Broc: Well, was it at the start of this month or was it I think it was the end of March? OpenAI announced that they are shutting down Sora AI video generator. So Sora is a video generator. And it started small. I think it started just in the Chat GPT app. And you could make some goofy videos, but then they launched it as like almost they were trying to make it, I think, like TikTok. I don’t know if you guys ever tried to use it.
Zach: There’s a weeks back.
Broc: But it’s literally I went on it once and it was like almost like TikTok, but it was pure AI. Pure Sora videos. And people could upload their their likeness to be used. I think by anyone.
Zach: Disney had like a ridiculously big contract with them that went kaput when they announced the closure.
Broc: Yeah. So that was like a huge thing and I think it started with a ton of users, like it it had a ton of downloads at the start. But this is saying that towards the end they were losing a million dollars a day.
Zach: How much processing power goes into just creating this slop.
Broc: Yeah, like we haven’t even talked about what AI does for the environment and for communities that are near data centers, but that’s a different topic. So this is this is i in my eyes, this is like step one of this thing just crumbling. Like this is the first s sort of card that they have to show. I don’t play cards, I have no idea how that works, but they’re they’re finally showing their their true colors. Like, yeah, this doesn’t make any money and people think it’s cool at the start and then the user base just dwindles because people are like, this is getting really old. I wanna go learn to play guitar.
Zach: Well this would be an expensive novelty product. Like for them to maintain like all that processing and server power costs a lot of money. Like it costs a lot more money to generate a video than it does to generate a wall of text that someone’s gonna use. So it’s like if they’re not getting any return on that I can see them pulling the plug much faster than on like one of their regular blog chat, chat GPT things that people use endlessly.
Broc: Yeah. So this begs the question, why should we care about all this? Why should the average person care about what we’re talking about? In your guys’ opinion. Okay.
Joseph: Yeah, I mean like posterity, right? Yeah. Who wants their kids to be like the people on Wally, just sit in their chairs watching whatever they’re told to watch? You know, by and large, your very best friend, you know, like no thanks. So I don’t know. I I think it’s as simple as just this is one of the times I think forward is not forward. We should probably go back a little bit. And you’ve got the guys who are like, we need to be careful. This could be really, really bad, but they’re more talking like, We’re gonna have AI just take over and enslave us all. And I’m thinking, maybe we need to be careful because we weren’t careful with social media and now we have like massive amounts of super depressed kids with ADHD who can’t focus, have headaches, all the freaking physical th symptoms. And I mean, the Meta just lost like a seven million dollar case because they proved that their app was like was actually meant to addict And do everything it did to this girl who has all these problems. And it’s like everybody kind of knew that, but they ignored because, like, it’s fun, it’s cool, right? The truth is why we should care is maybe we should like care a little bit about others besides ourselves. And if we do that a little bit, then we should be slightly concerned. Like nothing about this improves anything. Like, I haven’t learned one single thing from AI. Now it’s o it optimizes what I do. Like I can use Claude and having a computer science background, I can go in and I can say, hey, code this for me. And then when it doesn’t work inevitably, I can say, hey, it doesn’t work because of X, Y, and Z. And I’ll say, okay. And then it fixes it and then it works. But I would never know how to do that if I didn’t have a background in computer science. Yeah. I guess my point is like, yeah, it’s a great tool to optimize my brain, my workflow, but I have yet to learn anything from it.
Broc: Yeah, where do you think the line is between optimizing and enabling? Like where’s the line between this helps with my workflow and this is making me like lazy? Is there is there a definitive line?
Zach: Ever heard the term vibe coding? It means like entire companies are taking their entire dev teams and replacing them with Vibe coders where you have these people who are well versed in these coding languages who have been in the infrastructure for years. A lot of them are architects in their field and they’re just being replaced with people who can prompt Claude and Chad GPT to be like, build this for me. Hey, it doesn’t work. Why doesn’t it work? Fix it. And they eventually kind of get there with this thing that’s about 500 lines longer than it needs to be because it’s got all this redundant code in there and you have like important comments that are conflicting with each other and stuff but it eventually works 99 % of the way it should but at the end of the day you kind of lose the heart and soul of it like it wasn’t actually created by a person and really nobody actually knows how this thing freaking works I think that’s part of the issue there is we’re gonna end up with an infrastructure that nobody is going to actually know how it works someday if it all goes that way.
