DEV Community: Jason Hand

How will AI tools like ChatGPT and GitHub Copilot affect Developer Relations?

Jason Hand — Tue, 14 Mar 2023 16:14:03 +0000

As the world of artificial intelligence continues to evolve, it's increasingly evident that the AI tools are rapidly changing the way we work in developer relations and community building.

Tools like ChatGPT and GitHub Copilot are transforming the landscape of our work by automating tedious tasks and repetitive work, providing valuable insights and analytics, and even helping generate content on our behalf.

However, the integration of AI into our industry also raises new questions. Can AI truly enhance the work of our DevRel practitioners and allow us to focus on different challenges, or will it end up diminishing our creativity and impact on supporting developers?

How do we make sure AI is used ethically and responsibly, and what impact will it have on the future of not only DevRel, but software development in general? Join us as we explore the exciting world of AI in developer relations on this episode of the Community Pulse.

Welcome back to another episode of the Community Pulse. Today we're excited to speak with two amazing guests who have a ton of expertise to share with us today, Chris DeMars and Rizel Scarlet. Welcome to the show. Let's hear from each of you, starting with Rizel.

Rizel, before we get started, could you please give us a little bit of an introduction about yourself? Sure. Also, thanks for having me on the show. I'm super excited. But yeah, my name is Rizel. I'm a junior developer advocate at GitHub. I've been here since September 2021. And super pumped and excited about Copilot. I always try to do advocacy for it because it's something that at first I was like, what is this? I'm not really into it. But then I really found the love for it and how it helped me. Before that, I was a software engineer for a bit. And yeah, that's me.

All right, Chris, tell us about yourself. Hey, y'all, thanks for having me. And I have to give a shout out to PJ, my home boy in Buffalo, for asking me to be on this with y'all. So I really appreciate it. Yeah, my name is Chris. I'm a developer advocate lead at Split, Split Software. Split is a feature management and experimentation platform. I've been there since September. And I talk about accessibility and CSS and started talking about feature flags and AB testing, testing and production and all that fun stuff. So yeah, stoked to be here. All right. Glad to have you both here. Yeah, exactly. Very happy to have you both. Thank you.

I would like to start with our first question, which is kind of like level setting for people who aren't super, like familiar with tools like chat GPT and Copilot.

Let's go with you, Chris, can you tell us a little bit about how these tools are used in DevRel and DevRel communities? Yeah, so I think, especially like in the chat GPT side, you know, I have a writing background, and I'm a published author. I mean, it's been a long time since I've been a published author, we can say, in quotes, published author. But even with my writing background, you know, trying to come up with conference talks and titles, like you might be super passionate about a topic like I'm when I when I put a new talk together, like I'm really passionate about it.

I feel like people really need to know about this. Or if people have like a misconception about something that I'm passionate about, like I want to be able to drop that knowledge on them. And you know, a lot of the times I have trouble trying to come up with like a good abstract, like at least like four sentences, right. And even with that writing background, it's still, it's hard to come up with something because you don't want to take too long writing an abstract.

But you also don't want to just like, just off the cuff, you know, do it from whatever. Yeah, they call it the burden of knowledge, you know too much about it. So it's hard to distill it into right. And, you know, you don't you don't write the title before you write the abstract. So I found myself because I saw somebody else do it. I was like, Oh, this damn like chat GPT can actually like put together like a really nice, intelligent sounding conference talk. Let me try this.

So I did it. I said, you know, I asked chat GPT to like, give me five abstracts, conference abstracts on technology A plus technology B. And it's been out for me. And I read through them. And I was like, Oh, wow, like, these are really, these are really powerful. Like, this is way better than the stuff I could write. And then I had it generate titles for those. And it was like, it's super clever, because I'm really like big into clever titles. And being a conference, like, you know, being a conference organizer in the past, and being on selection committees currently, like, I'm always looking for a catchy title.

And it did that. I was like, Oh, this is pretty sweet. So if you struggle with writing, I think chat GPT can definitely help out in that sense when it comes to creating abstracts for conferences. Right on result. Yeah, um, I think AI tools are just having a really big impact in DevRel communities. And I say this because like, when I was a software engineer, I was like, wow, I would love to be a developer. Like, I still love it. But like, I had this different image of like, what being a developer advocate would be.

I'm like, I'm spent, I'll spend all my days just learning and teaching. And like, it's just going to be so fun. I'll have all this time to absorb all this knowledge and be able to create all this content. And even though I kind of do, I kind of don't have all that time, because it's like, I have to do like this podcast, and then right after this, I have to do a stream and then people are complaining about being on this podcast. Just checking. I'm happy to be on this podcast. But I think I I miss, I miss imagined or misenvisioned.

