Latest / DevQuestions with Tim Corey / 300. What is the Best Way to Learn Software Development? Has AI Changed That?

Transcript

300. What is the Best Way to Learn Software Development? Has AI Changed That? · DevQuestions with Tim Corey

• 0:00
  SPEAKER_01: Has the right way I do software development training changed because of AI?
• 0:04
  SPEAKER_01: As you probably know, I make my living training software developers.
• 0:08
  SPEAKER_01: I have dozens of paid training courses, training people in C Sharp, SQL, Azure, Git, and even AI.
• 0:15
  SPEAKER_01: But with the rise of AI, is video training obsolete?
• 0:19
  SPEAKER_01: Isn't it better to just have an AI create a customized training plan for you and then just follow that?
• 0:26
  SPEAKER_01: These are the questions that I've actually wrestled through.
• 0:29
  SPEAKER_01: In this special 300th edition of DevQuestion, I'm going to go through the things that I have evaluated in my conclusions on AI's place in training software developers.
• 0:53
  SPEAKER_01: Software development is more than just writing code.
• 0:56
  SPEAKER_01: So let's talk about the rest of it.
• 0:57
  SPEAKER_01: Specifically, let's talk about AI's place in training software developers.
• 1:03
  SPEAKER_01: Now, right up front, I want to talk about my perceived bias in all of this.
• 1:08
  SPEAKER_01: I make courses for a living.
• 1:10
  SPEAKER_01: Some people assume that since that's the case, then I'll conclude that courses are the right way or the only way to learn software development because I have a financial stake in the matter.
• 1:20
  SPEAKER_01: And that's a fair thing to question, but that's not how I do things.
• 1:25
  SPEAKER_01: My goal has always been to make learning software development easier.
• 1:30
  SPEAKER_01: So if it's an easier way, I want to help you take it.
• 1:33
  SPEAKER_01: I learn the hard way.
• 1:34
  SPEAKER_01: I struggle through dead ends, gaps in my education, and the feeling that I would never be good enough.
• 1:40
  SPEAKER_01: I don't want that for you or for those other people I train as well.
• 1:45
  SPEAKER_01: I want you to be the best software developer that you can be with the least amount of effort.
• 1:52
  SPEAKER_01: So if I think there's a better way, I'm going to change what I do to offer that better way.
• 1:58
  SPEAKER_01: So with that up, you know, out up front, let's talk about is there a better way?
• 2:05
  SPEAKER_01: So let's answer some questions that I've wrestled through and that you have either probably thought about or should think about when it comes to learning software development.
• 2:16
  SPEAKER_01: So, question number one: do you even need to learn software development when you have AI to do it for you?
• 2:25
  SPEAKER_01: Now, the answer is yes, you need to learn software development.
• 2:29
  SPEAKER_01: But let's talk about why.
• 2:31
  SPEAKER_01: Because yes, it can seem like you just ask AI for the answer and it gives it to you.
• 2:37
  SPEAKER_01: But as we saw last week in last week's episode, Dev Questions, that AI, when it writes code, writes code 1.7 times worse.
• 2:47
  SPEAKER_01: Uh had 1.7 times more errors than if a software developer writes it, which means that AI is not good at writing code.
• 2:57
  SPEAKER_01: So if you don't learn how to do software development on your own, then you can never check the AI's work.
• 3:05
  SPEAKER_01: And the AI's work will work.
• 3:07
  SPEAKER_01: And that's the scariest part.
• 3:09
  SPEAKER_01: It is going to work, usually, sometimes.
• 3:12
  SPEAKER_01: And sometimes that doesn't work.
• 3:13
  SPEAKER_01: But when it does work, you may think, well, then it's been successful.
• 3:17
  SPEAKER_01: But that's not the job of a software developer.
• 3:20
  SPEAKER_01: A software developer's job isn't to just make something that works, because we have to think about things like security and performance and long-term support.
• 3:31
  SPEAKER_01: And AI does all three of those things poorly.
• 3:35
  SPEAKER_01: It's not 99% of the way there.
