.png)
Develop Yourself
To change careers and land your first job as a Software Engineer, you need more than just great software development skills - you need to develop yourself.
Welcome to the podcast that helps you develop your skills, your habits, your network and more, all in hopes of becoming a thriving Software Engineer.
Develop Yourself
#277 - RAG: The Only AI Skill Web Devs Need to Learn in 2026
I thought Ben was a troll when he slid into my DMs after a LinkedIn argument. Turns out, he’s building some of the most practical AI systems I’ve seen. In this episode, we talk about how that disagreement turned into a friendship—and why Retrieval-Augmented Generation (RAG) might be the skill every developer needs in 2026.
Finally - if you want to go deeper with RAG, I have a tutorial I made just for you: https://parsity.io/ai-with-rag
Shameless Plugs
Free 5 day email course to go from HTML to AI
Got a question you want answered on the pod? Drop it here
Apply for 1 of 12 spots at Parsity - Learn to build complex software, work with LLMs and launch your career.
AI Bootcamp (NEW) - for software developers who want to be the expert on their team when it comes to integrating AI into web applications.
Welcome to the Develop Yourself Podcast, where we teach you everything you need to land your first job as a software developer by learning to develop yourself, your skills, your network, and more. I'm Brian, your host. Today on the Develop Yourself podcast, I'll be talking with Ben, CEO of Loby Software. And how we met was really interesting, actually. I wrote a post on LinkedIn and it was about the shortcomings of using AI tools and coding. You kind of challenged me. Ben said, Hey, you know, maybe you should approach it this way, or you're making a common mistake. Um, I got a little defensive, and then he then you slid in my DMs and said, Hey, let's jump on a call. And we did, and it was a great conversation. I really appreciated that. It's very, very rare that you meet with somebody and then like you disagree a little bit, and then it's like, hey, let's talk well on the phone. And we did, and it was great. And then I've learned about you. You're you're a great thinker. You have a really interesting story. You are deep into retrieval augmented generation, which is really why I wanted to get you on the show to get your take on it. And I also want to just talk about like how you one got into software and then got into entrepreneurship, and then a bit about RAG and how it works in practice and why this is likely one of the most important and maybe profitable uh skills you can develop nowadays. But welcome to the show.
SPEAKER_02:Thanks. It's a pleasure to be here. Uh, pleasure to to talk to you for the first time, really face to face. We've talked on the phone and in DMs, like you said. Um, I thrive on conflict. So anyone I meet who I disagree with is immediately the first person I want to talk to. So uh we disagreed, and I think what was a smart way, so it was easy to reach out. Um and and yeah, um, as for my journey, it's very it's very unconventional. I've only been in software a couple of years. Um, I took a very long road to get here, started out um with a uh degree in audio engineering and spent some time in recording studios. Uh worked very long days doing that. And uh when I was burned out, I would also manage restaurants, which is what I've done since I was about 16 years old. Um so I always kind of had like this underpinning of like restaurant management to pay the bills and the events something artistic broke me. And then I was spending a lot of time in recording studios on the side doing things like that, working, you know, 16-hour days pretty much my whole life, and uh got out of that eventually. Uh, started a film production company uh that started with my audio background and kind of evolved into more knowledge about uh film and video and did commercials, ran a commercial soundstage in Atlanta for a couple of years. Um, did some commercials there, did some indie projects for people out of California, little things like that. Uh, hated it, absolutely hated it, was completely burned out doing 16 to 18 hours a day. Uh started another company uh with a small army of dog walkers and pet sitters in Atlanta that I ran for about seven years. We did really, really well with that, uh, but kind of hit a ceiling, kind of maxed it out. Um, gave it to one of my sitters who couldn't afford to buy it, but promised to keep all our clients. So gave it to them and moved on and moved into software about a year later. And uh wish I had done it sooner. It's it's what I was born.
SPEAKER_00:That that's that's I've heard a lot of stories. That's gotta be way up there. That's about 10.
