webinars and then providing the CliffNotes version.

And then I think back, oh right, CliffNotes and SparkNotes. People have been doing that for ages. Like, just give me brass tacks. What are we here for? So, I sort of take that as, like, that's just human nature. I shouldn't be surprised at that. 1 of the things that does come to mind, though, is over the last couple of decades, we built an infrastructure around.

Around the existing technologies, a key example, I have, if you are a nurse, you can get certificate. You need to get certain number of certification hours. 1 of the ways that you get those certification hours is by attending, say, hour long webinars in order to fulfill. And if we've got an infrastructure for certification based off of webinars, and all of a sudden you introduce this new thing, you Now we've got to really think how that certification works.

And I think those are the types of little details that are going to be unraveling and evolving over the next few years in particular with this new technology that, again, I can't even think of all of them, but there's one more that I've just added to the list,

Dave Dougherty: right? Yeah. Yeah. Lots, definitely lots to think about and the more these things improve.

The more intense the decisions will have to be for sure. Cause I also think, you know, brand safety, right? I mean, I, I can't take the PR hat off. I, you know, once it's in, once it's ingrained, it's ingrained. Yeah. But, yeah. Yeah.

Conclusion and Final Thoughts

Dave Dougherty: Lots of thing about hope you enjoyed the conversation. Appreciate the time.

Write any comments, reviews, let us know what you're thinking on this. Have you experienced it yourself? The, the video AI and we will see you in the next episode of enterprising hearts, appreciate it.