FINALLY Good AI News From Google Boss - Engineering Manager AMA

All right, we're kicking things off here. Let's see if this stuff works. There's always a fun surprise every time I stream with Reream, but hang tight. Okay, looks like that's working. We'll get on over to Instagram. We'll go live. Nice. Okay, I think I think it's working and Substack actually has proper support now on Reream, which is super cool. So, finally Substack folks won't have to watch me from the side. Um, hello on Instagram. Good to see you. Let me get my chat refreshed on my side as well. Hello. Hello. We're just getting started here. Um, okay. I think we're ready to rock and roll. So, welcome to the stream. We do these every Monday, 700 p.m. Pacific. This is an AMA style stream if you haven't been here before. And so, if you have questions, leave them in the chat. Happy to kind of drop whatever I'm talking about to go answer your questions on software engineering, career development, that kind of stuff. Hey Devin, good to see you. YouTube is good to go. Thanks for letting me know. Um, but yeah, like Substack, I I said this on the last stream like minutes before I was like, "Wait a second, they have support." Um, and I didn't want to risk it right before. Um, so we're trying it out. It seems like it's streaming. That's what Reream is telling me, but today's new surprise is that it looks like Tik Tok revoked my live streaming ability. I don't know why. It's just like streaming's not allowed. Like, press here to apply again. And then when I press it, Tik Tok's like, "You're not allowed to stream." Like, "Okay." Um, so that's cool. Kind of dumb, but uh it is what it is. I find like Tik Tok's a ridiculously difficult platform for me. Um, I don't know. I'm I'm assuming it's because my Tik Tok videos, like my short form videos in general, aren't like super super exciting. It's like a quick quick educational topic in a short. So, uh, probably people don't love that kind of thing, but it is what it is. I'm not going to go making skits and stuff like that because I have other things to do. But today's live stream topic is going to be on the Lex Freedman and Sundar Bachai interview. And I talked about this on Code Commute. I talked about this on Dev Leader. And then I wrote my newsletter article on it. Um, I thought it was a super cool interview. I'm not sure people have seen it. I also messed up like moments before pressing uh to stream. I was like maybe I should see if I can get my computer audio working uh through the stream and I think I probably could. So like maybe we'll try it but also maybe not. Um I don't know. I'm a little nervous. I know. So Devon's on the stream. So, if Devon is willing to test it out, then maybe I will uh if he lets me know and I trust Devon, so maybe I can try giving it a go. But I what I would like to do is play the clip and we can all watch it together because if you haven't seen it, I think it's a really good clip because a lot of well I'm sure you've connected to the internet in the last uh year and everything that you see on the internet's basically software engineers are going to be replaced by AI. Everyone's freaking out about AI, especially if you're a developer like you know there goes your career. I've had people talk to me on social media that are like I'm dropping out of college um because they're like I'm not gonna get a job because of AI. Like this is it's getting wild. And uh Sundara Pachai in his interview with Lex Freriedman kind of gave what I feel like is a different perspective than every other tech leader is is uh is saying. And to be clear, I'm not I'm not suggesting that every tech leader that I've heard of like um like SA or um my goodness Zuckerberg like I don't hear them saying like you know developers your days are numbered but they do talk about like uh and we we'll get into this like the metrics for AI that's uh writing code right versus human. So that's coming up a lot, but this is the first time I've heard like a a tech exec saying like, "Nope, we're hiring more." Like I don't like I don't think it's going to replace developers kind of thing. So maybe we'll give it a go and I'll try to flip over to it. Um, apologies for not like prepping in advance to do this, but give me one sec. I have to turn on my desktop audio. And what I don't know, I don't think my desktop audio is picking up my own microphone, so that should be okay. Um, I have to assign some channels here. You guys can't see what I'm doing, fortunately, because I barely know. I don't touch this stuff because every time I touch my audio setup, something stops working and we don't want that. So, let me go ahead and change that. I think that should work. Now, what I'm going to do is go over to YouTube in another tab and then if I search for Lex. So, we'll just do a quick little test here. I think it's this one. I have to ask you on the programming front that I I just stopped it but because it's going to be a little bit delayed but is getting really good at programming. Looks like Gemini both the Agentic and just the LLM has been incredible. So a lot of worried that they're I'm going to stop it. I'll see if Devon's still on the stream or someone can respond and say they could hear the audio. I hate this stupid chat delay. I wasn't showing it, by the way, but sounds good. Okay, cool. Rock and roll. Let me get this pulled up. And what I'm going to do is I will put the the link in the chat so you folks can check it out after, too. Um, but let me get flipped over here. We'll go full screen. And I guess I kind of want to hear it, too. How am I going to do this? I don't want it to come through my mic. So I have to ask you on the programming front. Uh AI is getting really good at programming. Gemini, both the Agentic and just the LLM has been incredible. So a lot of programmers are really worried that their jobs, they will lose their jobs. Uh how worried should they be and how should they adjust so they can be thriving in this new world where more and more code is written by AI? I think a few things. looking at Google um you know we've given various stats around like you know 30% of uh code now uses like AI generated suggestions or whatever it is but the most important metric and we carefully measure it is like how much has our engineering velocity increased as a company due to AI right and it's like tough to measure and we kind of rigorously try to measure And our estimates are that number is now at 10%. Right? Now across the company we've accomplished a 10% engineering velocity increase using AI but we plan to hire engineers more engineers next year right so you because the opportunity space of what we can do is expanding too right and so you guys heard that right just want to pause it there because that's an important part. Um, they plan to hire more engineers next year, right? So, I'll keep going, but I just wanted to make sure I paused and repeated it for those of you that blinked and missed it. I think hopefully you know for at least the near to midterm for many engineers it frees up more and more of the you know even in engineering and coding there are aspects which are so much fun you're designing you're architecting you're solving a problem there's a lot of grunt work you know which all goes hand in hand but it hopefully takes a lot of that away makes it even more fun to code frees you up more time to create problem solve brainstorm with your fellow colleagues and so on right so that's that's the opportunity there and second I think like you know it'll attract it'll put the creative power in more people's hands which means people create more that means there'll be more engineers doing more things so it's tough to fully predict but you know I I think in general in this moment it feels feels like you know you know people uh adopt these tools and be better programmers like there are more people playing chess now than ever before right so uh you know it feels positive that way to me at least speaking from within a Google context uh is how I would you know talk to them about it I still I okay so pause there because he's going to go into like a bit of a uh he'll change this subject a little bit But uh I already see some comments in. Uh Ryan says, "I mean, you can say that now, but uh that to me is more of a PR statement. We're still 6 months away from 2026." Not saying he's wrong, but just saying. Uh Ramsay says they're already hiring. Um yeah, I mean it's like I don't I don't I haven't looked at Google job openings and all that. I haven't done the research, but um it's like I feel like, you know, even Microsoft has done layoffs and Microsoft is still hiring. Um, I haven't looked at like uh Amazon. I know Facebook is hiring. I've seen lots of people on my my LinkedIn posting like engineering management roles, software engineering man uh software engineering roles. Um Anthony says programming and chess go hand in hand. I would love to hear more about that. That sounds interesting. But we'll keep going. Uh by the way, like folks, jump in the chat if you have thoughts on on what's going on here. being able to allow anyone to just be able to code uh code code sites and uh and all with AI scar just scares the crap out of me. Um yeah, it's it's interesting, right? Like to to Ryan's point, I I think like I really think that it's cool that and I mean this genuinely, right? I've said this on Code Commute videos. I've said it in social media posts, but I really do like the fact that that using AI tools like whatever it is, right? Whether it's Chat GBT or um some of these like builder sites and services that that allow people to that don't know how to code to make something. I think that's super cool. But there's always a butt. the the challenge is like when people are confused by it when they go, "Oh, well that means I can just, you know, push this to production