Latest / DevQuestions with Tim Corey / 234. What Is Next For Software Development?

Transcript

234. What Is Next For Software Development? · DevQuestions with Tim Corey

• 0:00
  SPEAKER_01: What's coming next for the software development industry?
• 0:04
  SPEAKER_01: We live in an ever-changing world where technology is constantly evolving.
• 0:09
  SPEAKER_01: As software developers, we need to protect our future by planning for the future.
• 0:15
  SPEAKER_01: In this episode of Dev Questions, we're going to talk through what is coming in software development and how to prepare for that future.
• 0:30
  SPEAKER_00: Welcome to the DevQuestiPast with Tim Corey.
• 0:34
  SPEAKER_01: Software development is more than just writing code.
• 0:37
  SPEAKER_01: So let's talk about the rest of it.
• 0:39
  SPEAKER_01: Specifically, let's talk about the future of software development.
• 0:43
  SPEAKER_01: Let's start by going over the big buzzwords in our industry in 2024 going into 2025.
• 0:50
  SPEAKER_01: Then we'll talk about what is hype and what is real.
• 0:54
  SPEAKER_01: From there, we'll talk about what we should do to prepare for these changes on our industry so that we're setting ourselves up for success.
• 1:03
  SPEAKER_01: So let's start off with the buzz in our industry.
• 1:06
  SPEAKER_01: What's the things that are in our industry?
• 1:08
  SPEAKER_01: And of course, you know, number one, you're probably already yelling it at the screen or whatever you're listening to this on.
• 1:14
  SPEAKER_01: And that's AI.
• 1:15
  SPEAKER_01: That's, you know, no-brainer.
• 1:17
  SPEAKER_01: Everyone knows AI is here, but also coming.
• 1:21
  SPEAKER_01: And everyone's going to be putting AI into everything.
• 1:25
  SPEAKER_01: So that's number one on our list of things that are changing or the buzz in our industry.
• 1:30
  SPEAKER_01: We'll come back to what you'd do, but that is that hype or is that real?
• 1:35
  SPEAKER_01: But let's go on to number two, and that is data mining.
• 1:40
  SPEAKER_01: Now, this is very close to number one, because scraping data from sources to create LLMs or large language models, AI, and protecting your data from being scraped, these are important things to think through.
• 1:53
  SPEAKER_01: This is something that companies are starting to work through.
• 1:57
  SPEAKER_01: Because scraping data can actually kill a company.
• 2:01
  SPEAKER_01: And we'll talk more about that as well.
• 2:03
  SPEAKER_01: Number three, cloud versus local hosting.
• 2:07
  SPEAKER_01: What's the right way to host your web application, especially?
• 2:11
  SPEAKER_01: And this has become a more vocal thing to talk about is the idea that it's sometimes going backwards.
• 2:20
  SPEAKER_01: We started off with local hosting, then we had VPSs or virtual private servers, and then the cloud.
• 2:25
  SPEAKER_01: And the cloud is kind of like this big important thing that everybody decided to put everything in.
• 2:30
  SPEAKER_01: And now companies are going the opposite direction sometimes.
• 2:34
  SPEAKER_01: So we're going to talk about is cloud hype or is the local hosting hype or is VPS the way to go?
• 2:43
  SPEAKER_01: So we'll talk about that.
• 2:45
  SPEAKER_01: Number four is web everywhere.
• 2:48
  SPEAKER_01: Turning desktop applications into web applications and then taking those web applications and putting them on the desktop and mobile.
• 2:56
  SPEAKER_01: So here's a great example of this, and you're gonna yell and throw something.
• 3:02
  SPEAKER_01: If you work in a corporate environment, you probably work with Outlook.
• 3:06
  SPEAKER_01: Or more accurately, New Outlook, which New Outlook is web-based but still a desktop application.
• 3:14
  SPEAKER_01: And it quite frankly works horribly.
• 3:16
  SPEAKER_01: There's a lot of problems to it.
• 3:18
  SPEAKER_01: But that's a more common trend because things like VS Code.
• 3:24
  SPEAKER_01: VS Code is actually a web type application, meaning it's HTML, CSS, and JavaScript that's that's kind of running it, but then it runs inside of a tool, I believe it's Electron, that allows it, turns it into a native desktop application that can run on Windows and Mac and Linux and also run the web.