I don't even know if that's a word. But I didn't envision like how, how busy it would be and like how less time I would have to really sit down and like, like how Chris was saying, like, write an abstract or write, write a blog post or something like that, especially like when we're going from conference to conference. So I think like tool AI tools like these are really helpful in us like getting the work done faster, more efficiently, and then without us losing our creativity, because sometimes when you're like, Oh, I gotta do this really quickly. You like kind of sometimes don't put in all the effort that you need to. So I'm not saying I'm mad at being on this podcast. I'm excited. But I want to be able to give my 100% everything.

And I can't because I'm one human. So I feel like AI tools have been helpful in like, allowing me to be able to like, give a little bit more, you know. One of the things we were just talking about before we went live was, if we, you know, those are great examples of how the the tooling was, you know, ChatGPT specifically is so great at just sort of unblocking us for certain creative tasks, just like Chris is describing. I've used it for similar stuff, like I've been doing a lot of work in YouTube shorts. And sometimes I'm just feeling less creative than usual.

And it's, I'm sort of, when it comes to creativity, I enjoy working in groups and like brainstorming. And I feel like, even though my team has often been busy and unable to sit down and have those creative sessions with me, I still have something I can now rely on to at least give me some options to think about, which has been a great experience. And I think, you know, the obvious first use case that comes to mind for those creative experiences, our blog posts, titles, tweets, things that, you know, it's just, it's, you're really trying to like win the attention of algorithms in SEO.

And sometimes it's better to rely on on a machine learning model to help you figure those things out. However, in just the, you know, little time that ChatGPT has been out there and been used, I've seen books now written with ChatGPT being, you know, I'm sure a big contributor to the, to the text that's in there. Chris, to Chris's point, it's now showing, they're showing up in a lot of conference talk submissions. And I guess where I ultimately want to go with this is we can see the benefit as content creators trying to put some submissions out, CFPs, and that kind of thing. But also me as a person who has, you know, put on plenty of events and done a lot of reading of said submissions, we're now looking at a situation where we have just a lot more to go through to like figure out, are these talks, you know, what we want in our program?

You know, is this fit the content and the theme and everything else? And you used to go from like seeing two or three from a person being submitted, assuming that there's no limits, to now it, you can submit as many as you want. And it's just creating this picture, I think, where a lot of conference organizers are a little bit stuck on what to do with this. You know, we'd love to hear more and, you know, maybe sometimes it is just the matter of language and how it's described in the abstract or a catchy title, like Chris said, like that goes a long way. I have used very catchy titles specifically to catch the eyes of the CFP people that are reading it. So let's start with Roselle.

I would love to hear your thoughts on, is this a problem for our industry where we now are just creating so much content? It's hard to get through all of it. And yeah, so let's just start with that. Wait, can I clarify on your question? Are you saying, is ChatGPT harming all the content that we're putting out? Or are you saying, do we have too much content on our plate? I think from my perspective is, if I'm now going to be reading a lot of submissions, should I put some limits on either, I don't want any of them to be submitted with ChatGPT's assistance, or now do I just have to put a numerical limit on how many people, on how many submissions I take?

I guess. I don't, in my opinion, it's not, ChatGPT is great. I love using it, but I also recognize that we can now put out X in whatever variable you want content at a much faster clip, which still, at the end of the day, requires humans to consume. It just seems like it's creating an off-balance situation. Yeah, okay. I'm glad I got clarity. I just didn't want to start rambling on a completely different topic. You're free to do that as well. But I think with these AI tools, there needs to be a certain amount of balance. I've never been a conference organizer, so I don't know. And even without ChatGPT, I pushed the limit of, so sorry, conference organizers. If you say I could submit 10 talks, I'm submitting all 10 and hoping you pick one of them. Which probably is not the best practice, so sorry.

But even without, like, ChatGPT and stuff like that, I was already doing that. But I do think, and even when I talk about Copilot, there needs to be a balance of, like, how much you're inserting yourself in there and then how much you're using the tool. Because, for example, you can just use Copilot and keep tabbing and keep letting it auto-suggest things to you, but sometimes it might not be completely correct.

