The Changelog: Software Development, Open Source - In the beginning (of generative AI) (Interview)

Episode Date: February 2, 2024

This week on The Changelog we're talking with Joe Reis about data engineering and the beginning of generative AI. We discuss phone hacking via frequency, the role of a data engineer, this AI hype cycle we're in, build vs buy, the disconnect between data analysts and the business, ethical considerations around AI-generated content, and more. We also discuss the tension between AI and traditional engineering, as well as the inevitability of AI integration into pretty much everything.

Welcome back. This is the changelog. I'm Adam Stachowiak. And this week on the changelog, we're talking to recovering data scientist Joe Rice about data engineering and the beginning of generative AI. We start off with the origins of phone hacking using frequencies, the role of a data engineer, the AI hype cycle we're in, build versus buy around AI tech, the disconnect between data analysts and the business, ethical considerations around AI generated content. We also discuss the tension between AI and traditional engineering, as well as the inevitability of AI integration into pretty much everything. A super massive thank you to our

friends, partners, and the home of changelog.com, fly.io. Launch your app near your users. Fly transforms containers into micro VMs that run on their hardware in 30 plus regions on six continents. And as you know, Antarctica is coming soon. Wink, wink. Launch now for free at fly.io. dot IO.

Hey friends, this episode is brought to you by our friends at Sanadia. Sanadia is helping teams take Nats to the next level via a global, multi-cloud, multi-geo, and extensible service fully managed by Cinedia. They take care of all the infrastructure, management, monitoring, and maintenance for you, so you can focus on building exceptional distributed applications.

And I'm here with VP of Product and Engineering, Byron Ruth, and David Gee, Director of Product Strategy. So when you think about connectivity being the first thing to consider, someone pushed back on this and say, we'll think about it later. What competes with a mindshare of connectivity? Just like an HTTP developer, you actually just download and run the NAT server. Whereas an HTTP developer, if you're building an HTTP set of endpoints, you typically have to implement or use an HTTP library. And then whether it's a Go standard library, Python, whatever it is, and you're actually implementing endpoints that register into the HTTP server.

And then now you have to go deploy this HTTP server and ensure that it's performant. So it's a slightly different model, but you download the NAT server. It's a standalone binary. It runs on the majority of platforms. And then you have a handful of client SDKs across all the major languages. You download that, and we even have a higher level API that is akin to what HTTP developers have of defining a handler, for example. We just call it our services API. And you basically have a few boilerplate things

that you register your handler in the NATS context. And out of the box, it actually supports sort of a general request reply setup. And then you get all of these other benefits out of the gate. But the experience and the onboarding is arguably just as simple as any other HTTP onboarding, with the exception that you're technically deploying a client application that implements these NAT services in addition to the NAT server. But that's where the Saniti Cloud, it's already managed instance, and we even have the demo server for you to just try it out. It's a public endpoint that you can literally connect to. So you can still build a simple client application, use the demo server as the endpoint, and then you can

play with that and use that as sort of the server deployment. Well, if we talk about it just from the central view of applications, forget networking, all that kind of packet-based stuff, you were calling them HTTP developers, which kind of stalk instead of API devs. I mean, what do people do? They either glue it together at a primitive level, so the primitive being HTTP. They move up the stack in their mind's eye and they go, oh, we're going to do some gRPC, which is kind of still point-to-point.

So it's a lot of point-to-point stuff versus broker-assisted connectivity, which is way simpler. You connect to an endpoint, you get told about other endpoints. It's like connecting to a hive mind. What we're trying to do is move people away from coordinated point-to-point connectivity to easy connect to anything securely and connect to your other stuff securely instead of having to coordinate the whole you know rat's nest of where to connect to them then you've got to negotiate well what do we do then now we've

got to get the schema information and can we even connect to this thing and does it even work and you know what version is it and all this stuff and what we're trying to do is transform that and flip that to unify it to make it much simpler so i think we're trying to go from a rat's nest of point-to-point connectivity in the application space to making everything on net. And it's like connecting to a hive mind. And what we're kind of asking people to do is think about applications the same way you would video conferencing. So if me and Byron are going to have a chat, we might do a huddle on Slack or jump on a Zoom or something. But if we want a colleague to join, we ask them to join the same course. We can have a point-to-point conversation

by the same medium, or we can have a party line by the same medium. So, you know, it's request, reply, or pubs up, but it's on the same platform. We don't care about what Zoom server we connect to. We connect to the service and we coordinate our communications

over the fabric. There you go. Yesterday's tech is not cutting it. NAT's powered by the global multi-cloud, multi-geo, and extensible service, fully managed by Synedia, is the way of the future. Learn more at synedia.com slash changelog. That's S-Y-N-A-D-I-A dot com slash changelog. All right, we are here with best-selling author, data engineer, and architect,

recovering data scientist. I'm just reading your LinkedIn here, Joe. That's fine. We're here with Joe Reese. Hey, welcome. Hey, what's up? How you guys doing? Good, good, good. Awesome. Very well. I like recovering data scientist. You have that trademarked in your profile. I'll bite. What does that mean? Recovering data scientist?

I mean, I made a lot of bad choices in my life. Things I don't, things I, you know, not proud of. Data science, maybe one of them at one point. So, you know, I went to recovery and, you know, I'm very, very proud of where I am now. I'm just kidding. So, yeah, I think it kind of speaks. You're really selling it there. I was like, I was digging it. Yeah, I was going to of where I am now. I'm just kidding. So, you know, I think it kind of speaks. You're really selling it there. I was liking it. I was digging it. Yeah, I was going to turn this into a TED Talk.

I was going to say, you might try to cry. You can start crying. Yeah, you too can achieve your dreams in life. So, no, I mean, it was, it's, I think it was sort of a tongue-in-cheek joke back in the day. Data science, you know, was the sexiest job of the 21st century, I think. And me and other people had been in the data for quite a while, actually. We were hired as data scientists and just quickly realized that in a lot of cases, the foundations weren't set up. You couldn't really get data to do any science upon, for example.

It's really difficult to do your job. So I think we started out, I think we were calling calling ourselves me and my friends at reform data scientists and recovering sounded a bit more fun so uh just stuck with it yeah that's the story but you're finding more of them these days you know uh they're starting to come out of the woodwork people who are a bit jaded on data science which can maybe dig into if we want to yeah so i feel like the name of people who work with data in such ways just kind of morphs and now maybe that was like a phase when like data scientist became a proper noun and became a job title that you could make good bank with for a while maybe you still can but you probably still can and now there's like data

engineer and now there's like data engineer, and now there's like, and then there's AI engineers coming, and there's prompt engineers, prompt engineering. It sounds like data engineering is a thing that you ascribe to, since you have an O'Reilly book, Fundamentals of Data Engineering. Is that a title now? Is that a role? Or is it a verb, something you do? How do you see it? More of an adverb. I'm just kidding. So no, I mean, it is a title. But as you point out, it's interesting. A lot of titles have morphed.

If you look back to what happened with data science as an example, right? It was like it was one thing. And I think back in the day, it was kind of like you were jokingly called it the other day on a podcast, sort of the crossfitter of data. Like you're not really good at anything in particular but more of a generalist maybe good at like stats you can type code you might be able to type in general um i don't know i mean something you just do the uh put that in your anyway yeah 10 words per minute. Right.

Yeah. So, I mean, but that ended up being kind of an agglomeration of every title out there in data at the time, right? So, it was like, if you were an analyst, suddenly you're a data scientist. You know, if you're doing machine learning, which is more how I got into it, then you'd be doing, you'd be a data scientist and so forth. But I see the same thing happening with data engineering, where it started out, I think, as a, you know, sort of a big data engineer, I think, was maybe how that started back in the day when big data was still a thing.

And that was another kind of a bygone era, right? 2009, 2010. Remember data mining? Is data mining something still people talk about? That just gets folded into data engineering or data science now, depending what you're doing, right? So it's all data. I mean, it's just, but you got to look at

how job titles start, right? It's kind of like, you know, what you're doing right so it's all data i mean it's just but you got to look at how job titles start right it's kind of like you know if you're a company hiring people you want to get the sexiest catch-all title you can if you get too specific like you know oracle data miner or something like that you'll probably get a more specific candidate but people want the sexy titles and so i think that's kind of how data engineering came about but i you know as i dug deeper into it me and my co-author matt housley what we realized was uh i was writing the book and before like if you were describing from first principles it actually is a thing like it's a real thing so it's not just a made-up title but we can get into what that real thing is or at least how we imagine it but yeah it's so it's but title wise though um you know even as i put on the last

chapter of the book i feel like titles are just one of these things that are constantly evolving, you know, in our field where you're talking software or data or whatever. So it's just the titles today probably won't be the title tomorrow. Everything's iterative, you know, it's always a, it begins somewhere and goes somewhere else.

