Just Vibes? Software Engineering Doom or Productivity Boom?

Transcript is auto-generated and may contain errors.

Hey folks, I'm driving into work. It is Wednesday morning. Um going to take the topic uh from Reddit, like some inspiration from Reddit, I suppose. I'm not going to read it out directly, but the this is from experience devs. The um the thread was really like about AI hype or not, like when's it going to when's it going to die out? Is it going to die out? And I thought this would be kind of interesting to talk about. Um, it seems like when I do videos on AI or perspective on AI, um, probably just cuz it's very trendy right now, more people seem to watch. Uh, so selfishly like that's kind of helpful, but also seems like people are interested in talking about it. So, uh, probably if people are watching it, then it's cuz they want to see it. So, figured we'd dive into that.

Um, and then just wanted to remind folks like if you're new to the channel, I take questions in the comments. So, if you have a question about software engineering, career development, that kind of stuff, leave it in the comments. Um, otherwise, you can look for Dev Leader on social media. That's my main YouTube channel as well with more edited down videos, tutorials, podcasts as well, u, and live streams every Monday at 700 p.m. Pacific. You can message me. Um, I am basically on every social media platform. Uh Nick Causantino on LinkedIn. Send me a message, ask your question. I'll keep you anonymous that way. Um and yeah, I try to drive the channel based on questions that come in. Otherwise, I go to Reddit and dream up other topics. So, um AI bubble, you know, is it just hype? Um I have a couple of different thoughts on this.

I think um let's let me start by saying I think that when I talk through this I'm going to be kind of I want to talk about different angles that might sound like they're contradicting each other but part of me wants to say yeah there's a lot of like AI bubble going on a lot of AI overhype but at the the same token I want to be able to say like it's not going away and and talk about it from from that angle. So, I wanted to start with the hype um because I think that this is mostly what people are seeing, observing. It's a real thing, right? Like, uh there's a ton of stuff being posted literally everywhere. Um and we have to kind of think about what's going on, right? So, when we think about why we see so much stuff trending, there's a few reasons for that.

one, um, you're going to have stuff that comes from big companies trying to sell things. Of course, it's just like I mean, it's just reality. So, you might have the perspective, and you're allowed to have whatever perspective you want that like, hey, it's just overhyped. It's this company selling snake oil. Um, you might truly believe in all of it. I'm not I'm not saying right or wrong. I'm just saying you're literally allowed to have your own opinion. But the reality is we have companies that are trying to sell things. They're businesses. They absolutely will be promoting products. Um I I personally do not believe that companies like big companies that are promoting their AI products, I personally don't think that they're like sitting there going, "Oh, can't wait to fool these people. Like can't wait to get a bunch of idiots to pay us money." I don't like I think I don't know.

Personally, I think that it's a bit of a naive way to look at things. Like, you might not like a company. You might not agree with their beliefs. You might not agree with their values. I just don't think a lot of companies structure themselves that way. Now, I'm not saying there aren't any companies, especially because you'll have people that are like, "Hey, we can ride the AI bandwagon. You know, we can we can surf this wave." And, you know, basically, it's kind of like crypto and stuff, right? But that's why I'm trying to address like big companies. I'm talking about like you know uh Meta, Google, Microsoft, Apple, like whoever open AI any company that is doing something with AI that's big and an established company and not just like something that someone you know started in their basement and uh these companies they have their brands, right?

like they're not just going to sabotage all of that and like try to like trick you into AI things. So my point is that these companies believe in the products and services that they're creating whether or not they're overly ambitious or exaggerating or whatever else. I mean, TBD, I'm not I'm not saying no one's doing that, but I don't think that these companies are trying to fool you into like this thing isn't going to do anything what we're saying. We don't even believe it, but we're going to trick you. Okay, that's my take on things. You're totally allowed to disagree with that. That's I mean, that's why we all have different opinions. But these companies are trying to sell things, okay? So, they're going to be promoting it. They're going to be talking about all of the good things. It will be disproportionately good things. It would be ridiculous if it was the other way around.

