Hey, everybody. Welcome to the Lyas Snippet. Um, this one's gonna be an interesting one because I'm hitting record without having any clue what I'm about to say. Uh, often when recording a podcast, like of course, it's not scripted. Uh, usually not scripted, almost never scripted. Uh, but, you know, I'll have a general outline, um, a line that I wanna get from here to there, right?

Uh, but, you know, I'll have a general outline, um, a line that I wanna get from here to there, right? Uh, this time I thought it would be fun, though, to hit record and just see what happens. And it might be incredibly boring. Uh, but I don't know. It'll be close to, uh, how it might be if we were having lunch together and this subject came up. And of course, the subject, the topic is, um, ai, which I would imagine is on all of her minds.

And of course, the subject, the topic is, um, ai, which I would imagine is on all of her minds. And here's what I find interesting. Depending on who you talk to, you get a wildly different perspective, which I find interesting. And I don't know if, is this a reflection of our personality that maybe some of us see the benefits for us? Some of us lean towards, uh, caution. Um, I can't help it. Like my, I am getting

Some of us lean towards, uh, caution. Um, I can't help it. Like my, I am getting so much use out of it in my day to day. It, it's making me better. I like, there's a lot of talk about AI making us dumber. And I, I can see that happening currently, though, the way I'm using it. I feel like it is making me better. It is like the perfect peer programmer where I can say like, here's what I'm thinking.

It is like the perfect peer programmer where I can say like, here's what I'm thinking. Let's brainstorm with me, gimme some ideas of, uh, if this would be a good way to go about it. Uh, so I find that incredibly helpful. But at the same time, there's just this undercurrent of fear that I have. Um, and I don't, I don't know why I feel that way and other people don't at all. And it could be this, it could be that I'm 40 years old now,

that way and other people don't at all. And it could be this, it could be that I'm 40 years old now, and I was raised in the eighties and nineties, and I grew up on Terminator movie, like Terminator 2, even to this day, is one of my favorite movies ever. And that idea of AI turning against us is just ingrained. It's like, it's part of my, uh, identity to be concerned about what happens when, uh, technology evolves to the point that AI, uh, turns on you.

to be concerned about what happens when, uh, technology evolves to the point that Ai, uh, turns on you. And I know it's a movie, I get it, but still, like, I can't get beyond that. Every time I see a video, uh, from one of Elon Musk's companies where they're showing robots doing flips, I'm like, what are you doing? Have we not seen the movies? Do we not see where this is gonna go? Um, so yes, I get it.

Have we not seen the movies? Do we not see where this is gonna go? Um, so yes, I get it. I know it's a movie. I know it's not real, but man, I can't help it. Like, my default is, oh, here we go. Like I, I, I almost imagine we are going to benefit from AI right up to the point that like, all right, time's up, you're done, you're out. Uh, you know what I mean? So like, let's talk about in the programming community, how is it going

Uh, you know what I mean? So like, let's talk about in the programming community, how is it going to affect things? And uh, again, everyone has their own opinion, right? Some people think it's over. Some people think don't learn to code anymore. You're done. Uh, other people, I find, um, other people take the opposite view. Like, um, the code it writes will never be, uh, up to a certain standard.

Like, um, the code it writes will never be, uh, up to a certain standard. And it's always junk. It's always going to be equivalent to a college, um, effort, which is just mind blowing to me. 'cause you can see the trajectory, right? We can see what it was like in 2023 is that when OpenAI first announced ChatGPT I think it was 2023, we can see the hallucinations in terms of code.

GBTI think it was 2023, we can see the hallucinations in terms of code that it would produce back then versus now, where it's just not nearly as much, not even close to as much. And that's not to say, um, it doesn't get things wrong, but I find less and less am I saying, wait a minute, that's not a thing at all. Where did you get that from?

that's not a thing at all. Where did you get that from? And then it does that spiel of like, oh, I'm so sorry. Thank you for catching that. Let me do it again. And then it gives you another, uh, example. And you have to respond and say, Nope, that's not part of the API either. None of this is real. You're just making things up and then you just end up in a loop. I don't get that nearly as much, um, to be honest,

This is just a simple explanation with no code.

I don't even know how to calculate how good it's gonna be at that point. Um, the other day I heard secondhand a quote from, uh, Peter Thiel, so I didn't hear him say it, but I heard, as this often turns out in podcast, I heard somebody else reference what he said. So now I'm gonna tell you what this person said that Peter Thiel said. And it was something like, what is it? Think Jeffrey. Um, things are,

And it was something like, what is it? Think Jeffrey. Um, things are, things are always worse than you might expect and better than you can predict. Uh, that was horrible. What is it? You, you get the idea though, like whatever you anticipate in terms of how good things can be or how much things can grow, you are wildly underestimating how big it can get. Uh, but the exact same thing is true in reverse.

you are wildly underestimating how big it can get. Uh, but the exact same thing is true in reverse. So if things are going badly, what you imagine rock bottom might be is not even close to what rock bottom actually can be. Um, and I, I get the feeling that's how it's gonna be, right? I, I get the feeling that many of us are just going to become triple developers. The, the, the speed at which you can produce is going

to become triple developers. The, the, the speed at which you can produce is going to triple, uh, quadruple right up into that point when you realize, oh, wait a minute, I'm kind of the weak link in the chain here. Like right now, currently in 2025, I am the strong link in the chain, right? I can review things and see like, Nope, that's not right. That doesn't work. This is horrible. This looks like it was written in 2015,

That doesn't work. This is horrible. This looks like it was written in 2015, and I can kind of clean it up, right? I'm sort of the project manager, but like, it's not gonna be very long at all, maybe even a year, uh, at this pace where suddenly I'm the weak link where it's suddenly like, oh, wait a minute. I don't, I don't have a role here. Like, uh, I, I, I joked the other day on, uh, social media

I don't, I don't have a role here. Like, uh, I, I, I joked the other day on, uh, social media that like, my job is just going to be to agree with whatever AI produces. So I'll say, Hey, this doesn't work. And then AI will say, well, that's because blah, blah, blah, blah. And then I will say, yes, agreed to make it so fix it. You know? Uh, so we are going to benefit from it right up until the point when we are completely, uh, unnecessary.