and it's it's bulletproof." And the reality is like it's absolutely not. Um, it's really not. But I think that it's so cool that it can lower the barrier for people that were like because I've been doing this for a while, right? And I don't just mean like software engineering. I mean, one of the things that I enjoy doing, let me just go back to my my own face so I can so I can rant here and talk with my hands. But one of the things that I've been doing for a long time is in my career, I've been trying to help other people become software developers. So, I have talked to many people whether it's online, whether it's, you know, people that I worked with that weren't in a software developer role that want to get into it. Whether it's like people I used to go to, you know, high school with and stuff where people are like, "Hey, I think I want to do this. I want to try." But then they talk themselves out of it because and this is before AI. Like I'm talk I've been doing this for like over a decade trying to help people. And the common thing is like I'm just not smart enough so I don't I'm not going to try. And I'm like you you will never know. You will never know if you even like it if you don't try. Cuz that's the other thing. You might try it. Maybe. Maybe you try it and you're good at it and you hate it. I don't know. Like these things are possible. I used to be good at math and I didn't enjoy doing math. So, it's in the realm of possibility. But the point is, I've talked with so many people over the last decade plus that will literally talk themselves out of it. Won't even start. Nothing to do with AI. Just the fact that they don't think that they're good enough. They don't think that they're smart. They don't think that they have some prerequisite. They never get started. And I really like the fact that people can use tools to at least go, "Holy crap, like I made something." Well, and if it's AI made it, whatever. But they they they went from having an idea to having code or a something that's running. I love that. But like I said, the the caveat here is that it gets dangerous when people go, "Oh, therefore, like either you jump to ridiculous conclusions like therefore we don't need software developers or like haha look what I made in, you know, a 20 minute session versus what you've been building for a year." And it's like these are very different things. One of these you might be able to go have users and that pay money for a service. The other one is probably about to fall over at any given moment. But I do really like the idea that it kind of lowers the barrier, so to speak, and gets people at least interested. I realize for some people, I've seen this even online part of this conversation where people go, "Well, I don't like that because it means there's going to be more competition." And it's like, I don't know what to tell you. Like, then forget the AI thing. You could have been mad at me for trying to help more people get into software development. There's like there's always going to be more competition. Instead of getting mad about the competition, like focus on the thing that you can control. I'm just going to read some of the comments here. Um being able to allow anyone to just be okay, we read that uh normal hiring and layoff cycle of big tech is more worrying than AI. Yeah, I don't like that. Um you know, I it's kind of weird for me to talk about because I am at Microsoft there have been layoffs. I like read online there's going to be even more layoffs. like it's specifically with like sales and marketing I guess but like I don't know I don't I don't like the idea of layoffs. Uh I responded to someone on social media somewhere but for the context of layoffs because I do work at Microsoft this question comes up where people are like well what's going on? Right? I've had people after rounds of layoffs people message me on LinkedIn and they're like hey like are you okay? Um and I can't like Microsoft is a huge company. I can't speak for you know, most parts of Microsoft I cannot speak for at all. I work in a part of Microsoft that's called Substrate or within Office 365. It's the platform that the other services can build on top of and within our organization like I don't I don't have the details for how how layoffs occur, right? That's way above me. What I have heard in some context is that when it comes time for a decision where layoffs are happening mo like at least from the times I've heard of it, it is not that the CEO goes guess what folks time to lay off. And then he just like says boom boom boom that's happening. My understanding of how this works is that there is some communication to the different orgs within Microsoft or this probably works this way different companies and there is like the ask that you need to be able to have your your profit center more profitable. So you can do that in two ways at a high level. One is you make more money and the other one is you lower your costs. And my understanding is that it's up to and I don't know what sort of level in the org. It's up to the business leaders of that org to go how are we going to make that happen? So it's actually independent. Now in recent layoffs in certain situations I don't know if there were mandates where they were like you know this group of people they're are targeted and they're getting exited. I don't know but my understanding is that it works that way that the business leaders go make that decision about how to go organize that. Now with that said I have also heard from leaders within my organization that they are trying to use creative ways such that we do not have to have layoffs within our organization. And the other thing to add a caveat to there is that uh and I've heard one of the leaders in our org say this when they go listen when people talk about layoffs it's basically impossible for me this is the leader saying this to say that there are going to be zero layoffs and they said that's the case because there is always like you know not mass layoffs but there's always situations where regardless of these big sweeping layoffs where there's there's individuals getting laid off like so to to say like there will be zero. They're like I can't I can't say that because it would be a lie. But they said that within our org, they look for creative ways to make sure that that doesn't have to happen. So in some cases, going back to the uh sort of the the cutting costs versus increasing revenue, it's like okay, like we can slow down our costs if we're not hiring as much, right? That's one thing when you're forecasting if you are not growing. So if there's supposed to be an aotment for how much you can, it's like, well, if we dial that back. So different creative strategies. I realize that doesn't sound very creative, but different strategies that they're employing to try and reduce, at least for our organization, the potential impact of that. So anyway, all of that to say, I hate the idea of layoffs. Um, it's it's a mess, right? Even even more ridiculous is when you see people getting laid off, right? So, business units getting chopped completely out and then it's like you also see job openings at the same company. And I've lived, not I'm part of the layoff, but I've lived through periods of people being laid off and there were job openings and literally was in a conversation that was like a panel with one of our leaders in our org and someone was like, "Hi, just curious like why aren't people moved over? If we have open positions, why wouldn't we just move them?" And I don't know the to this day I don't know the reason why that doesn't happen but I assume that given the structure and size of the company that it's basically I want to say it's infeasible even though in my head I'm like it's got to be very possible. Um but even the leader in that conversation was saying like you know if you if you have names and stuff like absolutely get their resumes let's try to hire them back in. I just don't know why it wouldn't be an internal transfer. So all that like Devon totally agreed the the cycles for this kind of stuff I think are ridiculous. Um I'll just say it that way, right? Um Ryan says, "Until AI is uh truly aware of the security, making security number one priority, then it just scares me how many insecure." Yeah, totally agreed. Get some messy stuff there. Hey Andreas, thanks for joining the stream. Um yeah, and Ryan's just kind of repeating around the barrier being lowered in security uh just not not being a thing, which is wild because like why like why isn't that a focus? Like is it just that AI was trained on so many things like the the overwhelming majority of stuff is not secure. So that's what we got probably. Um Devin says there's been a a ton of small two to three user purely internal apps that uh I never got to that would have been great candidates for vibe coding. Yeah, I think that's um like I I think I I don't did I I think I shared it on the stream. I didn't get to it on this past weekend because I'm just falling behind on stuff. But uh like I vibecoded the codecute.com site and then I made YouTube videos for it. I I should mention this. Um my my editor is currently um not MIA is not the right word. He's uh a little bit removed right this moment because he's got some family stuff to take care of. So he gave me a heads up. But so I didn't have YouTube videos that went out like last week. I think I had one on Monday. Uh I have a bunch of videos that have been sent to him to edit, but I haven't published any because he's I'm just I'm not going to distract him. He's got some family stuff and I will keep making them, but they're just not getting edited and I don't have time to edit them. So, sorry for that. But, um, Technav, welcome. Good to see you back. Vibe code the world. Um, yeah, one of these things is not like the