So you can't completely rely on that. You want to make sure you're double-checking what it's suggesting. You still have to use your own log, no, not, what am I trying to say, logic and knowledge to actually make sure it works and make sure it's doing the right flow that you want and it's not introducing any, like, vulnerabilities. Same thing with ChatGPT. I don't really use it, like, everyone's like, they use it to write all these blog posts and stuff like that. I've used, like, Copilot and ChatGPT to write blog posts before and I've noticed that, like, less people want to read them. Like, I feel really strong in my writing skill and I really like my writing style and I think other people do too. So I think that using, when I rely too heavily on AI tools, it removes my voice and the way I would present things. So I think it's good in, like, idea generation and leading me in the right direction but not for it to completely remove, like, who I am and what I bring to the table as a developer advocate because we always talk about DevRel is about authenticity and stuff like that.

If we're just relying on these AI tools, it takes away that authenticity. So maybe as a conference organizer, yeah, maybe we should say, like, some kind of limitations, like, yeah, you can use ChatGPT for, like, idea generation or helping you with your title but don't rely too heavily, like, just like a little warning, don't rely too heavily on it for the abstract and stuff like that. I don't know, I'm not a conference organizer, so. This is great. Chris, what about you? Got any thoughts on, if you put on your event organizer hat, who has to look through all these submissions now? I think, so I think this could go, I would say that there could be, like, a couple disadvantages to this, right? One being, like, Rizel was saying, like, if you're going to accept a bunch of CFPs, I'm going to submit everything I have and hopefully one of those gets picked.

Now, not all of them are going to be written in ChatGPT, but still, like, if you don't put a cap on it, then that's a lot of content to go through. I think it would make sense for conferences, especially now with ChatGPT, maybe putting a limit, like, you can only submit three, or you can only submit five, because then that will bring that down. But also, I could see it as a disadvantage to people that are submitting talks that don't use ChatGPT, that are writing it on their own, because now you've got a handful of these talks that are probably amazing, and they sound amazing, because they're written with AI, and you have all these other talks that are written, you know, off, you know, off the brain, and it might sway conference organizers one way or the other, picking talks that are only written, like the abstracts are only coming from ChatGPT, opposed to somebody that's just writing them with, you know, the knowledge of writing that they have. So I could see the disadvantages of both ways, but then also, you know, conference organizers could say, listen, if you are submitting a talk with ChatGPT, we're not going to accept it, and we'll be able to tell, like, there's a clear distinction between something I would have written myself, and something ChatGPT could have written, because then you get that voice, like Rizel was saying, your voice comes through in your abstract, your style, the way you talk is going to be translated into words. So conference organizers could do that, too.

So I could see a handful of different disadvantages with it, which I've never actually thought of prior to this conversation. That's amazing. I mean, we're all trying to learn and we're thinking through this because we're all figuring it out. It's very, very new. And I think the conversation, at least the context of it, has been a lot about, as DevRel professionals, how we use it to do our job to kind of eliminate some of the steps to make things easier. I'm wondering from the content consumption and, like, end users, how do you feel that we can use these tools for maybe efficiently, like, choosing to make content more accessible for people trying to understand something? For instance, I know, like, if you're able to say, like, show me a video on how to use, I don't know, Ruby on Rails or Go in the style that I like to learn, or I know that a lot of search engines are going to be incorporating it and say, tell me the top three ways of doing certain X, Y, and Z.

ChatGPT and other tools like that using AI will at least give me a way to consume and learn in a way that is, I would say, conducive to the way that works better for me. Do you feel that in your roles of a way to making that or exposing that type of interface to help people kind of consume whatever content that you're creating as well? And we ended with Chris, so let's start with Roselle. Oh, dang it. I was like, let Chris go first.

Okay, let's see. So we're saying from the consumption end, how do we make this, how do we use AI to make things more accessible for other people? Yeah, like building it into your docs. So that, you know, sometimes you want code snippets, sometimes you want an exhaustive, like, before recipe where the rest of potatoes are like raised and grown and why they have developed over time before you get to the recipe. Some people like that.

Some people just want the recipe. So like, do you feel that that is, there's an opportunity to incorporate that stuff into the way that people can now like, choose to consume content? Yeah, actually, now that you bring up the doc stuff, I've seen that the Astro team, who like created this framework called Astro, basically, they developed an AI or an AI tool called Houston AI. And basically, what happens is like, you type in the chat of Houston AI, and you ask it questions about Astro, and it searches through the documentation and helps like return, like solutions for people. And I think that's such an amazing thing, because especially when I was like, learning to code back in 2018, documentation looked really scary to me and confusing. So why not have an option for it to be more conversational, where you're like, Hey, how do you do this with Astro?