Joseph: Yeah, a few years ago, a lot of the new like when JavaScript or when jQuery came out as a part of the JavaScript library library, a huge draw to it was that somebody was like, JavaScript needs to load faster. And so they came out with like Ajax, which was I don’t know, Pop Quiz. Does anybody know like the first big site to make a Ajax just cool? And everybody was like, my gosh, when they saw this. Twitter.
Broc: I don’t know what Ajax is.
Joseph: Do you guys remember when Twitter was like the first platform where you could be sitting there reading and a new tweet would pop up and you didn’t have to refresh the page? You guys remember that? I used Twitter.
Joseph: No, so like forever, if you wanted to see anything new, even on the page, if you were in a like you had to hit refresh where you didn’t see anything new. That’s just standard internet, like you have to refresh. But then along comes Ajax, where it can load things in the background and essentially it’ll auto-refresh without refreshing the browser. And Twitter implemented this to where then you could actually see tweets just popping up. And it just made It made the internet feel like it was alive, like you were interacting with it. But that was built because some developer was like, you know what’d be cool is if JavaScript ran faster. I know jQuery. Okay, well, what can we do on this new jQuery model? well, we could, you know, write in the ability in the background to do things without the person having to request it. Cause normally you would do like on page load or on submit. Those are all jQuery term or JavaScript terms. So on these things, like if you do this, then this will happen. Pretty standard stuff, but they basically said, well, or you can code it in to just say auto countdown every three seconds, refresh browser in the back end, and then just feed it down below. And this is all new stuff and it’s crazy. But it was built on like, well, this would be neat. So like essentially, like Zach’s saying, if you don’t have developers who are like, y you asked about where the line is. Well, I think the line is that like a developer would know exactly what they want built. So they say, Code this for me, and then they can look at it and be like, that’s sloppy as hell, like or heck. Let’s trim this down. We can use this different library, that’ll make it quicker, etc. And they’ll actually be able to build some cool things based on ideas that they have. But then if you get these vibe coders where their boss is like, I need to build this, then they’re just like, Okay, I need to build this. And it’s like, great, I built it. And they activate fatal error. well, what did I do wrong? Hi, I’m giving a fight a fatal error. Okay, let me check your code. This and like Zach said, five thousand lines of code later, you might have something working, but then if it breaks, heaven forbid it, they’re not gonna know they wouldn’t even know where to go to the files to look at the line of code.
Zach: So Broc have you ever heard of this thing called Y2K? This was back during that which a lot of people say that it was like a myth that was fake news whatever it was it was a legit thing and the reason why nothing bad happened is because they actually fixed it before something bad could happen basically from like the 70s through the 90s you were you had people building these software suites and a lot of them got lazy and then I’m gonna put in a two-digit year integer and
Broc: Yeah.
Zach: They got to the point like, what happens in the year 2000s when everything rolls over? All these computers are to think it’s 1900. Oh crap, we should fix that. At that point, they had to go back and pull in these dudes out of retirement who built these things back in like the 80s and 90s who had since then like finished their careers and retired because they were the only ones who knew how it was built. My fear, and I’m almost a little excited about it, is that we’re going to get to that point someday with all these things that people have built through like vibe coding and they’re going to need to do some kind of major infrastructure overhaul on it and they’re not going to have somebody to pull out of retirement to ask who built it because nobody built it nobody knows how this thing works they there’s going to be some underlying thing that some major infrastructure uses that nobody’s gonna have any idea how to fix where to fix it how to get the permissions to fix it and it’s just going to be just like that like with Y2K we were able to stop in course correct it’s gonna be like hitting a brick wall with something because we’re like, yeah, too many things run on this thing that was built in 2030. It’s now 50 years later. Nobody knows actually how to fix it unless we take the whole thing offline and rebuild it from the ground up, which is going to just destroy all these databases. So it’s like, what’s stopping that from happening? And my fear is like, maybe something’s already set in motion for that to happen.
Broc: Zach it’s very optimistic of you that you think we’re gonna last till twenty eighty. That makes me happy.
