Latest / DevQuestions with Tim Corey / 133 Will Artificial Intelligence Take Away Developer Jobs?

Transcript

133 Will Artificial Intelligence Take Away Developer Jobs? · DevQuestions with Tim Corey

• 0:00
  SPEAKER_01: Welcome to the Dev Questions Podcast with Tim Corey.
• 0:03
  SPEAKER_01: Join us each episode as we tackle the questions you are asking about a career in software development, understanding the industry, and new technology.
• 0:12
  SPEAKER_01: If you're just starting out or you want to grow stronger as a developer, this is the place to get your questions answered.
• 0:19
  SPEAKER_01: Now, here's your host, expert developer and online educator, Tim Corey.
• 0:25
  SPEAKER_00: Will artificial intelligence take away developer jobs?
• 0:29
  SPEAKER_00: This is a question that's been asked a lot recently, especially as Chat GPT and GitHub Copilot and other tools like this have come out.
• 0:38
  SPEAKER_00: So I want to address my thoughts on if artificial intelligence is going to take my job away or your job away.
• 0:46
  SPEAKER_00: So here's the short answer.
• 0:48
  SPEAKER_00: Probably not.
• 0:49
  SPEAKER_00: But things will change.
• 0:50
  SPEAKER_00: And that's important to note.
• 0:52
  SPEAKER_00: It's probably not going to take your job away or even future jobs necessarily, but things are going to change.
• 0:58
  SPEAKER_00: So let's talk through change and let's start with some history, because who doesn't love history?
• 1:06
  SPEAKER_00: This has happened before.
• 1:07
  SPEAKER_00: And you might say, Tim, artificial intelligence has never just taken over the world, you know, except in the movies like Terminator.
• 1:14
  SPEAKER_00: Yes, but we have had major shifts before.
• 1:19
  SPEAKER_00: For example, um, and there's a lot I can pull from, but I've pulled just three.
• 1:23
  SPEAKER_00: Uh the first one is cars, automobiles.
• 1:27
  SPEAKER_00: When the automobile was first invented, before that point, if you wanted to transport things a short distance, um you'd want to use like a train, you would use a horse and wagon, a horse and buggy, some type of horsepower.
• 1:42
  SPEAKER_00: Um but now with the advent of cars, we kind of eliminated all those jobs, right?
• 1:50
  SPEAKER_00: So the car killed off a particular industry.
• 1:54
  SPEAKER_00: Yes and no.
• 1:55
  SPEAKER_00: And it's here's where it's important, and we're gonna talk about this more, is that if you look at an individual job, then yes, a certain change will have eliminated a specific job.
• 2:08
  SPEAKER_00: Let's take, for instance, a fictional person who had worked for 30 years as a uh a milk driver, where he would have a couple of horses and a wagon and he would deliver milked places, right?
• 2:20
  SPEAKER_00: Let's say he's been that job for 30 years.
• 2:23
  SPEAKER_00: All of a sudden, the advent of the automobile, and suddenly trucks are available, and now he doesn't have his job specifically gets eliminated because those trucks can go faster and further, and it takes less work, and you don't his need to maintain the horses, you know, take care of the horses in the stable and feed them and clean them and fix their gear and all the rest of stuff.
• 2:49
  SPEAKER_00: So his job specifically gets eliminated.
• 2:53
  SPEAKER_00: And in his place, another job that's very, very similar gets put in place, and that is a a driver who drives a truck.
• 3:04
  SPEAKER_00: Well, it's not quite the same because a person driving a truck might replace two workers who used to drive the horses.
• 3:13
  SPEAKER_00: I'm making this stuff up, but you can see how, yes, there was a replacement here, but all of a sudden we have a change, right?
• 3:23
  SPEAKER_00: So a job is eliminated, a job is created, the net balance is pretty close the same.
• 3:28
  SPEAKER_00: It might be that more jobs were eliminated than created, or at least at first.
• 3:33
  SPEAKER_00: However, now you're talking about you need mechanics and you need a um people to build the trucks.
• 3:40
  SPEAKER_00: And you all of a sudden there's jobs on the other side that are created.