That's for sure. And sometimes it's posturing, right? Like I'm not a developer. I'm not a programmer. I'm a developer. I'm not a developer. I'm an engineer.

I'm not an engineer. I'm an architect. Yeah. not a developer. I'm an engineer. I'm not an engineer. I'm an architect. Yeah, and you almost want to get hired for the title you want to. That's why even with the specificity of the title, like Oracle Data Miner, for example, you might be like, yeah, I might want to work for Oracle, but do I want to be an Oracle Data Miner? I don't know.

Maybe, maybe not. You kind of get attracted to a title potentially. It could. It could. And then that becomes your identity. Yeah, exactly. Right? Because that impacts your ability to give talks

and stuff like that or show up on podcasts. Like, what do you do? I'm an Oracle data miner. No, what do you do? That's my title. Okay, that's me. I do that. We struggled with this, Joe,

back when we were naming the Practical AI podcast. We wanted to start a podcast in this space and Daniel Whitenack and Chris Benson wanted to do it with us. And Dan is Data Dan. I mean, he's a data scientist, or he was. Now he's a founder, and I don't know what he calls himself now. But we didn't know what to call the thing. Do we call it, do we focus around data? Do we focus around?

And we knew that AI was a burgeoning term, even though we didn't think it necessarily applied to the things that was burgeoning around. But that seemed to be the smartest noun to pick. And then of course, we wanted a show that was not just about the hype and the theory and all this stuff, but a practical show. And so that's how the name came out. But it wasn't easy for us to pick that because of this moving target because evolving and i feel like we picked a name that was well prepared for this current hype cycle but also probably didn't exactly apply to some of the stuff that's been discussed on the show yeah like when did the podcast come out um 2018 27 2018 oh you're way ahead of the curve that i mean we were way like five years we've been doing it for like five years i think yeah ai was i think it was just starting to become a term so back

i remember when i had a um i was working at a machine learning startup back in 2012 and we actually forbade ourselves from using the term ai because it had just come off a few like other ai crashes you know and so i think back then it was predictive analytics was the cool thing to call it. And then machine learning was sort of becoming the thing. Andrew Ng's course had come out maybe a year or two before that. And that, you know, it sort of blew up and then machine learning was becoming cool.

But then AI, I think around the time you, that was kind of actually a pretty bold move because it made, you know, practical ML probably would have been the more zeitgeisty. For sure. I think machine learning was hot right then. Yeah, sure was. We didn't know exactly what to call it.

I think we settled on that. And then we did the comma-separated list of keywords behind it. So it was called PracticalAI, colon, DS, ML, AI, data science, comma, machine learning, something like that. So that people could find it when they search for machine learning. Because you have to do those gross things in order for people to find stuff. You do, you certainly do.

For the record, I was against AI in the title. I was all for data. I was like, this has to have data in it. I can recall the arguments. It's a data show, it's data engineering. I didn't see where the puck was going. And thankfully, Dan and Chris were way wiser than I was because they saw where the puck was going and thankfully Dan and Chris were way wiser than I was because they saw where the puck was going and they're like no we want to talk

about practical artificial intelligence we want to talk about how it actually applies to developers today even though it's not quite here yet where is it going to really land and how can we actually use it and that was such a wise move on their part and that's honestly that's part of the wisdom of the crowd so to speak when it comes to

you know Jared and I run the show around here and Dan and Chris are not owners necessarily, but they were partners in the long-term endeavor. And we could have certainly forced our hand, in quotes, which we would never do. We don't do that. And push for, no, it's got to be data. This is not right. And thankfully, we were like, no, that makes sense. At least, I don't know what your position was, Jared, but I was like against AI.

I thought like, that's not coming. That's not here. What's going on here, you know? Okay, so you didn't like it because you were like, this is like fantasy land. Like, that's not it. Well, I just thought data was there. The machine learning aspect was there. And I just didn't see that the true umbrella term,

and this is where they saw it, was like the true umbrella term is all this is artificial intelligence. Even though what we think of artificial intelligence is going to be what it is is what these things are realistically so the umbrella term of ai really made more sense and they said we're not talking about just data we're talking about machine learning data science engineering it's all these things is really encompassed in this umbrella term of artificial intelligence and honestly that was a really wise you know suggestion back in those days because It's all these things is really encompassed in this umbrella term of artificial intelligence.

And honestly, that was a really wise suggestion back in those days, because obviously we're now where the puck went to. We skated to that puck with a name and we met it. Right. Which, interestingly, at least it's interesting to me, after November of last year, which was ChatGPT's release. And here we are. Well, not last year, I guess it was two years ago now, as we are in 2024. But 12, 13, 14 months ago, the audience of that podcast, thanks to nothing but the hype cycle, because the podcast was already good, still good. It was doing the exact same thing that they didn't change their content in order to match. But by picking that name, the audience of that particular podcast doubled. The audience doubled.

The rest of our shows have continued to do their slow and steady. It just went. And so there you are. There's prescience, right? Like you just picked the right word and happened to write. I was like, we're riding the wave now, guys. We're riding the wave. It really is a wave.

I mean, so I traveled the world last year giving talks. And inevitably AI was thinking i mean i've been in the ml field for a while right and data and um sure ai was sort of coming up as a term uh in popularity especially around you know kind of the but it always been confusing because you're like was that really ai like in the ai ai sense or is it kind of just like you're throwing this term on because it's kind of cool right so i still i don't know i think it's funny because like a curmudgeon friends of mine and myself who is also a curmudgeon and i guess we just reluctantly accepted that uh it's now ai um you know because that's the term everyone uses it's like uh you know one could argue it's probably not, but it doesn't matter. It is what it is, right?

So like when I traveled the world giving talks, it was interesting, like because increasingly throughout the year, AI was at every conference more and more. Beginning of last year, 2023, it was starting to become a thing. And then you get to the middle of the year, right? You go to Snowflake Summit, go to Databricks,

and it's like ai is the conference it's everywhere everywhere and i get to dubai um what was that october or something like that uh is that a conference called jitex which is like 170 000 people have this conference by the way it's massive yeah that's huge that is a lot of people you realize okay so like we live in the west right western society and what you realize when you look at the name tags and the locations of where these people, the attendees are from, it's like China, Iran, Pakistan, Africa. Like you realize the world is, there's a totally like almost parallel universe happening. Right.

And it's big. It's bigger than what we have here. Bigger than what we have here. Like a long shot. But every, every company was, had some AI story there at that conference. Every, everybody was an AI company. It didn't matter whether you were selling moths or something. You probably have an AI angle to it.

And I was at CES a couple weeks ago. Same thing. AI was the story of the conference. Everything had to have an AI pitch to it. And the other crazy thing was just the sheer amount of Chinese car companies. If you think American car companies are doing cool stuff like Tesla, I don't think it has much on what China is doing right now in terms of either self-driving or electric.

There's just a lot of companies that I've never heard of. Here, you know all the companies. There was actually a mention that I believe in Snacks recently, Jared, where they were saying that the, and I'm going to try and find the name, but there was a Chinese car manufacturer that was actually beating Tesla. In sales or in tech? Yeah, in sales.

BYD, I think it was. That would be them, I think. That's actually the one that, who is it, Warren Buffett invested in them many, many years ago. But you realize, I mean, I saw the technology, and I'm like, holy crap, it's just a tale of two worlds at this point. But, you know everything's ai these conferences um it is what it is yeah the cat's out of the bag you can't put it back in oh no i even uh you know when i was in dubai they have

a whole ministry of ai there they have a minister of ai you have to call him his excellency wow yeah he must be really good at it. Yeah. I hope he is. I just feel like anybody who gets called Your Excellency has to be pretty good at what they do, you know? Hopefully. I should have put that as my name for the show. You should add that to your LinkedIn list of accolades. Yeah. All right. So, Joe, we have your take on AI.

I think, you know, the sentiment resonates. If we were to call you His Excellency, like what would be your particular domain? Like what's your view of the world? I know data modeling is involved, but what's exciting to you? What's interesting? Like what are you into right now? I think the boring stuff is exciting to me, if you know what I mean. Yeah. Like labeling, data labeling. Data labeling is just the hottest thing in the world. You just love it. You just love to do that. I just love to label data all day. Just dream about it at night. What do you mean by boring though? Like that's what I think of as boring. Well, like I think the boring stuff is what excites me, right? I think like

right now, especially in software and in data for practitioners, there's no shortage of great tooling at this point. I don't think you could say any of the software data that we have a lack of great tooling. I mean, especially compared to 10 years ago, if you're a, you know, a developer, it's like, what are you going to go back to the means stack? You want to go do that again? Or, uh, you know, you know what I'm saying? So it's like, things have improved quite a bit. Things always improve, but it's the same as happening in data.

