Open transcript
So all the cool kids are vibe coding. Are you? So it all started with Andre Carpathy. As far as I can tell, he kind of invented this term, definitely popularized it. He's saying there's this new kind of coding I call vibe coding. Basically, the idea is instead of writing the code yourself, you rely a lot on AI to write the code for you. You kind of go along with where it takes you. Ideally, you would even use kind of your voice to speak it into existence as opposed to typing it out in English. And some people took that to heart. Interestingly, like with most things online, we've kind of polarized to a couple of different extremes. On one side are people that are just like embracing this idea of vibe coding. On the other side, we have people that are kind of like aggressively against it. So today, let's take a look at what's happening. So really fast, this is Peter Levels. He's the creator of the FlyPeter app, that video game that you might have seen that's now making something like $87,000 a month from sponsorships. So here's kind of what that's looking like. So he's built this using vibe coding and basically having the AI write most of the code, create most of the graphics and basically pretty much do everything that you see in this game. You're able to fly around, shoot things. It's a multiplayer and it's kind of getting big. So recently he added the kind of the HUD, the cockpit view. So he said he wouldn't shot an actual working cockpit heads up display. So meaning this is something that he was able to do sort of in one shot with these models. So he's using 3.5 Sonnet Max, which is now in the cursor. By the way, I'm building a resource for all this stuff on my website. You can find the link down below. So it's absolutely free. I just want to kind of have like the big resource that has all the tutorials and all the different games that people are building. And just kind of like, if you're interested in this thing, this vibe coding phenomenon, then sort of there's going to be just one page that has everything on it that you need to get started. You might have seen the video where I use the Manus AI, the AI agent to research all these games that are being developed and create kind of this little website with some of the games listed as well as, you know, all the information that you would need to get started, kind of the AI tools that were used, the creation process and any resources and stuff like that. The creator sort of mentioned that they've used to develop it. That was just a quick test project. But since then, I've reached a literal army of agents researching the various games. So hopefully that'll help, you know, collect all the resources. By the way, if you have a game or a resource tutorial, whatever that kind of goes along with it, let me know in the comments. I'll try to add it to the resource page. So I made a little app with, I think it was Grok3, where you can kind of like hold your hand in front of the camera and it tracks all your fingers and you're able to play music and stuff like that. Somebody had a similar idea, but for flapping your wings, your arms, to play Flappy Bird. Brilliant. I'm sure it's great exercise. Okay. That's pretty cool. Now this new vibe coding thing is really taking off. So again, here's Peter Levels. So he's saying that he's organizing 2025 Vibe Coding Game Jam, where everyone is able to sort of submit their AI coded game or the game they coded with AI assistants. There's some requirements. We'll go over that towards the end, but he managed to get two very interesting people to participate as judges, or at least to provide some input. One of them is Andre Carpathy, the one who seemed like a start at this whole thing and started the ball rolling. And the other person he managed to get on board, shockingly, is John Carmack and the guy that, you know, invented Doom. All right. So he did Keen, Doom. He was part of Oculus. Next thing that he's trying to release is AGI. That's the next project that he's working on. He's got a unique approach to it and he might do it. He's done it before. If you know anything about the original Doom 1 and 2, kind of like that development, kind of how they went about doing that, it's a little bit mind blowing. To say that it's ahead of its time, that it was ahead of its time, I think is an understatement. So it does seem that John Carmack is showing some interest in this idea of Vibe Coding. Now, of course, there's a lot of blowback to this idea, a lot of sort of criticism. We'll get to that towards the end because I want to kind of highlight both sides of the story. But it seems like the game jam will be extended by another week. We'll see if it happens or not. Also, with John Carmack's feedback, Peter Levels decided to remove the multiplayer component of the game, which of course opens it up to a lot more people. It's a lot easier to create a single player game than it is a multiplayer game. So if you're interested, check it out. Again, I'll have a link down below that's going to have a resource hopefully with a lot of the stuff that you would need to get started. Definitely, if this is something that's interesting to you, get into it. It's kind of a worldwide movement