Save Your Sons: What Parents Need to Know (DF#197)

Ryan Maruyama [00:00:00]:
Welcome back. We have been gone for a long time, but we are back and a lot has changed since we have been not recording this podcast regularly. And a lot has changed in the world of AI.

Hannah Maruyama [00:00:11]:
Yeah, a lot has changed. The AI tools that are available and then the models themselves have improved drastically. And while that's not actually getting rid of work or jobs, it actually is moving around the market quite a bit. And I want to talk about a few different things that a lot of people have been saying. You know, basically every time somebody goes in and just says, oh, your kids shouldn't go into X, Y, Z job, they're usually wrong. And, and that is something I want to talk about today, especially because I want you to kind of share a little bit about how you've been working with AI and also what you see coming as the future because you work so heavily with these tools. And so today I want to talk about this clip I saw of Jensen Huang, who's the CEO of Nvidia. And he was talking about how the fear mongering around AI and around a lot of the tools and the data centers themselves, a lot of that is doing a huge disservice to our country because that's a huge job creator.

Hannah Maruyama [00:01:00]:
Not only that, but it's really the equivalent. And this is my view of saying, oh, we shouldn't build trains, we shouldn't build skyscrapers, we shouldn't do X, Y, Z. Like a lot of the work that went on after the New Deal where our country's infrastructure was being updated, we updated highways, we updated power grids, we updated public transportation systems. And so it really is the equivalent in my mind of people just saying, oh, like, we shouldn't improve the infrastructure of our country. That I think is the wrong view to take. And what's happening, and I think that it is going to keep people from going into fields where there is demand, where there is good pay. And I want to get your view on what young adults can be learning going into this market, especially for those who are interested in computer science or software development. Because one of the things that's being said a lot is your kids don't need to learn how to code.

Hannah Maruyama [00:01:47]:
And that, I think is going to cause a huge software development shortage.

Jensen Huang [00:01:52]:
If we scare this country into thinking that AI is somehow a nuclear bomb so that everybody hates AI and everybody's afraid of AI. I don't know how you're helping the United States. You're doing a disservice. If we scare everybody out of doing software engineering jobs because it's going to kill every software engineering job. And we don't have any software engineers as a result of that. We're doing a disservice to the United States if we scare everybody out of radiology. So nobody wants to be a radiologist because computer vision is completely free. And no AI is going to do a worse job than a radiologist.

Jensen Huang [00:02:26]:
And we misunderstand the difference between a job and the task. The job of a radiologist, patient care task, to read a scan. If we misunderstand that so profoundly and we scare everybody out of going to radiology school, we're not going to have enough radiologists and good enough healthcare. And so I'm making the case that when you make these, make a premise that is so extreme, everything goes from zero or infinity, we end up scaring people in a way that's just not true. Life is not like that.

Ryan Maruyama [00:03:00]:
The one thing that you have to think about, like with Jensen Huang is like, consider your source. He is saying all of this because his company is the number one in the world for AI hardware.

Hannah Maruyama [00:03:11]:
The chips. Yeah, totally.

Ryan Maruyama [00:03:12]:
Well, for the GPUs. And so like. And GPUs as of right now are like the most efficient way to run these AI models. I mean, I think that Google, their last model, was like exclusively trained on TPUs. You have to remember that Jensen Huang needs us to believe that AI is foundational and that AI is here to stay and that it is not going to ruin the world. Because the better AI does, the better Jensen Huang does.

Hannah Maruyama [00:03:41]:
Totally. He's got billions of dollars worth of incentive to say this. However, I think that there is a lot of validity to it, given what we've been seeing. There's a ton of fear mongering around AI as there usually is with any type of new technology. That's kind of humanity's default response is for a while everybody's afraid of whatever the new technology is. Like. That's just our nature, I think, to be afraid of things, you know, down to cavemen being afraid of fire. At first, I think that this is just kind of how we are.

Ryan Maruyama [00:04:06]:
This is going back to the beginning. But I was saying this years ago. It's very similar to the atm. When the ATM came out, everyone thought that bank tellers jobs were going to go away. And in fact their jobs increased. The amount of bank teller jobs increased total. Because, yes, the ATM took a small portion of the transaction volume that a bank teller could do, but it grew the size of the pie, it made it cheaper and more economical. To, to build and stand up branches like bank branches all throughout the US and all around the world.