Our tools are wonderful, but what I feel like is everything, you know, tools are abstracting a lot of workflows. And increasingly, this allows a couple of things. One, it allows engineers of all disciplines really to, I think, do more, quote, enterprise-y things, you know, work with the business a bit more closely. So that's one aspect of it, which we would talk about. The other part of it is I feel like there's a massive gap

in terms of education and skills versus the potential of the tools and the capabilities that they promise. So, you know, when I say boring, I think that, you know, what I'm excited about is one helping upskill, you know, at least a lot of data practitioners, you know, so they can better leverage the tools that they have. And two, I'm excited for what comes next, because I feel like a lot of, at least in data proper, a lot of the classical, at least from a technology standpoint, things like analytics, which companies still struggle with for some reason. I mean, that should be a solved problem from a technology angle, right? But why is it we keep struggling with it from an implementation standpoint, right? So there's some things I'm pretty excited about. And again, these are not sexy topics. The sexy thing would be for me to say yeah i'm like totally stoked on uh ai and building

skynet and all this stuff and like why are you even bothering with um you know talking about these other topics but to me i find them interesting uh part of it is i think that you know this current ai uh cycle we're in it for at least for corporate implementations i feel like if we can't get the basics right and the boring stuff right, things like data management quality, all these things that are going to be powering AI, I feel like

a lot of this hype cycle is actually going to come crashing back to reality quite soon. Yeah, so kind of a garbage in, garbage out kind of thing. Literally, yeah. What about the skill set and the tooling of a data engineer what what are some of the skills and tools that in quotes a data engineer or one who you're trying to upskill is trying to get to yeah i mean as we describe it in our book

um we'll take a step back and sort of describe what we wrote the philosophy that we took so we wanted to uh the types of books that we saw, me and my co-author, that we really didn't like about data engineering were like these data engineering on technology X, Y, or Z, or programming language, say Python or, you know, on AWS. Nothing wrong with that, obviously. I think it's great. But the thing is, these are very ephemeral types of books and lessons. You know how fast things change.

Yeah, they're stale pretty quickly. If they're too niched, it's just very stale. Yeah. Yeah, so we want to take a step back and really understand from a technology agnostic level, what would be the immutables of data engineering if you were to describe it from first principles? And so really, that's what we try to describe, right? So it's things like, you know, getting data from source systems. So the systems that may be an application, you know, developer or engineer, however they

want to call themselves, right? Right. You know, systems that are more upstream from quote data, helping you understand what those systems are, and then obviously, ingestion mechanisms and methods to get that data, you know, storing it, processing it, transforming it, you know, querying patterns and so forth, and finally serving it for downstream use cases like analytics, machine learning, and so forth. So really, you know, I would say that framework is really the framework that we feel

a data engineer should think in. Obviously, this is not a linear progression. Things can loop back on themselves. I think this really was sort of the agnostic framework that we'd, I guess, come up with. And then, of course, undercurrents, right? So things like security, you can't just pigeonhole that into any one section of that uh data engineering life cycle as we call it security you can say oh security only belongs in storing

data like everything else is don't need to worry about it there because that would be like quite reckless so um data management orchestration everything else architecture so that's that's sort of the framework that we developed for data engineers to really think about and reconcile their job. Because up to then, there really wasn't a framework to think about data engineering as a profession. It was just, I think that the advice was, oh, just go learn Spark. And it's like, that's not really good advice.

It's like telling a surgeon to go learn Scalpel or EKG. It's like, that's... Right. It's one specific tool.G. It's like that's... Right. It's one specific tool. Yeah. It's not advice. So we really want to take a step back and I think just provide, I think, a mental framework.

But in a way that if you kind of, you know, as well as you can predict how things go in technology, you know, aren't going to change in five to 10 years, right? So, and I still see this as true. Like there's not a situation where, you know, the various types of storing,

you know, data for machine learning and analytics use cases. I don't know whether that doesn't really change. The details may change. You know, vector databases are more popular, but that was, they've been popular. They've been kind of on the rise for a bit before the AI boom.

So the whole point is just being able to rationalize how all these things fit together. So that's a framework that we want people to think of. Now, tools and technologies, I would say what we try and tell people is, you know, the early part of the book is full of best practices. Like thinking like a data engineer, thinking about architecture, right? Something that's not discussed at all. Typically, you have to go figure out on your own and then how to choose technologies, right? So build versus buy.

Oh, yeah. You know, open source, which is paid, all this stuff, right? So it's just it's a good rubric for a good practical way of thinking about it. Now you've got my attention there, Joe. I like those topics for sure. Let's talk about them. When do I build and when do I buy? Just give me the answer.

It's a simple, straightforward decision, isn't it? So if you use my affiliate code for... always buy just go buy for sure um um yeah i i am actually a fan of so i always use the the comparison of uh i like to ask you know how often do you go and make your own car tires from scratch haven't yet yeah i mean i haven't either there's people that probably do it you know yeah tempted truck tires are pretty expensive yeah they do it you know yeah tempted truck tires are pretty expensive yeah they are but you know if you want to locally grow your own car tires i mean go for it um you know but the shops that do this and people that can install them for you i mean i i can install car tires just fine but do i do that every day no because i got other things to

do and can you you can install a car tire yeah i saw this video uh speaking of ai and automation really it was more like ai that leads to automation i think it was volvo and it was showing off their plant and how efficient it is at creating the tire and putting on the rim in like minutes whoa and how quickly it can go from raw material to tire on rim and the automation process and i was like yeah and so i guess juxtaposing that to what you're saying i'm never going to create my own tire when i after watching that kind of video i'm gonna dig it up and throw in the show notes it was pretty it's pretty awesome but you would want to make a tire and i guess you would want to build when it's to a competitive advantage for you and that's what i mean like yeah if i was maxis or good year or if i was like a tire brand

i would for sure obviously be building my own tires but as an individual who has one use uh well actually technically four uses right because one per or uh four tires per vehicle unless i'm driving like something different it's got six economies of scale yeah or a motorcycle do you have like the big truck with like the two tires on the back end not dually no yeah those are those are fun yeah maybe in nebraska they have them over there in the farmland that's what they say there's lots of duallys there they do oh there's lots of duallys around here yeah yeah but it's the equivalent i think it think that's how I view build versus buy for technology where, and this is where I'm excited because I feel like, you know, again, as tools become more

abstract, there's a lot of options you can, you know, hopefully pay for or use open source, whatever, right? But the fact is, it's already been done for you. All the heavy lifting. Yeah. Thinking through this rubric of avoiding as much undifferentiated heavy lifting as possible and really focusing your attention on things that move the needle for your business and that's the comment earlier about working with the business like helping the business solve problems i feel like this is i think really awesome because there's a big disconnect right we see this in engineering all the time like you know the engineers aren't serving the business the data team isn't serving the business and they're off nerding out doing whatever they do this is a common complaint complaint that I hear from business stakeholders is,

you know, engineering team doesn't understand what we do, but then the engineers always complain, Oh, we don't understand what the business does. Why do they get to keep bugging us all these stupid questions and requests to help them. Right.

It's like, just go. Everybody has their perspective, right? Yeah. I was talking with a friend of mine this morning on a, on a podcast about, uh,

you know, just why is it that, cause she wrote a book on becoming a data analyst. And her question was, why is it that, you know, data analysts seem to have a disdain for the business people? That's an interesting perspective. Business people feel the same way about you probably too, if you have that mindset. So, you know, so that's what I'm excited about. And that's how i view it so what's up friends this episode is brought to you by one of my good friends one of my best friends actually one of our good friends tailscale and if you've heard me on a podcast you've heard

me mention tailscale several times in unsponsored ways because i just love tailscale and i reached out to them and said hey we're talking to a lot of developers. I love Tailscale. I'd love to have you guys sponsor us. And they reciprocated. So what is Tailscale? Well, Tailscale is a programmable networking software that is private and secure by default, which makes it the easiest way to connect devices and services to each other wherever they are. You get secure remote access to production, databases, servers, Kubernetes, VPCs, on-prem

stuff, anything. It's fast, like really, really fast. And you can install it anywhere. Windows, Linux, BSD, Mac OS, iOS, Android. It is an easy to deploy, zero config, no fuss VPN. It is zero trust network access that every organization can use. So what can you do with Tailscale? You can build simple networks across complex infrastructure.

They have ACL policies to securely control access to devices and services with their next gen network access controls. You can replace legacy VPN infrastructure in just a few minutes. Thank you. and more. You can save time with a trusted and proven networking solution that just works. You can transform networking security with a modernized set of solutions built on revolutionary protocols designed for today's mobile and multi-cloud era. You can securely connect to anything, no matter what operating system, hardware type, or configuration is in place, such as your GitHub runner to a database on-prem. You can authenticate without authentication using TailScale's app connectors. You can send files securely to any node

on your TailNet using TailDrop. You can code from your iPad even with TailScale and VS Code Server. There's just so much you can do with TailScale. Like the limits are literally limitless. And that's why I love TailScale. I've got 28 machines personally that I manage on my

tail net. And if you've heard me talk on the podcast, you've heard me mention Tailscale at least once or twice, but today you can try Tailscale for free up to a hundred devices like me and three users for free at tailscale.com. No credit card required. Again, it's free up to a hundred devices and three users all for free at Tailscale.com. No credit card required. Have fun. I'm more curious about this whole disdain. Oh, you are? And what the root cause is for that, really.