• 3:44
  SPEAKER_00: And so there is a balance here.
• 3:46
  SPEAKER_00: And it feels like at first that yes, but less jobs are created than are destroyed, and those specific people lose their jobs.
• 3:55
  SPEAKER_00: And the specific people, yes, can lose their jobs.
• 3:59
  SPEAKER_00: But as a whole, we don't actually eliminate that many jobs because as we have more efficiencies, typically you don't say, let's do the same amount of work, but let's be more efficient about it.
• 4:13
  SPEAKER_00: Typically, what happens instead is we can now do more.
• 4:17
  SPEAKER_00: That milk delivery company can all of a sudden deliver to more houses than they used to, or more businesses than they used to.
• 4:24
  SPEAKER_00: And so we start to see more drivers being hired and more trucks being purchased and maintained.
• 4:30
  SPEAKER_00: And so we have a bit of a balance.
• 4:33
  SPEAKER_00: Let's continue on and let's talk about the specifics of what I can see for AI for developers.
• 4:39
  SPEAKER_00: So, next example, airplanes.
• 4:42
  SPEAKER_00: When airplanes came out, all of a sudden fast travel was a real thing.
• 4:47
  SPEAKER_00: And so when we had, you know, in the 50s and 60s, where we had these jets that could take us long distances, it eliminated the need for or reduced greatly the need for boats and trains.
• 5:01
  SPEAKER_00: So maybe specifically a porter or a um some type of person on a boat that would take care of passengers.
• 5:10
  SPEAKER_00: That person's job was eliminated.
• 5:12
  SPEAKER_00: But then a uh you know, a host or hostess for the airlines was created in its place.
• 5:19
  SPEAKER_00: And really, those are close to the same type of job, or at least the same level of education, et cetera.
• 5:26
  SPEAKER_00: And so one industry goes down, the other industry raises up.
• 5:30
  SPEAKER_00: We have airports now, we have uh people that support the airport where we used to have train stations and people that supported the trains.
• 5:40
  SPEAKER_00: So we have this idea of one comes up and one goes down a bit.
• 5:46
  SPEAKER_00: Computers are the same way.
• 5:48
  SPEAKER_00: Uh we used to have accountants and filing clerks and lots and lots of papers.
• 5:52
  SPEAKER_00: And all of a sudden we we brought computers in and it eliminated a lot of those repetitive jobs.
• 6:00
  SPEAKER_00: And in some ways, I heard a lot, especially growing up, the idea of a paperless office.
• 6:05
  SPEAKER_00: And all of a sudden, we wouldn't need filing cabinets, we wouldn't need filing clerks, we wouldn't need people who dealt with paperwork.
• 6:12
  SPEAKER_00: And that still hasn't happened because we still have those people and we have a lot of people who are dedicated to working on a PC.
• 6:21
  SPEAKER_00: So we have eliminated some roles, but really replaced them with different roles.
• 6:27
  SPEAKER_00: So there is this idea of remove an ad, remove an ad.
• 6:32
  SPEAKER_00: But the other thing I want to point out is that these things are inevitable.
• 6:39
  SPEAKER_00: When you invented the car, it wasn't like you could look at it and go, that's cool.
• 6:44
  SPEAKER_00: We're gonna ignore that.
• 6:45
  SPEAKER_00: In fact, we will try to do that.
• 6:47
  SPEAKER_00: They said it will never work, it will never replace the good old horse and carriage.
• 6:53
  SPEAKER_00: This is just a foolish side venture.
• 6:57
  SPEAKER_00: You can see what happened.
• 6:58
  SPEAKER_00: Airplanes, same thing.
• 7:00
  SPEAKER_00: You know, yeah, it's fine, it's a fun little flukey can float in the air.
• 7:04
  SPEAKER_00: Who cares?
• 7:05
  SPEAKER_00: It's never gonna replace a boat or a train.
• 7:09
  SPEAKER_00: Computers.
• 7:10
  SPEAKER_00: Yeah, but it's never gonna be as good as my accountant is.
• 7:15
  SPEAKER_00: Your accountant's gonna start using computers.