SPEAKER_02:So, how'd you learn to code? I went to a boot camp. I went the boot camp route. Um, it was an at-your-own pace thing. It was it was very unstructured. There was uh some mentorship and guidance from uh third parties outside the organization. I owe a huge amount of what I know to my mentors who I met through that program. Um, but really just, you know, had had had a great curriculum, had a really what I believe to be sound and solid approach to building actual enterprise software, not not just, you know, local machine programs that run. And then when you get into an actual job, you're completely unprepared. It really prepared me for the actual job. Um, and then I'm a learning obsessive. So I I just did that the five, six hours a day that I had to do it to kind of get through the course, but then also did another five or six because that's the schedule I'm used to keeping and just couldn't get enough of it.
SPEAKER_00:Gotcha. Man, okay. What was the name of this this program? This is probably bad for my business, but I have to know the name.
SPEAKER_02:This was this was Bloom Tech, Bloom Institute of Technology.
SPEAKER_00:Yeah, yeah. I've heard good stuff about them. I mean, I I've heard good stuff, but any any big boot camp, you're gonna hear a lot of bad stuff too. It's just the nature of this industry.
SPEAKER_02:There's good and bad, and and it was a small program with small staff and all of the good and bad that comes with that too. But I I mean, all I can say is my experience with it and what I learned from it, and it prepared me. I I think I think it's like anything. If you take take things on with a good attitude and and you embrace the grind and you take responsibility for how well it prepares you and and you're accountable for what the end result is. I think you're gonna get good results out of anything.
SPEAKER_00:Yes, I I would also say that you're the kind of person that I like just hearing your story. I'm like, you're the kind of guy I just see like you're gonna succeed probably. You you had an entrepreneurial spirit spirit, um, you have like a background in that, you seem like you're a hard worker, you know, maybe beyond what the normal average person would ever do. Just those two two things alone let me know, like, okay, you're kind of like I would cherry pick a guy like you to have in parsody, for example, as a success story, because I'd know that you'd probably be successful.
SPEAKER_02:Those things are true beyond the bounds of what's healthy for most people. So um yeah it's a it's a it's a constant grind as to finding that balance between how hard you can work without going over the top. But so far, so good, man. I've been blessed.
SPEAKER_00:Seriously. So now you're CEO of a software company that you're building a very specific type of product, it sounds like this is what led us to a lot of the conversations we've had recently. Um, we've talked a lot about RAG, and I've talked a lot about that on the show, but really briefly, can you tell people like what is RAG RAG?
SPEAKER_02:Yeah, so very briefly, uh RAG stands for Retrieval Augmented Generation. It's kind of the AI methodology that's really in vogue right now for a lot of enterprises. A lot of enterprises have existing knowledge bases or existing data sources or or or a lot of times digitized or not. And they want to uh make those things more searchable so that they can implement some AI strategies for finding that information quicker than they're accustomed to finding it and using it in some way in their system. Uh and and it's really as simple as it sounds. You you are generating a response from some kind of AI, LLM, with information that you have retrieved to augment that response.
SPEAKER_00:I that's a really succinct answer. I love that. We're both building with this. I've I'm at the you've had more experience than me at this point. I was at calling you asking for your advice recently. Um and I and I've been building with it like building for a couple of AI startups in San Francisco. And it feels like this is the use case that actually makes sense for most companies. Now, I don't know if I'm being super biased or anything like that, or maybe you are too. But what's your take on this? Like, why we've heard so much AI hype, and it's all been around mostly for developers. We think, oh, code generation, or like just weird use cases, like, oh, now there's AI in Notion or your email. I'm like, yeah, that's cool. Why do you think that enterprises have like been really quick to adopt rag? And and and do you think that this like really is going to be like potentially the use case for AI going forward?