Like, how do you handle front matter and stuff like that, rather than like reading through these documentation, it'll return like the results to you in a more conversational way. So I do think there's more opportunities for other projects to do something similar, and make, like, reading documentation more accessible for folks. Um, but that's, that's like the one example I could think of. I'm not sure. That's awesome.

I mean, you even said the word read documentation. So it'd be nice if it does text to speech and gives audio props for those who may have visual impairments. So this is definitely could be a way of accessing something in a different format. Yeah. Oh, wait, and I would be remiss, sorry to interrupt you to not mention Hey, GitHub. Like, um, I don't know if y'all are familiar with it. But Copilot has like a voice activated, like, version, where, like, you can be like, Hey, GitHub, type this out for me. And now we've updated even more, where it could be more conversational. You can you could type in the little Hey, GitHub thing and be like, Hey, can you help me figure out how to do this? Or you can speak it out to say, Hey, GitHub. So that's also another option. That's amazing. What about you, Chris? How do you see any ways that this can make it easier for user to consume technical content? Yeah.

So coming from the accessibility space, I think, I think that's really good. So like, our our help docs at split, really robust, a lot, a lot of great information there. But any docs, like anything like if you can type in a search bar, you know, you know, help me or find this or give me like, a code snippet for that. The one thing you don't, the one thing you don't want to do is make your users think right. And your user is going to be the developer or anybody that's just reading the docs. Right. And it in the accessibility piece of that, too, is like, I don't know about you, but if I'm trying to look for something in the doc space, I get really like my anxiety gets a little higher. If I can't find what I'm looking for, I get frustrated. And I don't understand something. And then you know, I then I'll get distracted in the ADA, the ADHD piece will kind of fall into that. We want to try to eliminate all of that stuff. Because when we talk about accessibility, we're not just talking about visual, hearing, mobility, cognitive, and temporary. We're also talking about hidden impairments.

And a lot of times that's overlooked. And two big pieces of that are anxiety in ADHD. And if we can eliminate that, when it comes to searching and reading, doc and then like the docs space, I think that's extremely powerful. And I think that should be baked into more docs across all different types of organizations, and in your products and projects. Yeah, keep you from going to a forum and someone calling you stupid. If you know it's a, like it's an automated support system, then you can kind of reduce your anxiety knowing that someone's not going to say, you should have known this already, or, oh, you're new.

So yeah, that's definitely a really good point. I know for me, you know, much to what everybody's already saying, like docs sometimes are just scary. You go look at them and you're like, where do I start? And my eyes dart to the biggest boldest text first. So I, even though like somebody went to a lot of effort to put in a really great intro or summary of like what I'm about to find in this doc, I almost immediately find myself halfway down the page. And that's just the way my brain works when I go to read.

So then I find myself skimming and scrolling and not finding exactly what I want, you know, as quickly as I can. So I think this is a great use case where we can start looking at the consumption of our technical documentation and retrieving that documentation in different ways. I know, you know, Copilot does some interesting things around that as well, like suggesting some different, you know, code snippets or whatever. ChatGPT does the same thing. And it'll even tell you, like try to explain the code to you, you know, here's what this is going to do, which I think also helps. And, you know, docs in general, they go stale. And I think that that's another area too that maybe this could help is it takes some effort for a person to go back and review docs, you know, from even a year ago and make sure that this, well, you know, the screenshots on here still current because we're always changing the UI. Like there's just, docs are hard to stay fresh.

And I think anybody who's been, you know, around DevRel and been involved in technical documentation, creating it and reading it, have run into frustration with docs. Another area I kind of want to pivot to and talk about before we run out of time is this idea of metrics. Anybody in DevRel has had lots of conversations. How do we show our impact? What are we doing? Is it working? All those types of things. And aside, I don't want to get into the merits of what metrics are good and what aren't valuable within our industry, but one of the first use cases I had with ChatGPT was to report out on a project that I'm doing. And that required me, you know, doing some collecting of different numbers and crunching some numbers and then putting that into a summary and sending that out in an email.

And I will say ChatGPT cut the time in, you know, into a fraction of what it normally would take me to do that. And a lot of that was nothing more than copying and pasting some data out of cells and asking it to tell me, what does this mean? How can I explain this to my team in a simpler way that makes sense? I'm wondering, have either of you thought about this side of the conversation? What, you know, ChatGPT, of course, is just one, but there's a lot of AI tools out there that are trying to sort of shortcut some of the things that we either don't like doing, it's not our strength, or it's just difficult in general. Sharing numbers and metrics is hard.