• 3:47
  SPEAKER_01: So web everywhere is going to be a big deal.
• 3:51
  SPEAKER_01: So we'll talk more about that.
• 3:52
  SPEAKER_01: And then number five is what to do with passwords or better passwords.
• 3:57
  SPEAKER_01: So it used to be that username and password, and then we'd you know hear things like, well, make sure your passwords eight characters are longer, make sure your passwords are longer than eight characters, make sure it has an uppercase and lowercase and a number.
• 4:08
  SPEAKER_01: And now we're getting into make sure you have two-factor authentication turned on.
• 4:13
  SPEAKER_01: Or some companies are even moving to passwordless systems, how that works.
• 4:19
  SPEAKER_01: So knowing how to deal with authentication and knowing how to get beyond just a login and password, which really isn't secure anymore in most cases, knowing how to do that is going to be important going forward.
• 4:32
  SPEAKER_01: So those are the five things I think that are kind of the big things in our industry going forward as we look at software development.
• 4:39
  SPEAKER_01: Now, let's try to separate the hype from reality.
• 4:43
  SPEAKER_01: And let's start with the big one, AI.
• 4:46
  SPEAKER_01: Companies are trying to put AI into all their products.
• 4:50
  SPEAKER_01: I know of managers who are actively asking the question: how can we use AI more?
• 4:55
  SPEAKER_01: How can we stick AI into more things?
• 4:57
  SPEAKER_01: And on one hand, that's a good thing because AI can bring more efficiency.
• 5:03
  SPEAKER_01: It can bring better search results, it can give you more information more quickly, which means that people can do their jobs more quickly.
• 5:12
  SPEAKER_01: I use AI search a lot because of the fact that it allows me to synthesize things and it gives me a better answer usually than trying to Google search or trying to, well, Bing search, I guess.
• 5:25
  SPEAKER_01: But rather than trying to find this search on a search box on the internet, I go to an AI, I actually use ChatGPT a lot, where I go there and say, hey, how about this?
• 5:36
  SPEAKER_01: Or using Copilot in your code to kind of work along with you and do a pair programming thing.
• 5:42
  SPEAKER_01: Those things can make you more efficient.
• 5:44
  SPEAKER_01: So on one hand, figuring out how you can use AI more is a good thing.
• 5:49
  SPEAKER_01: If you're looking at AI and saying, hey, how can we be better able to do our jobs?
• 5:54
  SPEAKER_01: How can we do our jobs more quickly?
• 5:55
  SPEAKER_01: How can we do our jobs in a way that's more efficient?
• 6:00
  SPEAKER_01: That's a great thing.
• 6:01
  SPEAKER_01: But on the other hand, AI can create an absolutely horrible mess.
• 6:08
  SPEAKER_01: So when you start relying on AI to make decisions, you create massive problems.
• 6:15
  SPEAKER_01: There have been court cases over this.
• 6:16
  SPEAKER_01: There have been companies that have promised things that they never would have promised because an AI had access to their chat system.
• 6:24
  SPEAKER_01: Hallucinations cause problems, assumptions are built in, sensitive data is exported to third parties.
• 6:32
  SPEAKER_01: There's so many things that are causing problems with AI that that can be a major issue.
• 6:39
  SPEAKER_01: And so with AI, I can see this as being both a good thing and hype at the same time.
• 6:46
  SPEAKER_01: So that's one of the things we need to think through is it's not just, you know, is this reality or is this hype?
• 6:54
  SPEAKER_01: It can be both.
• 6:55
  SPEAKER_01: And AI, I think it's both.
• 6:58
  SPEAKER_01: So the bottom line is that you need to be cautious about AI.
• 7:02
  SPEAKER_01: Anyone who starts making sweeping statements about what AI is going to do for them is probably not technical enough to actually understand how AI will do that.
• 7:12
  SPEAKER_01: AI isn't magic.
• 7:13
  SPEAKER_01: And I'm using AI here as a stand-in for all LLMs and other similar tools.
• 7:19
  SPEAKER_01: So AI is not magic.
• 7:22
  SPEAKER_01: It's a great support tool, but it can't be relied on for decisions.