SPEAKER_02:Yeah, I think this uh going forward, it's hard to say, but this is the frontier right now, and I think it should be. I think you're spot on. I think the reason is is most simply put, because it's the easiest thing to get started with that actually impacts a company's bottom line. When you're talking about RAG, you're talking about the one thing AI is most good at, which is which is taking a massive amount of information that no human can possibly go through, going through it automatically, and generating intelligent, informed, dynamic responses, no matter how vague or specific that information may be. And that's something humans have been doing for a long time and wishing for a long time in every business I've ever been a part of have run or run, it's all that menial labor that hurts your bottom line. Because it's like if we could just free these people from these menial tasks that a robot could do, we would make a lot more money and free those people up to do other things or even shrink our workforce or some of the other things that you're seeing. It's scary to say, but it's real. But yeah, it's true. Yeah. And and and repurposing those teams into more tasks that impact your bottom line and getting them away from those menial tasks that are really just costs on your on your PL sheets is a big deal for companies. And it's it's easy to get started and it gets powerful really quick. And it and it's instant gratification. It's so quick to see how powerful it is for people who aren't technical. And I think that's the reason it's getting embraced as well.
SPEAKER_00:Yeah, that's an excellent point. And I'm curious, what is the stuff you're doing now? Like if you can talk a little bit about what you're actually building with RAG, I have to be possible.
SPEAKER_02:I it's it I have to be careful. Um, I it's it's uh it's a product for um various facilities right now within uh the healthcare industry. Um it is going to eventually expand to other industries as well. But right now, it it very simply put, it's it's what we've already talked about with RAG, which is hey, we have these massive documents that we can't memorize or read or even understand half the time or interpret. That's another power of RAG and an LLM is the interpretation of that information. Um, and we we need to find it quicker and we need to show people who don't know this information or don't even know how to look for it or don't even know where these documents exist. We need to get that in their hands faster so that they can better do their jobs. Um, and and and what I can be a little specific about that's that's unique is in my particular implementation and what we're building, you learn very quickly the difference between RAG in an academic sense and the and RAG in an enterprise sense, because one of the things we had to do early on that really threw a curveball in our initial approach, which was very naive, what's called naive RAG, um was maintain some structural integrity within these documents and and have references and have links that people could click to and easily access the source document where that information was. So starting to I've done this. So starting to chunk and and and I know we're getting ahead of ourselves with some of this terminology, but yeah, starting to get that information in the database in a way that allowed you to actually retrieve it in a way that not only is just raw, naive retrieval, but is very intentional retrieval in a way that benefits your use case so that you can turn around and use that information to accomplish a business goal is where RAG starts to get very complicated and also way more powerful than than any of us really even fully understand right now so far.
SPEAKER_00:And there's no there's no real experts, uh, is what I've noticed too. I haven't met any like, oh yeah, I've got this figured out, and here's the exact way you should build this together. It's not like somebody that's like a SQL expert or like you know, React or something like that. There's tons of consultants for this kind of thing. In this field, it's kind of like you're just figuring out as you go along. Now, what I'd like to do is if you can take us like through like zero to one, like I think a lot of people have heard us say rag a lot, and I think they're like they're not really probably understanding like what it is we're talking about. Like, talk about how you get information into a system that is uh that can support rag, support retrieval augmented generation.
SPEAKER_02:Sure, without getting how technical do you want me to get? I I don't know who your audience is.
SPEAKER_00:Let's go kind of like yeah, that I this this audience spans the people that that are really new to code to people that are like more senior, but at this, but I think this is so new that most people don't even get the idea of like you have a document, and how does it and first getting the document and then putting it into a specific database and then having to use that database and then feed it to a large language model?
SPEAKER_02:Yeah. So uh a lot of the knowledge bases people already have are just digital documents, they're PDFs, they're spreadsheets, they're things like this. The the goal really is to take this information that may already be digital or not. Hopefully, you're working with a digital knowledge base already, taking some of that information and breaking it into parts in in a way that is well, let me start at the beginning. You're gonna take that raw text that's parsed from that PDF, and you're gonna create what's called an embedding. That's that's gonna be you converting that text into some uh algebraic meaning, some array of of number patterns, to put it simply, that are gonna get stored, that kind of represent the um lexical or semantic meaning of this text. Uh it's it's gonna be stored mathematically, and you're gonna break uh each part of that text as vectors that get translated to vectors along the way into smaller pieces that are more easily searchable, that are more easily uh assembled after the fact once you do search and retrieve them. Uh, but in but in essence, you you're creating small mathematical chunks that represent the meaning of that text that are searchable by uh linear algebraic algorithms that AI uses to discern meaning.