Ryan Maruyama [00:04:47]:
And so they did that. And then because there was more bank branches, there was then total more bank tellers. And that's a second order effects that most people are not good at seeing. Pretty much nobody saw that coming. Pretty much nobody is good at seeing those types of things. And I think that we are in the same place with AI where I mean if you just look at the fear mongering coming from the providers themselves, coming from the big players, Anthropic, OpenAI, Gemini, Google, they have been saying that software engineers like this is going to take software engineers jobs in six months. They've been saying that for the last like three years. If you look back, literally that was said years ago.

Ryan Maruyama [00:05:35]:
And it's just a talking point that they keep repeating and repeating and repeating. And the fact of the matter is you can look at the overall software engineer jobs and it is exploded.

Hannah Maruyama [00:05:46]:
Yeah. It's up 15% from August last year. Yeah.

Ryan Maruyama [00:05:49]:
And that is because software has become cheaper to make. Everyone has software. We're going to get to a point where your local mom and pop bakery employs a dev. They probably employ two or three devs. It's like, I mean that's still a small team but like they probably employ two or three devs because of their online ordering thing or their marketing funnels or something like that.

Hannah Maruyama [00:06:13]:
Or their payroll.

Ryan Maruyama [00:06:14]:
Right, exactly. Or to create automations for their bread timers or something like that. The amount of software is going to increase by a very large amount from that. You can see that that's already happening too because if you look at the popularity of GitHub. Right, which is bought by Microsoft, if you look at the popularity of GitHub and you look at the amount of repos that are on GitHub that has exploded. That has exploded. And the reason why is because dummies like me can make software.

Hannah Maruyama [00:06:44]:
Well, you're, you're actually a power user. I mean even I can use it, which has been wild because you know, you've got me in VS code, which is wild to me. It's absolutely wild that I'm using, I'm, I'm reading things in Markdown and HTML, which I actively avoided even while I was actively working in, in with a software team, even while I was actively working in SaaS. I did not touch those things because I wasn't a dev. But now anybody is a dev. Anybody's a dev. But that means you need even more devs to compensate for the fact that people who are developing software are non technical.

Ryan Maruyama [00:07:14]:
Yes. And so say anybody is a dev, anybody can write code.

Hannah Maruyama [00:07:18]:
There we go, There we go. Well, because do you think that there's going to be maybe a different even title like, or maybe it's just product. Right. More people are going to have product people. Because that's more of what I'm talking about where you're not the dev and you're not even really an architect. What would you say?

Ryan Maruyama [00:07:32]:
I believe that there's going to be a split where maybe, yes, the smaller jobs or whatever get quote unquote wiped out. The role might be intern right now or like a, not even a junior dev, but like a junior junior dev. Somebody that literally just came out of a boot camp that doesn't know anything. Those people largely gonna be very difficult to break into the industry right now. If you have no experience with the tools, if you have no experience with the AI tools that everybody's using, then yes, it's gonna be very difficult for you. On the flip side, if you have extensive experience with the tools that you don't know that much code, you can probably find a job. Cause you know how to use the tools better. Now on the other side, on the higher side or more experienced side, the people that orchestrate these AI agents, the people that orchestrate these AI teams, those guys are going to be definitely well paid and there's going to be more and more and more of them.

Ryan Maruyama [00:08:34]:
People much smarter than me have painted a picture with development. When you're hired, you basically are hiring their agents. When you hire somebody, you're not hiring them, you're hiring what they can do. And what they can do is they can bring their agents. And basically, you know, agents are just their AI. They can bring the team that they've created over the years of actually doing this, the type of work that you're hiring them to do, they can plug that into your workflows, into your processes, into your organization, and then they can be that much more productive. And so you're going to see like there's 10x coders that are in the tech industry now and soon to be 100x coders.

Hannah Maruyama [00:09:15]:
And for those of you listening who don't know what a 10x coder is, that's a coder who is worth 10 regular coders. So this, this coder just pushes a lot more code. And you find these people a lot of times in organizations where they just do most of the work and it's just one person. And this person Just happens to be really good at their jobs.