• 7:27
  SPEAKER_01: The key here as a developer is to both embrace AI as a tool to help you and help others do their work faster and better, while at the same time not falling into the trap of putting AI where it doesn't belong or rely on AI to make decisions for you.
• 7:46
  SPEAKER_01: Be extremely cautious of any new uses of AI.
• 7:51
  SPEAKER_01: So when I have people work with AI and they say, well, I don't know how to do this, I'll ask, you know, GitHub Copilot.
• 8:00
  SPEAKER_01: GitHub Copilot may give them an answer.
• 8:04
  SPEAKER_01: But if you don't know what that answer is doing, you shouldn't implement your code because you don't know when it's going to break.
• 8:11
  SPEAKER_01: You don't know if it has any flaws in it.
• 8:13
  SPEAKER_01: You don't know if it is really doing what it's supposed to do.
• 8:16
  SPEAKER_01: You need to know what your code does.
• 8:18
  SPEAKER_01: You can't just say, well, it must know because it's telling me this.
• 8:22
  SPEAKER_01: I've had every different AI, so it's not just Copilot, but I've had every different AI tell me bad advice, things that aren't true.
• 8:32
  SPEAKER_01: Because the actually the most of the information that AI bases its knowledge on is old information, because it has more of that than it does new information.
• 8:42
  SPEAKER_01: And so you're going to have these weird cases where you're actually being told the wrong things that that thing has changed.
• 8:48
  SPEAKER_01: Breaking changes are a big deal to AI.
• 8:50
  SPEAKER_01: But also the hallucination, sometimes it says, well, this is a way of doing it.
• 8:54
  SPEAKER_01: And it's like, no, that's not how that works.
• 8:57
  SPEAKER_01: Or it may just say, I'm going to try and extrapolate what could be.
• 9:03
  SPEAKER_01: I've asked about APIs that are not documented.
• 9:06
  SPEAKER_01: There are APIs that I use that just don't have great documentation.
• 9:11
  SPEAKER_01: There's some good documentation on parts of them, but not the whole thing.
• 9:15
  SPEAKER_01: And I push things to their limits.
• 9:17
  SPEAKER_01: I try and use all the things that otherwise would be called edge cases.
• 9:21
  SPEAKER_01: So I was looking for, you know, what's the documentation on this piece that no one seems to talk about, but I know it's been used before.
• 9:28
  SPEAKER_01: And sometimes AI is good at that because it will find where people have used that in source code in GitHub.
• 9:34
  SPEAKER_01: So anything open source has been probably scraped by GitHub Copilot.
• 9:40
  SPEAKER_01: Therefore, it can look through that and say, oh, hey, I see someone else use that API here, and this is what they did.
• 9:46
  SPEAKER_01: Super helpful.
• 9:47
  SPEAKER_01: But I had it give me this whole long instruction set.
• 9:50
  SPEAKER_01: It was very clear.
• 9:54
  SPEAKER_01: Step one, two, three, four, five, with where to go to find these things, how to do it, how to set it up, and so on.
• 10:01
  SPEAKER_01: It looked great.
• 10:02
  SPEAKER_01: And I followed it to a letter and it didn't work.
• 10:05
  SPEAKER_01: And I did some more research and found out that was never a thing.
• 10:08
  SPEAKER_01: But it was confident in it.
• 10:10
  SPEAKER_01: So you gotta be very careful to balance the it's helping me versus it's driving the bus.
• 10:17
  SPEAKER_01: Okay.
• 10:17
  SPEAKER_01: Don't let it drive the bus, but let it help you.
• 10:20
  SPEAKER_01: So there's that balance with AI.
• 10:23
  SPEAKER_01: Now, the next one, data scraping, is going to be a big deal.
• 10:28
  SPEAKER_01: LLMs need more and more data all the time.
• 10:32
  SPEAKER_01: And they want to keep current because if you ask an LLM, hey, what about this?
• 10:37
  SPEAKER_01: And they're like, I've got information that's three years out of date.
• 10:41
  SPEAKER_01: That's not gonna be helpful.
• 10:42
  SPEAKER_01: And that you see the problem with ChatGPT, and they've kind of fixed that by continuing to put new information in and hooking up to the internet.