SPEAKER_00:I don't know how good or bad that explanation is. That's pretty good. I can I could give a short example too of like what I'm I'm building, which is not really private at all. So I'm at a company at doing TikTok influencer discovery, right? And so we'll go scrape the web or find TikTok influencers and we'll look at them, do some visual recognition using AI as well, and then we'll say, this person is you know a blonde-haired woman that makes content about children. She lives in Argentina and she posts about this frequently. And we did this for hundreds of thousands of people, or maybe a million at some point. Now we've taken we've done what you what you said, same same idea. We've taken this text data, and now we've we've said, now we're gonna store this in this vector database as numbers, as this embedding full of numbers, and we'll attach some metadata to it too, like the person's name, right? And then now when you search, you say, I want a blonde-haired person in Argentina. If you try to do that in um, you know, in TikTok, you're just you're not gonna get much. If you try to do it in Google, you're gonna get even less. It's gonna be just useless. If you try to do it in Chat GPT, same thing. So now we have this proprietary data, and you can search it. We say, boom, here's that woman you were searching for, or here's five women that are similar to this one that you were searching for, based on how we've like chunked them, chunked them up and stored them in this in this vector database. And that is rag, essentially. Now it gets way more complicated, obviously, because you're me and you are dealing with a lot of the other things. But how did you learn this? I learned this on the job because I kind of had to. And I was working with somebody that was way smarter that had done this um before he was like, you know, these are PhD level people that were that knew about AI long before AI became cool. How did you how did you pick this up?
SPEAKER_02:The same by total accident. I I had a friend, I had a friend who worked in the industry who's become my partner on the project. And cool. Basically, I I I they were basically asking me what I do, and I was communicating into the, you know, hey, we take problems that businesses have and solve them. Like, like, give give me an example of a problem you have and and let's talk about it. And as she started laying out a scenario in their industry, I was like, man, this is like a very simple problem to solve on its surface. Because it was it was this problem of like, we are governed by way more information than we can handle. Um, and I don't know how to search it, I don't know how to get it fast. And it was, and for me, it was like, hey, tech can solve this. Um, and and she didn't really believe me. It was like, you know, tech can't do this, you know, people have to do this because it's so specific. There are so many rules. So I literally just went in Chat GPT, went and created a custom GPT, gave it some very specific instructions, uh uh, but also bare minimum, right? And and took one of the documents that she had, put it in the the built-in knowledge base inside of Chat GPT, just to just to show her in a period of two hours, hey, okay, let's give it some instructions, let's give it this document. Hey, here's a question that only this document can answer. And for her to watch the LLM spit that out was like a light bulb for her, where it was like, holy shit, like this is incredible, this is real. And it it showed me very quickly that AI was a use case for this. I I had already built some things with AI and played around with the open AI API and things like that, and was starting to work with some of those things, chatbots and stuff like that. But as I explored this use case, I stumbled across RAG and was like, this is a picture perfect use case for for this process. This is what it was made for, exactly.
SPEAKER_00:Yeah, that that is really cool. And I I've done this for my own personal life too. I'm I'm uh on my kids' school board and I'm the parliamentarian, and there's all these rules that I have to know that I'm terrible at this stuff. And I'm like, there's no way I can commit all this to memory. The other woman had done it for like nine years. And so I did the same thing. I dumped all that information, all the documents, the bylaws, into a uh custom GPT. And I was thinking about making a little bit of a rag, you know, app around it too, maybe if I if I feel like it. And now I can like look up things in meetings. I'm like, okay, here's the bylaw that tells me this. Like, what do we do for budget, all this stuff? And yeah, it's really useful for this for this kind of thing when we have to summarize tons and tons of information. So I know people are listening to this and they're thinking, okay, I should do rag. What is the tech stack you're using? Um, I don't because I don't think I know exactly the tech stack you're using, but I'm curious, what is the tech stack you're using to do this? I uh to build these apps.