other. Uh, Hamza says, "A no different from the rise of other big 10 companies like those in Fang. If you think back when those companies started growing, they created a wave of new jobs. That trend has continued ever since. Therefore, more and more jobs uh will still be there. You just need to adapt what's coming and attitude into your tools. I to like HMS, I totally agree with that. Um yeah, I I think that makes sense. Deon says, "You already did lead code to get your position, but how does the other internal team know that you still know how to balance a binary tree?" It's true. That's why they got to rehire them. They're like, "You know what? It's been too long since you um you balanced a binary tree, right? You reversed a link list." So, um yeah. Cool. And then Ryan says, "Yeah, it's a good on the security thing for AI. Love to hear feedback from any of the GitHub copilot, Claude, Gemini, LM teams. Why security? Isn't a focus on their coding focused AI agents?" Yeah, honestly, I just feel like it's probably the way it's been trained. But I mean, these are companies that are building I'm assuming building more tailored models given that they're going to have tools around this stuff. So like why wouldn't they just like keep reinforcing that? I don't know. And maybe they are. Maybe it's just because we're so early and we're all kind of going through this being like, "What the heck?" Anyway, let's go back to this video. We'll keep keep it going. I just know anecdotally a lot of great programmers are generating a lot of code. So, their productivity, they're not always using all the code just, you know, there's still a lot of editing, but like even for me, it's still programming as a side thing. I think I'm like 5x more productive. I don't I I think that by the way I don't know how people arrive at these ridiculous numbers. It's like 10x developer. He he says I'm at least five times more productive. I don't know, man. To be honest, I feel like people don't understand how math works because that's like absolutely insane. Like I use AI a lot and to even say that I'm twice as productive would be like I think a pretty ridiculous claim. Google overall like their CEO just said 10% more productive. I mean it's a big organization of course but like five times. Okay. Let's keep going though. Sorry. That's uh even for a large code base that's touching a lot of users like Google's does. I'm imagining like very soon that productivity should be going up even more. The big unlock will be as we make the agent capabilities much more robust, right? I think that's what unlocks that next big wave. I think the 10% is like a massive number like you know if tomorrow like I showed up and said like you can improve like a large organization's productivity by 10%. When you have tens of thousands of engineers that's a phenomenal number. Uh and you know that's different than what others site as statistics saying like you know like this percentage of code is now written by AI. I'm talking more about like overall productivity the actual productivity right engineering productivity which is two different things. and and which is the more important uh metric and but I think it'll get better right and like you know uh I think there's no engineer who tomorrow if you magically became 2x more productive you're just going to create more things you're going to create more value added things and so I think they you'll you'll find more satisfaction in your job right so and there's a lot of aspects I mean the actual Google codebase might just improve because it'll become more standardized more um easier for people to move about the codebase because AI will help with that and therefore that will also allow the AI to understand the entire codebase better which makes the engineering aspect that's I've been using cursor a lot uh as as a way to program with Gemini and other models is like it one of its powerful things is it's aware of the entire code base and that allows you to ask questions of it allows the agents to move about that code base in a really powerful way I mean that's a huge unlock think about like you know migr erations, refactoring old coal bases, refactoring. Yeah. I mean, think think about like, you know, once we can do all this in a much better, more robust way than where we are today. I think in the end, everything will be written in JavaScript and run. Okay, we're going to can Lex for saying that in the end everything's going to be written by JavaScript. No, come on. Give me a break. Um, Ryan says, "Uh, honestly, I think the desire to to building a truly uh agi AI agent is ridiculous right now. I personally think the coding agents that you are going to be building, deploying, um, should be focused on the programming language that the person's coding." And that's interesting. Um, is that because of like this like the I don't know like the skill set or like the domain? I I wonder that it's maybe it's just more effective that way. Curious to hear, Ryan, your thoughts on that. Uh this is why I've been super supportive of the whole MCP agent design and all each agent is focus on. Okay, I think that's exactly probably what you were saying. Uh whoever A is just A, sorry, I don't know your name. If his job is just writing code, then yes, he can become 5X, but software engineering is not coding. Yeah, I guess maybe if because I don't know what Lex does in his free time for whatever he's coding. So yeah, maybe he can output 5x more code perhaps. Um, epic techn is it aware of the whole this for cursor. Is it aware of the whole code base though? Like it misses the mark more often than not. Yeah, this is okay. Flipping back to my face so I can rant a little bit. I rant a lot. I don't know if you guys know that. Um, so when I've been using like a speech, okay, VS Code absolute worst for AI agents. Doesn't matter what model. Absolute worst. The default when you're talking to VS Code like to to co-pilot in VS Code when you press enter it is only the context that you add. So by default it's only the file you're looking at. I don't know who is using an agent only in the file that they're looking at. I don't know like I don't think that's ever happened to me where I only want the agent to look at the one file. The if you controll enter I think it's control enter then it will basically add the codebase tag in. So in theory now it's got context of the whole codebase but you would think that if it had context of the whole codebase that it wouldn't be so terrible but it is. It's absolutely terrible. I can't get anything done with an agent in VS Code regardless regardless of the model that's being used. It's an abomination. I can't stand it. I work at Microsoft by the way. I'm saying this outright. It is terrible. Using chat is a different story. But even so, like when I want the context of my codebase, it's just not doing it. Cursor is better in that regard. I don't know why because it's literally the same thing. You pick from the same sets of models, but when it's pulling the context of the codebase, it seems better. It's still not good because as Epic Technav was saying in the uh in the chat, it still misses like in my opinion way more often than not. Now, the AI bros will say, "Well, it's because your prompts are bad. You're not prompting it good enough." And I just have to say that if I have to spend more time prompting than I would have to just actually go write the code, back to what Sundar was saying, it's not a productivity gain. Then now I do have a lot of room to grow in terms of my prompting without a doubt, but like come on, it's not it can't be that bad. I have found that Visual Studio in my opinion and so I'm a C developer for those of you that that don't know that. Uh so when I write in C, it's going to be in Visual Studio. I don't like VS Code for for C because I've sort of grown up on Visual Studio. I do find that Visual Studio does a better job with the context. It still messes up. It's still not like amazing, but it does a better job than the other two from my experience at least. Again, in C. The thing that is completely different and blows all of these out of the water by like I don't know, a landslide. It's nuts. Using GitHub Copilot with pull requests is without a doubt what I expected programming with agents to be like the whole thing for it. It has context to the whole repository just like the other ones do. Why is it different? I don't know, but it does such a good job with using the context of the codebase. Can it mess up? Absolutely. Does it mess up? plenty. But like when I compare what I'm doing in my IDE to like making a GitHub issue and then like going to sleep and waking up, man, it is wild how much better it is from my experience. Now, going back to the AI Bros and the prompting, I'm I have written some of the crappiest, most generic, mundane, lowdetail GitHub issues and it still does a good job. Like, that's the ridiculous part. I don't have to give it good instructions and it still does a good job. Now, the more ambiguous you are, then the more you should expect it's going to do weird stuff or not meet your expectations, but that's I think a given. It's just very interesting that I don't have to give it absurd structure in GitHub Copilot with the poll request. Um, I'm just checking the chat. Uh, Devin says, "Agree with Epic. I'm still waiting to be able to point out an existing class or component, get it to emulate the structure and style and creating a new one. So, I'll give you an example. Uh earlier before this stream, um in my codebase for brand ghost, I have I I don't use entity framework core. Again, if you're not a .NET developer, you might not know what that is. Entity framework core is an OM. So, you use it to basically uh map your your data to to objects in memory. It's an object relational mapper. So you don't have to go write SQL