• 3:37
  SPEAKER_01: And it's not, well, the next model will.
• 3:39
  SPEAKER_01: We have proven year after year after year when those predictions of, well, the next model, those are all been false.
• 3:48
  SPEAKER_01: They have not gotten significantly better when it comes to the actual output.
• 3:53
  SPEAKER_01: Yes, AI passes new tests and new assessments, and it can memorize facts better than humans can, but that's not really the true assessment of a developer.
• 4:04
  SPEAKER_01: A true assessment of a developer isn't, have you memorized all the syntax?
• 4:09
  SPEAKER_01: The true assessment is, can you get the job done?
• 4:12
  SPEAKER_01: And by done, again, I don't just mean, does the app work on the happy path?
• 4:18
  SPEAKER_01: What I mean is, is it secure?
• 4:20
  SPEAKER_01: Does it stand up to users giving bad input to it?
• 4:24
  SPEAKER_01: Does it have critical flaws?
• 4:26
  SPEAKER_01: Can it be exploited?
• 4:27
  SPEAKER_01: All its other things.
• 4:29
  SPEAKER_01: And so AI doesn't do that yet.
• 4:32
  SPEAKER_01: And I don't think it ever will.
• 4:34
  SPEAKER_01: It is not proven to be significantly better at doing that than it was two years ago.
• 4:39
  SPEAKER_01: Yes, it can sometimes write some better code and it feels like it's making better choices, but it really isn't.
• 4:45
  SPEAKER_01: It's actually just hiding some of the problems that's creating.
• 4:48
  SPEAKER_01: So no, I don't think that we can rely on AI to do our work for us.
• 4:54
  SPEAKER_01: And even if you could, why would you be a software developer if you don't have to know software development?
• 5:02
  SPEAKER_01: Like at that point, you're just a project manager.
• 5:04
  SPEAKER_01: And that's a valid career.
• 5:07
  SPEAKER_01: I have been a PMP, a project manager professional, been certified in that.
• 5:12
  SPEAKER_01: And that can be a valid career.
• 5:14
  SPEAKER_01: Not really a career for me.
• 5:16
  SPEAKER_01: I don't enjoy just managing projects.
• 5:18
  SPEAKER_01: I did it because I was managing a large project and it fit and I had managed lots of projects, but from a software development perspective.
• 5:27
  SPEAKER_01: And at the end of the day, I want to be a software developer.
• 5:30
  SPEAKER_01: That doesn't always mean writing every line of code.
• 5:33
  SPEAKER_01: Yes, AI can write a lot of lines of code for you, but you're still doing more than just managing a project.
• 5:41
  SPEAKER_01: But if AI does all the code writing for you and makes all the decisions about how that code should be written, well, then you're not valuable if you're saying you're a software developer, which means you get replaced by an AI.
• 5:55
  SPEAKER_01: Now, again, AIs are not good at that, but if you're even worse at that than the AI, there's no purpose for you existing in that space.
• 6:05
  SPEAKER_01: So yes, you need to learn software development and not rely on just AI to do it for you.
• 6:13
  SPEAKER_01: Okay, so question number two: can AI be a good trainer?
• 6:17
  SPEAKER_01: Because, okay, maybe AI doesn't do a great job at always writing good software.
• 6:23
  SPEAKER_01: In fact, it's much worse than a software developer does.
• 6:26
  SPEAKER_01: But maybe it could be a good trainer and teach you how to do write good software.
• 6:33
  SPEAKER_01: But again, if an AI creates 1.7 times more bugs and more problems in code than a software developer does, then why would you rely on that thing to train you?
• 6:46
  SPEAKER_01: Because it's going to train you even worse than if you were trained by a good software developer and or a software developer, let alone a good software developer.
• 6:57
  SPEAKER_01: So I don't think that AI can be a good trainer because you don't know when the AI is wrong.
• 7:04
  SPEAKER_01: You can't validate.