SPEAKER_02:Yeah, so I I'm one of those people who definitely believes in a different or near different tech stack for every product. It's like best tool. I I I think the best I think the best tools I've found for that. I'm using a Next.js front end for the user interface. I'm using a Nestjs backend, which a lot of people aren't familiar with, but it's a really powerful Node Express wrapper framework for the API layer. And I'm using Fast API and Python on the on the uh business uh back end or whatever, the RE to do all the RAG stuff. Um is what I'm using. And it's really as simple as that. It's a pretty simple stack, but then I'm using Pinecone for the vector database. So the RG right now is really industry standard. There, there are several very like clear leaders in the space right now that'll change, but right now that's the case.
SPEAKER_00:Um, so that that's that's what I'm using Pine Cone, Quadrant, I think Weave8, maybe I forget. Those are the ones I'm familiar with. I've I've used two of those three so far. Um yeah, I got to use quadrant and pine cone. I'm actually even using a simpler version. I'm I'm using just TypeScript in like Next.js and Quadrant, and that's it at this point. Really, really bare bones. Uh and we have like a whole Python layer too for the other like visual extraction. But yeah, I think that and that may surprise people. They may be thinking, what? And I'm like, yeah, you don't you don't have to like go get a machine learning engineering course or anything like that. You can do a lot of really cool stuff with full stack software development skills that you already have.
SPEAKER_02:Yeah, that's the I think that's the other answer to what you asked earlier about why it's so popular is because the learning curve is is the learning curve to get started is very low. Um because it's like you said, you can do it in any language, you don't have to have Python. We we chose that because we're planning on extending pretty quickly to the point that we have some transformations and and reads that need that kind of stack, but a lot of people don't, and and really under the hood, it's it's pine cone and open API, open AI's API. It's it's really the heavy lifting is done by the people who have built these tools. They can know all the math. We don't have to.
SPEAKER_00:Exactly. I I met a guy, I shouldn't even say this. This guy said he was going to build his own LLM. Yeah. Um trained it on billions of parameters. And I'm like, well, yeah, what who who are you gonna compete with, my friend? OpenAI's won, you know, uh Anthropic, they've won. They let them play the trillionaire fight out there uh in the headlines. There's no, we're not gonna win that. So yeah, it was kind of silly. So I think, yeah, we already have a good commercial option right now. Now it's more about how do we build on top of them. And I feel like a lot of developers are late to the party. I feel like for once in my career, I I feel like I'm really early to a really, really big thing. This almost feels like, and I'm trying to think of the last time I felt like I really missed something, and I think it was probably with the revolution of a cloud and and maybe React was the last time I feel like I really missed the boat. And this time I'm like, finally, finally here. And uh my my impression though is most developers have no clue about this stuff. But I don't know. Like, what is your experience?
SPEAKER_02:The same. And this is the boat. You're right. This this is the AI use case that's a sure thing, in my opinion. I I think there's a lot of AI, you touched on some of it earlier, code generation and things like that, that are are very uh popular ideas because of confirmation bias, and people very badly want those things to be true because some of those things are so expensive. They're like, if we can offload this, that'd be great. Yeah. Uh some of those things I don't believe very strongly in. RAG is here and it's here to stay, and it's it's gonna make a huge impact for companies. If you're not learning this yet, you better start because it's it's coming in a big way.
SPEAKER_00:Uh yeah. And I hate to sound like some salesy guy because I do want to actually do a course on this, like a whole program, and have you as one of the guest speakers, as a matter of fact. But like, this is like one of those things I'm like, oh, I'm so excited, I want to tell everybody because I'm like, this is this is the cool thing. And uh, I'm in San Francisco Bay Area, and I feel like every freaking AI startup that hits me up is doing some version of this. Now, some are very, very complex and some are less so. Yeah, and I'm doing a fairly complex version of this now. So there's a lot under the hood, like you've said. Um, but before I let you go, because this is another thing I'm genuinely curious about too. You're how do you you're an entrepreneur beyond knowing all this and learning on the side? You're you're and which is super impressive. And how do you where do you find customers? Because this is such a new technology, and you're not in what I consider like a big tech hub. How are you finding people and saying, I'm charging you for this technology that you probably don't really understand that's gonna solve a problem for you? What does that look like?