Companies trying to sell you AI products and they're like, "Yeah, but it's always terrible." Like, it's not going to happen. So, just as an example, I work at Microsoft. Obviously, we have things like co-pilot. There's a huge push for AI at Microsoft right now. I have been in conversations with people that are in product areas at Microsoft in leadership positions that have outright said like absolutely we want everyone you know to be using AI and trying it out. We want your feedback because we know it's not perfect like we know it's not but the only way that we make things better is by getting feedback. It's like any product right? So they're fully acknowledging like sorry this I thought this guy was going to walk and it's literally the exact opposite color light would have been a disaster. Um, so like that's the kind of and I'm not saying other companies aren't doing this, right?

But I mean I'm just trying to give you an example of like I I also do see the other side of it where where companies are saying, "Hey, look, no, it does make mistakes. No, it's not perfect." Like, yeah, there's challenges we got to sort out, but you will disproportionately see positive things coming from people trying to sell AI products and services. It's just kind of the reality. That's one part. Okay. So, that's going to be making headlines. There's always going to be new things. There's it's just the way it is. Now, you add on top of that, you have like do we like like can we think about how the news functions? A lot of people and sorry, this is probably going to sound like conspiracy theorist but like I don't really think I'm a conspiracy theorist kind of person at all, but um let's talk about the news for a second, right?

I know it's called the news and I realize that we want to believe that the intention of the news is to deliver us the news. Um, the news is a business too. So, uh, the way that the news makes money is that you click on the article or you watch the video and there's ad revenue. There you go. Like that's that's it. Is there news in there? Yeah. Like that's that's why you're clicking it is to get information. But they don't like news companies don't just make money because here's news. Like they it's just not how it works. They're making money because they have a business model that requires you to consume the content. How do they get you to consume the content? They need catchy headlines. They need catchy cover photos. It's basically like a YouTube video, right? Like thumbnail and title. People got to click They need to get your attention so that they can serve ads.

It's like, so what I'm trying to get at is that you're going to see tons of stuff in the news, tons, that's going to get your emotions going. This is this is how things work. I'm saying this as someone who makes YouTube videos. I've made over 600 YouTube videos now. I'm not saying they're all like I'm amazing at this, but one of the things when you make YouTube videos, if you're trying to learn how to get better, is literally that the content in your video will not even matter if your title and thumbnail are not going to get people to click it. Those are the only two things that people see before deciding to watch your video. It doesn't just accidentally happen, right? It's not like you type random URLs into Google maybe that or into YouTube. That might be a fun feature, but like it doesn't work that way.

You scroll through and you go, "Oh, that one." Based on the title and the thumbnail. So, how do you get more people to click things? way you get stuff that's going to get them curious or get emotions going. It's the same thing with stuff in the news. So, what I wanted to do here was explain that you're going to see stuff coming up in the news for that exact reason. They're going to be talking about AI. They're going to be talking about things like layoffs. They're going to be talking about things that are going to get you scared. They're going to get you excited. Like, anything to get emotions going. And then who else is doing that? every single content creator that can possibly incorporate AI somehow into the content they're making. What am I doing literally right now with this video? I said it at the beginning, right?

More people seem to watch the videos where I talk about AI. Guess what I'm doing right now? I'm making a video talking about AI in hopes that more people will watch the video, right? Um, what I'm not do, and I'll probably have some ridiculous title that's going to hopefully make you click, but I'm not making thumbnails for this channel really, so I'm missing out on that. But point is, you have content creators. So, you have companies that are blasting AI. Um, sorry, there was something on the road that was like a big piece of plastic. I had to swerve around it. Um, I only said that because I don't know if you could hear the I don't know what you call it. What do you The shoulder the shoulder is like grooved so it makes uh rattling sounds. Anyway, my point is that you're going to have everyone making noise about this stuff.

Yes, it's going to be a huge hype train that everyone is trying to jump on because it's going to draw attention. So, 100% there is it's overhyped. Absolutely fully agreed with this. Is that going to fizzle out? Yes, it will at some point. It's not for the rest of our lives. We can't be talking about AI is the most amazing thing cuz there will always be things. AI is going to be like it's it's not going to like it's not going to die off like this year or something, but people are going to be talking about it for a while and then it's going to be like okay, we've heard lots about it. It will become sort of like this normalized thing and then there will be new things and more things that will keep coming up. Okay. So, no, I don't think it's just going to disappear, but it will slow down.