queries. I like writing the SQL queries and having them. So I use Dapper instead. It's a different OM. And so I have repository patterns in my codebase. They're everywhere. And so I like that. Personal preference. You can disagree. It's not what it's about, but they're everywhere. And they all follow a very similar pattern, right? So create, read, update, delete. Maybe the odd one has a couple of specialized like repository methods and then there's caching with fusion cache. It's it's the same across like I'm I don't know like 30 plus different files. There's lots of repositories. Earlier today before this stream I said I need you to make a repository class and I said use Dapper use fusion cache and then I said model it after the other repositories in the codebase. Maybe I needed to give it a specific one or a list of like 10 of the repositories. I don't know. But it didn't follow any of the patterns. It used Dapper cuz I told it. It used fusion cache cuz I told it, but like nothing else was the same. It even decided to like to like reduce I don't know why it does this sometimes. It like just reduced all of the lines. So it left no empty lines between things and it wrote the SQL queries like on a single line instead of like I don't know if you've read SQL queries that don't make your eyes bleed. They're generally like a little bit have some verticality to them. Imagine reading an essay and it's all one line. Like, no, you're not going to do that. But for some reason in this class, it was like, I'm doing it. That's how it goes. And then it didn't even compile. So, I don't understand. There's so many repositories. Like, why could it not just get it? But anyway, it's super frustrating. Let's go back. We're almost through this video. Sorry for my tangents and ranting. Chrome. I think it's all going to that uh direction. I mean just for fun Google has legendary code coding interviews uh like rigorous interviews for the engineers. How can you comment on how that has changed in the era of AI? It's just such a weird uh you know the whiteboard interview I I assume is not allowed to have some prompts. Such a good question. Look, I do think, you know, we're making sure, you know, we'll we'll introduce at least one round of in-person interviews for people just to make sure the fundamentals are there, I think that'll end up being important, but it's an equally important skill. Look, if you can use these tools to generate better code, uh like you know, I think I think that's an asset. And so, uh you know, I think uh so overall, I think it's a it's a massive positive. Cool. So, what do you guys think? Curious to hear. Um, I have my thoughts because I made made videos and stuff on it already, but do you guys do you think that that's more feasible versus the other side that we've been hearing everyone else talk about? Like basically all of the media, right? All of the media. There's articles, there's videos on like, you know, AI taking developer jobs. It's always accompanied by the graph that shows like basically the peak of developer jobs and it's falling off a cliff kind of thing. Um, you know, are developers cooked or or do we believe what Sundar is saying? Is there somewhere in the middle? But would love to hear what you guys have to say. um you know I'll just I'll keep rambling and until there's some chat messages but the uh you know when I've talked about this this has been sort of my take um what Sundar was saying and so I have there's a lot of like confirmation bias that I get when I watch this and I'm like see like I I must be right but you know it's no one has a crystal ball like even Sundar says it I'm like pointing with my chin like at Sundar on my screen and you guys can't tell what I'm doing but Sundar what's up? Um, so like he says, I don't know, but this is what I I foresee, right? So it's the same obviously for me. I'm sure he has a much better idea than me, but um my take is that uh it's not it's not going to like eliminate jobs. I think it was kind of said in the chat already today and it's like it's going to change things. It's not going to eliminate jobs. Steon says, "Can a CEO really say what he thinks? Anything he says will probably affect stock price?" Yeah, I mean, it's true, right? That's why I think that a lot of people, a lot of like tech leaders are doing the opposite, right? We got AI to sell you. We got AI this, we got AI that. Like, all we're going to do is talk about how like how it's revolutionizing everything and anything to draw attention to it. Like especially because I think that there's a lot of people that run businesses where they're like, "Hold on, you're saying that, you know, one of our biggest costs is headcount." And you're telling me we don't need as many heads to get the same amount of work done. I think that's probably what crosses a lot of people's minds. So you have these tech leaders that are talking about this in a particular way as well as the media that's all doom and gloom on this stuff. Now I've talked about the media side of this on code commute and I don't like getting too tinfoil hat here but um I think that the way media works in general is like and this is the same for social media. It needs to evoke emotion in you. If you are afraid especially it works very well to get your attention. It works extremely well if you're curious, right? And if you don't believe me, this is why you see so many stupid YouTube thumbnails where people are like, "Right, they do that because it works." And no one likes it. Everyone's like, "That's so stupid. Why do you make stupid thumbnails where you're making the dumb YouTube face?" They do it because it works. Analytics show it works way better than anything else. Why? because it makes you curious. It draws your eyes. Having a face on a thumbnail, there's stats that will show that there's like statistically more likely that you're going to go um I think even if you're reading stuff like on social media, if there's a face, you will pause longer. Humans pause longer to look at faces. Just a thing we do. Same with text on thumbnails. If you put text on thumbnails, it's better um I don't know what the right word is. better if you're scrolling through stuff and kind of like what it'll catch your eye more if it's on the left side. If you put text in the bottom right, it will not catch your eye as much. Even the thumbnail that I have here is a bad example. I put the text in the center. It should go up here. Right? All these little things are designed to draw your attention in a particular way. Right? So, when it comes to news and media, they've been doing this for way longer than social media has been around. It is the it's designed to make you like like borderline panic the whole time. Turn on any news channel, how much good news do you see? Approximately zero. Because that's not what gets people to watch the news. So you have like I don't think that the where my tin foil hat comes off is I don't think the media is necessarily designed to like I don't know like cause bad things. I think it's designed because it's a business. They have ads on the other side of it. They need you to go watch or click and read. So they will draw you in especially if it's fearful. So, we have a ton of that put in front of us as software developers, right? It's coming from all angles. Everything in the news, plus you have like tech executives that are kind of talking about this stuff. Then you add in that you have all of these new people who have never been able to code going, "Oh my god, I can do your job." Because they had code that got written and happens to compile and or run. So, it's like it's coming from all angles. So, I really wanted to to cover this video. That's why I've done like, you know, code commute video. I did a dev leader video. I did a newsletter and I'm doing the live stream on it. I wanted to keep talking about it because I don't think we see enough of this from the other side where it's not someone going, "Yeah, guess what? Like, you basically got five months until all of the code is written by AI and we don't need you." Because that's that's what we hear from everyone else. Um, Andre says, "I like to see an AI agent dealing with company policies and spaghetti code." Yeah. Like I don't know. Um, I don't know who's going to be better with spaghetti code, AI or or people. Um, AI might complain less, but um, Jose says, "Devs are delusional. 