SPEAKER_02:Uh it's a sales pitch. Uh, because you're right. I'm I'm about an hour west of Atlanta. So so Atlanta's a tech hub, but I'm I'm not connected, I'm not going out to dinner and bars and meeting people in Atlanta, you know, every night. Um it is it is a sales pitch. Now, we also don't only do AI, we we build all kinds of software for all kinds of people. Um so we we do other things too. I'm I'm building a mobile app right now for a client that's that's pretty uh you know not simple, but it's it's it's a mobile app, is what we're used to. Um it's not AI. Uh but it's it's really a sales pitch. It's really yeah, you have to love it. You have to know how to communicate it in terms that aren't specific to a tech audience. Um, and you gotta find companies with problems, and and you gotta you gotta walk in the door with a solution to a problem you know they have. And and sometimes people are soliciting for that, but more often it's the scenario I described with this other person who became my client, which is just like, tell me about a problem you have, okay, I can solve that. And here's how and and when they don't believe it, show them how, as fast as you can and as simply as you can, and and prove it to them because because they're gonna have to see it to believe it, especially in the case of AI, because there's so much skepticism, there's so much fear. How much of this is hype and how much of it is not. If you can, if you can walk into the door of a company, so to speak, and and say, I know you have this problem, I know how to solve it, and let me show you what that looks like, then that's a very good step in the right direction. It's very hard to get people to pull money out of their pockets and give it to you, but that's that's how it has to start is solving their problems and showing them that you actually can do that and showing them you can do it quickly. One of the business differentiators for us is MVPs in 30 days. Whatever you have, whatever you do, we're gonna we're gonna help you scope out an MVP that can be built in 30 days. And in 30 days, you're gonna be up and running and generating revenue. And if and and that's been big for me so far is is so much of our competition is well, it's gonna take five to eight months and sixty grand. And and software developers have been scamming people for a long time. They sit around, they do nothing for six of those months. I hate to say it on a software pod, but it's true. That's true. It's true. They sit around for six months, they take your money, they don't do anything, and then the last two months they push it out the door anyway. Um, and if someone's working like that, I've seen this happen. Yeah, oh, for a hundred percent. I I worked with a company that had a legacy product that they went through five different teams that did that. Um that that that's the key though. Just care about people, care about the problems they have, and and come in with creative solutions about how to solve them. Don't give them work to do. Don't don't tell them how they need to solve their problems or ask them how they need to tell them what what they can do.
SPEAKER_00:I like that. And probably don't lead off with the tech either. I like how you say, like, what's a problem you're you're you need to solve? Like because I I've seen this a lot when I first did some, I've done some freelancing throughout the years too. And I was naive and I was like using buzzwords. I mean, I realized they don't know what the hell I'm talking about. They don't care if I'm using React or you know, Rag. They they don't want to know either. They're like, I don't know what the hell you're saying, you know.
SPEAKER_02:They're the longer the longer you can keep anyone from hearing you say anything technical, the better. If I can get through the entire life of a project and never use a tech term, that's like a perfect scenario. Yeah. Because they don't care. They don't want to, and they shouldn't. They don't care. Yeah.
SPEAKER_00:Yeah.
SPEAKER_02:I always cut through on that. There's a there, I get a lot of clients who come in and and say they start trying to, they've worked with tech teams before, they start trying to communicate in tech terms. And I tell them immediately right off the bat, hey, translating your business language into tech language is my job. I don't want I want you talking to me in the terms that your business uses. Give me your domain language, let me worry about how to translate that to tech. That's a big thing, too.
SPEAKER_00:Dude, that's yeah, that you're you are spot on on that. That's a super important one. That's some great nuggets. Um, before I let you go, I want to get a couple hot takes from you real quick. Sure. I got those. Cool. I yeah, I bet you do.
SPEAKER_02:Um boot camp or self-taught? Self-taught. I I believe very strongly. I won't go on and on about it, but I believe very strongly. And you'll see if you follow me on LinkedIn, you will see me ranting and raving and arguing with people about this all day. I think every single thing you need to learn to do this job right now, you can learn for free or very, very cheap on your own if you have the discipline and and and self-motivation to do that. Boot camps are way too expensive if you can summon the structure and discipline for yourself.