Yes, this is overhyped right now. Now, the other side of this that I wanted to talk about though is that I'm so far I've been talking about this from the perspective of like people sharing stuff online, sort of like marketing and social media and that kind of stuff, right? Like that's the part that I think is absolutely overhyped. when it comes to the actual technology. Um, this is the part that I'm my stance is like this this part is not just going to like disappear. Uh, fully fully believe AI is here. You have the biggest companies in the world extremely invested into it. just to I'm going to maybe draw a little bit of a a comparison here and I don't mean to be insulting to any any single company but I kind of have to do it here for this comparison.

This is a little bit different than meta being like metaverse right you didn't have every huge company in the world going this is the thing right now you have billions and billions and billions of dollars from big the biggest companies in the world being invested into this space. Okay. What I like I don't know from a technology perspective. What I'm not saying is like anything that's ever been uttered in terms of promises of how great it will be that will realize all these dreams. Like I don't know that. I'm not you know deep invested into AI research and how things are advancing that way. I I don't know. I see the practical application of these things as they come up. So it's not for me to comment on that. But I do know that companies are investing billions and billions and billions of dollars.

So it's either they all were like, "Oops, I guess this was wrong and this stuff doesn't work and no one's seen any value." But I think at this point they're saying people see enough value and there's enough promise here that we need to heavily invest if we want to be able to reap the the benefits of this going forward. So no, I don't think that this is a bubble in terms of oh these companies are just doing this and then all of a sudden it's going to disappear. Like I think we're I think we're a few too many dollars past that at this point. Just from my perspective. Okay. Again, you're absolutely allowed to disagree with this. I would love if you have a different opinion on this if you think that all of these companies have just made a mistake and like this technology is not going anywhere.

Like that's what the comments are for. I just ask that you be respectful. Otherwise, you probably get I don't know made fun of by other people for being an But, you know, if you have comments where you're disagreeing, like, please share them. It's totally cool. Different perspectives are absolutely welcome. But, um, I don't think the techn is going away. The, um, the part that I think is interesting, and I wanted to spend a little bit of time on this, is like I think what people see on the outside of companies in terms of decisions being made and things like that. Um it's it's sometimes it's a little bit opaque and then we have to make assumptions about why businesses are operating a certain way and then you have things like social media everyone's got their own opinion whatever right I read it here I read

it there someone said this so on so forth but like if you take a company Unless like you've talked to the board of directors and the CEO and the people that are literally making those decisions, unless you've heard it directly from them and you were in like the decision-m meetings with them, you don't actually know the reason. You have an opinion or a belief about why and it could be right. It could absolutely be wrong. Right? But you don't actually you don't actually have the data that that backs that up. And that's that's okay. But like I guess my point is that you need to accept that. So I think too many people go, "Oh, well this is exactly what's happening and now it's fact." And it's like it's not fact. It's it's just your opinion and you're allowed to have it, but you need to acknowledge it's an opinion, right?

Just like I've been saying in this video, I I think that the the hype that we're seeing is going to die down. That's my opinion. I don't think the technology is something that's just going to disappear. That's my opinion from my lived experiences, right? You might disagree with that. But when people start to say, "Well, companies are making these decisions based on these reasons." I'm like, "You don't actually know that. It's an opinion." And you're allowed to have it, but it's an opinion. So, the biggest thing that's coming up right now, I feel like in terms of uh like the fear-mongering and all that is like, "Hey, well, because this is for software engineers, right? AI is going to replace software engineers. We have vibe coding now. anyone can be a software engineer, right? Like, which is hilarious. But, um, what's really cool is that you do have people that were non-technical before able to create solutions that that involve code.

This person's going under the speed limit in the fast lane, and I have someone in a modified Mustang tailgating me. It's like this work pickup truck in front of me that's going 10 miles an hour below the speed limit. It's like it's one thing to be like speeding dangerously. Yeah, this guy's doing what I want to do. Uh man, like but when you're going under the speed limit, it's it's equally as dangerous because then you have other drivers that are trying to do the speed limit and then they have to drive in certain ways to get around you. Kind of stupid. I don't know. Anyway, what was I even saying? Um, it's very distracting when you have like really, really bad drivers. That's super dangerous. I'm looking at all the people that are trying to go around this guy because they have to. He's going, we're in the fast lane.

