Phil Hawksworth 0:19 I’m very, very happy to be here and excited to get into this conversation. I’ll be honest, I’ve been a bit hesitant to embrace AI for you know, for a variety of reasons. So it’s an odd choice that I’d be involved in this conversation, but maybe I’m a bit like some of some of you, maybe some of you have been hesitant as well. Maybe you, some of you are running headlong into it, so there’s a good opportunity to talk with people who know, frankly, much more than I about that, and that’s what we’re going to do. So I’m really chuffed that I’m going to be able to have a conversation with Eve Porcello, who’s the founder, well, an engineer who’s been around the web for a very long time. I don’t want to to age you, Eve by that comment, but I’m just leaning on the fact that you’ve got lots of experience accomplish engineer in the React space and Graph QL and lots of lots of materials you’ve been building for so long. I’ll let you introduce yourself in a second. But, but yeah, Eve from from Moon highway and and other places, is is here joining us, as is Kira Corbett as well. So Kira is also working as a software engineer in all kinds of areas, lots of interesting stuff that I’m hoping we’re going to get a chance to dig into. But also works at Moon highway, also doing lots of developer education. I won’t preempt your own Introductions any more than that. So why don’t, why don’t we let you both give yourself a better introduction, a more thorough introduction? Eve, why don’t we go to you first?

Eve Porcello 1:48 Sure. So thank you so much for the intro. I am old. I’ve been working at Moon highway for about 13 years now, 1213, years, and we do all sorts of developer education, whether that’s in person training or tutorials or written materials, all sorts of things that take people from skeptics to reluctant users of various technologies, and that’s where we’re at now with AI, nice and Kira,

Kira Corbett 2:21 yeah, thanks for the introduction. I am a developer at developer educator at Moon highway, and I come from a background in like robotics and human computer interactions and more recently, transitioning over to AI and how we’re interacting with AI as humans.

Phil Hawksworth 2:36 Very cool. Very cool. Well, I did kind of tease a moment ago that I was a bit hesitant with AI, and I’m I’m not a complete Luddite. I’m not completely opposed. I’m embracing it more and more. But I think different people, different things, bring us to using AI in different ways at different times. So maybe a good place to start might just, you know, before we we dig into more of like, some of the practicalities and some of the lessons, and what have you, is just understanding, you know, where you both started your journey when it comes to, you know, exploring the use of AI as part of building the web. Because, you know, one of the things I’m particularly interested in about is, is building for the web. So could you maybe talk a little bit about how you kind of first started exploring AI with regards to, you know, building things on the web. Why don’t we go to? Let’s mix it up. Let’s go to, let’s go to Kira

Kira Corbett 3:26 first. Yeah, I would say I started. I mean, you could argue AI has started long ago, depending on your definitions and definitions people refer to, but I would say for myself, personally, I kind of started exploring with AI when machine learning models were coming out, and I would be that, I would be like, That person in high school or college, just like training things behind the scenes on these, like large computers, actually, but I didn’t really understand necessarily, like what the use case were or anything. I just like to dabble around, and didn’t keep up with it, and then kind of revisit it quite a bit, actually, once more of the larger GPT models started coming out, and I was like, Okay, this is actually powerful. We can take a lot of these, like pre trained models, we don’t all necessarily have to be trading everything all the time, and make some really cool SaaS products out of it, or even, like, really helpful tools to make you code faster. So it’s kind of where I started dabbling with it.

Phil Hawksworth 4:22 Nice, that’s great. And Eve, how about yourself?