SPEAKER_00:I wouldn't disagree with you either, but it this is not my hot take time, so I'm gonna refrain from my hot take time. Yeah, self-taught TypeScript or JavaScript? JavaScript.
SPEAKER_02:I I think that so I I will delineate. I think that if you are working with a team, uh TypeScript is really valuable. I use TypeScript on the back end um for my API layers and things like that. When I'm building on the front end, I use JavaScript. To me, uh, the more velocity you can steal, the better. And and TypeScript will slow you down so much. And it type I heard it said by one of my mentors best TypeScript is for developers and JavaScript is for users. All the TypeScript you write is going to compile down to JavaScript at the end of the day, anyway. Users do not care. If you're working by yourself, do not use TypeScript. Don't use it, it'll slow you down. It's all about getting it built fast and out the door. And then if you want to convert it to TypeScript when you bring on a team, do that. But but I always start with JavaScript unless I have a real damn good reason to go with TypeScript.
SPEAKER_00:And you can you can do it incrementally too. It's not all or nothing, which I think a lot of people forget. Like, you don't you can do a little bit of both. I I like your take, and that was my take for a while, too. Like front end, I really don't want to add types, back end. I feel like, yeah, I probably need to add types now. I'm using Next.js, and I feel like I'm just kind of locked into using TypeScript everywhere now. But yeah, I've definitely gone through type gymnastics like everybody, yeah, which is always uh a pain. Yeah, all right.
SPEAKER_02:I am not a social media guy. I deleted all my social media accounts a long time ago. I think that I here is my hot take on this. Someone needs to come out with a platform that puts LinkedIn out of business because right now it's it's it's a dinosaur, it's doing nothing for anyone that it claims to be doing. It's not finding anyone work, it's not providing any networking on any real level. Someone needs to come in and replace LinkedIn. Now, if people have found a way to use X for that, that's great. Um, or other social media platforms, but I think the person who invents the next job search platform that solves the problems LinkedIn has is going to be very, very rich. It ain't gonna be me. I don't have time or desire, but someone should do that. I'm putting it out there. Someone does.
SPEAKER_00:LinkedIn is trash. I I it's I used to love it too. Like I I still kind of like it, but I've I've just I'm not into it like I used to be because I used to actually meet people and make genuine connections, like meeting you. Like this kind of thing happened more often just a year ago. And then I think, what the hell happened recently? And a lot of bigger creators are are leaving the platform, I noticed, for that very reason. Because like I'm just talking into a void now, and it's just like corporations have taken it over, and you can't find a job. It's a social media site that's like kind of like a job site that you got to treat like a social media site in order to find work. It's the strangest way of finding jobs I could imagine. Like, so it's like Instagram for jobs, yeah. It's super weird.
SPEAKER_02:It's ripe for it's ripe for replacement.
SPEAKER_00:Last one prediction for the next year.
SPEAKER_02:Yeah.
SPEAKER_00:Are we cooked in the industry or do you see things taking a turn?
SPEAKER_02:No, some people are cooked. So you're you're talking about in terms of employment and and layoffs and things like that. Yeah. This is a long conversation. I I think my personal belief, and this is from someone who's new, from someone who wasn't here for when this happened, so it's purely anecdotal. But my experience as an outsider who's who's also worked with dev teams in the past as an outsider, my experience is that a lot of people snuck into this industry when it was new who don't have the work ethic to sustain a career in this industry, both companies and individuals. And I think those people are cooked because I think there is a rude awakening coming for people who are phoning it in and the people who work their asses off and really deliver and really charge fair rates in 2025 with the advent of things like AI assist and stuff like that. I think those people are going to dominate and own this industry. And I think that it's very true that one developer is gonna you is gonna be able to do what two or three or five or eventually 10 used to be able to do. And and in that sense, a lot of people are cooked. But the people who mean it and love it and live and breathe it and and know how to stay ahead of the curve, they are going to uh reap the rewards of that rude awakening. They are they are gonna be the the people that inherit all of the fallout and and turn it around and and really start doing some cool new things that have never been done before.