He's going 10 miles an hour below the speed limit. Like, it's not okay. And there and if you can't tell because you can't see, there is no one remotely close to being in front of me now that I've passed. Okay? So it's not like he was going slow because there was someone else. So on the, you know, decisions companies are making live vibe coding, we got AI replacing software engineers. I think sometimes people are seeing like they're they're kind of reading what they believe is the writing on the wall which is okay we have AI that can program things we have some CEOs and executives saying like oh like software like software development is done like you know anyone can do this now that that kind of stuff I think is 100%. Um what I don't think is though is that um software development is changing.

I think it is already forever changed and will continue to change for the record. That's my opinion on it. But it's not it's not dying off. The proof we have is like you have the people that are doing the vibe coding kind of thing. And apparently I was corrected on this. Apparently, vibe coding is when you're not even correcting the LLM. You're just like blindly taking anything it does. And I'm like, I don't even know how anyone's doing that. Um, if I'm, you know, air quotes vibe coding, I'm certainly reading what it's saying and then going, "Oh, no, no, like change this, change that, or did you consider this?" And someone was telling me, "No, no, that's not vibe coding." So, I'm like, "Okay." Like, uh, sorry. I'm just trying to do a better job. But um the uh the reality is that people that are building solutions like this, they're okay for for quick things.

I think it personally I think it's great and I mean this genuinely I think it's super cool that a barrier has been lowered where people can create things. I love it. I think it's super cool. What I think is not super cool and is basically super stupid is to go, "Hey, look, I have something that's been created. Therefore, we can replace an entire group of people. You're now obsolete." And the problem with that is that you actually don't have any domain experience to be able to make a statement like that. Like, it's just illinformed and dangerous and quite frankly, it's wrong. as someone who's been doing this for many, many years. Um, now with that said, will the tooling get better? Absolutely it will. People are putting too much money into this kind of stuff for it to not get better. Okay, it's going to get better.

We see it getting better all the time. Does it reach a point where you literally don't need someone who has any experience as a software engineer to be able to go build things? Perhaps we're or not anywhere close to that right now. If you have zero domain experience as a software engineer, right? you don't have that skill set and you try vibe coding your way into a production app. The the AI is only doing what you're telling it to do. If you don't know how to ask it questions about security and scalability and all these things, you're basically just going to run into those problems. It's not out of the box just doing that for you. Now, could people put together prompts and better instructions to get it there where like it's forcing you down these things? Yeah, I think so. But that's not what's happening.

And you're still using information from like a software engineering background, right? So anyway, my point is I think that's another overhyped thing. Uh I think that the people that are going to benefit the most I think two groups. One is the people with little to no technical experience that can at least see some things come together. I think that's super cool. I think that's super powerful, but they have to understand the limitations of that. That's one group of people. The other group of people I think is the the actual domain experts, the software engineers, the people that understand this stuff. And the reason why is because now they have at their fingertips someone that they can basically pair program with, pair a design with. They can bounce ideas back and forth. They can ask this thing to have things explained in different ways. I think that's super powerful, right?

Like I've been I've been programming not professionally but uh what for what 22 years now and professionally for like 15 years I've been in the industry and I am still finding that like there are better ways and more effective ways for me to use an LLM to go design things. Good example is recently I'm trying to build a notification system in brand ghost. Now it's really trivial to generate a notification in memory and then pick any type of library you want to go send that notification, right? You want to send it to a Slack channel, you want to send an email. Easy. This stuff's already been done a million times, right? What hasn't been done a million times is I need in my system a way to be able to batch notifications and have them sent out in groups at certain intervals depending on users configuration.

And I also need to have support for potentially different end points like should that be a push notification, should it be an email, like something like that. I need a little bit more customization. So I I sat down with Chad GBT and built this thing together with it. And what was really cool was that I could give it, you know, here's what I want to do. And it kind of sets me off in a direction. And I read through what it's saying and I'm going, okay, is like that feels that feels like highle direction I want to go in, right? I have some ideas in my mind for how I want to approach it. And then it's like, okay, let's go let's go do the next thing. So I'm like, okay, let's try it out. And then it tries to write up some code or talk about the design.