70 90% reduction in five years." Devs are delusional about about what that you think there's a 70 to 90% reduction in five years or the devs that think that are delusional. That sentence needs a little bit of uh restructuring. I think uh Deon says where Sundar is right is using AI as a force multiplier. Yeah, the companies that understand that as opposed to a way to cut headcount will do well. I agree. This is my um my take, right? That that exactly is my take. And so I'm just going to by the way, I didn't share this at the beginning, but this is the newsletter article I wrote about it. So let me just plug this for a moment. Um if you're if you're enjoying this live stream and you want to join next week or the week after, I write a newsletter article called Dev Leader Weekly. It'll actually be issue 100 this weekend coming up, which is nuts. Um, that's really cool to think about. That's 100 100 newsletter issues. So, uh, it's at weekly.devleer.ca. It's an email newsletter. You absolutely do not have to subscribe to it. You can check it out like a blog post. It goes out on Saturdays. So, if you're like, "Hey, live stream was kind of cool. I wonder what's going to be the next topic." Just go to dev or weekly.devleer.ca. You can check Saturday, Sunday, or Monday. um or just show up because it's gonna be awesome no matter what. Um but yeah, so I wrote this newsletter article and um one of the things I talk about in here and I think I talk about it in the YouTube video I made is this idea that like and I still I still don't know the right words even though I I wrote a little bit more about it. Um, but I think that I hear a lot of people talking about AI and um kind of like replacing developers as in if we think about software engineering work as in say you're at a company you have a product or a service and the bucket of work is this big. So if you have a bucket of work that's this big and you have x number of developers if you add in AI then in theory what you should be able to do is remove developers because you have this much work right if you have something that you're adding that can do that work you don't need a developer to do it so it's kind of like a zero sum kind of thing or like it's bounded it's constrained there's a finite amount of work to And when it's finite and you add workers, even if they're AI, then you don't need the other types of workers. And I think this is like it seems to be the number one framing that I see coming up everywhere. To Devon's point, it's like that's how you would reduce headcount, right? You could say, "Let me get rid of the people because they cost way more." For those of you that don't know, in software development companies, like I don't know a maybe not every software development company depending on what else they're doing, but employing people is a huge cost, right? It's a huge cost that you incur for having people on. So, back to the thing I said earlier around layoffs, again, I'm not in favor of them. I'm just giving you some framing. If your options are to increase revenue by $100,000 within this year or you can reduce your costs by $100,000, you could probably reduce your cost by $100,000 like that. And I'm not saying it's good. I'm just saying it's a little bit easier. So I often think that that's what happens is like it's a it's a easy way to make that math work. Now yeah and then so Deon to my to the next point I want to say Deon said it in the chat tell me with a straight face that you've ever worked at a place where the backlog was empty. So the reason that I think that this whole framing is flawed personally this is my take on it. You're absolutely welcome to disagree. I'm not going to like blast people in the chat or whatever. Like it's it's an AMA. I want people to share their perspectives and stuff too. But my take on this is exactly as Devon said, the work is not finite. I have never worked. I've been, you know, in the industry. I had two years of internships plus this summer will be 13 years. So like 15 years of writing software as a engineer and a software engineering manager. So 13 of those years were as a manager at startups. All of my internships were at very small companies. I worked at a startup after school for eight years. Now I work in big tech for last five years. I have never ever seen a backlog that's empty. I've never seen a backlog that has a finite amount of work. And sorry, I'm using terms like backlog here. You might say, well then you might be doing agile wrong up the back. Like no, I just mean the amount of work that we want to get done is never overall shrinking. There is always more stuff to do, new priorities. I would wager most of the time in my career has been spent trying to prioritize whatever the newest high priority thing is, the newest P 0 thing, trying to say, hm, we have to shift our focus to this thing, but we're already doing this thing over here. And it was the P 0. Oh no, like what do we do? How do we start getting focused to this one now? It's always that. It's never like, "Oh, thank God we got this one done and we only have one thing left before the service is done or the product is done for forever." It it just it doesn't happen. So, in my mind, I just don't see that as like a thing where you get rid of all the developers and then you go, "Cool." like the AI turned through the code and now now we're done. And like even if you're saying, "Okay, well I I also don't agree it's going to be done." I'm like, "I don't know, man." Like why wouldn't you just like why wouldn't you augment your developers so that they can go do more stuff and then you can use the AI to do all the like Sundar was saying, all the crappy stuff that no developer wants to do. Could you imagine? Let's imagine a world where AI agents are very very very good at refactoring legacy code and migrating patterns in your codebase to the new standards that you want. They're very good at taking the brittle code and they can say, "Great, we're going to take this apart. We're going to make sure that there's test coverage on it. We're going to restructure it this way." And imagine, I know we're not there, but imagine that we were. And you could be like, "Cool. On the weekend, I'm gonna let this stuff run for our company and it's gonna go through and it's going to because we're assuming, we're pretending here that it's going to do a very good job refactoring and testing all the legacy code. Basically, you have stuff that can constantly migrate your patterns and practices to be the very best. Imagine that we had that and then you could just focus on building all the cool stuff and sorry maybe you actually like the you know the tech debt and all that other stuff which is totally fine. So pick something else that you want it to do, right? There is so much in terms of a code base over time that no matter how much effort you put into it, you could say, "Hey, wherever, you know, we're cleaning up as we go. On every commit, we're going to add some cleanup." You can, right? You can. I can still basically guarantee that over time your codebase rots. It just will happen and it's because it's really hard to prioritize some of the legacy refactoring and stuff if we're never going to have to go touch it or rarely touch it. There's a question on Instagram. This is from Aban, I believe. Sorry if I have your name incorrect. Um, what is your say on new developers using GPT or AI for everything and ignoring cores and basics? I think the, you know, the obvious answer that I want to give here is I don't think it's a good idea. Um, and it's because they're accomplishing different things. So, I think that if you want to use AI to be able to generate code or to to make features or products or services, whatever, it's a tool that can do that. It is also a tool that could help you learn. But I think what happens is that we have people that are junior or just getting started out and what's happening is they like every every human wants a shortcut. Doesn't matter. Pick any you name an industry, a domain, you name it. We all want shortcuts, right? Imagine for work if you could just, you know, snap your fingers and you could have the highest paying job in the company. Imagine you could snap your fingers and you could be retired and you didn't have to work. We want that shortcut. Imagine, I'm just going to make up a bunch of examples now. Imagine you wanted to be a professional basketball player and you weren't like me that's 5'4 and has no basketball skills at all. If you could just snap your fingers and had a shortcut, you would do it, right? If you wanted to go to the gym to get shredded or you wanted to get big and strong, if you could snap your fingers and do it, we all want a shortcut. And I think what happens when you have really junior developers, they're going, I want to be a good developer. Good developers can make things. So look at me, I can also make things. I'm using AI to do it. The problem is that you're not actually learning anything or should say you're not learning anything. you're you're minimizing your learning. You are optimizing how much code gets output and you are minimizing your learning about what you're doing. And I think that's the biggest risk, right? It's the same just to give you an example. Imagine you're trying to put your resume together. Okay? So, and this this is a I think something that is probably relevant for a lot of people. You're you know early in your career. Maybe you haven't started yet. you're trying to go for your first job and you're like, I don't have experience. And then one of the things that, you know, everyone including myself will say is, well, you got to go build projects. And you go, okay, I'm going to build the projects that way I can stack them up for my resume. So, what you do is you follow the tutorials online and you basically copy the code over from the tutorial and now it's running and you go, great, yes, I have it. I'm adding that to my resume as a side project. And you repeat that. I have five side projects and all of them you basically followed the tutorial. You copied and pasted the stuff from the YouTube video, the GitHub repository, the the you know the blog entry, whatever. At the end of that, you have your five side projects and not a single one of them you understand how it actually works. You you prioritize the wrong thing. You said, "I wanted to have something on my resume." But when I'm talking about building side projects, I'm saying you need to build them so that you can learn the things, right? For me personally, if you have side projects on your resume and they're not complete, they're just like things that you were learning about, in my opinion, hell yeah. I care about that. I'm interested in seeing that you didn't have some experience with X and you were like, I'm trying to learn X. I tried making a thing. Cool. That tells me that you are interested in learning. You're putting effort into it. I'm interested in that. I think that's a great quality. That is the kind of thing that if you were in the interview, I would say, "Tell me about all these things that you're building. It sounds interesting. What did you learn?" So, you can use AI tools for this stuff. But people generally don't because they're just looking for a shortcut. So, what you could do