to this little thing that's taking place that's got a lot of people very interested, knowing how to code and knowing how to do software development is not required. You don't need too much previous experience. You should be willing to learn and be flexible. Learning can be a little bit uncomfortable. That's totally fine. It might take you some time to figure stuff out. But keep in mind that a lot of people are figuring this stuff out kind of in real time alongside with you, you know, independently right now. So what sort of things are people creating for this Vibe Coding Jam? Well, here's Ian Curtis, first person game with some great physics. I gotta say, it's looking pretty good. He's also gonna have some tutorials on how to do this. I'll add those to the little resource I have at the bottom as well when that's available. And you are able to play this yourself, actually. And I gotta say, I mean, it feels smooth. It feels very, very good. The bounce off and yeah, the physics are phenomenal here. It does seem you're able to launch them into the stratosphere. And I don't think they come back down, but that's fine. And interestingly, we're also able to change the lighting and other aspects of it. Arm position and all sorts of stuff, interestingly. So you're able to add more and more customizability to the game, which is pretty interesting. So Min Choi put together a compilation, 10 wild examples. So here's Fortnite means Minecraft on 3JS. I mean, this is looking phenomenal. Looks like this one was Vibe Coded. They're saying it took 37 hours using Cursor. This one used WindSurf. Then we have a Catan-like video game. It looks like you get to use builders to collect resources, etc. Anybody remember Bomberman? He used to run around, set bombs everywhere that would explode in these sort of down these little alleyways. Somebody recreated that. By the way, this is Danny Lemonsetta. He's been putting together some pretty wild examples. This is a Hexen heretic kind of a similar game. It's looking pretty good. He has Mech and Cheese, which is a vampire survivors style game, I think. I believe, if I recall correctly, Elon Musk was very impressed with this one. So this one was coded using Grok. I believe he's using Grok for a lot of these. And it looks like Bolt New is also doing a similar hackathon with 1 million plus in prizes. It's hosted by Greg Eisenberg, who is on YouTube. So I don't know too much about this company, but this is definitely looking pretty cool. And so it looks like Peter Levels and Logan Kilpatrick from Google. So he's working on the Google AI Studio. He's on Twitter. He's on YouTube. Great follow. So it looks like both of them are going to be the official judges on this one. Here's another online first person shooter game built with Vibe coding. It looks like he'd spent two days and about 20 hours on it. 3D Pokemon with the Replit Agent. So I'm seeing more and more people use that for different things. It looks like this Replit Agent is now able to also build 3D models within it. So I gotta definitely take a look at that. This is another entry into the Vibe coding jam by Levels.io. So an 80s retro futuristic endless drive game. AK on Twitter. Great follow for AI science papers and stuff like that. So he has a little tutorial on how to Vibe code games, deploy them for free entirely on Hugging Face. So I was thinking about creating something for this competition. I unfortunately did not have too much time to do it. I made a few 3D games as sort of a quick test projects, but my thing was always those like idle incremental games where you kind of let it run itself. So I've been building something like this where you're able to progress, fight various different things, level up your skills, level up your abilities, etc. And it's been going pretty well. This might not be the thing that everyone's interested in, but you know what? If they do end up extending the Vibe coding session, if they do end up extending the Vibe coding session in seven days, I'll probably finish this thing up and submit it. I went through a number of different variations. I started out trying to do a party of five battling against monsters, but that became a little bit unmanageable. I used Mid Journey to create some of these characters that I needed. And I gotta say, I think the Mid Journey is definitely looking really good for that. I used Cursor to do a lot of it. And I would say it took me maybe two and a half hours of actual, you know, in front of the screen presence to basically outline most of the starting game mechanics and kind of how I want the game to progress. And I was able to do it without touching the code in any way, shape or form. So there's no code involvement on my part. It was all through natural language. I did run into a few issues where Claude decided to just like rewrite everything for some reason, which kind of broke the game and it couldn't fix it. So you got to make sure that you're able to revert to a previous checkpoint. Cursor has some built-in checkpoints, but using something like GitHub is probably a better idea. So something to think about before you get started. But if you're just starting out, I wouldn't even worry about any of that. I would just