• 10:51
  SPEAKER_01: So that's gotten better, but at the same time, data scraping, it's gonna be a big deal because the problem with data scraping is that they're using that data without permission and without attribution.
• 11:06
  SPEAKER_01: And that can negatively impact a company's bottom line.
• 11:09
  SPEAKER_01: So here's a really simple example.
• 11:11
  SPEAKER_01: It's a fictional example, but here's a simple example.
• 11:14
  SPEAKER_01: Let's say your company creates dessert recipes for people to try.
• 11:18
  SPEAKER_01: So you give away the entire recipe for free.
• 11:20
  SPEAKER_01: It's just post online, like a blog post.
• 11:22
  SPEAKER_01: Um, hopefully, you don't tell a story about your childhood way to do it.
• 11:26
  SPEAKER_01: Um, but you give away these recipes for free.
• 11:28
  SPEAKER_01: But you have a sponsor link that will allow people to buy all the ingredients with one click.
• 11:35
  SPEAKER_01: So here's a you know, great dessert recipe for uh chocolate chip brownies.
• 11:40
  SPEAKER_01: Okay, so they have this recipe and they're looking at this and saying, okay, it takes this and this and this, and there's a button there that says, you know, have all these ingredients sent to me home.
• 11:49
  SPEAKER_01: Okay, again, fictional example, but that might be useful.
• 11:53
  SPEAKER_01: So that might make it easier for the consumer, and it's gonna pay for the work you put into hosting the site and creating the recipes.
• 12:02
  SPEAKER_01: Because you know, every time a person pushes that button, you get a small percentage of that.
• 12:06
  SPEAKER_01: And that's gonna pay some bills.
• 12:07
  SPEAKER_01: And that could be a nice little side business for someone.
• 12:10
  SPEAKER_01: And that's kind of a small microcosm example of what a lot of companies do because advertising pays for a lot of different company activities.
• 12:21
  SPEAKER_01: And you don't get the advertising unless you go to that site.
• 12:26
  SPEAKER_01: Now, an LLM will take the recipe and give that recipe to people without giving you credit and without giving the income from that one-click purchase or the ad revenue.
• 12:37
  SPEAKER_01: So you do all the work, but now you don't get any of the profit.
• 12:41
  SPEAKER_01: That means you go out of business.
• 12:44
  SPEAKER_01: And not only that, the LLM probably won't give you exactly the exact same recipe because it doesn't want to just steal.
• 12:51
  SPEAKER_01: So what's gonna do is remix some things.
• 12:53
  SPEAKER_01: But does an AI know how to create great chocolate chip brownies?
• 12:57
  SPEAKER_01: No.
• 12:58
  SPEAKER_01: So it's gonna put some weird stuff in there and it's gonna get people to be frustrated with those recipes.
• 13:03
  SPEAKER_01: So doing that scraping may put you out of business.
• 13:08
  SPEAKER_01: Now, this is a tiny example of what companies are gonna be facing in the coming months.
• 13:12
  SPEAKER_01: They're already facing it.
• 13:14
  SPEAKER_01: As a software developer, if you work for a company that might be affected by data scraping, you're gonna need to work to come up with solutions to prevent your business from going away.
• 13:25
  SPEAKER_01: So you need to figure out how do we do this?
• 13:27
  SPEAKER_01: Is it a paywall?
• 13:28
  SPEAKER_01: Is it a non-pay wall?
• 13:32
  SPEAKER_01: Like you have to log in first, or you have to do something else to get this.
• 13:37
  SPEAKER_01: Maybe you have an app now instead of a website.
• 13:40
  SPEAKER_01: There's a lot of different ways you could go about this that actually would improve the user experience while at the same time protecting yourself from data scraping.
• 13:50
  SPEAKER_01: So that's another big deal in this industry going forward.
• 13:54
  SPEAKER_01: Number three, we talked about uh where to host your web application.
• 13:58
  SPEAKER_01: That's always been a debate.
• 14:00
  SPEAKER_01: It used to be between self-hosting and some type of web hosting, like a VPS or virtual private server.
• 14:06
  SPEAKER_01: Now we have a third option that is the cloud, like AWS or Azure.