is pick anything that you're building. You could use AI for it, but you should be questioning it. You should be getting it to quiz you on things, testing your understanding. You should get it to give you like point you in the right direction, not give you the code, then you try it out. Then, when you're stuck, then you ask it for help. All right? we have all of these opportunities to use AI this way for learning and we don't because why would you, right? You're just we're all looking for a shortcut. So, I'm just trying to encourage people to to think about that a different way. So, hopefully uh again, sorry if you're I'm getting your name wrong, Aban. I hope that is helpful. That's my take on it. Um the other thing that I'll quickly add is that I do wonder, I mean this genuinely, I wonder if there is a world where we fast forward and all of the people that started coding like basically purely using LLMs and vibe coding their way to the future. I genuinely wonder if there is a future where that happens to be more effective. I have a hard time believing that, but I will be the first to say that I can't say that it's impossible. So, I'm keeping an open mind. I just don't I don't encourage people to do that. Hope that helps. Hey, Unstoppable. Um I don't know if you joined the stream and left, but if you're still there, hello. Um do I maintain a prompt library? I don't. I probably should. I'm starting with like my co-pilot instructions. It's been something I've been doing more. Um, but I've already found like across a few repositories like there's some parts of the co-pilot instructions where they're they're very similar and um so I'm like, okay, do I need to like do I have need to refactor my prompts and stuff, my instruction to something more common? Um, so I don't know yet. Uh I if I take it the way that I would code, I need to sort of duplicate it a bunch of times to see more patterns to go, okay, every single time I write co-pilot instructions, here's how they look or I'm doing this prompt enough, right? Like it's literally just like code in that sense. I'm doing it enough. Maybe I should have it saved so I can reuse it later. Um so that's my take. Like I probably will do more of that. I just haven't started organizing prompts that way. Um, I don't know how to pronounce your name. I'm sorry. Uh, Reto, again, apologies. I'm terrible with names. I'm trying to make myself say them out loud, though, because this is one of those things. Just total tangent. Um, I I genuinely believe that when you feel like you're bad at something, if you avoid it, you will just keep getting worse because you never practice it and you'll keep fearing it or resenting it. I've done this for too long with names. I am very nervous to say people's names incorrectly because I don't want to offend them. So that's why you will hear me try to pronounce them and I'm very sorry if I get it wrong. I don't mean to offend. I just don't know how to pronounce them. So what kind of projects do you think are outstanding as a junior dev? And do you have any advice which fields beside webdev? Um yeah like this is a it's a tricky question, right? So um on the second part any advice for which fields beside webdev what do you enjoy like in any of these fields that are related to programming like jobs exist I'm not minimizing that it's difficult to get jobs right now but there are jobs like if if you really like embedded development not a lot of people talk about it you could do that right like you could do mobile um you could make you know there's phone apps and stuff you could make that instead there's there's a still a market for desktop development still right a lot of people say oh well no it's just going to the browser like sure lots of it has but there's still a market for desktop development so I think that for me to answer that question for you it's really difficult because I would do this is going to sound funny I would do what my mother always told me which is like you have to do something that you love because if I told you I don't have stats on this but I said if I told you you know what web development is going to be the best to to for you to make the most money I don't know like is that what you're trying to optimize for even like it's not really for me to make that decision it feels kind of crappy for me to to say that right so um there's there's lots of different fields I think one of the things that's super cool about software development is You pick any single industry you can possibly imagine. Pick one. Anything. Pick the most random industry that you can think of. It can benefit from software. And if you're like, "Yeah, but they don't make software for that." Guess what? You can go make a ton of money then because no one's using software there yet. So, go do that, right? Like, every industry can benefit from software. I will stand behind that. you name one, throw it in the comments if you think that you're going to trick me and it can't benefit from it. There's something that you can benefit from. Um, and then what type of project do I think is outstanding as a junior? I have I probably have a different take on this. I think a lot of people are like, "Oh, I have an app that has like, you know, thousands of monthly active users." Like, that's super cool. I think that's great. Probably demonstrated a lot of things, but I'm not focused on you making money from it. I'd be more interested like, oh, a thousand, okay, like cool. Like, how did you scale to that? Like, did you need to design software in a way that lets you scale to that? So, I'm less curious about or less, you know, impressed by you making money off a project. I just want to see that you were trying to learn things, that's all. So, when some people say like, oh, like, you know, to-do apps are dumb or like, you know, whatever whatever template app, I'm like, look, if you're just following a tutorial, copying and pasting the code, sure, it's not helpful. It's just it's not like it's not really going to help you learn and it's not going to be impressive for a resume. But there's nothing to say that you couldn't take a to-do app and then start building tons of interesting complexity into it to try different technology and stuff out. Like I built a to-do app that can sync across different platforms and it's like uh it has, you know, AI built into it so that when I write my notes, it like kicks off automated workflows. Like you could do so much cool stuff with that. And the the meta point in answering this question is really just like I just want to see that you're learning about different things because as a software developer on my teams, I expect that you will have to go learn things and ideally quickly because you're interested. That's what I want to see. Uh LLM's killed Stack Overflow perhaps. Yeah. If the dev can't explain the code, reject the PR. Yeah. I mean, how do you test for that? I I don't necessarily disagree with that. I would say if you have a team of developers, I can't explain the code. Reject it for sure. Um Devon says, uh layer on more and more instructions is probably the way to go, but there's an element of you just need to prompt better, bro. Yeah, some of that should just be built in. Exactly. This is my the whole AI bro argument is like, yeah, prompt better. Ju make the tool better. Like I don't know what to tell you. Um Gaming Yun, welcome. I love dev leader. Well, I love you too, man. Or I don't I don't shouldn't assume your gender. I don't know these things. Um friend, love you too, friend. There we go. Epic Technav embedded software and C++ job market is booming. Is it? I actually don't know, but that would be super cool, right? I I don't have Oh, and um and then Technav says, "I had some friends transition to C++ and find success." Very cool. Good to know. Um, I see I don't know who's sending these messages on LinkedIn cuz the the names don't come up. Let me check so we have a little conversation. I think it's supposed to say AI is going to slaughter webdev. Um, okay. Is it source? Trust me, bro. Um, Rao. Uh, oh, right. I got the pronunciation. Nice. Okay. And I love hardware stuff like IoT embedded, but the market is so small in my country. Um, you think nowadays it's more important to good, uh, domain knowledge rather than my technical skill. Uh, no. I think like honestly it's a balance of all these things. So I trust me when I say this and this is from my own experience as someone that that thinks this way and also from talking to tons and tons of people. This is by the way this is like not my full-time job but this is just what I do because I I like to be able to try and help especially for people getting started as software developers or people that that think this way. We try to like optimize everything which is great. It's very good but the problem is that it can be paralyzing right so in a situation like this like and Raido I'm not I'm not I don't mean to pick on you or single you out. I think it's just a good example of a comment. So thank you for sharing. Um the the comment is do you think nowadays it's more important to uh to get good at domain knowledge rather than my technical skill? Like no. Um and then would I say no domain knowledge like also no and you didn't mention soft skills like also include those like you need all of them right and the reality is that if you are spending time building things basic if you're trying to do the things that you would be doing on the job and you are practicing those things in as representative as a manner as possible then you are actively practicing the things that go into software development. That's I think that's what you need to do, right? Like so ideally if you could work with at least one or two other people building something that you're interested in in a in a domain that you're interested in using technology