Ryan Maruyama [00:09:31]:
So to bring it back, I think that we are in that same inflection point. We are scared of the ATM machine. We, because we don't understand that there's gonna be a lot more bank branches.

Hannah Maruyama [00:09:41]:
Sure.

Ryan Maruyama [00:09:41]:
And like you're scared of your child being a software engineer now, but you don't understand because you're not in it. And like you don't get it because of whatever your job is that everyone is going to have software, everyone is going to need software. It is going to be standard for mom and pop shops to have a custom built piece of software that fits their workflow. The days of buying and procuring software out there and you can see it in a lot of startups, they've, they've crumbled because they, they might have done one thing really well and, and there was like connection between this piece of software and this piece of software and they just did that and their businesses crumbled because it's just not necessary anymore

Hannah Maruyama [00:10:26]:
because everybody can do the exact same thing without having to buy their tool.

Ryan Maruyama [00:10:29]:
Right, exactly. And so where I think there is a competitive advantage in the market is I think the big players like Salesforce, like the CRM guys for the customer facing stuff or like for the people that understand that mostly for those guys, I think that they're probably safe for a really long time. If anything, their moat's going to deepen as they integrate AI into your data stack. And then you're just going to be like, oh well this is so much easier than building my own customer and hiring a dev. Right, exactly.

Hannah Maruyama [00:10:58]:
So you're just going to have the two. It's either of course I have software because everyone needs software or of course I have a dev, everyone needs a

Ryan Maruyama [00:11:03]:
dev or those guys get. So those bigger companies make it so modular, they make the design of their software so modular that you can kind of like WordPress, like you can change and alter your instance of their software to how you see fit. And that's easier than just like building Salesforce from the ground up.

Hannah Maruyama [00:11:22]:
And I think this just for, for those listening to this, I think a good way, if you're a parent, you're listening to this, you're going, well I, you know, I just want this to be true, but I'm not really sure. Well, if you look at social media is a good example of the last like big tech turning that we had was there was, nobody had social media. You know, 20 years ago it was uncommon for businesses to have every type of social media. Now, for the most part, most businesses are on at least, you know, at least two platforms, if not three or four or five. You know, you have people that have to manage these things. Social media jobs like social media marketers, social media coordinators, marketing managers, digital marketing managers didn't exist. You know, they didn't exist 15, 17 years ago because Instagram didn't exist. Just for context.

Hannah Maruyama [00:11:59]:
And so all of these jobs that now people are going into college to get degrees to get, which you don't need, all of these jobs that they're going after, these high paying social media jobs, they didn't exist because there was no technology to support them. In the same way. That's exactly what Ryan is saying about AI is that as we actually start to feel its effects, because I did just see this clip of Mark Andreessen saying, like, yeah, AI is not the thing that's taking jobs. It wasn't good enough until December of 2025 to even do most of most of the activities that businesses need it to do. And then I also just read a report that said that something like 90% of Fortune 500 companies have like, not even really deployed it at a meaningful level because they just don't know really what to do with it. And then you and I both know there's a lot of security concerns as well that they could get in deep trouble for. So they're not really sure like where, where they're going to be using it or who's going to put it, where it's supposed to go, who's going to manage it. So those things are complex and I think a lot of people just think, oh, AI tools are on the market.

Hannah Maruyama [00:12:52]:
So of course businesses are just using it and laying off people. And it's like, no, that's really not, that's really not what's happening because if anything it's going to cause them to hire more people. That's definitely what's going to happen as a result of this.

Ryan Maruyama [00:13:02]:
One of the things that I wanted to talk about with that clip as well, when Jensen was talking about radiology, it brings me back to a conversation. I've covered it on this podcast, but it was years ago as well, a conversation that I had with a radiology oncologist. And at the time it was the image models accept images. They're not creating images, but they accept images and they can reason about it. At that time, he was confident that AI was doing a better job of spotting those things than he was. It was already prevalent back then. But that doesn't replace him as the doctor.

Hannah Maruyama [00:13:45]:
No, just. He just got better tools.