• 14:12
  SPEAKER_01: And you know, the cloud does a lot of stuff, and sometimes people forget there's still VPSs, there's still small hosting accounts, and there's also the ability to host on your own.
• 14:24
  SPEAKER_01: Now, recently there's been some very public talk about people leaving the cloud and going back to VPSs or even local hosting because of the cost savings from the cloud.
• 14:37
  SPEAKER_01: Now, knowing how to navigate these waters is going to be important for anyone who has any say in the matter.
• 14:44
  SPEAKER_01: Like with all decisions in software development, the answer is going to be it depends.
• 14:50
  SPEAKER_01: You need to know how to accurately compare the various options.
• 14:54
  SPEAKER_01: Just looking at the price tag isn't a good metric.
• 14:57
  SPEAKER_01: Again, we're talking about the idea of going deeper than that, having a much better knowledge than just, I know these basic things about the different options, knowing having a much deeper knowledge of it.
• 15:09
  SPEAKER_01: So knowing what you get for that price is going to be important.
• 15:13
  SPEAKER_01: For example, um, you could go to Walmart and buy a Power Wheels car, which is a little um, you know, car that kids drive that's you know, little battery in there and they can drive around the driveway.
• 15:25
  SPEAKER_01: It's gonna cost$200.
• 15:26
  SPEAKER_01: Or you could buy a UTV, a utility vehicle that is like a little vehicle you can drive off-road only, uh, might even have a little bed to put stuff in.
• 15:37
  SPEAKER_01: Those usually cost$10,000 or so.
• 15:39
  SPEAKER_01: Or you could buy a small pickup truck for$20,000.
• 15:43
  SPEAKER_01: Each vehicle will drive, each will take you down the road.
• 15:46
  SPEAKER_01: That doesn't mean the cheapest option is the best option for you.
• 15:50
  SPEAKER_01: It might be the right option for you, but you might need off-road capabilities of the UTV because the UTV can get in small spaces and work offline, it's lighter or off-road and it's um lighter and smaller.
• 16:03
  SPEAKER_01: Or you might need the safety and protection of the pickup truck so you can drive it on the highway as well as go off-road.
• 16:11
  SPEAKER_01: So your choice when it comes to web hosting is very similar.
• 16:16
  SPEAKER_01: It comes down to what do you need to do?
• 16:20
  SPEAKER_01: What do you need to add on to this besides just the main thing?
• 16:25
  SPEAKER_01: It's not just driving, it's the stuff around driving, because all three of our previous options were driving, they all drove.
• 16:32
  SPEAKER_01: So it's the stuff around that, the power, the ability to work offline, the ability to be safe, the ability to go on the highway and some of the things.
• 16:42
  SPEAKER_01: The same thing is true with web hosting.
• 16:44
  SPEAKER_01: So you're gonna spend time focused on, you know, if you host it yourself, you're gonna spend time focused on security updates, testing new operating system patches, network management, firewall maintenance, network security, physical backups, and much more.
• 16:58
  SPEAKER_01: Are you good with that?
• 16:59
  SPEAKER_01: Is that okay?
• 17:00
  SPEAKER_01: Is that because you've got to add that into the cost of hosting yourself?
• 17:04
  SPEAKER_01: Now, most of those things are included in a VPS, and even more are included in cloud hosting.
• 17:10
  SPEAKER_01: So knowing your different options is being really important rather than just saying, oh, this is what the industry is doing.
• 17:16
  SPEAKER_01: That's the trend.
• 17:18
  SPEAKER_01: Yeah, but some of that's hype.
• 17:20
  SPEAKER_01: And some of that is for that company or for those companies, that was the right choice.
• 17:25
  SPEAKER_01: That's not your company.
• 17:27
  SPEAKER_01: So knowing how to know the difference is gonna be important.
• 17:31
  SPEAKER_01: Now, the next one is turning web applications into desktop and mobile applications.
• 17:36
  SPEAKER_01: That's something you really need to need to know how to do.
• 17:39
  SPEAKER_01: It used to be that companies would essentially duplicate all their efforts to create multi-platform native applications.
• 17:46
  SPEAKER_01: Simple example, we want to have an app for a company.
• 17:49
  SPEAKER_01: We're gonna create, you have an Android team to work on the uh Android version, or have an iOS team to work on the iOS version.