try to see how far you can get. If something gets messed up, see if you can revert it. If not, it's not a big deal. This is all part of the learning process. If you've never coded before, have never done any of this before, and you just wanted to kind of get your toes wet, probably the easiest way to start is to just go to Claude.ai. So that's Anthropics Claude, one of the main competitors to OpenAI. And it's very, very good at coding. Ideally, you have the paid version, but you don't need it. You select Claude 3.7 Sonnet with the extended mode on if you have that. And I just typed in BuildMeA3D racing game in the browser as an artifact called Claude Cart. So the artifacts, usually, you know, when it gives you the code, you need to take that code and then be able to run it somewhere. With these built-in artifacts, as they're called, you're able to just see the game within this browser window in Claude.ai. So if you just click on this, it kind of opens up right there. And there's our little drifting racing game. So I'm able to go around the track, use the spacebar to drift. And once I have that, so as you can see here, pushing the spacebar like really makes it spin out of control. So it might be a good idea to iterate on this game a little bit more, which I did saying make the drifting more gradual, maybe just a quarter of what it does now. So it updated it like this. So here's kind of the updated version. And now whenever you hit spacebar, the whole game crashes, right? And you get a little error message. So you just type in that error message into Claude and it troubleshoots it. It fixes it. And we get our final version, which is this. Now you're able to drift around corners like a pro. And I got to say, this feels very good. So at this point, I would say this is an example of a simple game that was completed in maybe 10, 20 minutes. It didn't one-shot it. I had to troubleshoot some errors. Oh, somewhere along the way, I also said add some smoke effects and stuff like that to make it a little bit more realistic. But literally, if you wanted to try this out, just go to Claude.ai, make sure you have, you know, Claude 3.7 Sonnet. If you have it, if not, you can use 3.5. Extend it is better for coding. Type in your prompt in here. Hit go. Now, of course, this Vibe coding approach is catching some flack. Some people think this isn't a serious way to approach code development, software development. A lot of people kind of dismiss it, make fun of it, saying that this is basically people that don't know how to code are trying to do it, but they can't really get to, like, the really advanced, sophisticated stuff. Plus, it kind of lets people kind of create stuff without necessarily having the skills that are needed to create high-quality stuff. Peter Levels got into a fight with one of the very well-known game developers. Some heated words were exchanged. So there's a bit of a debate going on. Both sides have some good points. I don't know who's wrong and right, but the one thing that this reminds me of is kind of back in the 90s, you saw the emergence of the digital camera that was good enough to be used by kind of regular consumers, and it allowed a lot of new kind of entrance into, like, the photography field. Anybody could buy a camera, point and shoot, capture images very easily without knowing too much about how film operates, how cameras operate. They didn't know how develop film, select film. They, in a lot of ways, were clueless as to what real photographers did or kind of, like, they didn't have those skills. And a lot of the professional photographers, with a lot of expertise, they kind of looked down on the idea of the digital camera, saying it was somewhat useless for them, right? So they were able to select the right film for the job, to develop it properly, to create the right exposure, etc., to really capture the exact sort of visuals that they needed. Their high-level abilities allowed them to do much better and high-quality work than the person that just bought a digital camera and didn't really know too much about photography. A lot of them didn't really see the point in digital cameras. So kind of 1999 or somewhere around there, maybe it was 1997, that was kind of like the peak of the sales of film cameras, of film. Kodak Gold, I think, was the number one selling film at the time. And I'm sure at the time, a lot of the sort of the professional people, people with a lot of skills and expertise in the area, they probably scoffed at some of the digital cameras, right? Because they knew how to select the right ISO, how to set the right aperture, what shutter speed they need, what exposure settings they needed to capture, like, the best quality photography possible. Whereas the digital cameras more and more allow people to just like push one button, like if you can point the camera in the right direction and then push the button, then you would be able to capture the image that you were looking at. And yeah, maybe it wasn't as good, at least initially, but as cameras improved over time, as that technology improved over time, that gap started to close. And over time, the sale of film and film cameras just cratered. By 2005, 2006, it created something