Eve Porcello 4:26 Yeah, kind of the same journey. I started to play around with things a little earlier on, with like TensorFlow and things. And then, kind of thought, This isn’t really ready. I don’t need to spend too much more time on it. And then, like in the last year and a half or so with chatgpt, it feels like AI is the law. You have to start to use it a little bit, and I’ve started to become more bought into the idea that AI can really help us do our jobs more effectively. And there’s all sorts of really cool tools and ways. It kind of goes beyond the chat interfaces of that we’re all familiar with, and there’s a lot of cool applications for web developers, but also just as like a person living their life too. So

Phil Hawksworth 5:15 yeah, and I think people have, I think people are mostly back. There was a little bit of a glitch on the stream. And I don’t think that was brought about by anything we did. I think we were fine. I think we nailed that. I think people are people about but yeah, as we’ve kind of introducing our, kind of our route to kind of the first experiments with AI, you said something there Eve that that really get kind of in my, my, I Twitch a little bit, which was, AI is the law, and, yeah, that feel that feels like maybe that’s one of the reasons that I’ve been a bit hesitant, the fact that it feels like we’re rushing to to use it maybe, or rather, it’s just getting incredibly popular. And I have a feeling that maybe I’m carrying a little bit of baggage from, you know, a few years ago where, you know, I was working at an agency, and there was lots of conversation for every client that came through the door that, oh, yeah, we can solve your problem with AI. And what we meant by that when we said that was with a chat bot, and what we meant by that when we said that was a very narrow decision tree with a conversational kind of, kind of UI to it. And I don’t know. I think maybe that made me hesitant about AI, because I thought, well, actually, is this even AI at all? But it feels like now we’re past a watershed, and it really, we really are using ai, ai for real. So when, when did that happen? I mean, when exactly do you think that that started to take root, and it really was turning into using AI proper. Was it the advent of things like, I don’t know, open AI or some other particular service? When do you think it it happened? When would we cross that boundary?

Eve Porcello 6:53 It feels like that to me, right? Like the chatgpt moment was where everyone was like, Ooh, this can start to actually make my life easier. People started copying and pasting those long paragraphs that became everyone’s email. And then it was like, Maybe we should push back on that a little bit. But it was this moment where it was like, I can actually put in some pretty loose inputs even, and get some pretty competent outputs from this. So then that led to, I don’t know, more people adopting copilot. I feel like, when copilot came out, everyone was like, this is not going to be what we’re going to use, and this is just scraping all of GitHub. And I don’t want to allow this. And this is bad to this actually gets my work done quite a bit faster, and I’m out for a walk or a bike ride earlier than I thought I might be. So I started to see that people as it’s hard to, I don’t know, it’s hard to push back on the idea that it is actually helping in many cases. So

Phil Hawksworth 7:59 right, yeah, yeah, absolutely. And I don’t know the, I think the ways that we use it, not just you know what we use it for, but actually, you know, the processes we use, the techniques we use, and you know how, how we hold it, I think is, is key. And I have a feeling that that’s probably something that the two of you have done lots of, lots of work with, because, you know, you’re, you’re you’re building courses, you’re building courses, you’re building materials to help people kind of start their journey. So I don’t know, maybe we I could ask you a little bit about that, about, you know, the tools and techniques that you’re finding as are beneficial to teach people and kind of help them start their their journey, the kind of things that maybe I should have asked you to help me get started sooner. So, yeah, what kind of tools and techniques. Are you? Are you finding valuable to teach people when they’re starting their their AI journey? Kiera, how about do you want to take that one to start with?

Kira Corbett 8:48 Yeah, well, I have to say at first, like, I very similar to yourself, and even also working in, like, robotics and things like that, I was pretty hesitant to, like, adopt a lot of AI tools at first, because I’m like, I just like my whiteboard and my pen and paper, and I just want my simple text editor or simple IDE all these sort of things that you just already have, all these workflows already implemented. But then, like, you know, here and there, you start hearing someone pulling up some sort of AI tool or something like cursor co pilot, and you might try it out a little bit, and, you know, it kind of gave me a little bit of a different perspective. I saw how other people were using those workflows, and started slowly adopting them into my own workflows. Cursor is a great example, because you have your, let’s essentially, like chat GPT built into your Visual Studio code. And I think, you know, you can use a lot of these tools, like when you’re teaching with students, and even in the courses, like students could just go copy paste stuff or put in really simple, loose inputs and get, you know, quite a bit of output. But I think the benefit with using some of these tools is treating it like, almost like a rubber ducky, you know, like having another developer there with you and kind of walking through the code or asking questions. Other tools that we’ve like are. Starting to explore and even show people as like tools that you can literally draw out, like a UI, and it’ll spit out some code, and it’s oftentimes like, really great places to start if you’re working on a huge project. So yeah, I think it just takes time. Sometimes adopting workflows like change isn’t easy for humans, and some people are really like, all for it, and some people are like, maybe a little more curious or a little more hesitant,