• 17:56
  SPEAKER_01: Okay, so those are two separate code bases.
• 17:58
  SPEAKER_01: Maybe they pull from the same APIs, but they're two separate code bases, two separate teams, two separate ways of doing things.
• 18:05
  SPEAKER_01: That's a lot of duplication of efforts.
• 18:08
  SPEAKER_01: And then you get into we're gonna deploy it to the desktop of Windows versus Mac versus Linux.
• 18:13
  SPEAKER_01: That's more dedicated teams and more code bases that are different.
• 18:19
  SPEAKER_01: Now, what companies are doing now is they're saying, hey, you know what?
• 18:23
  SPEAKER_01: That's too expensive and it's not giving us enough bang for our buck.
• 18:29
  SPEAKER_01: So those days are falling behind us.
• 18:32
  SPEAKER_01: In order to save money, standardized applications are becoming more common.
• 18:38
  SPEAKER_01: So this could be as simple as a progressive web application where you create a web application and say, hey, you know what, this actually installs on desktop and mobile.
• 18:47
  SPEAKER_01: Or you can even use um something like PWA Builder, where it deploys natively in the store, but it's just a PWA.
• 18:55
  SPEAKER_01: Um, that's potential and that's a really simple option.
• 18:59
  SPEAKER_01: And maybe that's all your app needs because web is getting more and more power all the time.
• 19:05
  SPEAKER_01: Or it could be that you take web code, HTML, CSS, and JavaScript, and run it natively.
• 19:14
  SPEAKER_01: Um, either a wrapper like Electron or a WebView with something like Blazor Hybrid.
• 19:21
  SPEAKER_01: So these things allow you to take code that you know maybe you're you're writing a couple things different.
• 19:28
  SPEAKER_01: For the most part, though, it's the same code and you're deploying it natively to the different platforms.
• 19:34
  SPEAKER_01: Knowing how to work with these systems and how to optimize them so they work better than the new Outlook is gonna be important.
• 19:41
  SPEAKER_01: And yes, I'm throwing Outlook under the bus a lot because Outlook sucks.
• 19:45
  SPEAKER_01: But knowing how to work with these systems and optimize them is going to be important because you want to still give your customers a great experience, but at the same time, having one team with one code base that's efficient and can deploy to a number of different platforms is really important.
• 20:03
  SPEAKER_01: Now, finally, the last option or last thing we have to talk about here and trends is password management and security.
• 20:10
  SPEAKER_01: It's becoming an increasingly important field.
• 20:13
  SPEAKER_01: As more and more data breaches happen, a simple password on its own is becoming an unsecure solution.
• 20:20
  SPEAKER_01: The next step has been two-factor authentication, although even that is not necessarily fully secure.
• 20:27
  SPEAKER_01: Other options like passwordless systems are also becoming more common solutions.
• 20:31
  SPEAKER_01: Knowing how to secure the data of your organization is incredibly important.
• 20:36
  SPEAKER_01: Whether that means replacing your simple auth system with a better implementation or implementing a new way of authenticating users, it's important that you know going forward how to do things like that.
• 20:51
  SPEAKER_01: Protecting your data is paramount in organization.
• 20:56
  SPEAKER_01: So, yes, it's a trend, but it's not hype, it's reality.
• 20:59
  SPEAKER_01: You need to know how to protect your users.
• 21:02
  SPEAKER_01: There's so many data breaches that happen all the time.
• 21:05
  SPEAKER_01: You want to make sure that when a data breach does happen, you say, yes, a data breach happened.
• 21:11
  SPEAKER_01: They have access to your password, but don't worry, you're all two-factor authentication, and we've reset all your passwords.
• 21:18
  SPEAKER_01: So pretty much they can't get in.
• 21:20
  SPEAKER_01: Versus we lost all your passwords and all your data.
• 21:24
  SPEAKER_01: That's not good.
• 21:25
  SPEAKER_01: So make sure that you're upgrading your security beyond just the basics of login name and password.
• 21:32
  SPEAKER_01: So, what can you do as a software developer to set yourself up for success with these emerging trends in the industry?
• 21:39
  SPEAKER_01: Well, number one, go beyond the basics.