like 95% or 99% of what it used to be in 1997. And the digital camera allowed a lot of new people to come into the field and just start taking pictures without needing to spend a lot of time learning and studying and figuring out how to do it. You could just get in there and start shooting. To me, it seems like this idea of vibe coding and various AI-assisted tools for coding, we might see kind of the same thing play out here. We're going to see tons of new people enter the field and try to code things up, to try to create software and games and whatever else they can think of. And of course, the people with lots of experience, lots of expertise in the field, they might not understand the appeal at first. They might say, oh, these people that are entering the field, they're noobs, they don't know what they're doing, etc. And of course, they will be right. Now, all these people that are trying to make games with vibe coding, they don't have nearly the experience that a professional software developer has. If you search for vibe coding, there's tons of people dunking on it saying, oh, we're leaking the API keys to the public. People are charging $1,000 an hour to fix your vibe coded mess. Here's a picture of two founders vibe coding the product. You know, here's somebody looking at the code written by Peter Levels saying, I've never felt more secure in my job prospects as a senior engineer. When asked to explain, actually explains it, which I'm sure a lot of people definitely appreciate this. He's saying, game networking is hard, hard, dark, and full of terrors, or horrors in this case. Peter responds, you sound like the typical gatekeeper guy who feels threatened for his future. You're not wrong for now, but with this mindset, you're going to get turbo wrecked in the next decade, 100%. People are saying, Peter, don't be all angry. Peter's point is that some people are doing gatekeeping, and that coding is for anyone and everyone. He's scared he's going to be unemployed, and with the attitude, he will be. Interestingly, Jonathan Blow, so he's a well-known, very well-respected American video game designer and programmer. He's jumping in and saying, it's guys like you making claims like this that are damaging to the reputation of AI. Is it pretty amazing that AI can generate this? Yes. Is it anything close to the quality of what you need to build for an actual playable game? No, it's not within 10 ballparks of the target. He's saying, if you've never tried to make a game, I'm sure it's fun to have a game-making experience. For those who have made games, even just simple ones, getting stuff on the screen is not impressive. It's easy. It's making the game good that is hard. Peter Levels is responding, and he agrees, but this applies literally to every art form or industry. The point is that AI lets me explore entirely new areas of creativity, which is really cool. So Jonathan Blow continues saying that no one's saying it's bad for someone to exercise their creativity and play with AI to make something they wouldn't otherwise be able to make. They're saying that this is very, very far from what an actual game programmer would make, quality-wise. It's not gatekeeping. It's understanding that this is actually a very deep discipline where a lot of skill is required to make something good. Anyways, they kind of go back and forth. And again, this is one of those things that happens on Twitter and elsewhere online where people kind of project some stuff on each other and kind of go back and forth, and maybe some of the finer points of the discussion are lost. Whatever the case is, I think the point is that the experienced developers, they do have a strong point. AI currently is not at a level that it could replace somebody with a lot of experience. Somebody that's spent a lot of time figuring out how to do this stuff, etc. Right? So they're right in a lot of the stuff that they're saying, but also this idea of going, looking at this, you know, flying game that Peter Levels created and going, oh, like my job is safe as a developer. I think some people are going to find those kind of statements maybe a little bit weird. Because again, like Peter Levels says, it's like, yeah, I started three days ago. It's like if somebody's building some sandcastle on the beach and a highly decorated architect walks over and is like, you fools, this is a very rookie buildings you have here. You know, I feel secure in my role as an architect. I'll never be able to approve this zoning. This will be kicked out by City Hall. Where are your load-bearing walls? This is nonsense. Right? It would be a little bit weird. Because again, no one's presenting these little games as sort of the next level software development projects that are supposed to put all the engineers out of business. Keep in mind that those 20-year-old digital cameras, they took blurry photos. The color would be off. If you didn't have the right lighting, it would look horrible. You needed the flashes to make it work. You had really grainy images, et cetera. If you didn't focus on the right thing, everything would be blurry. And again, the professionals would be able to use their settings ISO, shutter speed, aperture, et