Phil Hawksworth 10:24 yeah, for sure. There’s, there’s, there’s definitely things in there that we’ll, we’ll, we’ll circle back to, for sure. And I’m, I’m desperate to start talking about, you know, things to do with drawing a UI, and then it’s kind of unfolding in front of you, but, but I want to kind of give Eve a chance as well before we get into that about talking about, you know, some of the tools and techniques that, like the the first place, the first port of call for, for you when you’re kind of teaching people,

Eve Porcello 10:48 I think, yeah, as Kira said, just using it more like an assistant than a replacement. I think we experimented with generating some data for just a prototype that we were working on recently, and we found, oh, that does that in one second when I would have thought about every name and how to make every name super funny for my little demo that I’m trying to create. Or maybe you’re you don’t have your data sources quite wired up yet, and you need to have some mock data working with a tool like chat GPT or copilot for that is really incredible. Just to spit out some JSON and get started where you might take, I don’t know, you might take your energy for your coding project into creating something fake when chat GPT can do it in two seconds, and then you can get to the actual functionality, the things that you have to think hard about or optimize, or those decision making, things that really do still require an engineer, but a lot of the kind of heavy lifting of generating some mock data or doing some of the setup, things that might not be as critical, that’s really useful to collaborate rather than have chatgpt Just do it for you.

Phil Hawksworth 12:10 Well, what do you mean by that? When you say collaborate, do you mean you’re collaborating with the AI?

Eve Porcello 12:15 Collaborating? Exactly. Yeah. So, yeah, totally. So you can just ask, like, here the fields, generate some data, and then you can tweak it from there and tune it from there. But again, that’s way faster than doing that all from scratch. Yeah,

Phil Hawksworth 12:34 yeah. I’m kind of keeping an eye on the chat as we’re all talking, and I’m seeing some things in here that feel a bit familiar to me, because, you know, you this idea of collaborating with with the AI, rather than kind of handing it over, kind of brings me back to the thinking that I’ve had that is, it feels really intimidating getting started. And, you know, I’m noticing in the chat some of the things that I’ve experienced recently, and the tools that some people have using that, that I’ve been tinkering with. And, yeah, knowing how to, how to approach this, I found really intimidating when it’s like, well, what, what do I do first? There’s some magic in here. So, you know, I’ve been using cursor, you know, which is, you know, a fork of VS code with some AI magic sprinkled in is, that is the marketing language that they don’t have on their homepage, but I’m sure they’ll have next time. But that’s, that’s the way in to use that I found quite daunting, because it’s like, what, what do I do first? It’s like, when you first encounter a terminal, you know, you know, when there’s a blinking cursor and you don’t know what to ask it, it’s like, where, where do I start? But, you know, finding those kind of little breadcrumbs, those those those on ramps, was really, was really valuable for me, and the kind of tasks that I found myself doing were, as you say, you’ve kind of collaborating with it. It would start to auto complete things in a way that really surprised me. And I wasn’t actually asking it to do anything, and somehow it was kind of intuiting what I was, not only what I was doing, but also the way that I was approaching it. And that blew my mind a little bit, because it didn’t feel like I was doing something very, very deliberate. It felt like it was kind of accompanying me. And then that led me to feel like, okay, maybe I can engage with this a little bit more. Because until then, I was really thinking, Okay, this is a replacement for some of the tools I would use, but actually it feels more like an enhancement. So, so yeah, that’s kind of got me, got me rolling a little bit, but I’m very cautious, I think still about some of the potential kind of pitfalls of this. I guess I don’t want to just sacrifice all of my critical thinking, and maybe, maybe I’m too precious about it, and maybe I need to learn that I’m not the smartest being, and I could give over more responsibility. But I don’t know. Are there? Are there pitfalls? Are there? You know, what are the kind of places that we need to be cautious or exercise some restraint, or or what have you I mean. Have you had thoughts about that? Have you? Have you encountered that? Were you? It’s like, I don’t want to use the phrase AI gone bad. That’s too dramatic. It’s not quite Skynet. But where, where are the limitations and the kind of the places that maybe it’s been less helpful than than it could be. Have you, is that just my anxiety, or is that’s a real thing.