• 7:05
  SPEAKER_01: Again, it's come back to step one, where if you rely on AI to just write the code for you, well, if you don't know what that code says or does or how it works, or if it's the right code, you can't properly evaluate beyond it passes all the tests and it runs.
• 7:22
  SPEAKER_01: Therefore, it must be good.
• 7:24
  SPEAKER_01: And that's not good enough.
• 7:26
  SPEAKER_01: So, in the same way, with question two, can AI be a good trainer?
• 7:31
  SPEAKER_01: If it's training you, then you don't know enough to know when it's wrong.
• 7:37
  SPEAKER_01: And yes, it is wrong.
• 7:40
  SPEAKER_01: There are lots of times where it can create things that are wrong for you, give you a wrong training plan.
• 7:46
  SPEAKER_01: In fact, I did a study recently that said that 60% of the time, regardless of which model, 60% of the time, when you ask an AI, are you sure on a question, it will reverse course.
• 8:00
  SPEAKER_01: Now, what does that mean?
• 8:02
  SPEAKER_01: It means the AI is not actually giving you good advice, it's giving you the advice that it thinks you want to hear.
• 8:08
  SPEAKER_01: So when it thinks you might want to hear the opposite, it will give you that information.
• 8:13
  SPEAKER_01: That's not a good trainer.
• 8:15
  SPEAKER_01: A good trainer doesn't just let you choose what you want to hear.
• 8:20
  SPEAKER_01: A good trainer is going to train you in the way that's best, in the way that's going to best create the best output for you, the best outcome for you.
• 8:30
  SPEAKER_01: So, no, AI can't be a good trainer because you have to know more than the AI in order for it to be effective.
• 8:39
  SPEAKER_01: Now, if you do know more than the AI, then you can use it to kind of brush up on things.
• 8:46
  SPEAKER_01: Because you can look at it and go, wait, are you sure that's true?
• 8:49
  SPEAKER_01: Let's let's validate that and go back and look at things that maybe it's teaching you that either you missed or that are correct.
• 8:58
  SPEAKER_01: And you can know when to kind of push back on the AI and validate from other sources.
• 9:04
  SPEAKER_01: So, yes, if you're using it to refresh your memory, AI can be good because it can give you mostly right information.
• 9:12
  SPEAKER_01: And that's the key here is that AI will give you mostly right information and it will all look right.
• 9:20
  SPEAKER_01: It'll all be convincing, but there's gonna get little bits here and there that aren't correct.
• 9:26
  SPEAKER_01: Which means going back to, you know, when I was learning software development, I struggle with gaps in my education, things that I didn't know that I didn't know.
• 9:35
  SPEAKER_01: Well, this is gonna create gaps in your education as well, but in a different way.
• 9:40
  SPEAKER_01: You're gonna learn things that are wrong, but you don't know that they're wrong.
• 9:45
  SPEAKER_01: So you're gonna go off and do these things, not realizing that you're creating more problems in your software because you didn't know that what you learned on was wrong.
• 9:56
  SPEAKER_01: So that's a really scary place to be.
• 9:58
  SPEAKER_01: That's why you should always trust your trainer.
• 10:00
  SPEAKER_01: Because even humans are wrong sometimes.
• 10:03
  SPEAKER_01: And it's not like we're infallible.
• 10:05
  SPEAKER_01: But when you trust a good trainer, then you are trusting that they have the experience to teach you the right ways of doing things.
• 10:15
  SPEAKER_01: That's why, for me personally, first of all, I have almost 30 years of experience in the software development field.
• 10:21
  SPEAKER_01: I've seen a lot, I've done a lot, I've learned a lot.
• 10:24
  SPEAKER_01: But I also don't train you in everything because I have deep experience in a lot of areas, but I don't have deep experience in other areas.
• 10:36
  SPEAKER_01: And so I have to spend a whole lot more time making sure that I am solid in area before I'd ever even consider training you.
• 10:45
  SPEAKER_01: I take it very, very seriously that I don't just regurgitate what somebody else told me.
• 10:51
  SPEAKER_01: That's important for a good trainer.