I'm wondering what, we'll start with Chris this time, what we should be thinking about with within AI and how can we help it tell our story and be advocates for us and to like share what's going on within our communities? Chris? I've never even thought about doing that, so whatever strategy you're using, hook that up to your boy, because I need to figure that out too, because I'm using Orbit and like just trying to pull from a bunch of places that I can't track with the Orbit plan that we're using, right? So yeah, that's, I mean, I like that, like if you could just give ChatGPT all these numbers and say, put this metrics report together for me, whatever, and be able to take that consume that for yourself and then hand that out, like that's, to me that's super powerful, because that takes a lot of the guesswork out of it, right?

Because now you're not, you're not having to go through tool A and tool B and then like search for Twitter spaces that you did just to get the insights and the metrics on those and that can be a pain. Yeah, I haven't, I haven't really thought about that. It sounds slick, it sounds like a slick, slick move, so you need to hook your boy up with your process for that. Well, first of all, I wouldn't recommend putting anything into ChatG, into anybody else's tools that might not need to go there, you know, any of the numbers that I'm sharing are basically public numbers, there's nothing under NDA type of thing, so I don't want to advocate for publishing your metrics into somebody else's system, but Rizel, what about you? What are your thoughts on how can AI help us share this story of what's going on? I'm gonna echo Chris, I didn't even consider this as an option, but it sounds amazing, like I would love to use AI to help me better track my metrics, because I don't even really know what I'm doing with my metrics, I'm just like cool, people viewed it, I think it's going good, so yeah.

This has been a great conversation, and I just want to take a quick temperature check before we move to the our checkouts, and so just really quickly, just we've talked about ChatGPT, we've talked about AI, we've talked about using it as a tool, we talked about assisting people, we've talked about even like giving us tools to help our job, but we've also been talking about consuming it, but not it being consumed by it, so how would you feel, I mean we all do a lot of public-facing work, how would you feel if you are part of the training set, you're part of the training data, and someone else is using things that are derived from your work? Just temperature check quickly, Roselle, what do you think? How would you feel? I feel okay with it, I don't know, at my perspective is like I'm an open source, and I want to share my knowledge with people, and part of sharing my knowledge with people is them seeing my work, that's how I learned from people, is like I looked at their talks or their abstracts and stuff like that, and I was like oh that's how they do it, I'm gonna kind of copy until I find my own style, so I'm okay with it.

That's awesome, what about you, Chris, what if someone submits an abstract that looks like you wrote it? That's awesome, what about you, Chris, what if someone submits an abstract that looks like you wrote it? Oh, so now you're getting into a whole other world of mine, so I'm cool with it, that's fine and dandy, but I'm very much into the paranormal ufology and cryptozoology world too, right, and so there's always talks about other people living your life in like a different dimension, and I saw a meme the other day, and trigger warning, this might be a little dark, but not, it probably isn't a little dark, or dark at all, if like when we pass onto the world, the next world, wherever that is, but it's actually just somebody taking off the VR set and saying how was that, and the person taking off is like wow that was awesome, and they might have only been doing it for like two hours, right, but they lived your life that whole entire time, right, and now they can take that and move on with everything that they've learned from you, so you're getting into like a very weird space, and weird is a is a compliment, if you do the etymology of the word weird, it's actually a compliment, I love being told I'm weird, it's a compliment to me, but you're getting into a whole other dimensional category, and I'm fully on board, let's rock and roll with it.

So I just want to make a quick comment on that, I guess anything, well Roselle, do you have anything else that you want to add on this one? Okay, before I jump in here, great, so Wes and I have talked about this a number of times, because as we were planning this episode, and this is this is one big part of it, I know I've been in in an adjacent around the AI sort of space now for a few years, in a lot of my previous projects, and it always comes up about being responsible and ethical with our AI practices, and what we're doing, and I think this kind of speaks to the heart of that, a lot of us want to make sure that we're doing right by everybody else with a lot of this stuff, and some of it is, I think it's a little bit personal, when we do put our content out there under the badge of our company, that's one thing, when we put stuff out there under just our own persona, and our own personal world experiences, dev rel experiences, whatever you want us to talk about, it does raise these questions of, is this content mine, is it somebody else's, is it just like you said, Roselle, I've never thought of it that way, but I do think I'm an open source human being, I wouldn't be doing things like this show, and wouldn't be in this industry at all, if I wasn't really just driven on helping, and putting out as much information as I can, tearing down all the walls, and removing the gatekeepers, and let's all just sort through this together, I don't necessarily have a problem with the natural language models pulling in my content, my ideas, now the one thing I have thought about though, is I think there's a difference between putting out things that is more of an opinion of mine, a lot of the things we talk about on here, we'll get into more opinion rather than just objective reality, but I think when it comes to like sharing content about how I built something, if there's architectural designs, or I've just built some piece of software somewhere somehow, and I've showed people how I did that, there's not a lot of opinion in there, I might have some thoughts about why I made certain architectural decisions, but at the end of the day, I put that out there hoping, expecting people to do something with it, and I think if these new tools make it easier for that material to be consumed, or found, and used, then that's to all our advantages, and so let the machines have it, as far as I'm concerned, that's sort of the point of this position in general, so as long as they're not twisting the truth, and leading people astray, and making people make bad decisions that harm others, I think that's kind of the main thing we have to watch for.