Kira Corbett 15:22 I think there’s a few avenues. I mean, there’s kind of the more obvious of, like, Okay, this is spitting out so much output. Can we just, you know, copy, paste it and, you know, not engage in that critical learning. But it’s almost like a tool. It’s like, if you’re training for a marathon, you can put as much or as little as you want into it. And kind of the same thing with like, AI, it’s like you can engage or not engage as much as you want with it at the time, at the moment. But then you look at other things too, where it’s like it might be really powerful in a use case, like how you’re describing Phil, you know, using it like an assistant and a tool. But then there’s some stuff where it can it still be a little clunky, like, how are all these services integrating? So if I’m trying to, like, totally automate all even writing code and deploying code, but not all that stuff is, like, linked together. It still feels a little bit siloed in some avenues when you start integrating things. But as a tool, I think the biggest pitfall could be just the critical learning piece that I’d be concerned as, like, an educator,

Phil Hawksworth 16:19 right? Yeah. And does that kind of that overlap with creativity as well, like create a critical learning and, like getting into kind of creative thinking as well, or are those kind of two sides of the same coin? Because I, I, I have anxiety about, you know, the the way the creativity kind of flows through this, and how that kind of is influenced by it.

Kira Corbett 16:40 I think that’s also a double edged sword too. It’s like, how much you engage with it. Like, I personally, I can’t even draw, like a stick figure. So for me, like AI is like, a really creative Avenue, because when I’m doing graphic design or video editing, or even photo editing, I actually turn to AI to help me, like, generate some ideas first and kind of get the, like, creative juices flowing. But then, like, it can be a double edged sword, because someone who might be like a photographer, videographer, for example, like maybe AI, is taking some components that they don’t want to use. So I feel like it could be a double edged sword. How you you see both sides of it? It’s just like a matter of how, how hard you’re going to trade for that marathon, or how much you’re going to engage with it.

Eve Porcello 17:22 Totally. I was also just going to add to that, when we’re thinking about learning and creativity, we’re reading about this study where basically the learning outcomes are so much better when you engage with others or you engage with the tool. So like a student who goes to a class, they might feel like, well, I’m going to create this paper. I’ve created it with chat GPT. I’m throwing it out there, or I’ve created this app with chat GPT, or the use of copilot. And really the learning starts to come in when you’re doing when you’re engaging with it, when you’re building a lab activity, when you’re kind of walking the process with the tool. So that kind of like second step is the important part, where you might be using the tool for you might be using AI for ideation. Then you need to say, how do we make this better? It’s kind of like the old joke. Is it a joke anymore about Stack Overflow? Like just copying and pasting from Stack Overflow? That’s the way that we’re going to get this app done. And anyone who has done that knows that copying and pasting, if you just throw that into an actual app, results are pretty bad usually. So same goes for the use of AI tools. You need to involve that critical and creative art of your brain to take that as a suggestion, but also fine tune it. Yeah,