• 7:17
  SPEAKER_00: So it was kind of inevitable that when it was created, it would be used, it would be taken advantage of, and it would change how we do things.
• 7:27
  SPEAKER_00: And because of that, it would eliminate jobs and it would create jobs.
• 7:32
  SPEAKER_00: And this is just the nature of how change happens.
• 7:35
  SPEAKER_00: And we could fight it, we can say it's never gonna change me, but the reality is it will.
• 7:42
  SPEAKER_00: Okay.
• 7:43
  SPEAKER_00: So let's talk now about software development.
• 7:48
  SPEAKER_00: So I want to start before we talk about AI, if it's gonna replace us, let's talk about what our job actually is.
• 7:55
  SPEAKER_00: As software developers, what is our job?
• 7:59
  SPEAKER_00: Our job is not to write syntax.
• 8:03
  SPEAKER_00: That's not our job.
• 8:05
  SPEAKER_00: Our job is not to know or use curry braces and semicolons.
• 8:12
  SPEAKER_00: We have to do that in our job, but that's not our job.
• 8:14
  SPEAKER_00: That's not our purpose for existing.
• 8:17
  SPEAKER_00: Because I can tell you, as a software developer with over two decades of experience, the language I use has changed multiple times over those years.
• 8:30
  SPEAKER_00: The syntax I use has changed dramatically over those years.
• 8:36
  SPEAKER_00: What I can do with that syntax has greatly increased over those years.
• 8:43
  SPEAKER_00: So it's not about the syntax.
• 8:46
  SPEAKER_00: Instead, our job as a developer is to create logic.
• 8:52
  SPEAKER_00: That's our job, to take the requirements and figure out the logic necessary to get the solution we need.
• 9:00
  SPEAKER_00: The outcome we need and the desired outcome, we have to match those up by creating logic, create that bridge between what you want to do and what it does using logic.
• 9:13
  SPEAKER_00: And so that's why when people talk about, oh, well, I want to skip around languages, learn one deep first.
• 9:19
  SPEAKER_00: Because it's not about the syntax.
• 9:22
  SPEAKER_00: If if you learn one language really well, you'll have some definite benefits in that language, but you will also have definite benefits in other languages.
• 9:33
  SPEAKER_00: It's not as big of a deal to switch languages.
• 9:36
  SPEAKER_00: Now, I'm not saying bounce around, but I'm also not saying always stay in one language.
• 9:41
  SPEAKER_00: What I'm saying is get a depth of ability in one language, because when you do, you'll be practicing more and more logic and less and less focusing on the syntax of it.
• 9:55
  SPEAKER_00: Okay.
• 9:55
  SPEAKER_00: Once you get the syntax down, then the logic is what you have to build up because that's the important part.
• 10:03
  SPEAKER_00: So when it comes to artificial intelligence, artificial intelligence cannot create reliable logic.
• 10:12
  SPEAKER_00: Now, will that ever change?
• 10:15
  SPEAKER_00: I can't say it won't.
• 10:17
  SPEAKER_00: But for now, even, even with the crazy stuff we've seen recently with Chat GPT, it cannot create reliable logic.
• 10:27
  SPEAKER_00: Instead, it creates an imitation of logic.
• 10:30
  SPEAKER_00: There's a really important distinction there.
• 10:33
  SPEAKER_00: Because even though it looks logical, even though it creates things where you're like, man, that looks so good, it's an imitation of logic.
• 10:42
  SPEAKER_00: And where did that logic come from?
• 10:44
  SPEAKER_00: It actually came from developers.
• 10:47
  SPEAKER_00: It came from developers saying, This is how I solve this problem, and this is how I solve that problem.
• 10:53
  SPEAKER_00: And it took the two and said, okay, can I mix those together?
• 10:56
  SPEAKER_00: And it tried stuff and it it guessed, but based upon lots and lots of input.
• 11:03
  SPEAKER_00: So it's not creating logic.
• 11:06
  SPEAKER_00: It is not sentient.
• 11:10
  SPEAKER_00: How about that?
• 11:10
  SPEAKER_00: It's not making a leap.
• 11:12
  SPEAKER_00: There's a big leap.
• 11:13
  SPEAKER_00: There's a big gap right now.