I mean, I think that humans not getting along. Well, I think to your point, it's kind of like tangential to the build versus buy, which is focus on the product. The product is the business, right? The product is a result of the business needing to succeed. And so if the analysts are upset with the business folks and vice versa, then it's because there's a disconnect on what the focus of the business is and what the individual roles contribute to the focus of the business, the product. So if you're not focused and you're not centered on what the goal mission is, that truly is a business issue because it's the business executives that should drive the mission of the business and what the focus of every individual and team should be focused on. So the disdain is really a lack of focus, right? I agree. I mean, because they're there to solve

a problem. And if they're solving a a problem then it should result in business benefit i.e better product faster product you know a more quantified product better data for marketing better data for sales so they can focus their efforts more better data for engineering so that they're building the right thing at the right time with the right purpose you know all those things so that's kind of how i look at it yeah i agree and then how often does it that you hear from engineers like, gosh, our CEO just keeps coming to us every damn second with some new idea, you know, or, you know, so, so lack of focus. It's it.

Yeah. I heard a crazy story. One of my relatives was an early engineer over at Nikola. Actually the car company that like car company. They roll down a hill. Didn't they roll one down a hill? It was really good at being put in. They roll down a hill. Didn't they roll one down a hill? It was really good at being put in neutral and rolling down a hill. Yeah.

That was the product. That's the future of cars. So cool. Fred Flintstone all over again, basically. Right. All we need is more hills. If we just had all hills,

we'd be fine. We should make more hills, especially in Nebraska. Well, we could definitely use some more, but he would tell me stories. Trevor Milton, I guess he's now off to jail, so it doesn't really matter. He'd come into the office every day.

It was just like a new idea. Like, I need this done. And people were like, we never heard of this before. Or he'd make an announcement like, hey, we're doing this. And people were like, what? We never heard of this initiative at all. Like building a car that works or something. I don't know.

But that's an extreme case, but it drove this relative absolutely insane. She's like, I can't stand working here. I'm like, just stick it out. Who knows what happens? Ended up cashing out for a lot of money. Thank God. But the amount of just stuff you would say,

I'm like, yeah, man, I don't know. This is like a different level of that. Because as engineers, we've all seen this, right? Somebody gets an idea for a feature, like, oh, this is going to change everything for a company. You got to make this a priority right now. And Adam, you're absolutely right. It's focus, right? That's 100,000% right.

But all too often, it's the hardest thing, isn't it? Yeah. When you're zigzagging, you're not focusing. No. Which is why we have a common refrain around these parts which we hopefully impart to our listeners which is the main thing is keeping the main thing the main thing that's right and you got to do that and if you fail to do that you find yourself focusing on all kinds of things which is not focus they're not the main thing and uh that's

a recipe for failure i've just seen it too many times. Yeah. I mean, how do you reel that in though? How do you, have you guys found it? Is there a magic bullet or silver bullet for a, there is a magic bullet.

Yeah. Slow down and check yourself. Now we're giving you more of our lines. Oh, we've got a phrase for everything. We got to, you can make a whole album out of this. We could.

Yeah. It's challenging. It's challenging, but it's easier for small orgs, I think. So here we are in a small org. And so it's easy for us to have a cohesive mission and vision and focus because there's just not so many people to keep on that same focus. And so we're kind of in our perch here telling people how to live their lives, but not in large, very complicated org structures in which these things are very difficult. But that's what podcasters are supposed to do. I mean, we're all supposed to pontificate.

Yeah, we're doing our job. You know, you do yours. If you need a coupon code, we can use Jared20 for 20% off BetterHelp. Your first month is free. Yeah, so. So yeah, it's easy for us because we don't have to actually do it.

We just talk about it, you know? It's what we do best. A little bit tongue-in-cheek there, but I think it goes back to leadership and leadership has to have the focus and they have to be able to like keep that distributed throughout everybody else. And sometimes you just got to remind, I mean, even in our small, uh, little org here, even myself, even inside myself, I have to remind myself, is this the main thing right now that you're doing? Is this actually driving value? Or are you just like doing your engineering thing and just doing some engineering because

you like engineering, for instance? And so, yeah, you have to be cognizant of it. And then in large teams, I don't know, maybe you've worked in larger teams than we have and can help us come up with some ideas. It's a good question. I think it's also up to the CEOs to really, there's an old trope where CEOs need to over communicate the mission, right? And I think that that's definitely true, but it's, but it's hard. I mean, I own an MBA,

people who have to run companies, it kind of sucks actually. So, but yeah, it's definitely the hard work, but it all trickles down because otherwise what happens is people just, if you're not intentional about the focus, then people will just figure out what they think the focus is. And then that sort of just goes on its own, right? We all seen this happen at companies. And so, but you know, engineers, everything kind of flows downhill to engineering, even an engineering led organizations, I would say that that's, you know, still is the case. So it's an interesting one. I think that, but it's something I'm excited about, again, it's just the boring stuff, like getting, I think, having the ability and privilege to have these discussions now, I think that just shows a sign of maturity

for the industry, the tech industry, whether you're talking data or software. And so that's pretty cool to see. And it's cool to have these discussions with people like yourself. I mean, you talk to a lot of people, and so do I, and this is sort of what I'm hearing out in the world.

There's a little framework I want to share, too, on this note of keeping things in the right direction that I learned probably a decade ago. And it's actually, I believe it comes from a restauranteur that was the salt shaker theory of leadership. And basically it boils down to constant gentle pressure. And so imagine that you are the leader of a restaurant and you like your table set a certain way. And the mission of the staff who is on the front side is supposed to keep that table set the right way. Salt shaker goes here. Pepper goes there. Napkins over here, etc.

Now, are you going to get upset and like scream at these people when they get it wrong? No. You're going to constantly apply gentle pressure. Hey, by the way, this is how it's done. This is how we do it. Hey, by the way, this is how it's done. This is how we do it. Every time there's a mishap, it's not freak out. Oh my gosh, you're fired. Oh, what are you doing? It's just constant gentle pressure on the mission, on the standards, on the requirements of excelling at the business we've decided to build. And they're participants and partners in that mission. And so you have to treat them like participants and partners. And that's where the constant and the gentle comes in.

But it is pressure. You can't let off. It has to be constant, but it has to be gentle. I think, that relates to the side of being respectful and with a version of love or empathy for the people that you've chosen to work with you. Now, in that constant gentle pressure framework, if it's a constant with a particular individual, maybe at that point the gentle pressure is no longer working and you have to apply a different thing, which is maybe there's no longer a reason for you to be here because you don't understand our mission. I've tried to be constant and gentle to apply the pressures of our business to you and it's not working. So therefore we part ways. But I think that's a lesson I learned a while back. And it's like every time I've been in a leadership position and I have to redirect the mission, it's been constant, gentle pressure, not you jerk or you bad person or any sort of like negative thing you can

ever think of never been like jared what's the last time i yelled at you like never right like i've never yelled at you right and you've never had to yell at me and vice versa like it's because of that it's that i respect who he is and if jared is out of alignment with what i think we're trying to mission towards i apply constant gentle pressure like here's where i think we're trying to go do you agree with that where are we? Where do we need to come back into sync to? Et cetera. And it's never been like, we're way off, man. And you're, this is not working. I mean, that only happens whenever you don't have constant dental pressure and you allow it to get so out of alignment that you've missed the opportunity for the constant dental pressure. Now you have to

apply brute force and it's not good for anything. mean it's just like a relationship or a marriage yeah yeah so it's uh but you're actually right a lot of things yeah applies to a lot of things there's some businesses i've seen where the uh you know the boss is um just berates the employees all the time and i know one of them well i won't mention them obviously but it says but you know it's basically just uh kind of a meat grinder of employees who just go in and out. And it's a very hard environment for people to work in. And the people that stick around learn how to play the politics game

and self-preserve. But it's tough. I mean, the Glassdoor ratings on this company and the reviews are just absolutely horrible. But you have to deal with the constant pressure, gentle pressure. That's the key. Because otherwise, like constant pressure without the gentle part is just, that's a vice that like literally collapses you like a black hole or something. So it's, yeah, but that's, that's just it. a lot these days too with um engineering teams and data teams uh especially with the you know kind of the great reset so to speak and uh you know kind of the jobs landscape really you know

i mean what was it the other day i saw like what ebay had posted like record profits or something and then also you know said oh yeah by the way we're letting go of like a bunch of people you know i don't know what the decision what the rationale behind that was but i think the optics of that right when i talked to engineering friends on social media they're like yeah it's like how do these two connect and all of a sudden do it that company is just leave even more of an environment of probably i'm guessing just distrust and feeling like you're um you know under not gentle pressure but constant pressure right so it's interesting interesting times interesting times indeed.