Phil Hawksworth 18:55 and yeah. I mean, I couldn’t agree more. And I think perhaps the kind of rush that people have had as you know, these tools started to emerge of doing exactly what you’re describing, like copying, pasting something into a box, and then taking the output and going finished. It did a thing for me, and then using that, I it was so tempting, because it on the surface, it kind of this, this passes the kind of squint test a little bit. But when you look a bit close, more closely, then it’s like, Okay, hang on. I’m starting to see patterns there, and I don’t know, I think that that rush of popularity and people’s finding it really accessible to do that, that might have been, in part, responsible for my own hesitancy in engaging with it. Because Hang on. This is, this is just more of the same guff from before, and it’s kind of a shame that that that was that we had to go through that learning curve. Because I think, I mean, it’s it’s just how we got from where we were to where where we are. And I think when people do start realizing how to use it as a tool with bit more finesse and to not replace some of the critical thinking, I think that’s when the opportunities get really. Exciting, particularly from, you know, certainly from my point of view, Jeff made a comment in the chat about it being really good for learning as well. And I think that kind of that hooks into what you’re talking about. I mean, I’ve certainly found that as I started using it, because I’m not very experienced with TypeScript. I’ve been writing JavaScript for a very, very long time, but I’ve never really invested the effort in learning TypeScript properly, despite the fact that one or two people have told me maybe I should. There are people out there who have opinions about that. And, you know, I started working on a project where I was using TypeScript and I was using cursor, and I found it incredibly helpful in terms of not just saying, oh yeah, this is the code you need to pop in, but also saying why, and if I had a problem, it would tell me what the problem was and why. And so I couldn’t agree more about this, this comment about it being really great, great for learning. And I don’t know that that took a bit of a mind shift for me. I mean, is that, when you’re teaching people how to start using tools for AI, is, is that kind of part of it? Is there a mindset that you’re trying to get across, or is it, is it, you know, do you teach people more about here are the tools, you know, get stuck in was because it feels like it’s a, it’s a bigger, a bigger kind of mind shift. Really, it’s not just about what you know, it’s about, you know, your approach. I think, is that something that you talk about?

Eve Porcello 21:17 Yeah, I think it’s like we wouldn’t just shut off the internet for our students and tell them they can’t look at documentation pages or tutorials. So we’re saying, like, use this to help you, but you’re not going to be done there. So it’s kind of a, as I said, collaborative thing, where you’re just saying, let’s get some help from this. And also, how do we use, and Kira has done a lot of awesome work with this, is like, how do we use these developer tools to boost our productivity and just do more so it’s less, like, let’s just do the bare minimum. We want to, want to do what is actually possible with all of these tools. So maybe she could speak more to that. But there’s this is also it’s going to get better. So even people who are using things and they’re like, Well, that didn’t really quite spit out what I wanted, I don’t think this is ready. I don’t know, give it a year and maybe it won’t be the case anymore. So I think this, these tools will continue to explode and shift the way that we’re working as developers. Just like when, I don’t know, Adobe software came out, it didn’t ruin graphic design or just like when, like Dreamweaver or WordPress, it didn’t kill the field of web development. It just made it different.

Kira Corbett 22:44 Yeah. And one thing too, I feel, like, with people who are a little hesitant with starting using AI, and like, you know, even in classes, is like, how can we use AI on like, a daily basis? And one suggestion I like to provide people is kind of replacing, not replacing, but like, kind of using AI like your Google or like your search engine, because you can kind of get a little bit of feel for some of the capabilities that is possible with AI. And also, you know, there’s a lot of developer specific ones too. Like, there’s this very, really cool tool called Find, and so it’s like, if you’re going through, like, still some of the Stack Overflow articles, it can help, like, source the best one, because we do have a lot of oversaturated information on the internet. So it does make it a little bit easier sometimes to use AI, like your search engine as a developer, or, like, learning about certain things in code. And that’s one step. I suggest to people who are sometimes a little hesitant, just see what happens. And you know, you might not like it, you might actually find it might be a little bit helpful too.

Phil Hawksworth 23:42 Yeah, the idea of it being your, you know, your search engine, I think that that really resonates and and, you know, it feels like it’s really well suited to to that kind of activity, where the content that you’re training it on is very specific, or very, you know, very contextual. I mean, I think that’s, that’s probably why we’re seeing it being used so so often now, for you know, products that have documentation you know, and help forums you know, being able to be able to have an interact, an interaction with, you know, a conversational interaction with, with an AI that’s trained on your documentation and your you know, all of you, all of your content relative to your product, and you’re seeing that in Help forums and Doc sites all over the place, and that feels like a really that’s felt like a real step forward for me, particularly as a product gets rich and detailed, and finding the information in the Docs is tough. It just feels like a really good application of that. So yeah, it’s nice seeing, seeing that happen. You kind of touch on the idea of, you know, this anxiety that jobs get replaced. Do you, do you think we’re Do you think we’re moving past that kind of mindset at the moment? Do you think? Do you think that we are reaching the point that people are more open to. People more like me and cautious. I mean, I’ve never really felt like AI is going to replace me as a web developer, but I know a lot of people have thought that. I mean, do we still encounter that kind of viewpoint, or are we beyond that? Yeah,