• 11:15
  SPEAKER_00: And that gap is from where we are with artificial intelligence, where it's taking machine learning and learning from us and then making some inferences based upon what it learned from us.
• 11:27
  SPEAKER_00: But a big gap from that into learning on its own, where it's making its own and testing its own and validating and verifying its own inferences until it gets the right answer every time.
• 11:39
  SPEAKER_00: Okay.
• 11:39
  SPEAKER_00: There's a big gap there.
• 11:40
  SPEAKER_00: I'm not saying it won't be filled.
• 11:42
  SPEAKER_00: In fact, I think it will be filled five, 10, 15 years from now.
• 11:46
  SPEAKER_00: I think that that gap may be filled.
• 11:49
  SPEAKER_00: But in that case, so where we are right now, let's talk about first, and then we'll talk about in the future.
• 11:56
  SPEAKER_00: So right now, no, we're not going to have AI replace our jobs because AI can't do reliable logic.
• 12:05
  SPEAKER_00: And our jobs are to create logic.
• 12:09
  SPEAKER_00: Now, again, when you're talking about specific jobs, there may be specific jobs that do get replaced.
• 12:20
  SPEAKER_00: So there are some companies that will say, well, if we use this Chat GPT or whatever the next iteration is the next year or two, we'll use that as a companion or a co-pilot, maybe poor choice of words, a you know, a uh pair programmer for every developer in our department.
• 12:42
  SPEAKER_00: And we have 10 developers, let's cut that down to five, because now we have five developers with pair programmers.
• 12:48
  SPEAKER_00: So it's kind of the same.
• 12:50
  SPEAKER_00: And some companies may short-sightedly do that.
• 12:54
  SPEAKER_00: And it might actually even work.
• 12:57
  SPEAKER_00: So, yes, those five jobs will have been lost, but that isn't kind of the end of the story.
• 13:04
  SPEAKER_00: So developers will get faster at their jobs, mostly, but they'll get faster at their jobs using an AI companion.
• 13:13
  SPEAKER_00: Uh, GitHub Copilot is a great tool for making you faster at your job.
• 13:18
  SPEAKER_00: You say, This is what I'm trying to do, and it goes, here's some code for you, and it just puts it in.
• 13:24
  SPEAKER_00: And then you get to validate.
• 13:25
  SPEAKER_00: And here's where the the mostly comes in.
• 13:28
  SPEAKER_00: I saw a comic, I believe, on Twitter, um, where a person said, you know, before Chat GPT, before GitHub Copilot, there was a developer that spend two hours, you know, banging their head against the desk trying to figure out how to write the code and another six hours debugging the code.
• 13:47
  SPEAKER_00: And now with ChatGPT, they can get their code done in five minutes and spend 24 hours debugging it.
• 13:54
  SPEAKER_00: That's kind of where we're at.
• 13:56
  SPEAKER_00: Because while it will mostly make you faster, there's gonna be some serious quirks with it.
• 14:02
  SPEAKER_00: It's gonna add more bugs, it's gonna trip you up in ways you hadn't expected.
• 14:07
  SPEAKER_00: So while developers will get faster, they're gonna be slowed down at some point as well.
• 14:14
  SPEAKER_00: So one of the great illustrations I've heard about specifically Chat GPT is it feels like an intern.
• 14:22
  SPEAKER_00: I think that's a great illustration.
• 14:24
  SPEAKER_00: It feels like having an intern working for you.
• 14:27
  SPEAKER_00: They may not know everything.
• 14:29
  SPEAKER_00: In fact, they don't know half of stuff you need to, and they may get stuff wrong, they may break stuff, but they'll also offload some of your work.
• 14:38
  SPEAKER_00: And so there's some benefit as well.
• 14:42
  SPEAKER_00: So some developers, most developers, should get faster using Chat GPT or AI in general.
• 14:49
  SPEAKER_00: Now, some companies are gonna hire less developers or let some go because of this.
• 14:55
  SPEAKER_00: And so they might say, hey, you know what?
• 14:57
  SPEAKER_00: This is a way to cut our budget.