cetera. They would be able to select the right film. They would know how to develop it in those little chemical baths, et cetera. So I don't have a crystal ball to see the future. I don't know what's going to happen with vibe coding, AI coding, et cetera. But I think we've seen this thing play out in the past with certain results. And I think the digital cameras replacing a film could be a sort of a model through which to kind of view this thing. So digital cameras were at first not as good as the professionals using film. But over time, they replaced film. They didn't replace professionals. I think we're going to see the same exact thing play out with AI coding. At first, it's not going to be as good as a professional that's doing everything themselves and has a lot of expertise doing that. But over time, AI is going to make a big dent in this idea of typing out code letter by letter. That does not mean that the professional software developers will cease to exist. They will stay with the times. They will learn how to use these AI tools. And it's going to allow them to produce bigger and better stuff to 10 extra output. As OpenAI's chief product officer, Kevin Wheel, said, you know, when he's working on some engineering project, there's times where you really need to focus and do it yourself and make sure everything's right. But there are times when you just need to do something that's fairly straightforward where you need to just type a lot. In those cases, it's nice to have the AI kind of automate that for you so you can focus on the higher level tasks. The second thing that digital cameras did is they removed the skill barrier to entry. You no longer needed to know how to do 10 million different things to take one photo. You just needed to be able to point the camera and press a button. So digital cameras and then the digital cameras we have on our phones just exploded the amount of pictures that were being taken from maybe hundreds of millions or whatever it was to then approaching a trillion by 2007. By the way, think about who takes the most picture, like what demographics sort of just outputs the most pictures. It's probably not the professionals. It's probably not the people who do that for a living. It's younger kids, young kids, high school, college, etc. take a lot of pictures on their phones and they send it to each other. They share it. It's basically a form of communication. They have more time on their hands and they use a portion of that time to take pictures. I think this is the first time I saw this idea kind of mentioned in this post. I don't recall if this is the first person that I've heard mention this. The idea is that most code in the future will be written or generated by the time rich, right? So most code will be written by kids slash students rather than software engineers. This is the same trend as a video, photos, and other social media stuff. So here's Niki De Beers saying basically, you know, if you've been following this explosion of vibe coding stuff, he's saying we're close to somebody building a creation ecosystem or app store where anyone can be a game developer. The only gap that needs to be closed for this ecosystem to exist is abstracting away the remaining pieces that are too complicated for the layman. Authentication, server networking, and deployment. Once that's achieved, the only thing left for a creator to do is dream up their concept and click publish, right? So right now, there's this kind of debate that's going on, right? So a lot of the software engineers, people in the field are saying, well, software development is really hard and it requires deep knowledge and skill and vibe coding isn't there yet, right? And so kind of the opposing side is vibe coding allows anyone to create software. And so people are going back and forth and yelling at each other who's right, who's wrong. The reality is those are both true statements. Imagine switching software development for any other thing, like for example, writing. The people that spend 10, 20 years mastering the art of writing are extremely, extremely good at it. But anyone can write anything and post it on Twitter or elsewhere, create a blog, whatever. So the barrier to writing and publishing has been removed. That doesn't mean that there's no more gap in talent or skill, expertise, etc. It just means more people are allowed to participate and practice and get better. Same thing with photos, right? A three-year-old can now take a pretty good photo, right? If they're able to hold the camera phone steady and click that button, it might come out really, really well. There are still people doing photos, you know, for a profession. We've seen the same thing play out over the last couple decades, right? With, for example, music used to be needed a recording studio. And there was a moat, if you wanted to be able to produce and record music. Now anyone can publish it online. We did same thing with photos and video and writing. And now that same wave is coming for coding, for software development. And just like there are still professional musicians and professional writers, there's still people that do talk shows, that are professional entertainers, that, you