You know? Uh, so we are going to benefit from it right up until the point when we are completely, uh, unnecessary. And here's what's interesting though. Again, it's like people, people disagree on this so much programming, friends, I have wildly disagree with me. And I think it is useful to some extent to consider like motivations and personality and I don't know, just sensitivities. Like nobody wants to hear that your job might become superfluous or irrelevant,

Like nobody wants to hear that your job might become superfluous or irrelevant, but that doesn't mean it's not true, right? So some people may feel like they are the special breed of programmer where it's like, okay, maybe, maybe AI can produce all of the, the crappy crud apps that the rest of you plebes are, are working on. But what I do is, is extra, right? What I do is serious stuff that cannot be replicated by AI. And I feel like every argument about that, it, it's,

What I do is serious stuff that cannot be replicated by ai. And I feel like every argument about that, it, it's, it's almost like, um, the God of the gaps argument. It's like every time you reach for that and then AI grows and then proves to the point where it can do the thing you said it couldn't do, then you're just going to stretch it out a little further and say, well, I can't do this. And then six months goes by and then it can do that, right? I feel like that's the trajectory for

And then six months goes by and then it can do that, right? I feel like that's the trajectory for that particular argument. Um, but I don't know. I think we're all just trying to figure out what it means. And some of us feel excitement. Some of us are just benefiting from, like, I am 100% benefiting from it. Uh, whereas I very much was on the verge of hiring a second programmer to help me out for Lyas.

This is just a simple explanation with no code.

where suddenly your top performing developers are even more performance right now you have less need to hire more people. So we can see that the juniors especially I think are going to suffer a little bit of what, is there a benefit to hiring a junior when e even right now, like what AI can contribute is almost equivalent to a junior. You know, maybe, maybe even better. And I know nobody wants to hear that. I don't wanna hear that, but I think, I think it's true.

And I know nobody wants to hear that. I don't wanna hear that, but I think, I think it's true. So is the reality, not is the reality right now that no AI is not gonna take programming jobs, but you know, what's gonna happen is the programmers who leverage it really well are going to go to the top of the pack, and then everyone else, and especially newcomers and juniors might suffer. And I don't wanna say like, look, it is against, it is against my business model to suggest that, right?

And I don't wanna say like, look, it is against, it is against my business model to suggest that, right? Um, I focus on education. So as you can imagine, like this whole AI thing, I think about it all the time and what that means for a video-based education platform, um, when AI education is freely available, like how do we need to adapt? And then what does it mean in a couple years when it's like I can just ask AI to generate a video with some kind of, um,

And then what does it mean in a couple years when it's like I can just ask AI to generate a video with some kind of, um, fictional character that looks human, who will explain my exact predicament, my exact struggle in video. Like what happens at that point when I can generate a video tutorial on the phone? And it's, here's what's crazy. I'll, I'll, I'll give you an example, this kind of outside programming, but I think about this, I do this thing

I'll, I'll, I'll give you an example, this kind of outside programming, but I think about this, I do this thing with my children now when I put them to bed, uh, my daughter's eight and my son is six, where, um, they have like a bookshelf full of kids books. But what we've been doing lately is I will ask AI to write a five minute story starring them, and then I will, I will pull them and I'll say, okay, give me, give me three things that need to be in this story.

and I'll say, okay, give me, give me three things that need to be in this story. So like last night, my daughter said, um, there needs to be unicorns. Of course she said, unicorns, uh, my son said there should be a helicopter. And then, um, what else? And of course, my son is six. He said, poop needs to be in it. So we put poop helicopters and unicorns into the story, and it, it was great and they loved it.

So we put poop helicopters and unicorns into the story, and it, it was great and they loved it. So already we're at a point with children's stories where it's like AI can on the fly, generate a great story that they're probably going to enjoy quite a bit more than something, um, written. And then of course, it's five seconds to apply, uh, to split it in the book form and then generate images to go along with it. And then maybe I can upload photos of my children so

and then generate images to go along with it. And then maybe I can upload photos of my children so that illustrations of them are included. Well, like, this is not profess at all. I can do this right now. So in the future, it's just gonna become easier and easier and easier to do that. Um, now will that happen with like TV and movies too? Like, will I be able to say, I feel like watching a quiet space movie, like my kids make fun of me.

Like, will I be able to say, I feel like watching a quiet space movie, like my kids make fun of me. They always say, dad wants to watch Quiet Space movies, and it gets on my nerves because it's 100% true. When I break down what I like to watch, uh, a movie I really like is called Moon, and it's like the definition of Quiet Space movie. So it annoys me when they say that, but they're, they're not wrong. But yeah, could I say like, generates a quiet space movie

but they're, they're not wrong. But yeah, could I say like, generates a quiet space movie that is an hour and a half long, that deal, and it's like, it's gonna be able to do that. And at the start it's not gonna be very good, but then it's gonna get better and better. And then what happens when it's like, like, wait a minute, this is like legitimately as good as anything I will see at the theater. I don't know. It's not even farfetched at all.

as anything I will see at the theater. I don't know. It's not even farfetched at all. So like, I don't know how you cannot be wildly excited and extremely concerned at the same time. What do you think.