Eve Porcello 25:15 I think we do. I think we totally do. I think there’s economic shifts kind of went along at the same time as AI’s explosion, and so where the lines are between all of those things are that’ll be interesting to see how that all shakes out. But yeah, I just think, like the AI conversation, it really is about like using it most effectively to help you as a developer, but also to help your creativity, to think about what AI might be able to help you build as a user experience. So thinking about personalization and using you talked about the documentation example, like, how do we find the best thing, or how do we create a learning platform where this is actually boosting us? Or thinking of popular sites like Netflix, how do we use the machine learning that’s already been there behind the scenes and incorporate more just better results, faster results? And I think if you think about how you can harness this technology. Then it makes you less replaceable. To say the quiet thing out loud, we are constantly having to evolve and learn. Just like 10 years ago, our stack might have been totally different. We had to learn a new one. We’re going to have to keep learning in this career, and that’s what’s fun about it, but also, like, we don’t have the luxury, maybe, to stop doing that. Yeah,

Phil Hawksworth 26:47 I’m going to interject just because I’ve seen there’s a comment in the chat from Jacqueline, who I used to work with at netify, actually. And Jacqueline says, As a docs writer, I admit I’m nervous, and, you know, I did use the example a minute ago of or it’s great for helping people find find support. And I haven’t I have thoughts on that. But do I What Would either of you kind of respond to that with? Have you got thoughts on that for people who write documentation? Otherwise, I’m going to start spouting my own opinions that’s coming next.

Kira Corbett 27:19 I mean, I think it’s one of those things. Like, I personally, I think it just depends on different businesses, different projects, is whatnot. But the thing that I think that’s powerful about AI, is that, like, if we can automate some of the tasks, like, yeah, automate some of the tasks so that we can focus on more of the human to human interactions, I feel like that’s powerful. But when, like, it comes to writing, like, I feel like that’s where you gotta let some of that creativity and critical thinking still come in. Like, we, I mean, you could have aI right for you, right? Like, it’s like, we’ve seen it done with chat GBT, but there’s something different about like tuning it to your personality and having your own creativity and flow with it, because at the end of the day, like aI under the hood is a lot of just probability and mathematics and a lot of training of yes or no, of what’s going on, and that doesn’t always replace, like, our human style and personality and those things too.

Phil Hawksworth 28:10 Yeah, I Yeah. I couldn’t agree more. I don’t want to just just be echoing exactly what you’re saying, but I’m giant plus one to that. And I, you know, I see this in other areas as well. You know, I, a few months ago, I was at an event to give a talk, and one of my fellow speakers there was was telling me how they were really excited that they were going to give a talk that they generated with chat GBT, and it’s like I fed it a blog post, and I asked it to write me a talk, and sadly, I wasn’t around to see the talk, but I was kind I’ll be honest, I was kind of horrified, and just I wasn’t, because I felt like my job was at risk, but, but I think, you know, Kira, you kind of talk about, like the human aspect of it, and it’s not just about what we’ve got to protect human jobs. Protect human jobs. That’s an expression. That’s a weird one, isn’t it? You know, the human characteristics? That come forward through through that medium are so important. And, you know, it might seem more most obvious for, you know, endeavor, an endeavor like writing a conference talk where you’re presenting, but I think it’s just as important in places like documentation, where the way we communicate is just as important as what we’re communicating. And I think, being able to to write things that resonate with people and that are pitched in a way that that cut through the, you know, the complexity, and kind of make a point eloquently and in an approachable way. That’s that’s a very difficult thing to do well, and technical writers are wonderful at that and, you know, and and the writing in one place on one brand for one product isn’t going to be the same necessarily for another. So I think, I think there’s a whole host of nuance and expertise which is so important in that arena. So I don’t That’s why, kind of my, my muscle twitch reaction to when I see things like that, because I. I, I, yeah, I just don’t think we can automate all of that away and expect the same level of connection to be there. Yeah,