And then I'm able to look at that going, no, like that's absolutely not going to work. like glaring holes here or yep that looks pretty good but um you know the schema for this database table like it's got some things like I'm not really a big fan of or sometimes it would do things and I'm like hey like uh what was a good one um there was like a state column and then like a boolean processing column and I was like why would you do that like isn't processing just another state um and usually when I do this kind of thing and I ask the LLM, I'm like, "Hey, like why did you do that?" Usually it's like, "Oh, great point. Let's not do that." And then I'm like, "Okay." But like it did it for a reason. I'm kind of curious like why. In this particular case, it didn't back down, which I thought was super cool.

And it was telling me like, "Hey, that that state is used to track this information." The processing uh and it's sort of like a final state. The processing flag is more about like an intermediate state about which job processor actively owns that. So it's more transient. And I was like, "Oh shit." Like interesting. So it was trying to say you could do this in different ways, but that's why I've split it into two different columns. And I'm sitting there going, "Cool." Like I'm I'm very glad that I have someone some robot that I can talk to you about this. That's interesting. I didn't think about that. We go building some other things. Uh, one of the things that it kind of flopped on was I was telling it about I want to minimize round trips to the database when I'm building some of this stuff.

And so we were building it and in my head I could see the direction it was going and I'm like I think we're going to do a lot of round trips here. So was suggesting code was I was kind of like vibing it into the into the code base and um but in my head I'm like I know I'm not going to settle on this. Once I saw it kind of come together more I said great see what it's doing and then there was like this triple nested loop on the outside and on the inside was a query. So what this thing was going to do was for every user that we have on our platform, it was going to make five times for every user uh was going to do another loop for the different types of notifications we could do. And I was like, man, this is like absolutely insane.

So if we had 10 users, it would already be doing like I don't know like 150 queries or something crazy. 150 200 queries. And so I was like, "Hey chat GPT, like remember what I said about minimizing round trips?" I'm like, "I'm pretty sure we could reduce this into like one loop on the set of results." Like do you think we could do that? could we flatten it down? And it was like, "Yeah, great point. We can." And then it did it. It collapsed one of the loops. And I said, "Hey, I think that's a good start. Do you don't you think we could collapse this next one down as well?" And sure enough, it was able to do it. Right. My point with walking through this example was like it was extremely helpful for me to be able to have an AI tool that I could bounce ideas back and forth with.

Um, for something like this, this was a bit of a daunting thing that I wasn't super excited to to work on. Knew it had to get done. Um, but like I'm like, ah, like starting on this is going to kind of suck. Um, it helped me get the momentum. Uh, helped me kind of think through things different ways. And I think what really saved my ass was the fact that I know how to build software systems, right? I could sit there and ask it questions and challenge it. I could catch things like a triple nested loop to do a query. So, it's not going to do a million queries. People that are vibe coding that kind of stuff in that have zero programming experience. Does it work? Yes, it will work for one users. it will work for 10 users. It's going to suck. But then you get a 100 users and all of a sudden you're like, why is everything falling apart?

And this is just for the notification system, right? So it's just like yeah, I don't know. It's I don't think we're there yet in terms of just like replacing people. So, I think that that's a bit overhyped. Um, but I do honestly I do think over time we'll see that uh becomes more and more and more capable. I'm excited for that. Uh, I think some people are still nervous about it. Uh, but I'm I'm genuinely excited to see that because uh I don't know. I think if you stay on top of this stuff, you see where the gaps are. You see how you fit into the picture still as a software developer. I think there's still tons of opportunity for us. So, I'm not I'm not concerned. But marketing hype, absolutely. I think that will die off. I don't think the technology is going away. I think the technology will continue to get better.

I don't think that AI is replacing developers. I think that when people see that there are uh that there are layoffs, that there are hiring freezes, and they're going, "Oh, it's because the AI is replacing developers." I don't know. Um that's one take. I think that there are other takes. Um I don't know. I don't know a good way to say it without revealing more things, I guess. But um maybe maybe acknowledge that there are potentially other ways of looking at things versus like oh it must just be the AI replacing people. Something to consider different perspectives. Um but I think that's all for today. If you have questions and stuff that you want answered outside of this, just leave them in the comments. I'm happy to go through. Look for Dev Leader on social media. Send me a message. And otherwise, if you have questions or comments about anything that was covered in this video, would love to kind of hear your thoughts in the in the comments.

If it's something that warrants a whole other video, I'm happy to go make a response video to that. Um, be kind to each other, please. Thank you. We'll see you next time.