• 10:54
  SPEAKER_01: So when it comes to AI, it's just regurgitating.
• 10:58
  SPEAKER_01: And it's not even regurgitating good, consistent training.
• 11:02
  SPEAKER_01: It will be different every day if you're not careful.
• 11:05
  SPEAKER_01: So, no, I don't think that AI can be a good trainer.
• 11:10
  SPEAKER_01: Now, number three, can AI create a good training plan?
• 11:15
  SPEAKER_01: And this is starting to get in the territory of maybe, because you can more easily validate a training plan.
• 11:24
  SPEAKER_01: And still there's there's reasons why you might not want to rely on the AI because it's not making good decisions about which thing to learn when, but it can give you all the things you probably need to learn, probably.
• 11:39
  SPEAKER_01: And this is where you can start to say, well, that might be valuable, or at least it might expose to you things you're like, wait, I didn't know about that.
• 11:47
  SPEAKER_01: Let me go learn that.
• 11:48
  SPEAKER_01: And that's not a bad thing.
• 11:50
  SPEAKER_01: Learning new things is a good thing.
• 11:52
  SPEAKER_01: And if it exposes the things that you don't know yet, that can fill in some gaps for you.
• 11:58
  SPEAKER_01: But order is important, and that can make the difference between your training taking six months or six years.
• 12:06
  SPEAKER_01: When I was teaching in the classroom, I taught at a community college for five years or so.
• 12:12
  SPEAKER_01: And in that classroom, I had a book that I taught through.
• 12:15
  SPEAKER_01: And that's because I was required to teach through the book.
• 12:18
  SPEAKER_01: And so I took that book and I said, you know what?
• 12:21
  SPEAKER_01: I can't teach chapter one first.
• 12:24
  SPEAKER_01: And the reason why is because chapter one, it was on learning C sharp.
• 12:28
  SPEAKER_01: People that knew nothing about C sharp were gonna learn C sharp.
• 12:32
  SPEAKER_01: And chapter one was file new project for a Winforms app.
• 12:37
  SPEAKER_01: And what you would do is you would drag a button onto the screen, you would drag a text box onto the screen, you would double-click on the button, and you'd say message box.show, text box.text.
• 12:49
  SPEAKER_01: And what I'd do is you'd be able to say, you know, hello, Tim, in your first application.
• 12:54
  SPEAKER_01: The problem with that, and the reason why I did not teach that first, is because that while that's very simple, and you get an immediate win of an application that works, that's visual, that has the maximize and minimize and restore buttons that you know you can make put colors into, you can resize buttons and all the rest.
• 13:13
  SPEAKER_01: It's very visual.
• 13:14
  SPEAKER_01: But the reason I didn't teach that first is because it relied on object-oriented programming and events.
• 13:22
  SPEAKER_01: And those are things that you don't have a foundation how they work.
• 13:27
  SPEAKER_01: And so I had students come from other professors who were going through that same class, maybe failed out, and then came to my class to try it again.
• 13:36
  SPEAKER_01: I would ask them, what do you do when things don't work?
• 13:41
  SPEAKER_01: When the editor crashes, and I'd show them how to crash the editor.
• 13:45
  SPEAKER_01: I'd crash it by just deleting the event handler for a button or for the form.
• 13:50
  SPEAKER_01: And I said, What do you do in that situation?
• 13:53
  SPEAKER_01: And their solution was we delete our project and start over.
• 13:58
  SPEAKER_01: Why?
• 13:59
  SPEAKER_01: Because they did not understand the foundation upon which they had built something.
• 14:06
  SPEAKER_01: And so order was very important to me.
• 14:09
  SPEAKER_01: I would go through and teach up through chapter eight before I taught chapter one.
• 14:15
  SPEAKER_01: Because chapter eight got into object-oriented programming and classes and instantiation and events and how all these things work so that when I taught chapter one, the students go, oh, I see this is an event and an event handler.
• 14:30
  SPEAKER_01: And if I delete the handler but don't delete where I'm wiring that up to the event, that's a problem.