know, that do interviews, even though anyone can have a podcast. And it seems like everyone does. And now that same change is coming to software development, to coding. So we can kind of predict what happens based on the fact that we've kind of seen this happen before. Number one is that great talent is still valuable. If you're a great musician, your work is still valuable. If you're a great writer, your work is still valuable. If you're a great software engineer, your work will always still be valuable. The fact that there's less barriers to entry also means that we're able to develop more talent. We're able to find more talent in the music industry and in entertainment and in writing because the entire world is now able to participate, not just the people that have close connections to the industry, etc. It also means that the sort of the mediocre people, the people that maybe aren't as talented or aren't as hardworking, they're not as good at the thing. Yeah, they suffer. They make less money. And we also know that the product, the actual output becomes less expensive because there's more of it. In the past, when only a few people could write and read books, books were very expensive. Now we have a multitude of books that are available for free online, various texts produced by various people in the world. There's still books that cost money, but there's a lot more that are free and cheap. And in general, the cost of putting text in a book is less expensive. There's more free music. There's more free entertainment. The cost of the product, how much it costs to acquire it, decreases. Right now, Adobe, for example, they've earned looks like 5.7 billion in February 2025 doing software as a service. They get paid for having products like Photoshop and Adobe Premiere, etc. And certainly there's probably people and organizations that use all the features and all those products and are like heavy users for a lot of us, myself included. I probably use 5% of all the things in Photoshop. Photoshop has a lot of features. I simply don't need most of them. I just need a few that are really good. And Photoshop does a great job of handling that. Same thing with Adobe Premiere and others. But if I could custom code, create my own sort of artisanal software specifically for my needs, I probably would no longer need to pay Adobe for their services. The other day, for the first time in my life, I needed like a little piece of software that did this one thing. It was pretty simple. So I went online and opened up Google because I was about to type in like, I need software that does this thing for this particular use case. It doesn't really matter what it is. I just needed it to create some visualizations for me that I could use, which is how I would normally find some software, go to Google or Reddit or somewhere like that. And more recently, probably use ChatGPT or something like that to do a web search. But I stopped myself and instead I went to Claude and I told it to just make that software for me. 30 seconds later, I had that software running and it did exactly what I wanted it to do. Now, it wasn't super complicated, but it was custom. I could change it how I needed it. It was exactly what I needed and took something like 30 seconds to create. Probably faster than it would take me to do research and then download the software from the internet if I found the correct software that I needed and then go through the installation process. And they'd probably make you register and do all sorts of stuff. Instead, it was one prompt in Claude. I took the code, I ran it, and I was done. So to all the people that are in the software development industry that are saying that AI vibe coding is not as good as actual coding and that their job is safe, etc., and that a high level skill is needed to create high level, good quality software. First of all, I agree with everything you say 100%. Great engineering talent is in demand, will continue to be in demand. The skills that people developed as a result of years of training and learning how to do stuff, that's not about to disappear. It's very valuable. But also at the same time, and it's not mutually exclusive, is the fact that this AI coding, this vibe coding wave, it's coming. And it's going to be huge. Again, why did the number of pictures taken went from some, you know, fairly low number into the trillions at a rather rapid pace? Is it because all of a sudden more people needed to take pictures to survive? No, just the technology for the everyday consumer with little to no skills, their ability to take good photos started improving. They didn't need any extra equipment. They didn't need to go and buy a camera or carry a camera with them. It was always on them. It was always on their phone, in their pocket. And so whenever a thought occurred that, hey, it would be nice to take a picture of this, they pulled it out and took a picture. So what happens when it's just as easy to create a quick little software to do the thing that you want to do? I had the Manus AI agent, which is running on Claude 3.5, we believe. I don't think it's on 3.7 yet. I had one shot creating two reinforcement pipeline training projects using something like