What are you seeing though in terms of like, you know, you talk to a lot of guests. I mean, what are some of the big themes that you've been coming across? With regards to

the life of a engineer or, yeah. Yeah. Well, I think it's a lot of trepidation right now. There's a lot of people

who are hoping that the tide has turned or is starting to turn back towards the thriving life of software engineers because it's been a rough 18 months and they're still to this day, we just had Justin Garrison on the show talking about Amazon silent sacking

and really this trend of large orgs not even laying off employees but just either making their lives miserable via mandatory return to office things or removing roles that no longer exist and no leadership and how you're going to find a new role, but you're not fired, but they just want you to quit, like basically kind of making it so that you quit is definitely a thing that we're hearing about. And a lot of people have either DM or, you know, raise their hand or liked when that kind of a post goes out because they had a similar experience that Justin put out there. And we talked to him about that. Those are the big trends. I think there's slight hopefulness that, you know, the layoffs are kind of waning and the market is recovering. And there's a hope that the interest rates hikes are over and perhaps we'll have some cuts next year. And things will kind of return to the way that they were.

I don't think they're ever going to be what they were during the peak of the bubble, maybe someday. But that probably wasn't healthy. And so this is the ramifications of that. So do you want that to be the case? I don't know. That's what I'm hearing. Interesting. Adam, anything different?

Anything to add to that? I'd say it's pretty similar. I mean, I think that we all reflect on, I would say, better economic days when it seemed more hopeful. It's interesting because I think, well, even here in Texas, we're having a,

I live in Texas, Austin, Texas, and whether you want to call it a border problem or not, there's a lot of people coming through the border that are coming through illegally. We have to have the right way to have people come into the country. And there's cities that I live near that are being just overthrown by large amounts of people. And then I think, okay, that's happening, right? I see the headlines. I'm in Texas. I see a version of that. And I love everybody. I want everybody to be happy and succeed. And there's a right way to do things. And I'm not talking politics necessarily here. But at the same time, I'm like, well, I just really want to run my business, right?

I just want to worry about the change log. I just want to worry about the things that I, like, I want to live in this little bubble that is just like my life and not really worry about these things out there. So I think from my personal perspective, I wish I could just like, just do what I want to do and the world's problems would just somehow be solved, right? The wars, the whatever's happening elsewhere. I mean, obviously these are all terrible things and I'm not trying to be, you know, personally focused on it, but like I would prefer a more blissful life where I didn't have to worry about all the problems of the world. And it seems like the speed of the world today just seems to just inundate everybody. And there was a time when it was just less than that, or I was just benign to it or

in living in bliss. I don't know. The times of call waiting. Call waiting. Yeah. We're talking about that at the beginning, right? Because that was pre-recording. We're talking about the days of call waiting. So I had a, it's an unfortunate call back to before our listener was with us, but yeah, like the, the days of call waiting when things were just kind of simpler and you didn't know everything about everything all the time.

And that can make your life feel frantic and distressing and, you know, like worrying about stuff that you have no control over because you found out about it. And how can you not worry about it? Because it's worrying, you know? And so that does, I think affect the psyche of all of us it's easy to just doom scroll on social media like i got off most social media um except for linkedin yeah i'm like off

most of everything good for you yeah just read a lot of books you know that's kind of fun and um i always joke i even got rid of most uh meetings and phone calls too like only, I always joke that the only time I'm actually talking to people is usually on podcasts. Those are my meetings. Nice. Because I don't know, you just want to give yourself the mental space to just think. And the thing that I've been thinking about a lot too is just sort of what's coming down the pipe with AI in terms of AI generated content and like the, how useful is it going

to be really? And like, how much do I going to be really? And like, how much do I want to consume with that versus just actually probably just going, buying print books and just reading those instead. Maybe I just become a Amish or something at the same time. I don't know. But you know,

like the thing that set me off the other day was on LinkedIn. They had these AI generated articles, right? These, so you can get the top voice in whatever topic you happen to collaborate on an article, like this AI generator article. As I read through these, I'm like, who are what are these for?

These read very poorly. It's at best very mediocre coverage of a topic that is very uninformative. Right. What is the point of this? And then I think there's a stat that, what was by 2025 i.e next year like 90 95 percent of uh you know the internet um content is expected to be ai generated like what is it all for at the end of the day are you what are we doing here yeah what do you read are you gonna go read that that's the hard part for me because i 100 agree with you and i get quite curmudgeonly about it

and i think i even even just yesterday nick nisi from our js hardy podcast posted some ai generated rap song into our channel about me and typescript or whatever and yeah it was cute he said finally a good use of ai and he's just joking we're having fun and i my response was like this is c plus rap like this rap is not good rap this is it's barely entertaining it's funny that you applied it to me in TypeScript. Like that's what made it funny was your creativity. It's C plus content. And that's what you're talking about. Very mediocre, right?

Like it passed the class, but it's not going to change anybody's life. It's not going to be outstanding or excellent or any of the things that we desire in great art. Then I asked myself, but am I just looking at today's ai generated content and i'm correct you guys can correct me if i'm wrong but i think we're agreeing that it's c plus content but it's going to be b plus in 25 it's going to be b minus no that's that's worse than b plus it's going to be it's going to be a minus you know it's going to get to be where like now it's indistinguishable for magic and then we're not going to be able to distinguish.

And that's what we're just seeing it in its infancy. And it's going to be so much better quickly. And I don't know the answer to that. Is it going to be so much better quickly? Maybe, Joe, you know more than we do. I don't know. I mean, as a content, you know, let me ask you this as podcasters. What are your thoughts on how this impacts people like us?

You know, there's a chance for me to talk to peers like this. I'd love your take on this. Like, what do you think all this this impacts people like us? There's a fair chance for me to talk to peers like this. I'd love your take on this. What do you think all this does to people like us? I long for human touch. Yeah. And I think if I live in a world where I feel like I can't discern between AI generated, whether it's good or not, and the human touch and the why behind things, then I think I would begin to question the world more so because like, what is its goal towards me? Am I a transaction to transact or am I human to, you know, love and endure?

You know, like that's the kind of thing I think about. And I think I'm hoping that as a society, we will all feel that way. And we would generally just push back on like, well, AI generated content is good and all, but I care about humans. And so we just sort of have a rise or a revolt against anything generated, you know, that is strictly non-humanistic focused. I'm a humanist. Yeah, that's it. I'm a human. So I'm, that's why I'm a humanist. I'm not even trying to be like negative about that. That's just seems so weird to even say that in 2024 or in my year that i turned 45 in my life you know that's just crazy that i have to declare that so

i'm a humanist i want to be touched and loved by other humans not non-humans now i have a dog so that's maybe the unique thing i do care about animals and stuff like that so that's where i'm a non-humanist where i i transcend when i'm like I care about my dog. But you get what I'm trying to say. Like I'm not cool with an artificial intelligence thus far. That being said, don't you talk to chat GBT as much as almost anybody in the world? Okay, fine, Jared. You got me. Busted. Does it know that you're a humanist?

Have you told it? Yeah, I think it knows. That's because I do find it helpful at the same time i'm not i suppose okay that that does make sense but i think it's because i long for what seems like okay maybe seems like is better placed in there seems like human interaction and because chad gbt i mean it doesn't respond to me like it loves me back or anything like that. So I'm not in the movie. Her here. I'm sorry to wake you.

It's okay. I just wanted to hear your voice and tell you how much I love you. I love you too. Okay. Well, that's all. Go back to sleep, sweetheart, okay? Okay. Okay, goodnight. Goodnight.

I am still interacting with it. I'm just being polite because I'm just a polite person. I can't help it. It's my nature. But it's because I long for humanistic touch. I suppose I anthropomorphize the thing and make it feel more human than it actually obviously is not. Because the desire in my humanistic ways

is for human touch. So you want your podcasts created by humans? Oh yeah, for sure. Even if you can't distinguish? That's where I get lost because I can't distinguish it. So how do I know? Well, I think if I can't distinguish it, then where I get lost because I can't distinguish it. So how do I know? Well, I think if I can't distinguish it, then I'm thinking like, what's the, what is the motivation?