• 14:36
  SPEAKER_01: So if I have this error, I can go into the code and fix it.
• 14:40
  SPEAKER_01: Because they understood the underlying technology.
• 14:44
  SPEAKER_01: Order is important.
• 14:47
  SPEAKER_01: And when the AI gives you that order, it might not be the best order.
• 14:53
  SPEAKER_01: It might help you in some areas, but might also make you struggle through things.
• 14:58
  SPEAKER_01: So, yes, working in the territory of an AI can create a good training plan, but you don't know for sure.
• 15:06
  SPEAKER_01: And it might make it more difficult to learn, but it might.
• 15:11
  SPEAKER_01: Number four, can AI create a good practice project suggestion?
• 15:17
  SPEAKER_01: The answer is sometimes, yes.
• 15:21
  SPEAKER_01: Um, I have tried this, and I'm I want to be very clear here because I have said it before, I've said it in other areas, but just to be clear, I do not create AI-generated content.
• 15:34
  SPEAKER_01: I will not create AI generated content.
• 15:37
  SPEAKER_01: I do not let the AI create my outlines.
• 15:40
  SPEAKER_01: I do not let the AI create my uh text that I read off of sometimes or my you know my script or anything of that nature.
• 15:49
  SPEAKER_01: I don't have it create my code for me.
• 15:53
  SPEAKER_01: Now, I have used it in the past, and I continue to use it sometimes to take my suggestions, create a bit of code, and then I rework it kind of as a starting point.
• 16:03
  SPEAKER_01: And in this case, I do, I have used AI to say, hey, you know what, I'm trying to create a practice project for this.
• 16:12
  SPEAKER_01: Help me figure out something simple to do.
• 16:15
  SPEAKER_01: And it has sometimes given me some good suggestions.
• 16:19
  SPEAKER_01: Now, what I often find is I keep or try out about 5%.
• 16:25
  SPEAKER_01: 5% of the time or 5% of the projects that it gives me, I go, ooh, let me try that.
• 16:29
  SPEAKER_01: That might be a good suggestion.
• 16:32
  SPEAKER_01: And I throw the rest away because the rest are nonsensical or don't reapply or too complex or don't really test the thing that we're trying to work on.
• 16:42
  SPEAKER_01: So it can help you by brainstorming.
• 16:46
  SPEAKER_01: I think this is where AI can really can really shine is brainstorming, where you're trying things out and saying, hey, what are your thoughts on this?
• 16:56
  SPEAKER_01: I do this for my video titles, where I say, you know, I'm trying to communicate this.
• 17:01
  SPEAKER_01: Give me 20 titles.
• 17:03
  SPEAKER_01: And I don't think I've ever used a title the AI suggested.
• 17:08
  SPEAKER_01: Part of it is I don't like to do clickbait things.
• 17:10
  SPEAKER_01: I like to be very clear in my titles.
• 17:12
  SPEAKER_01: Now, I do want titles that grab your attention.
• 17:15
  SPEAKER_01: And that's a bit different than clickbait, though.
• 17:18
  SPEAKER_01: So clickbait is where you give a title that doesn't really match the content.
• 17:22
  SPEAKER_01: I want to make sure my titles match the content.
• 17:24
  SPEAKER_01: So I have to be very careful in how I word them.
• 17:28
  SPEAKER_01: And so I'll try 20 different suggestions.
• 17:30
  SPEAKER_01: And I'll go, you know what?
• 17:31
  SPEAKER_01: I'll take a word from here and a word from there and rework something into something that works and is more um tightly packed than I would typically do.
• 17:40
  SPEAKER_01: I'm more um effusive in my words.
• 17:42
  SPEAKER_01: I, you know, why use five words when 20 will do, right?
• 17:46
  SPEAKER_01: So that's not great for titles.
• 17:48
  SPEAKER_01: And so the AI can help me with that.
• 17:51
  SPEAKER_01: But I usually find that I am taking pieces from here and there.
• 17:54
  SPEAKER_01: Well, the same thing is true for practice projects.