Ryan Maruyama [00:13:47]:
Right, exactly. That doesn't replace his expertise. And because he has better tools, I think it's going to be the same thing there. And we're just. We should see more. Radiology screening is going to get more accessible because it is going to get cheaper. And then, you know, assuming that nothing happens, assuming that are the terrible healthcare system stays in place that we have in the U.S. assuming that that sets steady state and it's like horrible, then radiology screening, because it is cheaper, more insurance will cover it and therefore you'll need more people.

Ryan Maruyama [00:14:29]:
You'll need more people because there'll be more demand. Exactly.

Hannah Maruyama [00:14:31]:
Yes. It's very predictable. And then that's where you see Elon Musk's predictions about like, you know, healthcare is going to become like more widely available. All these services that now are super expensive are going to get less expensive. I mean, like preventative screening services. That's kind of what you're talking about as well. That's all going to get less expensive.

Ryan Maruyama [00:14:47]:
I'm saying earlier screens.

Hannah Maruyama [00:14:48]:
Sure.

Ryan Maruyama [00:14:49]:
I'm saying earlier screens because it will be cheaper and therefore insurance will be able to cover it.

Hannah Maruyama [00:14:55]:
Because it costs them less money in the long term.

Ryan Maruyama [00:14:57]:
Because it costs the provider less money.

Hannah Maruyama [00:15:00]:
Right.

Ryan Maruyama [00:15:00]:
Then that charges the insurance company less money, which then allows you to hopefully not die.

Hannah Maruyama [00:15:07]:
Right.

Ryan Maruyama [00:15:08]:
All I was saying about, you know, insurance at a steady state is that like, eventually that will happen if it gets better, if it improves, if whatever political opinion you hold, but if it just costs less money some way somehow overall, then we'll get to that realization quicker, more than likely.

Hannah Maruyama [00:15:29]:
That's a lot of what I wanted to get your viewpoint on, just because you've been in these tools so heavily. But really it was that take of the. I just keep hearing this line of like, your kids, you know, your kids shouldn't learn to code. And a lot of people that are saying this tend to be academics. I think the academics should be scared for their jobs. A lot of the reason is because most of them are not very good at their jobs. And I will flat out just say that. And because they're not very good at their jobs like anybody else who's not very good at their jobs.

Hannah Maruyama [00:15:52]:
Yeah, they should be worried that their jobs are going to. Large parts of it are going to be automated away. And a lot of that I think is market correction. Because as AI does move in and take parts of jobs that don't really work well now, then people are just going to have to find other parts of the market to work in. And I do think that that's just a correction because we have this artificial system that keeps these people in place. And I keep hearing this, yeah, your kids shouldn't code. Art's going to die. Software developers are going to die.

Hannah Maruyama [00:16:19]:
Yeah, I just don't think that things are moving that direction because if you look at like we were just talking about the job listings, the job postings for software developers are up 15% from August of last year. And in the same way, I think what we're going to see as we move into the future where people are talking about, oh, there's no, there's no jobs for artists. But you see that Disney is, is actually hiring hand illustrators because there's going to be so much AI slop that people are going to want human hand illustration instead. I saw this video recently of a young lady who was 19 years old in art school and she was crying because her professor told her that her degree was totally pointless and that learning art was totally pointless because of AI, which is a very academic view to take. And I say academic because they're insulated from the market and they don't know what's going on out in the market because they've never been in it and they never will be in it. And so for creative kids, for technical kids, really, it's just becoming obvious that college is really the problem here. Like it's taking too much time, it's costing too much money, and it's not teaching your kids effective skill sets for the market they're going to go into for the world that they're going to go into for the technology that is going to exist in the world that they work in. And so when you're hearing this advice from K through 12 schools or from their college, just take that with one or two or maybe like a ton of salt so that you are not having your child make emotional, stressed out decisions that do not reflect what's likely to occur in the market.

Hannah Maruyama [00:17:48]:
Just because technology is going to change the type of work that needs done and the type of work that needs to get done.

Ryan Maruyama [00:17:54]:
I think with the engineering specifically, take it with a grain of salt. I'm not technical. I would make the argument that a software engineer's value is no longer in the code that they write or in the code that they manually type.

Hannah Maruyama [00:18:13]:
It's here.