What's the end game with the, you know, is it because you want to create more? So I think where the lines get blurred, there's this kind of side conversation we've had behind the scenes, which is should we have an artificial intelligence help us with our transcripts? Should we have artificial intelligence help us with a literal recreation of our podcast in my voice in a different language? And we've been talking about the behind the scenes. We've seen that recent politician in Davos happen. Like that was super cool. I could somehow literally listen to somebody that spoke in a whole different language for a speech that would often just discard because not because I don't care, but because I can't care because i don't speak the language i can't understand right and so now we have this just to paint the picture for those who are catching up or listening is like there was a thing in davos there was

somebody who i'm not sure of all the details he spoke a foreign language in comparison to english for me and typically i would just not pay attention but because somehow some way i'm not sure how it happened that was ran through some sort of artificial intelligence thing and it was his voice as a politician just as a normal person staking his claim and sharing his ideas but spoken his native language translated to english in his own native voice so it was as if i heard the same person speaking english and now i can understand and i was like okay that's cool I can understand this person more. That to me, I mean, I guess when we start talking about generated, what is the purpose of generation? I suppose, like if it's to pull the wool over my eyes and entrap me or ensnare me in some way in a scheme, then I'm not for it. But if it's for my betterment

and it's for me, then I think I'd probably be more in line with it. But I think over time, even with the betterment side of it, and you can't discern it, it's like now you question everything because you're just not even sure if anything is for you or against you. Gosh, how doom and gloom is that? Have you gotten any new Apple Vision goggles yet? Uh-uh. Me neither.

I'm curious how that's going to work though, right? Because degenerative AI with that, I think it's going to work though right because i mean you know because degenerative ai with that i think it's going to be a very interesting uh use case i mean um i'm curious what the apps are going to be coming out with that but it's just uh i totally agree it's like the uh the human touch thing i mean i i'm about to go you know right after we record and go into a conference um you know with actual people and i feel like in person is going to be like a thing again. Yes.

Absolutely. And then tomorrow I'm going to Austin, right? For a conference. And it's like, that's people hanging out. We tried the whole experiment of doing Zoom conferences back in the pandemic. And that was barely passable. Being in a room with, you know, being in a building of 170,000 people, that's actually too overwhelming, but you kind of get the idea. It's like, yeah, that's a lot.

I don't think I would like that. No, it's like an expo at that point. You just feel like you're in a major city walking around, but you do get to meet people, you know? And I think that that's going to be more and more of what people, um, but I kind of wonder though, is that going to be something that only, it's sort of like, uh, you know, processed food versus organic food. It's like the human experience is going to be something that only, uh,

you know, people who have means are able to do or is that you know and everyone else is stuck with just ai generated uh drivel and that's sort of the equivalent of eating twinkies and uh that's right yeah well maybe more hopeful than that i've likened it to handmade goods oh yeah yeah which are still the world of the wealthy because you can afford to pay more for this kitchen table than the one that's at ikea but there's nothing wrong with the one that ikea makes that's the manufactured ai version of the content but there's still going to be a marketplace i think for people who are like you know what this was made by this guy who lives 10 miles up the road and it took him three months and i paid a premium for it but I'm willing to because it was made by a person

and I know the story. And like that is a thing that very much exists today inside of furniture making and stuff. And I'm wondering if that will be how content online becomes as well. It's a good question. I was talking to some influencers yesterday, people who get paid to do a lot of sponsored posts. And I think the biggest question is, okay, so what, at what point are we kind of automated out of doing what we do and so it's it's an interesting one i don't know where this all goes i mean i'm quite excited about it in some ways i mean it reminds me of the uh i think we're all about the same age i'm guessing but we're all around when the uh you know the web information superhighway uh came into existence uh yes those are pretty

fun days i think there were a lot of good questions about where it was all going to go. And so I feel like we're back here again. Yeah, where will it go? I don't know. If you're listening, you may remember the early days of the internet where open networks like HTTP and SMTP led to an explosion of websites and online communities. Building a fan site and connecting over shared passions led so many of us to careers in software.

Back then, it seemed like anything was possible because the internet of the 90s was built to democratize information, not consolidate it with a handful of big tech companies. Read, write, own. Building the Next Era of the Internet is a new book from startup investor Chris Dixon that explores how network architecture plays out in our online lives and the decisions that took us from open networks governed by communities of developers to massive social networks run by internet giants. Read, Own is a playbook for reclaiming control and for reimagining applications so users can own, co-create, and even profit from the platforms they use every day.

From AI that compensates creators to protocols that reward open source contributions, this is our chance to build the internet we want, not the one we inherited. Order your copy of Read Write Own today or go to readwriteown.com to learn more. I think I've heard of folks generating movies or alternate versions, like fan art in a way. Yeah. Even that is kind of cool to me because a human directed the AI to do something. And the AI produced it, but in such a fast amount of time in comparison to the human who would never spend their time doing the fan art version of it yeah you know or a scene like redoing a scene in a different style with different music via an ai's help i mean that's where the line really gets blurred there's some cool stuff yeah yeah there's a lot of creativity creative opportunity directed by a

human i think that's where i'm will probably always draw my line is like is a human involved in the mission and reason for this and is there reason goodwill for me right goodwill for me i think is key human involved goodwill for me and i think if those two remain true human involved goodwill for me or the the pluralized me of team or community you know then i think i'm probably gonna be okay with what's generated because it's about an x of time saved and ultimately a good output that is for the people i'm trying to serve already yeah absolutely i'll give you a good story like so i was hanging out with colin kaepernick a couple weeks ago the old nfl player yeah um we were having coffee um for quite a while actually and name dropping here no no but he i'm gonna tell you in a second why this is cool yeah no i mean he's a cool dude but like uh

a friend but like we were hanging out and he's working on something really cool so he had a publishing company which still has one well i'll rewind so he was trying to get books published with like the the big publishers right and and because of some of the things in these books they'd be like well this doesn't really fit what we're trying to do. So, he started his own publishing company, mainly to give underrepresented voices a chance to get his own version of a superhero in a cartoon, right? And so this is the stuff he's working on right now is like trying to giving, you know, underrepresented people the voice, the chance to like, create their own superheroes and all this really cool stuff and, you know, tell their stories. And I think that's an awesome use of, you know, generative AI is just giving people the ability to create whatever they want. I mean, my son's in the other room right now just drawing. He's sick today. But you know, generative AI is just giving people the ability to create whatever they want. I mean,

my son's in the other room right now just drawing. He's sick today. But, you know, so he just draws, he's actually drawing my friend who's stopping by in about a half hour. I told him to make a really horrible cartoon of him. But, you know, people, I think people want to be creative. And that's why you watch movies because you vicariously want to be a part of that scene, right? But now you have the ability to create that scene. I think that's what we're talking about. Create, you know, kind of mashups of movies that you watch and all this cool stuff. You know, when I was talking to Colin, I was like, that's actually a really cool idea in the sense where you're giving people the ability to, you know, tell their own stories in the way that they want them told.

You don't have to go through these gatekeepers anymore, publishing companies and mass media. It's like you can do whatever you want i thought that's pretty cool yeah my son's face recently since you mentioned your son and the scenario he was playing the adventure mode i suppose of minecraft and he's new to minecraft and so he didn't know how to make i believe it's called a portal or something like that and he didn't know you had to like build it and how many would you know bricks were required and how to use a flint and whatnot to like make it this i think it's just called a portal i could be wrong i'm probably it's a portal yeah yeah and so he's like dad can you hunt with this he's like maybe we can ask chat gpt what to do

i'm like that's probably great because minecraft so prolific, it probably has a decent answer. So I'm like, of course. I dropped the question to ChatGPT, and it basically gives me a step-by-step instruction. And my wife is like, that's cheating. And I'm like, no, it's kind of like not cheating. You know, you're just kind of getting over a hump, a hurdle in the process of learning something new. And he's like, was that AI, Dad? Because, like, I showed him, and we did it, and it worked.

And he's like, was that AI? And I was like, yeah. And he goes, like his face was like so huge because AI assisted him to do something in his life. And for the audience, I made a big O face basically. Because his face is burned in my brain because I was like, here he is. He's eight, right? And this is his real first experience with artificial intelligence at eight that's so cool my gosh that's crazy to think about and he was like the omg face like i don't know how to describe it otherwise but i was like this is interesting you know like what will it do for

the future and that to me is really cool you know having uh in colin's case and what he's doing that's super cool too you You're enabling the disabled, basically you don't have an ability and you're giving or offering a way to get an ability. And that's curiosity. That's creativity. And that's all positive human behaviors. That's just going to create a better world in the future.

I'm for that kind of stuff. Yeah. That's the kind of content I think that's cool. Right? Like, you know, you contrast that fact having election coming up and it's gonna be deep fixed lore.