SPEAKER_00:I'm seeing I don't know if I agree with all that, but I but I agree with with a lot of what you're saying, especially the idea that remember those day in the life TikToks where people like bootstraps. Okay. Anyway, well, there were there was a whole thing on people like I drink a coffee at work and I get paid 400k a year. And I'm like, yeah, and this in this attracted a lot of people like at Parsi, that's one of the things we we we we we are a mentorship program, you could almost call it a boot camp, but it's a mentorship program for people that want to get in the industry. And one of the things we always look for is like if you're just looking for a way to like just make money, there's a ton of ways to make money. Yeah. Um, if you don't find some intrinsic joy in this, yeah, you're likely not going to do well. I've yet to meet somebody that I respect or that I think is really good at what they do that's like, I don't like programming. It's like you kind of have to like it. Like if you don't like this stuff, you're you're not gonna do well. And if you're only in it for the money, there was a point when people were getting like crazy amounts of money for learning React. There's like that little tiny window. Yeah, and I do think that a lot of those people are gonna find themselves in a rude way, being completely honest. Uh, and yeah, that's a maybe a mean take, but I do see that as reality.
SPEAKER_02:And you gotta start thinking like businesses think. If all you can do is sit down and take instructions and write code, then you're done for it. Oh, yeah. You you have to start making that jump to to business domain knowledge and and do more than just write code because those days are coming to an end quick.
SPEAKER_00:I do feel like that. Yeah, that's my my job has dramatically changed over the years from actually being kind of like a a task taker. Yeah, like here's like a little nice little report of what you're going to do to being like, let's all figure out what we're building, how we're gonna build it, and then you also are going to build it. So now I spend a lot, I mean, a lot of time translating tech requirements or talking to the CEOs at the company or people that are non-technical and figure out what are we actually gonna do. And I think that's uh what a lot of developers are feeling because the the teams are getting smaller, they're getting rid of mental management and big tech firms, they're still hiring lots of developers at a lot of places, but they are expecting you to do more than just write code, like you've said. Interesting time, super interesting time in many ways.
SPEAKER_02:And I and I think it's uh I think it's an opportunity for entrepreneurs to enter this space. And because there is a big opportunity right now to go in and compete with companies much larger than you can compete with in any other industry. And and I think I think the the uh environment is very rich for that right now. So if you're an entrepreneur looking to get into something really exciting and fun and crazy, um, this is an option because I I think I think entrepreneurs are going to make a big dent in the tech landscape in the next few years here.
SPEAKER_00:I I I love what you're saying. I love this conversation, man. You're keeping thanks for keeping it real. Yeah, for sure. And uh, where can people find you online? It gives more of your hot takes.
SPEAKER_02:Well, I'm on LinkedIn. Uh Ben Johnson, you you can follow me on LinkedIn if you want to if you want to see. I I get closer to getting banned from LinkedIn every single day. So if you want if you want to see somebody get banned in real time, follow me on there. Uh, but then you can you can find me on my website at lobysoftware.com as well.
SPEAKER_00:Um that'll be in the show notes too.
SPEAKER_02:I I float around and and eventually eventually we'll have a podcast coming. It's gonna be called the Panic Loop Podcast. We'll probably start that up uh in the next six to nine months. We'll see how that goes. So just keep an eye on that too.
SPEAKER_00:Sweet. Yeah, let me know. Um, I'd love to help you promote it, maybe maybe even be a guest. I don't know. Absolutely.
SPEAKER_02:We will return the favor. Thank you, man.
SPEAKER_00:Yeah, appreciate it. Well, thanks so much.
SPEAKER_02:Appreciate it.
SPEAKER_00:That'll do it for today's episode of the Develop Yourself podcast. If you're serious about switching careers and becoming a software developer and building complex software and want to work directly with me and my team, go to parsity.io. And if you want more information, feel free to schedule a chat by just clicking the link in the show notes. See you next week.
Podcasts we love
Check out these other fine podcasts recommended by us, not an algorithm.