Ryan Maruyama [00:18:14]:
I would argue that what's going to differentiate good software engineers and bad software engineers is probably the same thing that differentiated them before, which is, do you understand what's going on and can you engineer a solution to fix that problem without messing anything up?

Hannah Maruyama [00:18:35]:
Right.

Ryan Maruyama [00:18:36]:
I think that that's it. I think that that skill set will remain constant and has remained constant since the beginning.

Hannah Maruyama [00:18:44]:
So for parents whose kids are 18, 19, 20 years old and are either leaving college or are considering what they're going to do after they graduate school, what type of experience would you suggest they get? If they are interested in like, let's say they graduate high school or they, they've gone to a year of college and they don't feel like they're learning anything useful? Let's say the parents go, hey, what can my child do? How can they learn this type of applied problem solving that you're talking about? Like, what can they do right now to find some way to build that skill set into the future?

Ryan Maruyama [00:19:14]:
Yeah, so I think that this is more of a thing not necessarily for software engineers specifically. This is like literally for, for anyone that is curious about AI. So this is definitely for them. And then I would also say for anyone that is thinking that AI is going to take over the world and it's all going to go to hell in a handbasket, this is double for them the biggest thing. And it sounds so simple, but it is to play with the tools. It is literally to use the tools and they're not expensive. The subscriptions to these like the bottom tiers, subscriptions to OpenAI, anthropic to Gemini, they're like 20ish dollars a month for the, for the first tier. And then just play with it.

Ryan Maruyama [00:20:06]:
Okay, that's not super helpful. Let's say that they're already playing with it. The first tier is like replace Google. Instead of asking Google, ask AI. Okay, awesome. And then what you'll find when you start moving down this continuum, what you'll find is that what you used to type into Google, that seven, maybe five word search, I forget exact number, but I think the average Google search is like five words or something like that. That will start to be like 10 words in OpenAI, in ChatGPT, in Claude. Eventually when you get more used to it, that's going to be like 30 words.

Ryan Maruyama [00:20:46]:
You're going to stop using the keyboard altogether and you're just going to start talking to it. You're just going to say things like, hey, I have two projects that are due next week. One's on Monday, one's on Tuesday. I have baseball practice at whatever, I have whatever at my dorm going on, whatever. An insert this Give it to the schedule. Hey, tell me when I can complete the rest of it. Awesome. There you go.

Ryan Maruyama [00:21:09]:
And then that's like a full two, three paragraphs of a prompt. And then it'll just start doing the work and it'll say, here, here's what you do. And then you just do exactly that. That's like the first level and that's, hey, I have these Google searches or I have these problems that I just need a little bit of solving. Eventually they're gonna get to the point where it's like, I have this problem and I wish there was a solution to it. Let's see if AI can solve it. And that's really where the fun starts to begin. Because that is where the way that I view it is that anything that a human can do digitally, AI will eventually be able to do.

Ryan Maruyama [00:21:47]:
Arguably, if you're a good enough developer, I think that you would, you could probably do that now. I think it just takes a lot of time and a lot of effort and a lot of money tokens. Yeah, like literally a ton of money to just make sure that your infrastructure is correct, the architecture's right, and that your prompts are right. And I think that just that testing and iteration probably doesn't. The cost benefit analysis isn't really there for it to do everything. But I do believe that eventually probably it'll get to that point as long as you architect it. Well, where I think when you're starting off is have your child think of a problem and just have them think in a workflow and just have them think, what are the nodes in this workflow? So just thinking of, sorry, this is if this is getting too abstract, but literally just think of a if this, then this, then this workflow and just think of all the different nodes in that. First I open this window and then I log into my school's website and then I click my assignments, those types of things and then, okay, cool.

Ryan Maruyama [00:22:55]:
Dictate all of that to ChatGPT and say, how can you make me something to automate this process?

Hannah Maruyama [00:23:03]:
I think I have a couple more examples of like hands on stuff you can do as well that you can incentivize your kids to do too. Because I know that sounds like while extremely useful, I think you have to connect it to some real world reward so that they actually want to do it. Otherwise it's like just buy tears, buy chat tears and they're just going to use it like a chat bot because they just don't understand. So if parents start having your kids Solve your problems. So for instance, if your child maybe, maybe and this is more for like a high schooler or like early, you know, like late teen. But have them either like plan a trip or have them complete or automate a household task that costs you time. Like have them figure out how to turn the lights, you know, your outdoor lights, like on and off, or automate the garage door or something with, from their phone. Have them figure out how to do that or something like that.