And like, that's the, the dystopian view of it but it's it is what it is and it's happening with or without our opinions you know so it's just like right yeah we're kind of just along for the ride i mean at the end of the day it's another tool and every tool can be used for good and creativity or bad and destruction you know You go back to the industrial revolution, factories, you look at the printing press, you look at guns, all these things, they're tools and they can be used for good and evil. And very powerful tools like these generative computer programs are going to be used for very good, good and very bad, bad. And so that's why it's both hopeful and dystopian, because people will use them for both. And like you said, Joe, there's nothing so that's why it's both hopeful and dystopian because people will use them for both. And like you said, Joe, there's nothing we're going to, we can't stop it. I mean, it's

like, it's the horses left the corral. I don't know that one. Yeah, there you go. Yeah. I think of the okay corral for some reason. Yeah. The horse is out of the stable. Hold on and hope that, you know, we use it for more good than we do for bad yeah it's interesting i mean even my kid though right he's just he likes to draw he doesn't want to use ai i already showed him everything and he's just like it's kind of boring he just does uh he continues doing his own stuff and uh yeah i mean i think to your point people want to uh i think get back to the kind of the human experience too i think i haven't noticed this as engineers too like i think it's really easy to to go off and use copilot and it's a great tool um i think it's going to invert the paradigm of how engineers think about writing code because you're going to um you got to kind of know what good

performant code looks like uh before you start using these things i was talking to somebody yesterday about it and they felt like um a lot half the time they had to go back and re-edit the code i think part of it is the prompts that you give it though too and so we've been using chat gpt using copilot whatever but um it's kind of like having a very convincing intern who's like this is absolutely the best code you're gonna get it's like right okay i don't believe you but that's like i like i like your gumption though man it's good right totally they're very confidently incorrect yeah i think that as long as humans are in the loop at this point everything's gonna be okay because we can say no that code's wrong or i ran it it didn't do what you said it was going to do but it's the

auto gpts which are very much just demos today i think and the prospect of that or adam's prospect of no human touch content right? Like the only thing the human said was, you know, auto generate a movie about this. And like the details of that had nothing to do with creativity. It was just like on autopilot. I don't know when we're going to get to there, but that's where I lose interest, I think. And that's where I think no one's there to say this code actually is completely incorrect and it's just going to execute in production right and shut off of the grid and then be like oh it messed up oh no we were joking the other day on my uh podcast um this podcast i do called the monday morning data chat on linkedin but we were

we're joking that actually the uh the world's not gonna end through skynet it's gonna end through a bot that goes around doing a drop table on every corporate system. Gosh, yes. Yeah, yeah. Old Bobby tables, right? Yeah. It just runs rampant. Every mainframe, everything is just gone. Just gone.

And that's really how it goes. And then I go back to the good old days of voicemail and call waiting, you know? I'm just like, no, no more data mining. No more big data. It's all little data. Actually, one of my friends who is a Bill Inman, he's the guy who created the data warehouse. He has a landline in his house. He doesn't have a cell phone. Really? Yeah. He's like, I don't want people calling me. So if you're at his house and it rings, he'll either pick it up. He'll just like pick it up and then slam it back down. Um, and that's the, uh, do

an answer button. Yeah. It's very loud. Yeah. That's pretty uh do an answer button yeah yeah it's very loud yeah that's pretty pretty cool but yeah he'll travel internationally without a cell phone wow i think it's bananas but that's bold these days yeah yeah i would feel so lost it provides so many different utilities besides a phone right yeah totally yeah like boarding passes yeah maps all the things. That's how you take a true adventure nowadays, is you leave the country without your phone. Oh, scary. I think I'm out of the country probably every couple weeks.

I mean, I couldn't imagine not having a phone. That would be weird. Why? But I should try it sometime, I guess. Maybe we'll take a walk in the wild side. But maybe not. No, it sounds like a bad idea that's actually really interesting to think about more so than ai is our addiction to our devices really yeah i actually think that my heartbeat changed when this topic came up because i think that i got you know excited and scared with the thought of what it might be like in the process of empathizing with this person

on this journey, on this international journey, sans phone. Like that to me is like, what? No, no. And screaming no several times really loud. I'm not going to kill you guys ears, but that's what I'm thinking in my brain. I'm like Michael Scott over here.

No. Bill went to China too a couple of months ago and left his phone. I mean, for a lot of reasons, you probably want to bring a burner phone there. But he went there

with no laptop, no phone. Oh, gosh. Yes. I was like, I hope I see you again, man. And he came back. But that's, yeah, that was interesting.

Do you guys do you have an apple watch used to mine broke and so i didn't get a new one yeah it's gen 4 though it's like super old gen 4 oh i had the same one yeah i had to upgrade my battery kept dying on it i can't find a reason to buy another one i just can't it still works like what am i getting really for it i don't know i don't know probably nothing actually it's bigger exactly yeah this yeah you get this one i mean i got this one the battery kept dying and i go running a lot and stuff but uh but have you ever okay so if you have the lte connection on your watch and a fun thing to do if you want to kind of like blend 2024 with 1990 is just leave your

cell phone at home leave your smartphone at home and just go out with your watch or just actually go out without a watch and go do something too but But it's like, it's sort of the best of both worlds. You can still take phone calls on your watch, but you don't have your phone with you. So you can't scroll anything. Yeah, right. It's very bizarre actually. Like I sort of live in this century or this year,

but I also live back in like 1985 or something. Yeah, that's funny that you say that because this is like a simplified version of it, but it's recent. I went to get my haircut this morning. And so I had to sit down for about five, I had an appointment, right? So I got there on time, but the person was taking a few minutes longer with the previous customer and I had to sit there. And so what do I do? Right. I'm in a state of boredom in a public environment where I don't know anybody. I'm not really trying to connect with anybody. So I, of course, go into myself.

I'm self-centered. Pull up my phone. Look at things that probably don't even matter. I didn't scroll social media, but I looked at email. I may have looked at Slack. I think I may have checked in with this thing that I have with Jared. We're trying to plan a trip and stuff like that. I may have checked in with things.

And I didn't interact with the environment whatsoever. I can't even remember the waiting room like was there magazines there did i get to touch taste and smell the environment around me rather than this phone and the answer is no because i had my phone with me and so as a result of this i suppose appendage now it's gonna be there like everybody in the world has a phone and you just said taking a trip without one is like what but i just pulled it out oh yeah no i mean i've been around the world a lot and i'm universal sign of like a person now or universal like it's just that somebody hunched over like the phone and no matter where you go anywhere in the world this is the universal like posture of a human that is sad and i'm i do that and that i'm sad about that, really. And we use it as a crutch to avoid, I suppose, humanity.

Oh, yeah. What kind of person am I? I'm like a humanist not being a humanist. What's wrong with me? Yeah, man. Come on, Adam. Get together. You're a phonist.

You're out there in the world with other humans, and you're talking to a robot on your phone, asking it step-by-step operations for zfs you know uh the shame i live in the shame well you're doing that too and you're engaging with your uh your device and it's just getting more and more signals to get you more and more uh using your device too it's the whole feedback loop it's like oh adam likes this when he's at this place so we'll just give him more of that next time he's getting his haircut you know i mean my barber i got my haircut yesterday my barber is telling me about somebody who just like it's constantly just on the phone while he's getting a haircut my barber charged him double he said because he's just like yeah you disrespected my time so man

sorry so you either talk to me or you pay double okay that's that's what he did he's pretty ruthless about that but i know i know other celebrities too that like they don't let you take a picture with them you know they won't let you uh have your phone around them it ruins their cameo rate you know they gotta have keep that cameo number up yeah yeah i remember uh true story so uh who is it uh maynard from uh that band tool he does jujitsu over um where my jiu-jitsu when he comes to Salt Lake because we share the same instructor. And it was interesting because

people were like, hey, can I get a photo with you? And he's like, nah, can't. He's like, but my name is James if you want to talk. Yeah, that's always an awkward moment. Yeah, they're like, why? Well, because you're not going to meet me. You're just going to take a picture with me and then go away. He's just done with moment. Yeah, they're like, why? Well, because you're not going to meet me. You're just going to take a picture with me and then go away.

He's just done with it. He's just had enough pictures in his life. But then he'll introduce himself and just have a conversation with you. He's like, I don't want your phone around me. I'll be happy to answer any question you have or give you advice or tell you about my life or whatever you think is interesting to talk about. But the photos is a no-go. Yeah. Are there a lot of celebs in salt lake or what's

going on over there you got all these these big i don't know man i just have random friends like if uh like i got a set list from uh maynard that's hanging on my wall there from the show we did here but it's like uh yeah yeah i mean i don't know i just have this weird thing where i just randomly meet people so it's like in the most random locations i don't i cannot explain it and don't take their picture you know pixar didn't happen joe locations i don't i cannot explain it and don't take their picture you know pixar didn't happen joe so i don't believe it you know he had his bodyguard there too so if you tried to do that i'm pretty sure his bodyguard would like do something to you or your phone so yeah but you know jujitsu so he's got to toil with you yeah true yes that'd be a

bad bad form uh well jujitsu is not karate, right? Jiu-jitsu is all grappling, right? So it's all the... Yeah, it's all grappling. It's all about, it's really to not be the aggressor. It's to be the offense, the defense to somebody's offense, right? It's the don't touch me, get away from me, and try to avoid a fight. Really just trying to trap them, to make them struggle and give up and go away. Yeah, I mean, it'd be really hard to be like,

to go on the offense with Jits because you'd have to like actively chase somebody and then try and get them into an arm bar. That'd be like really weird. Yeah. You'd have to not be Jits. You'd have to be a different art. Well, I assume that Joe would be doing defense if Maynard's bodyguards coming after

him. I mean, wouldn't he be? He's a pretty, he's a tough dude, man. I wouldn't want to mess with his bodyguard. I know his bodyguard. He's not, man. I wouldn't want to mess with his bodyguard. I know his bodyguard.