PyTorch to actually train these two little agents to play a snake game and use different approaches to train them to see which one worked better. It did it in one shot. It took it maybe five minutes, but I didn't have to sit there and wait for it. I told it, hey, build me this thing. And it went off on its own computer and its own virtual machine, took 10, 15 minutes, however long it was. It coded up the snake game, created a bunch of little obstacles, created two little agents, created the reinforcement learning environment, ran through, I think it was 500 sort of iterations each, 500 instances to train the agents, and then compared the outputs, wrote a below report, and then told me what the result was. I think one of the training approaches was, I think, like 36% more effective. And so think about that. That was from me saying one or two sentences and then just letting it go off and do its thing. Imagine having an agent like that on your phone that you're able to talk to on your hands-free wireless speaker or whatever that just sits there and codes up stuff for you, tests it, et cetera. And anyone everywhere can do it on their phone just by speaking basic English to it or whatever language they prefer, just natural language. What's going to happen to the amount of code that's produced on this planet, right? So we're not talking about necessarily like the code that we need to protect the national interests of the countries or whatever. We're not talking about enterprise grade software solutions. We're not talking about that. We're talking about the million different things that you can think of if you're able to have something code, something for you and apply some sort of an AI, some neural net, maybe to help it like run or operate. You can probably do it for your thermostat, for your microwave, for your car, just sort through your email, like just anything you can think of all of a sudden can be improved, automated, made it to be more intelligent just by speaking a few words. I think in that case, what happens is the amount of code produced will probably be steeper than what you're seeing on this line, where it basically takes off into the stratosphere. So whatever else you do, if you haven't made a stupid little game with assistance of an AI yet, don't sleep on this. Do it. It's not hard. I think you'll be shocked how easy it is. And if you have, then keep doing it. Keep practicing it. You don't have to do it, you know, 18 hours a day, but make sure maybe every week, spend a few hours keeping up with the latest developments. Because again, it's going to be a skill like any other skill. And again, over time, people will come up with solutions to take care of some of the harder tasks like deploying the software. There's still already, you know, solutions for that that are fairly simple, but it's just going to get more and more simple for all the new entrants into the field. Like it's going to be like a one-click solution at some point. It's going to take care of the security, the hosting, the whatever else you need. Basically think about how hard right now you have to think about taking an image, developing that image and storing it, right? How much thought goes into that? It's one-click, you point, click. Most phones actually take multiple images. A lot of the images that we see are composite images. They're not one image. They're sort of an image with different focus on different parts of the image that is made into a composite that we see, but it's actually not one image, right? They clear up your skin and make sure that the lighting looks good and whatever. And then save it into your phone. It's uploaded to the cloud. You're never going to lose it. Like there's a lot of complicated stuff that happens, but for you, it's just, can you point and click? If you can do that, you can make a great image that you're not going to lose. It really does seem like very soon we're going to have software development, at least again, for the simpler stuff. We're not talking about the super high-end software development. That's still going to be for the professionals, but there's a billion use cases for little simple things like this, whether that's games or little home automations or whatever. There's stuff that we haven't even thought of that's going to be extremely, extremely useful if you can just sort of conjure it into existence by saying a few sentences. But let me know what you think. Do you still think that it's one way or the other? It's either software engineering is hard and AI is bad, or AI is going to like wipe out software engineers. Do you still think it's one or the other, or do you think it's both? Good software engineers will still be good software engineers. And there's an AI wave coming that's going to completely change everything. I think those are both true statements. So, but let me know what you think. Oh, and check out the link. So I'll be working on that, on all the resources to get started. It's free, no opt-in required, although you can choose to receive the newsletter if you want. Make sure you're subscribed. Big things are coming. If you made it this far, thank you so much for watching. My name is Wes Roth, and I'll see you next time.