• 17:58
  SPEAKER_01: It's very important that you practice software development, not have AI do it for you, but that you say, hey, you know what?
• 18:06
  SPEAKER_01: I'm gonna create a small practice project to practice what I just learned and do it more than once.
• 18:13
  SPEAKER_01: So asking the AI, hey, what project could I do to test this can be valuable.
• 18:19
  SPEAKER_01: Now, I will tell you the AI is gonna tell you a full application.
• 18:23
  SPEAKER_01: You don't want to do that for practice projects.
• 18:25
  SPEAKER_01: Practice projects should do nothing essentially, they should just test what you're learning.
• 18:30
  SPEAKER_01: But if you're practicing a real world, putting multiple things together, you might get some valuable practice project suggestions that are small enough.
• 18:40
  SPEAKER_01: Number five, can AI evaluate my work?
• 18:43
  SPEAKER_01: Maybe you have learned from a good trainer and you on the training plan that that trainer put together for you, and you have done your own practice projects, but now you want someone to look at that practice project and say, is that a good job or did you miss something?
• 18:59
  SPEAKER_01: Can AI evaluate that work?
• 19:03
  SPEAKER_01: For practice projects, AI can be really valuable.
• 19:08
  SPEAKER_01: Because part of the issue with AI is the fact that it cannot get enough context.
• 19:15
  SPEAKER_01: So in a real-world application, you might have 100,000 on the very low end to a million plus lines of code in modules all over.
• 19:26
  SPEAKER_01: Well, when AI reads in your project, it does not read in the entire project.
• 19:33
  SPEAKER_01: It doesn't look at all of your code, it can't.
• 19:35
  SPEAKER_01: It does not have a big enough context window.
• 19:38
  SPEAKER_01: So it looks at part of it, which can be okay in certain circumstances.
• 19:42
  SPEAKER_01: But what that leaves out is do you have a method somewhere else that might be valuable?
• 19:47
  SPEAKER_01: Or have we already done this work?
• 19:49
  SPEAKER_01: Or what's the pattern for how we write our code for this situation?
• 19:54
  SPEAKER_01: Those are the kind of things that AI often misses in the real world.
• 19:58
  SPEAKER_01: But when you're creating practice projects, you're creating very, very, very small amounts of code, just enough to practice the things you've learned.
• 20:09
  SPEAKER_01: So can AI evaluate that work?
• 20:11
  SPEAKER_01: Yeah, it can.
• 20:13
  SPEAKER_01: Now, it's going to give you a whole bunch of false positives, though, as well.
• 20:17
  SPEAKER_01: For example, maybe it's a very simple example.
• 20:20
  SPEAKER_01: Maybe you're learning the if statement.
• 20:21
  SPEAKER_01: And you create a console application that has a hard-coded uh first name variable.
• 20:27
  SPEAKER_01: And then you say if first name equals Tim, then say hello Tim.
• 20:31
  SPEAKER_01: Otherwise, say hello person I don't know.
• 20:33
  SPEAKER_01: Okay?
• 20:34
  SPEAKER_01: If-else.
• 20:34
  SPEAKER_01: You've just tested an if-else statement.
• 20:37
  SPEAKER_01: The AI might say, hey, you've hard-coded a string, you probably want to ask for that in a console read line.
• 20:43
  SPEAKER_01: Well, yeah, in a real world application, sure.
• 20:46
  SPEAKER_01: But that's not what we're doing.
• 20:47
  SPEAKER_01: So you have to be careful to say, hey, I'm learning about the if statement.
• 20:51
  SPEAKER_01: Look at the if statement usage, and what could I do better?
• 20:56
  SPEAKER_01: Or is there other things I can do to test this?
• 21:00
  SPEAKER_01: So evaluating your work, yes, an AI can do a good job at that.
• 21:04
  SPEAKER_01: Just make sure you know you'll get some false positives as well.
• 21:08
  SPEAKER_01: Number six, what are the dangers of using AI?
• 21:12
  SPEAKER_01: This is a very important question to ask.