He's not, he's not one you want to mess with. The bodyguard should be that type of person that you don't want to mess with. Otherwise they're probably not very good. Yeah. Don't mess with them. It's the same bodyguard who's the bodyguard for like, uh, uh, Roger Waters and other musicians. So he's, he's around the block a bit, but like, but yeah, jujitsu is one of those things where this is developed because, uh, well, Brazilian jujitsu was developed, um, for like

the unruly streets of, uh, of Rio de Janeiro and places like that. And the guy who came up with it, or at least made it popular in Brazil, is Helio Gracie. He was really of the opinion that a smaller opponent should be able to out-leverage and out-maneuver a bigger opponent. And that's sort of the essence of Jiu-Jitsu, is really using leverage to a bigger opponent uh in a street fight um so that's kind of where it came from but it's uh it's pretty cool like i think everyone should learn it it's really nice to know so what i find interesting about the jiu-jitsu story is that it's not brazilian no it's japanese that's right it originates in japan but became famous as Brazilian Jiu-Jitsu. Yeah.

By Gracie, as you just said. That's so interesting, because when you research Jiu-Jitsu, you're like, well, you find the origin is Japan, but then its claim to fame is Gracie and in Brazil. And then obviously MMA is a version of that, but not really. MMA started in Brazil with Vale Tudo back in the day. They used to have these bare-knuckle fights where it's literally no holds barred. You can do whatever you want. That's some crazy people, man. I would never do that. It's nuts. Not at my age, no. Maybe younger. At any age. Gosh, forget that.

Let's zoom out for one second. There's one thing that I took a note of from something Reesey wrote. And you mentioned this, so I'm going to go back to it. It was your, you know, what to expect from data engineering in 2024. You know, some predictions, essentially. And you mentioned Copilot being a gateway to AI. And a lot of the incoming stuff, I kind of want to, not so much end,

but I think we're coming close to a version of an end. So I want to get your take on this aspect of AI-enabled workflows and the longstanding idea of engineers actually engineering. How's that going to play out? Will we always prefer writing code? Will AI overflow this thing and turn us into just generative people? I don't know like what's the play out there it's an interesting one because two tensions involved in the discussion that i had on like co-pilot is a gateway drug to ai for companies right because on one hand it means like

companies can legitimately say oh we're using ai to improve our workflows right i think that that is you can legitimately say that it does improve your workloads right that's cool on the other hand engineers want to engineer right so i suppose it's maybe in in uh in proper engineering what we do i sort of engineering but real engineers i'm just joking um you know mechanical engineers or civil engineers back when you have drafting tables right you're drafting stuff a hand and and then autocad comes out right now i can just do all the designs uh you know and have them all generated i think that's sort of a similar ish thing where but you're still designing but it just makes it a ton easier but the tendency i know is like a lot

of engineers like it's just sort of there's a certain like manly or bravado feel to like writing code in your mechanical keyboard that's super loud like your wife can hear you in the next room if you have one of those um but it's like it's loud it's cathartic and that's the experience i think that engineers want to have is like i'm writing the code i you know it's like the scene of hackers or matrix like stuff's going on your terminal and it's just you feel like you're doing something right right whereas i um yeah so i do that's the tension that's your industry you could do a lot of stuff with co-pilots and whatnot but like what's the tension between me feeling like

i'm doing real work i kind of make an equivalent thing to like a electric motorcycles now and the fact that harley had put on like the sound of a motorcycle to make it sound like right so like a harley uh motorcycle so that's sort of the tension because engineers i know it's like well that's not real software engineering not a real programmer if you're not clanking on a keyboard right what's the point so that's the tension but i think that it's it remains to be seen you know i don't think it's i think the inevitability is though these tools will integrate themselves so far into like the fabric of every corporation that it's sort of the inevitability that they're going to be there you know yeah i kind of see that as just being um a generational thing perhaps

and just being like one generation away from that being gone i mean i i like it to manual cars and automatic cars it's like you'll always have your enthusiast who's going to drive their manual but pretty much i mean i haven't seen i can name one hand, my friends who drive a manual and pretty much everybody else just drives an automatic. And it'll probably be like that with people that want to have the clicky keyboard and the hardcore engineering with their Vim and all the things like people like me, we're all going to die. And then the next round, like they didn't even know that life. They grew up like Adam's son at the age of eight, getting instructions from ChatGPT, and their very

first programming course was all code generated by GitHub Copilot. And so that for them was normal. So that just might just be a thing that passes with time. I think you're right. I think you're right. It's just a generational thing we're holding on to. Pull this mechanical keyboard from my cold cold dead hands kind of thing.

So Charlton Heston. Yeah, exactly. Exactly. Cause I, it resonates with me. I mean, not the mechanical keyboard, but all the other stuff does totally. Yeah. Yeah. It's good. Pull my Vim editor. You know, I use arch by the way, like all of these memes of, of the neck beard generation, that's probably just going to go by the wayside, you know? And for the audience out here, none of us have neck beards. None of us have beards, actually.

We're enlightened. That's right. We've shaved. We all just want to look younger is what it is, actually. Yeah, that's true. I would be grayer if I had more facial hair. So that could be the case. I'm still back on build versus buy no yeah we didn't even go there gosh

well adam took a whole different direction which is why there's two of us well i think we kind of answered in a way didn't we though it was good man yeah we uh we went a lot of different directions but it was it's a great great conversation with you guys yeah yeah it was a good ride i dug it joe yeah i'd love to have you on my show sometime. If the audience likes the podcast, happy to be back anytime. We can riff. Yep. Sounds good.

Well, it's what happens when you get three podcasters together. It ends up just being a podcast. It just ends up being a podcast. It ends up being a podcast. That's right. Strangely enough. That's a good point to end on i think right there awesome well thanks joe awesome having you here thanks guys yeah pleasure yep yeah thank you very much and as we close out real quick thank you to our listener darren smith for pointing us joe's

direction darren recently joined changelog plus++ and I was emailing with him, thanking him for his support and asked him, hey, what would you like to see coming up on the shows? And he said, Joe Reese. And I said, I'm on it. And that was not too long ago. So happy to fulfill this one. Sometimes it takes us years to do listener requests. This is like

weeks. So hope you enjoyed it, Darren. Appreciate you supporting us and listening. Sick. Thanks, Darren. ChangeLog++. plus it's better there you go yeah but no seriously i've been a listener of your podcast for a while so when i when i uh heard you guys were interested happy on the show i was like yep absolutely do this very cool so yeah big fan yeah you're welcome back anytime thank you very much yep be back what a fun conversation with joe so i guess the question is did we adequately cover build versus buy i know that subject in every facet of developer life is interesting and i suppose we covered it tangentially if not indirectly but i guess

the question like i said again is should should we have Joe back on for a particular deep dive on Build versus Buy? So if you have compelling feelings about that, you can let us know pretty much anywhere. X, Twitter, LinkedIn, Slack. That's free, by the way. Head to our free Slack community, changelog.com slash community. It's free to join and everyone is welcome. On Monday, we have a new album dropping and you're going to love it. I dare you not to dance. I dare you. I won't spill the beans here, but just wait for Monday.

Make sure you listen to changelog news on Monday or pay attention to the Slack community. Okay, so I want to give a massive thank you to our friends, partners in the home of ChangeLog.com, fly.io. Without their support, we would not be as good as we are because we would not be paid. They cover a lot of things for us. They love us. And if you want to love them, well, fly.io and tell me sent you. Also, our friends over at TypeSense love them as well. TypeSense.org. Use it in the open source or upgrade to cloud if you need those amazing cloud features. Jason and team are amazing we love them typesense.org and of course our good friends

at Sentry Sentry.io and use our code changelog to get $100 off the team plan Sentry.io okay those BMC beats keep on hitting hard

I love them I hope you love them cause we're going to keep playing them and that's kind of fun. That's it for this show. Thank you again for tuning in. Check out changelog and friends coming soon to you. We were traveling this week, so we're a little delayed on some of our shipping. So don't mind the schedule shifting, but it is coming and you're going to love it. So check that out changelog.com slash friends. That's it. Bye for now.