• 14:59
  SPEAKER_00: And so they will reduce the headcount because they can get the same amount of things done because they have that companion, that AI companion that will help them move faster.
• 15:12
  SPEAKER_00: I think that's a short-sighted idea, but some companies will do that.
• 15:17
  SPEAKER_00: And that's gonna be the case even if there wasn't an AI companion, because companies always want to spend less to get more.
• 15:26
  SPEAKER_00: That's always gonna be true whether or not we have something like an AI companion.
• 15:32
  SPEAKER_00: Now, some companies are going to hire new developers or create developer positions because of the fact that if you hire one developer, it's almost like hiring one and a half or two.
• 15:47
  SPEAKER_00: And all of a sudden, it becomes more feasible.
• 15:50
  SPEAKER_00: It becomes a little easier to do.
• 15:53
  SPEAKER_00: And so now all of a sudden, you can hire just one person and get more done with them.
• 16:01
  SPEAKER_00: And you couldn't afford two before, you kind of needed to, but now you can hire one and get the whole job done.
• 16:07
  SPEAKER_00: And so companies that hadn't previously hired developers are gonna start hiring developers.
• 16:13
  SPEAKER_00: And then some companies are going to hire people just to work with the AI.
• 16:20
  SPEAKER_00: People who can think through the logic of it and say, okay, this is how we ask for this.
• 16:28
  SPEAKER_00: This is how we tweak our requirements, this is how we communicate what we need to in order to get the right results out of it.
• 16:36
  SPEAKER_00: Because there are a lot of benefits to understanding the logic to think through of how it works, what information it has, and how to get the most out of that.
• 16:47
  SPEAKER_00: That's really a developer role, even if they don't write any code.
• 16:51
  SPEAKER_00: Examples of this might be a marketing firm.
• 16:55
  SPEAKER_00: Marketing is nothing to do with software development, right?
• 17:00
  SPEAKER_00: Well, yes, but if you had a marketing firm where they wanted to improve their messaging, reach more people, have uh catchier clickbait titles, whatever it may be, they could use Chat GPT to get those.
• 17:16
  SPEAKER_00: But marketers aren't necessarily the people who are trained to think logically, to think about how to get a software to do a job, how to ask the right questions, or how to tweak things.
• 17:32
  SPEAKER_00: And so a person could be a specialist in just getting the right answers out based upon what they want in.
• 17:42
  SPEAKER_00: And so all of a sudden, this person that's being hired is taking the requirements and getting the result.
• 17:50
  SPEAKER_00: The only difference is instead of writing code, let's say C sharp code or whatever, instead of writing that code, they are writing the right script for the AI, the right set of commands and the right tweaks and the way to dial in exactly what they want from the AI.
• 18:09
  SPEAKER_00: So it's still logic.
• 18:11
  SPEAKER_00: It's just that it's not a traditional developer role.
• 18:16
  SPEAKER_00: So I think that might come about as well.
• 18:20
  SPEAKER_00: And so is AI going to replace developers?
• 18:24
  SPEAKER_00: No, no, not really.
• 18:26
  SPEAKER_00: Um, it's not going to be this mass layoff in the industry, and all of a sudden, you know, AI is doing all the jobs.
• 18:34
  SPEAKER_00: It's just not going to happen.
• 18:36
  SPEAKER_00: It's going to take away some repetition.
• 18:38
  SPEAKER_00: And so, developers who were doing simple repetitive jobs, yeah, they might not have a job anymore, but then there's other jobs that will open up because there will be other opportunities to be a developer.
• 18:53
  SPEAKER_00: And no, it's not going to take away all entry-level jobs.
• 18:56
  SPEAKER_00: That's already a problem area where, you know, what we don't have enough entry-level jobs.
• 19:02
  SPEAKER_00: People want to hire a developer who is more skilled, right?
• 19:06
  SPEAKER_00: Well, if you are trained well, you have a good foundation in development, even if you are a junior-level developer, you can have the AI help you and be a better mid-tier developer just by knowing how logic works and using it to help you with syntax or helping you with a certain key piece.
• 19:30
  SPEAKER_00: And all of a sudden, yeah, maybe we don't have the entry-level developer junior positions where they just fix if statements.