Kira Corbett 30:08 totally. And like, you could argue too, that, like with you could train models to be more human like or maybe you have business expectations where it’s cheaper for a corporation to pay AI to write things, but then at the end of the day, it’s like, you might be building, like, a similar to, like, if you were to copy paste a bunch of Stack Overflow or, like, really hurry to get somewhere, a code base somewhere, you’re building code base on stilts. So you’re building on top of like, now you have all this tech debt. And I, I feel like I argue that when it comes to writing, creativity is like, you can have a lot of that replaced, but you might start building things on top of stilts and like, you know, you kind of end up in a sticky situation at a later point, even if it was, you know, more cost effective, for example, upfront.

Phil Hawksworth 30:51 Yeah, totally. And I feel like we see trends as well, right? I mean, we see, I mean, goodness me, as developers, we embrace trends very often because, you know, it’s, it’s exciting to follow the trends and the new ways of building things. And, you know, different we learn different things at different times, and then things become popular. And, you know, it’s a very trend driven environment, I think, but we see that in terms of the visual esthetic that comes across as well. And it’s really easy to spot sometimes, when you mean that there’s a lot of people have beautiful avatars of them looking a little bit like a superhero in a very oversaturated pencil drawing kind of style. And I kind of feel like, you know, we’ll ride that wave and then we’ll be on to the next thing as trends move on. But it just feels like a stage that we that we kind of have to go to as we go through, rather as we’re kind of being educated about, you know, what the creative use of these. These things are what? Okay, changing tack a little bit. I mean, what? What are the Are there any particularly creative uses of this that you’ve seen that you think that I would never have thought that up? You know, that’s that’s something that’s a really interesting execution of something, or a creative idea that has used AI along the way. I mean, I put you on the spot there, that was a very open ended question, what do you think?

Eve Porcello 32:05 I think a lot of the kind of testing tools that are out there are interesting for and there’s a lot of potential there. There’s a lot of there have always been testing tools that exist to support what you’re doing. But I think that’s an area where things can continue to improve. So that’s just built into cursor and copilot right now, but others will emerge. I feel like that’s a an area that, especially as you said, like TypeScript being more of the default with people in the JavaScript world, I think that will be super helpful. And then then I’ve seen just tons of video production tools, so transcripts and things that make our A lot of times a podcast or a video would just not have transcripts or not have assistive tools like that. And I think it’s kind of like because AI has made that easier and more cost effective. It’s something that is a requirement, even though it maybe should have always been a requirement, but, like, YouTube has them, and then Riverside and other production tools like that has made it really like you have to at this point, half those things? Yeah, Kira, I sure has a bunch more than me.

Kira Corbett 33:23 No, those are all very good ones too. I feel like I’ve seen a lot more things that were like, surprising with it too. Is just a lot more software services coming out too. A lot of people when AI was mostly when, like, chatgpt kind of was starting to become bigger, but that’s when a lot of people started building building GUIs on top of just like, pre trained models. And so it’s been really interesting to see like similar like video production or podcast production. It’s just been interesting to see all these different types of services and businesses coming out that are just powered with AI under the hood, but actually do, like, a very helpful task for a lot of things, or may have something infused into it, like there’s a productivity app called Monday, and there’s a lot of, like, automated tools in there, or even swell AI, where it’s all, like podcast editing. So like Eve said, there’s a lot of stuff that’s kind of evolved to that might, that maybe should have been there for, like, accessibility on the web, but now, like those things, we don’t have to be, you know, typing it up there, it can just be automated.