that you like boom boom boom checking all these boxes like that's ideal. Is that going to happen every time? Probably not. Personally, I didn't like to build stuff with people before. And that's because for me, coding has always been like a little bit of an escape. I can go like, you know, I can just ignore the world and kind of program. It's it's it's my spot to be in. So, I've always liked that for my side projects. And then probably good few years back now because I'm I'm old um but like I had a colleague that I was working with and we were talking so for those of you that don't know I have like a a hobby role playing game that I built on the side. It's never going to be finished. That kind of thing. Just sandbox for me to program in. And we were talking about like game systems and he was like oh it would be really cool like if you could do like this or that. We were talking about got some crazy ideas for like dialogue systems that basically by interacting with more of the NPCs of particular like races or cultures, the more that you're interacting with them and like building up rapport with them, if they speak different languages, the more that the text actually starts to come through and understandable so you can like literally learn the languages. And then we said, "What if it worked the other way and the NPCs could learn your language?" like the most ridiculous complex systems for RPGs. And I was like, "Hey man, like why don't you try building that? Like I think that's super cool." And that was like the first time that I really felt like I'm going to invite someone into like my little my bubble on this and we build stuff all the time. So, you know, I I would encourage people to try and emulate what happens in the real world. That's also why when people talk about lead code, I'm like, if you enjoy doing lead code, like, hell yeah, go do it if you enjoy it. It will help you with your interviewing, unfortunately, because people interview for that. I think that it does not have the carryover to real programming scenarios that most people do, especially if you're just like memorizing them because I don't think it's representative of mostly what we do. Um, gaming says for embedded systems, wouldn't one also need hardware knowledge? Yeah, of course. Um, it's a different framing, right? you get into like I would say just because I was talking about lead code some of the stuff you probably care more about from like a lead code perspective and embedded because you're doing interesting optimizations a lot of the time you have more interesting constraints right you don't have as much and this depends on what you're you're building right but like you don't have unlimited storage you don't have you know the fastest CPU in the world you don't have unlimited memory, you have resource constraints. There's probably not as many libraries compared to like React. Um, so you might be building some of your own stuff and that's cool. It's different. So, do you need uh hardware knowledge? I would say you can still learn it for sure, right? because you have to understand things about like you have and it depends what you're building right if you're just like I'm running something on a on a processor and it's not doing anything it's maybe abstracted enough these days uh I haven't done embedded in since like my internships but um you know most of the time you're building embedded stuff you're interfacing with like you have inputs and outputs whether that's lights or sensors or motors whatever like there could be a million different things. So you will need to understand that concept. But a lot I think a lot of the the stuff that you can go get like Raspberry Pies and whatever else for building they will give you the guidance around that and then you can use chat GPT to ask a question. So I wouldn't on the hardware front I wouldn't be like nervous if you're like oh I don't know anything about it. That's cool. I I will be the first to say five years ago I applied to Microsoft. I had only built uh not only but I had never done distributed systems. I'd never had code running in a data center anywhere. Never done it. I applied for Microsoft. I got hired into the deployment team for Office 365 where we have to deploy hundreds of services to hundreds of thousands of machines across the planet, right? Like you will learn these things. It's okay. Um, Epic Technav says, "For everyone here, stop copying tutorials. Just start building and find the code snippets and info as you go." Yeah. Yeah. This is This is so true. Um, you I mean, you don't have to do it that way, but I feel like I feel like you should the people, like I was saying earlier with the shortcuts, right? People want shortcuts. And if you're feeling like this is painful, this is hard or I'm stuck, like we don't like that. We try to avoid it. But that is literally the thing that's going to help you learn. Like you have to hit that. You have to get stuck because when you find the solution after getting stuck, I don't know. I'm not a doctor. I'm not a not a biologist or a psychiatrist or a anything with the brain, but I feel like there's probably some data on this, but I feel like if you are not getting stuck on things and you're simply getting solutions, you will not retain it effectively. Um, hit the Yeah, gaming says, "Hit the like button." Yeah, smash the bell. Do people do that like to get notifications? Is that a thing? I don't I don't even know. like, subscribe, com comment below for the algorithm. Um, and Nabon, sorry, I just going back to um to Instagram. Is it worth getting I think I see a 204 like I'm assuming Azure 204 cert certification is my last software engineering. I'm learning things from past year and I'm thinking about getting that. Uh my take on certificates for software development is like get them if you are interested. I would say my take is please do not try to collect them in hopes that that's going to be the thing that gets you hired. I have never and and sorry for anyone that like I'm not trying to minimize this. I have never been in a hiring conversation where someone said, "Oh, but they have so many certificates." It just doesn't just doesn't happen. So I would say if it's an interest to you absolutely pursue it right I am totally of the mindset that if you want to learn more right that's why if people like people are like oh you don't need college university to to be a software developer you certainly don't but if you wanted to go spend years and get more education further on that like I think that's awesome if you want to go to a boot camp because you want to to learn more I think that's awesome I reviewed someone's resume. It's one of the last ones I put up on my channel, I think, or it's not posted yet. I think it's up. Um, there was someone who had been to university and like 5 years later they went back to a boot camp, right? They're just trying to improve their their learning. So, I I would never tell people like don't if you're trying to learn more things, go do that. I think it's helpful. Um, yeah. So, and then Aban says, "It's a back-end developer certification, which is what I do and I want to do." Yeah. So, I think, you know, if you enjoy it, then absolutely, right? If you want to learn more, um, I know someone I used to work with who I was trying to help transition from a different role into a developer role. Um, he has said, and I've seen him, I haven't seen him post on LinkedIn in ages, but I have seen him engage in comments where people on LinkedIn were like, "Oh, certificates are dumb." and he was like, "Look," he's like, "I actually enjoy taking them because for me, it's a guided way to go learn about things that I would I would struggle with otherwise." So great. Like that's an awesome tool for him to go learn. But um Rao says, "What do you think about the future of blockchain?" Oh, I just don't care anymore to be honest. I'm so exhausted by blockchain conversations. Um, sure, like the technology is there, but like how many years has it been? Um, and what what's using it? Like I I'm not saying that it can't be useful, but the I think the big problem is that it's all it's never the technology when it comes to crypto. It is never the technology. It's unfortunately tied to everything's a token. Everything has a monetary value. And as a result, the only thing that you will continue to see until something changes is people gaming it. Every single time it is purely PvP with other people, people put money in and then people extract the money from the pool of money. So blockchain as a technology, I think there is benefits to it. I think there's use cases it solves, but when everything's just a token, it's like it's really difficult for me to to see past that. So, I like I'm just steering completely clear of it. Yeah. It's just pump and dump. Not a real environment to invest. Yeah. I like I know people that have made lots of money from it. I've seen people lose tons of money. I I just will not. Um Yeah. Hey, crypto bros and AI bros in a cage match over who gets the compute. Yeah, that's the real the real thing. Um cool. Okay, we're we're past eight, folks. I am going to wrap up. So, I didn't even read much for my newsletter, but I shared the newsletter link. So, let me get flipped over to full screen screen sharing. So, this is my YouTube channel, Dev Leader. I think most of you probably know because you're coming from there. Um, but this is the channel where I have, you can see that was the first video up there, right? So, um, you can check out my review. Basically, it's a lot of what we talked about in this live stream, but if you want to check it out, it is up there. You'll also notice I have resume reviews up here as well. So, if you're interested in having your resume reviewed, I'm a little bit behind to be totally honest. So, I do apologize if you've