• 21:15
  SPEAKER_01: Too many times, people will ask only the positive.
• 21:18
  SPEAKER_01: And the positive is, yeah, it can help you with project suggestions.
• 21:23
  SPEAKER_01: Yeah, it can help you evaluate your work.
• 21:25
  SPEAKER_01: Yeah, it can show you areas you may have missed in your training, and it could even supplement some of the things that you're doing.
• 21:34
  SPEAKER_01: But what are the dangers of using AI?
• 21:37
  SPEAKER_01: This is very much important.
• 21:39
  SPEAKER_01: So the dangers of using AI are that you start to rely on AI in a way that means you can't do the work without it.
• 21:47
  SPEAKER_01: And if you can't do the work without AI, all of a sudden you aren't nearly as valuable as an employee.
• 21:54
  SPEAKER_01: Because, well, anybody can do that work, right?
• 21:57
  SPEAKER_01: If if it's just the AI doing the work, And you're just going, yeah, yeah, yeah.
• 22:02
  SPEAKER_01: Well, that's not really valuable.
• 22:05
  SPEAKER_01: So the danger is that you start to rely on it in a way that your skills atrophy, where you can no longer remember how to do anything without an AI helping you.
• 22:16
  SPEAKER_01: And all of a sudden, you've become lazy to the point where you're not valuable unless your company pays for an AI.
• 22:25
  SPEAKER_01: And also you're contributing less to pushing back on the AI because you forget how to do things not using the AI.
• 22:33
  SPEAKER_01: And all of a sudden, you're out of a job.
• 22:37
  SPEAKER_01: And that's a real danger.
• 22:39
  SPEAKER_01: So especially when learning, I highly recommend that you don't use AI.
• 22:47
  SPEAKER_01: And that's not because I want you to fall behind.
• 22:50
  SPEAKER_01: You're not going to fall behind.
• 22:51
  SPEAKER_01: In fact, you're going to move ahead even faster.
• 22:56
  SPEAKER_01: Because if you use AI, you're going to not know what you're missing.
• 23:01
  SPEAKER_01: You're not going to know the gaps in your knowledge, and you're going to start to be reliant on that AI working and that AI helping you.
• 23:10
  SPEAKER_01: And all of a sudden, your skills are going to get less and less and less an atrophy, even as you're trying to build them up.
• 23:18
  SPEAKER_01: So when you're training, I highly recommend you don't use AI.
• 23:24
  SPEAKER_01: Use it after you know more than it and after you get to the point where you can do it on your own.
• 23:32
  SPEAKER_01: So that then you can just help use it to help you go a bit faster or make sure that you don't miss things.
• 23:39
  SPEAKER_01: But when you're just relying on AI, you're not super valuable as an employee.
• 23:45
  SPEAKER_01: So here's some foundational questions I think are probably better to ask yourself.
• 23:50
  SPEAKER_01: They're kind of, you know, separate from AI itself, just ask these questions and then see how you answer them.
• 23:56
  SPEAKER_01: So number one, what are employers looking for from a great employee?
• 24:01
  SPEAKER_01: Do you know that?
• 24:03
  SPEAKER_01: Or are you saying, well, they should be looking for me?
• 24:06
  SPEAKER_01: That's the wrong thing to ask.
• 24:09
  SPEAKER_01: What are they looking for from a great employee?
• 24:12
  SPEAKER_01: Because if you just look at yourself and say they should be looking for me, well, your feelings don't affect reality.
• 24:21
  SPEAKER_01: The reality is what you have to look at.
• 24:23
  SPEAKER_01: And so start asking employers, start talking to people and saying, hey, what are they looking for?
• 24:30
  SPEAKER_01: Look at the job postings, look for what they're looking for.
• 24:33
  SPEAKER_01: Talk to people.
• 24:35
  SPEAKER_01: I can tell you that the thing that employers are looking for is competent developers who can get the job done, who know more than AI, who can help direct the AI, and who can get good code written and into production?
• 24:56
  SPEAKER_01: They're looking for people who can make the decisions on what to use out of a various number of options, when to use it, and why.