Hannah Maruyama [00:23:49]:
And then if they do, they get a reward. Like they get to go to a movie or they get to go do xyz. Another thing you could do that's kind of fun is if you have multiple kids, you could have them create like presentations to prove a point or almost like a book report when you're in elementary school. But have them use AI completely to draft this story. And then, you know, you can give out awards for like the most entertaining one or the most convincing one. And whoever, whoever makes the most convincing presentation, they can put it on their phone and then, you know, put it on your, your, your smart TV or whatever, or even just a regular tv. But have them figure out how to do this or a laptop or whatever, have them give you a presentation. And then whoever does the best one now gets, you know, a trophy for that week.

Hannah Maruyama [00:24:28]:
It could be something small, but that way you encourage like competition and communication, which is also really key because that's the thing employers are saying is the most lacking in Gen Z. And if you can foster that with your kids now, it's really going to help their job prospects going forward because they use the tools and they're telling stories and they're communicating effectively in a competitive environment as well, that's good.

Ryan Maruyama [00:24:49]:
I think that's especially good for younger kids, but I think for older kids just finding something in their day to day work because that's normally what they'll be doing. Like, so let's back it up. When they work somewhere, assuming that they get all of their work done, an employer will never be like, I'm so upset that you took the time to make this process more efficient. Assuming that they are already taking care of the things that they need to take care of and they're doing it manually or whatever. But if they're playing around and they're saying, look, I have this problem every Monday I have to take this spreadsheet and I have to copy these rows and then put it over into this thing manually, how do I not do that? Okay, and if they can figure that out, awesome. It Sounds like at the end of it, you know, everyone's going to be a software engineer, kind of sort of. But that's not really what I mean, because exactly what you were talking about, presentations or other documentation, Google Docs or PowerPoint, PowerPoints, Notion Docs, Excel spreadsheets, whatever it is, AI can help in all of those things right now, currently. And so the workflows is definitely a more technical thing.

Ryan Maruyama [00:26:04]:
A lot of software uses workflows. I mean software, arguably all it just is workflows. And then, you know, marketing teams use workflows and everything like that. But if we can start thinking in workflows and have them start thinking in workflows and just think, okay, now how can I apply AI to this so that I don't have to do it? And yes, the answer is like, well, just learn how to code. Sure, that probably like you could probably over a few years learn how to do that. But the point is for 20 bucks a month, you don't have to. For 20 bucks a month you can just solve problems. Yes.

Ryan Maruyama [00:26:39]:
And you can take what they learned over the decade that they've been working on it, or however long, and then just solve your immediate problems today.

Hannah Maruyama [00:26:50]:
I think that that's honestly a pretty good summary of what parents can start doing with their kids now to help equip them going into the future and especially to make them more competitive in the market they're going to be going into. And a lot of them are very much under equipped right now from either from high school or from college. It's just completely failed to teach them anything useful. And so parents at home, now it's accessible to you to help them use the tools they're going to need to know how to use to be successful in the marketplace. And they can build those skill sets at home. They can even build those skill sets while they're working part time, like you said, at some, you know, at some entry level job where they're able to automate these tasks, or if you have a small business and your child is able to come into your business and automate some small tasks, incredibly useful, great experience. Especially if they can articulate what they did, how they fixed it and then why it benefited the business. Now if Your child is 18, 19, 20 years old and they're trying to figure out the right career for them and then how they can start it, we do that.

Hannah Maruyama [00:27:41]:
It's a very intense four, four week service. One on one, a real person talks to your child, walks into exactly what they need from work, what careers are going to fit that down to where you are and then helps them figure out exactly what they need to do in order to get that career, the best way for them to learn. And then a 12 month plan walking through exactly what to learn in what order, how to put it on a resume and then how they're going to go about getting that entry level job.

Ryan Maruyama [00:28:04]:
You can go to degreefree.com launch to apply and until next week, guys, hello.