submitted your resume and I have not made a video. I didn't get to it. Um, and I already mentioned on this stream my video editor is currently out, so there's going to be a big catch up. So, I do apologize. I will get to them, though. I promise. So, if you want your resume reviewed for free, just watch one of the ré review videos. Um, they're some of the least watched videos on my channel, unfortunately. Um, but I think they're some of the most helpful. Personally, I've had a couple of people now message me to say that they're getting interviews. Like, it's all I can ask for, right? It's free for you. Cost me money for the video editor, but I want to do that for you folks that are interested. But yeah, C tutorials, there's a podcast. I did a podcast interview on Friday. I did one today. I have one on Thursday or Friday this week. So, there's a few more queued up. Uh, which is cool because I took a little bit of a break from that. But that's that. Code commute is my other YouTube channel. I explained a little bit earlier that code commute is my vlog channel where um what I will generally do is throughout the week when I'm commuting, as you might suspect from the title, I set up a camera in my car and I talk about software engineering topics. And if I'm not commuting a lot, then sometimes I'll post up in my office right here with my my little mic. I think people were poking fun about the the tiny little mic. Hold got to hold it like this. Um when I'm driving, it's clipped onto my seat belt. But um this is a channel that's all driven by questions that you want answered. So kind of like on this live stream where people post in the chat. I'll answer it on the live stream, but if you want like a full video on a topic, just go leave a comment on a video. I'm happy to go answer it for you. Um you can pick any video you want. It doesn't have to be related and just leave a comment. Um, you can message me on social media as well. The comments are kind of nice because I feel like it probably helps the algorithm in some way. But, uh, if you have something that's more lengthy or you want to keep personal because you want to be anonymous when you ask it, just message me. You will be kept anonymous. But, this channel for me is a ton of fun. Um, people also asked for it, so I'll keep sharing it, but it's at spotifi.codemute.com, codemute.com um which goes to the Spotify channel. People were saying, "I like Code Commute, but I don't have YouTube Premium and I watch it when I'm doing chores. Could you put it up on Spotify?" So, uh I am currently uploading every week until I've caught up 20 episodes. I think I'm just over halfway now. So, it's still going to take a little while, but I am catching up on code. Um, so all the YouTube videos will make it here eventually, but I think I'm at five seasons now, five and a half seasons. Each season is 30 videos, so over 150, so just past. So you can check out Code Commute on Spotify. Otherwise, I have courses available on Dome Train. Um, primarily focused on C, but not only C. So these two courses here are 11 and a half hours for getting started in C. They're intended for people that have no programming experience, at least to start with the getting started one. And then the deep dive one goes a little bit deeper into that. It's really difficult when I read through the reviews. Some of them are like, "Oh, amazing. Five stars." And then some people will be like zero or like one star, zero stars, whatever. Like the getting started wasn't beginner enough or the deep dive wasn't deep dive enough. And I'm like, "Oh man, like you can't you can't cover everything uh in two two courses." But um I'm really proud of those. Uh I'm proud of all my courses, but it was really cool to be able to make these for dome train. If you're not a net developer, do train is run by Nick Chapsis, who is uh arguably aside from Tim Corey, I think. I I don't know their their subscriber count, but Nick Chapsis is one of, if not the biggest C uh YouTuber in the world. So to be able to have the getting started courses on his uh his course platform is a bit of a an honor for me personally. I think that's super cool. Otherwise, there are some courses on like career development soft skills. I partner with Ryan Murphy who's an engineering manager at Yelp. So we got some of those going on if you're interested. And then some more Courses. Still with me? Go buy all those courses. Uh if you check uh the sale has changed. This one's different now, but there are uh pretty regularly there sales. So, um I would say like, you know, just letting you know if you don't see a sale and you're like, I don't need it right now, like wait a little bit. There's probably going to be a sale. Um which is great. Great for you, not for me because I get less money for it. But I do appreciate if you want to check those out. Otherwise, the last thing that I will be sharing is brand ghost. Brand ghost is the SAS that I am building on the side. This is what I use for posting all of my social media content. Um, it's been kind of cool. I've shared this on live streams now and I've had a couple people from live streams that have signed up which is great. Um, you know, if you are you don't have to be a content creator necessarily. If you want to do some learning in public, it's to it's literally totally free. Boom. There's a free tier. You don't have to pay anything. You don't have to put a credit card into it. You can literally cross-ost and schedule for free across social media platforms. Um, if you see how I post on social media, like I know Devon for example interacts with me across every platform. Uh, aside from the comments that I post, like all of my content is done through Brand Ghost. So, I create it. Brand Ghost posts all of it. I shared my calendar. I think it's probably a good thing that I I share my Brandos calendar when I come on uh these streams. Just give me one second. I'm just pulling it up. But I share it so that you can see like I don't schedule this stuff. It just This is what Brand Ghost does for me. So, one sec. Let me pull this tab over. Right. So, this is my social media schedule across all my platforms. Um, let me do this. Move this over. Right. Like this is one day of posting. This is for a few different things though. This one's actually a test thing that's got to go away. Someone was adding some stuff on here they shouldn't have been. But anyway, that's a little bit inflated. Let's pick another day. Right. All this stuff prescheduled. This is done indefinitely, by the way. I never have to go schedule content again in my life. I just create it. So, basically, if you see a post go up on social media, it's through Brand Ghost. Sometimes I comment through Brand Ghost. So we built an aggregate social media feed so I can respond to your comments all from one spot, but sometimes I'm already on the platform and I'm responding to things. So it is not content that's written by AI. It's all my content. We have AI helpers and like suggestions and stuff like that, but um I just use it so that I don't have to schedule my content anymore. So, with that said, um if you're interested in getting started, it is totally free. Uh if you are more serious and you want some of like the more advanced functionality like I was just explaining, there's a paid tier. And I will remind people if you're like, ah, not really for me. If you have a business, okay, if you have a business and you're like, I don't have time for the social media part. I think I need to hire someone. I don't know. You could get started with Brand Ghost for free. You could try scheduling and crossosting. Then when you build some momentum, you could try out the paid tier because I think it's 30 bucks a month for the first tier. Then you're not paying a human being for it. As soon as you have some content added that you can repost in the future, by the time you hit that point in time and you have a bunch of content built up, you have to create less content. Some people don't know this. I haven't written a technical blog since April of 2024. Every single day I post a blog article link. Every single day because I have like 300 blog posts to link to. The same with my YouTube videos. Right now, if you check, every day at 2:00 p.m. Pacific, I post a YouTube video, a link to a YouTube video. To to be clear, it's posting still from like sometime like almost two years ago. It's got a lot of content to go through. So, uh it's really cool even for those YouTube videos like some of them are they haven't been watched in a long time. And I'll see a spike in views because it got shared on social media. So, if you are trying uh to get a little bit more um going on with your business, you can try Brand Ghost out. If you're curious how it can help, just message me. I'm happy to talk to you about it and get you started off. With that said, folks, we're all done here. Thanks so much for being here. Thanks for the regulars for for the constant support. I I don't know. Like, I can't tell you how cool that is like that people keep coming back. Like, I think that's super awesome. Um, my wife really likes to watch like um like law live streams where they just like they're streaming court and then she listens to like social media influencer lawyers like that are legitimate lawyers but uh they do live streaming and she loves it. So she's always like watching it. I'm like man like they're doing such a good job live streaming and I'm like one day one day we'll be there. Um so I appreciate all you being here. I can't do it without all of you. So hopefully we'll see you next week. And like I said, if you have stuff that you want answered in the meantime, head over to Code Commute, leave a comment on something. I will film you a full video just for you and anyone else who's watching Code Commute. So I will see you next time. Take care.