• 25:05
  SPEAKER_01: Because if you just go, well, the AI says this, yeah, but why didn't you choose those other things?
• 25:11
  SPEAKER_01: I don't know.
• 25:12
  SPEAKER_01: Because the AI said this.
• 25:14
  SPEAKER_01: That's not good enough.
• 25:15
  SPEAKER_01: So what are employers looking for from a great employee?
• 25:18
  SPEAKER_01: That's question number one that you need to answer for yourself.
• 25:21
  SPEAKER_01: Question number two, what steps do I need to take to get there?
• 25:26
  SPEAKER_01: Because you can always improve as a great employee.
• 25:30
  SPEAKER_01: So what steps do you need to take?
• 25:32
  SPEAKER_01: And then number three, how do I show off what I know?
• 25:37
  SPEAKER_01: Because you can know lots of great information.
• 25:40
  SPEAKER_01: You can know how to be a great employee for a developer.
• 25:44
  SPEAKER_01: But if you're the employer that's looking at you can't see that, they're not gonna hire you.
• 25:51
  SPEAKER_01: So you need to be able to communicate that well.
• 25:54
  SPEAKER_01: And it's not just I put it in my resume, not good enough.
• 25:58
  SPEAKER_01: So, how do you show off what you know?
• 26:01
  SPEAKER_01: Those are the questions you need to ask.
• 26:04
  SPEAKER_01: So, not everyone learns best from video training.
• 26:07
  SPEAKER_01: Okay.
• 26:08
  SPEAKER_01: Again, my role, I make my living off of video training.
• 26:13
  SPEAKER_01: I sell video training on IamtimCorey.com.
• 26:16
  SPEAKER_01: Not everyone learns best from that.
• 26:18
  SPEAKER_01: Forget AI for a minute.
• 26:20
  SPEAKER_01: Some people learn best from reading, some people learn best from getting their hands on doing.
• 26:26
  SPEAKER_01: But what I have found is that seeing how someone does something and learning why they do it that way and when to do it, that's really important.
• 26:35
  SPEAKER_01: I have not yet found a better medium for communicating this information besides maybe the classroom.
• 26:41
  SPEAKER_01: And the problem in the classroom is that it's expensive, both in terms of money and in terms of time.
• 26:48
  SPEAKER_01: It's expensive to go to a classroom or even attend virtually because it's, you know, it's gonna cost something.
• 26:55
  SPEAKER_01: There's only a certain number of seats in that classroom, so supply and demand.
• 27:00
  SPEAKER_01: And you also have to conform your schedule to that of the class.
• 27:04
  SPEAKER_01: It's also hard to revisit the lessons later.
• 27:06
  SPEAKER_01: That's why I create video training because it's like the classroom in most ways.
• 27:11
  SPEAKER_01: Now you can't ask the professor.
• 27:13
  SPEAKER_01: That's something that's a little harder.
• 27:15
  SPEAKER_01: But what you do get is the ability to watch that lesson over and over again at your pace, at your time, and for a much cheaper rate.
• 27:25
  SPEAKER_01: Okay.
• 27:25
  SPEAKER_01: That's why I do what I do.
• 27:28
  SPEAKER_01: Because it allows me to train the most amount of people, the most effective way possible, that is the cheapest.
• 27:36
  SPEAKER_01: So that's why I still do VA training, even though, as you can see, AI makes people think that things can be better or different.
• 27:45
  SPEAKER_01: But the reality is I don't see it creating better developers.
• 27:49
  SPEAKER_01: What I see it creating is people who think they're developers, but just aren't employable.
• 27:54
  SPEAKER_01: I want to see you employable and better employed.
• 27:58
  SPEAKER_01: So that's why I do what I do.
• 28:01
  SPEAKER_01: Thanks for listening.
• 28:02
  SPEAKER_01: And as always, I am Tim Corey.
• 28:04
  SPEAKER_00: Thank you for joining us for this episode of Dev Questions.
• 28:07
  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.