Phil Hawksworth 34:22 Absolutely, it’s a Yeah. There’s Yeah. We think about like, AI tooling and AI like features that we can use in web development. We, I don’t know. My mind often goes to like, the bigger, more, like, more flamboyant uses of it, and expecting to kind of just code to be generated out of nowhere. And certainly I’m experiencing a level of that, and kind of an A level of kind of amazing auto completion, or I’ll write a comment for a feature, a function, and then ta, da, there it is beneath the comments. And I was like, Wait, hang on, how that was what I was going to type. And that’s, that’s mind blowing. But I think some of the things that you allude to there, I. Just as if not more impactful than than that. It’s the it’s the things that we don’t always necessarily think of as, as as AI, because they’re, you know, a layer or two. They’re subtle, you know, I, I like boring things, you know. I like things that are simple and boring, and art don’t make you go, whoa. It’s magic. And it’s some of those things that become almost invisible and you take for granted, like, you know, like the layers in a figma file suddenly be getting, getting named the right thing, you know, without you having to do it. Or, you know, details like that, where you know, some of those bits of automation that are the the busy work, the administrivia that you know, aren’t necessarily about creative thinking. They’re more about hygiene and kind of putting in that, putting in the effort. And I love it when bits like that start to be done for me. And it just reduces the friction. So in many ways, you know, it feels like this allows us to be more creative, because it gets out of the way. I you know, a lot of the the the talking I’ve done about trying to simplify things, or kind of embrace a simpler stack, is because if you can simplify things, it gives you the ability to concentrate on the more creative pieces. And I don’t know it feels like there’s an opportunity for that for us with AI, goodness me, it’s nearly 20 past the hour. Did I sound like a radio DJ for a second before we go to the news and weather? I know we have to go back to Brian in a second, who hopefully will give us an update on the weather, but that’s not his main job for the day. But I kind of wanted to round out the conversation a little bit with just an eye to the future. I mean, we it’s kind of blown my mind what’s come along in the last couple of years, but I don’t have a clue where we might be going next. So I don’t know what, what are your thoughts, instincts, hopes, fears, you know, what’s, what’s, what’s coming, what’s coming up in the next couple of years? Do you imagine that might be interesting? What’s, what’s your thoughts? Kiera, do you want to go to you first, and then we’ll close with

Kira Corbett 36:55 you? Yeah, I think it’s going to really change the way we’re working, and how we’re working in the types of jobs we’re going to be experiencing. You know, I can’t predict the future, but I do think there’s going to be a lot of change, and that change might be a bit of a hiccup for like in people in industry, especially if we’re automating so many things, and then we might not have people interested in doing stuff that requires an actual human to do something, and those types of jobs too. So, yeah, I think there’s just going to be a lot of change, and it’ll be interesting to see how we all respond to that and adapt

Phil Hawksworth 37:26 interesting. Yeah. Eve, yeah. I

Eve Porcello 37:29 think web development tooling will continue to improve through this. I think you’ll still need to support that tooling, but I also see the need for developers to because of costs of running these large language models, maybe start to create their own and create very small language models that support their own tooling and use cases. So I think we’ll kind of like learn more about being the doing the most advanced stuff with our kind of JavaScript stack that’s a personal lens, but then also learning more about lower level things to really train our own models and tune those up so that we don’t have to pay open AI for everything. Though we would love to line their pockets more, of course, but it probably makes sense to start to think about our own absolutely

Phil Hawksworth 38:21 awesome. Well, I mean, I so, I confess I am, I am excited about kind of the developments that are coming coming our way now that I’m getting a slightly more healthy perspective on it, and I do feel like we’re, we’re getting beyond some of the misnomers of AI that might have been around in the past. And I think it’s, it’s an exciting, it’s an exciting future. Um, thank you both for hanging out and having a conversation. It was, it was great to great to chat with you both. I know that I’m very fortunate that I’m going to get to see you both again next week, because, and this is, this is bad form on an event, plugging another event, but, but you’re both going to be at netify, compose next next week, aren’t you? And hopefully you’re going to get a chance to actually build something in front of people which, who, who knows, who knows what’s in store for us, but who knows

Eve Porcello 39:06 what’s in store? Will there be costumes? Will there be too many props to bring on a plane? Who knows? Hard to say.

Phil Hawksworth 39:14 Expectations, Kira spaces, picture enjoying Kira, face to that. Reaction to that as well. But expectations, height, sky high for that, but, but enough of that for the moment. Thanks again.