• 21:42
  SPEAKER_01: People who know the basics of AI endanger their systems.
• 21:46
  SPEAKER_01: People who implement the basics of authentication do the same.
• 21:50
  SPEAKER_01: Basically, if you know the basics, you just know enough to be dangerous.
• 21:56
  SPEAKER_01: Go deeper to learn when to use something, when not to use.
• 22:00
  SPEAKER_01: It, learn about the pitfalls and how to avoid them.
• 22:03
  SPEAKER_01: Now, number two, focus on saving money.
• 22:07
  SPEAKER_01: Companies are looking to cut costs and improve efficiency.
• 22:10
  SPEAKER_01: Lean into it.
• 22:12
  SPEAKER_01: You could complain about that you could have a pity party session for yourself because your company wants to cut costs.
• 22:19
  SPEAKER_01: Instead, lean into it.
• 22:21
  SPEAKER_01: Use it to help implement some of these new things that are going to both save your company money but also improve things.
• 22:29
  SPEAKER_01: Investigate how to turn your existing web app into a PWA or a simple cross-platform app.
• 22:35
  SPEAKER_01: Identify how to protect your company's data.
• 22:38
  SPEAKER_01: Figure out how to use AI to speed up work without compromising results.
• 22:43
  SPEAKER_01: Compare the costs of your various hosting platforms for your specific situation.
• 22:49
  SPEAKER_01: Any of these five can save you money if you do it right.
• 22:54
  SPEAKER_01: Plus, they can also improve your output for your customer.
• 22:59
  SPEAKER_01: But you have to be proactive and push the best one of your company rather than just allowing people who have a base level of knowledge to make bad calls and implement things poorly like AI run amuck.
• 23:12
  SPEAKER_01: Okay?
• 23:13
  SPEAKER_01: Number three, don't just chase the hype.
• 23:16
  SPEAKER_01: Give new ideas time to settle before you even consider them.
• 23:20
  SPEAKER_01: I know that when you see new things, it's it's very um easy to say, I want to do that.
• 23:27
  SPEAKER_01: That's that's gonna be great, it's gonna save us lots of time.
• 23:29
  SPEAKER_01: Wait, take a breath.
• 23:31
  SPEAKER_01: Take time to think it through.
• 23:33
  SPEAKER_01: Make sure that you thoroughly evaluate a situation before you look to implement it in production.
• 23:40
  SPEAKER_01: For example, custom companies that rush to implement AI are experiencing issues such as unmet expectations where they promised the world and it didn't happen, or unexpected costs because AI can be very expensive to run, or data privacy and compliance issues because LLMs gobble up information and they don't necessarily keep it inside your company.
• 24:04
  SPEAKER_01: So a lot of that could have been avoided with a more careful approach rather than a we have to be first to market approach.
• 24:14
  SPEAKER_01: Number four is remember your primary job.
• 24:18
  SPEAKER_01: Your primary job isn't to implement trendy things or do new things with technology.
• 24:24
  SPEAKER_01: Your job is to provide your company with value.
• 24:28
  SPEAKER_01: That means making your company more efficient, serving more customers, better serving customers, or otherwise being a net positive to your company's bottom line.
• 24:38
  SPEAKER_01: Make sure the thing that you pursue will aid in those goals.
• 24:43
  SPEAKER_01: Product stability, data safety, and application reliability are big factors in achieving those goals.
• 24:49
  SPEAKER_01: So don't forget them as well.
• 24:52
  SPEAKER_01: There will always be new trends and ideas in the software development industry.
• 24:57
  SPEAKER_01: Knowing how to identify the trends and also safely implement them will be helpful and it's going to help you avoid the ones that'll be a distraction from your company's goals.
• 25:10
  SPEAKER_01: It's an important skill to improve to make sure that you can identify the good ones versus the hype and make sure you implement things properly.
• 25:20
  SPEAKER_01: Thanks for listening.
• 25:21
  SPEAKER_01: And as always, I am Tim Corey.
• 25:23
  SPEAKER_00: Thank you for joining us for this episode of Dev Questions.
• 25:26
  SPEAKER_00: When you're ready to learn to think and code like a professional developer, head over to IamTimCorey.com and enroll in a course.