• 19:41
  SPEAKER_00: But what we do have is entry-level positions that can use the AI to do more advanced stuff.
• 19:48
  SPEAKER_00: So I don't think we're going to see uh this revolutionary change in our industry where developers are out of jobs.
• 19:56
  SPEAKER_00: I don't think that's true.
• 19:57
  SPEAKER_00: Um, we do have currently a downturn or seeming downturn in developer jobs, but really we don't.
• 20:04
  SPEAKER_00: And this is where it's it's tricky because a lot of people look at the um the major, the big players, you know, Microsoft, Google, Apple, um, these big companies, and say, oh, if they're laying off, the entire industry is slowing down.
• 20:19
  SPEAKER_00: And that's not really true.
• 20:20
  SPEAKER_00: In fact, a lot of companies out there are hiring and ramping up and adding more positions.
• 20:26
  SPEAKER_00: It's just that, you know, this small shop that only has 10 developers and wants to go to 15 doesn't make as big of a splash as Google.
• 20:36
  SPEAKER_00: And so we think that Google is the only thing.
• 20:39
  SPEAKER_00: And that's not true.
• 20:41
  SPEAKER_00: So be careful when you think about where the industry is if you look at just the big players.
• 20:47
  SPEAKER_00: So, no, I don't think that AI is going to replace our jobs.
• 20:50
  SPEAKER_00: I don't think that we should stop teaching software development.
• 20:53
  SPEAKER_00: I don't think we should stop learning or encouraging new people to come along.
• 20:58
  SPEAKER_00: I absolutely think it's a great time to be a software developer.
• 21:02
  SPEAKER_00: I think that there's a lot of benefits to becoming a software developer.
• 21:06
  SPEAKER_00: And I do think there will be jobs available for software developers for a long time going forward.
• 21:14
  SPEAKER_00: Okay.
• 21:14
  SPEAKER_00: I do not anticipate in the next 40 years not having roles for software developers.
• 21:20
  SPEAKER_00: I think the more roles, not less, for software developers in the next 40 years.
• 21:26
  SPEAKER_00: So even when AI becomes self-aware, we'll still have needs for software developers in the industry.
• 21:35
  SPEAKER_00: I really don't think self-aware AI is going to happen.
• 21:37
  SPEAKER_00: Um, I don't think it's going to become this perfect solution.
• 21:41
  SPEAKER_00: It's still learning based upon what developers create in order to give out developers solutions.
• 21:48
  SPEAKER_00: And that is actually a big danger for AI because in the future, if it starts learning from itself, that can actually degrade further its ability to be viable in the industry.
• 22:01
  SPEAKER_00: So we might actually see this rising peak of AI assistedness and AI helping out.
• 22:08
  SPEAKER_00: Also, we might start to see a downturn in the effectiveness of that because of the fact that it's kind of feeding upon itself and self-reinforcing things that aren't correct and all of a sudden becomes less and less valuable.
• 22:22
  SPEAKER_00: And there's some definite hurdles to overcome there.
• 22:25
  SPEAKER_00: So our job is still safe.
• 22:28
  SPEAKER_00: You should, it's definitely still a good time to become a software developer.
• 22:32
  SPEAKER_00: Okay.
• 22:32
  SPEAKER_00: Hopefully that answered your question.
• 22:34
  SPEAKER_00: Thanks for listening.
• 22:35
  SPEAKER_00: And as always, I am Tim Corey.
• 22:37
  SPEAKER_02: Thank you for joining us for this episode of Def Questions.
• 22:40
  SPEAKER_02: Tim is committed to making it easier for you to become a developer.
• 22:44
  SPEAKER_02: If you would like to help make more content like this possible, please like, subscribe, rate, and share Dev Questions.
• 22:51
  SPEAKER_02: You can also send your questions to questions at IamTimcorey.com.
• 22:56
  SPEAKER_02: Until next time, remember, you are too smart and your time too valuable to waste it making all the mistakes Tim did.
• 23:03
  SPEAKER_02: When you're ready to learn to think and code like a professional developer, head over to imtimcore.com and enroll in a course.