Yeah, and I like how you brought up being responsible AI, because I do see how it could affect other creators differently, like artists, they're not necessarily putting out their art for people to learn like we are, they're putting it out to make sales, and stuff like that, so I can see how it's harmful, and how AI creators should be more careful with that type of stuff, just wanted to put that out there. That's great, well listen, we are just about out of time for this episode, I feel like we're just scratching the surface on this this subject, so I appreciate both of you being here, Roselle and Chris, to share your experience, I feel we're gonna have to get you back on soon, and do some more talking about these things, but why don't we switch gears here, and start talking about some of the checkouts, we do this at the end of every episode, where we have a little bit of time baked in to just share some of the things that are on our top of mind, both as hosts and guests of the show, within the realm of the podcast, so some things might be related to AI, some things might not, so why don't we just go around, and starting with Roselle, you want to share with our audience, anything that's top of mind that you'd like to put out there?

Yeah, first I want to share this stream that James Q Quick was hosting, he was featuring Corey Weathers, and they were talking about developer relations for Black developers, so that was an interesting topic, I'd even consider the fact that maybe DevRel is done differently for Black developers, or that we should, so definitely check that out, I thought it was a good conversation, I almost didn't show up to this podcast, because it got me, and then also there's this book called Finding Me by Viola Davis, I usually, I'm not super into books written by celebrities, but it's really well written, really deep, and it really shows her growth from how she started out, being poor, even living with rats and stuff like that, and then eventually getting what her dream was, and then also I made a discord recently for Black professionals in DevRel, or people, or Black folks looking to get into DevRel, so definitely check that out if you're Black, and thank you. How about Chris, what about you, you want to share some things with us?

Yeah, I got a couple things, so I wanted to share an article, and you can definitely just, if you go to Split, Split's website, split.io slash blog, you can find it there, it is called Three Ways Feature Flags Could Have Saved Jurassic Park, if you are a Jurassic Park fan, or a Jurassic World fan, I love the whole entire franchise, except for JP3, that was kind of meh, but it's a great article written by our director of DemandGen, Elizabeth George at Split, really, really cool article on feature flags and how it could have saved Jurassic Park, so if you're a fan of JP, you'll definitely dig that article, and then I'm also going to be on the road here in the next couple months, so if anybody wants to come, hang out and kick it, go to a talk, I will be, I'm actually speaking at a local meetup here in Michigan in a couple weeks in Southfield, but I'll also be at Orlando Code Camp in March, I will be at Dev Nexus in Atlanta, Georgia in April, and in May, I'm going to be speaking at Chain React in Portland, Oregon, so come hang out if you can. Excellent, all right, Wes, let's hear yours. So, of course, as a lot of people are, especially in our space, have transitioned over to Mastodon, and that moves into figuring out what you can and can't do there, but also what tools you have available to kind of use Mastodon and make it easier to interact with people, and so I've tried different tools here and there, I've tried Toot, there's ice cubes out there, there's a lot of different tools. One that I like that is web-based is Elk, so there's Elk Alpha, they're just still developing it, and it tastes kind of like the web interface for Mastodon, but makes it prettier, and it feels a little bit easier to use and more approachable than using the default Mastodon web client, because I think a lot of people experience Mastodon differently when they move to a mobile client, but this is a little bit different than experiencing Mastodon using a different web client, so if you go to elk.zone, you can now use your login for any server you're on and see this new kind of interface, and they also make a PWA version of it, so you can actually use it on mobile as a mobile application, so I like that type of experimentation and usage of Mastodon for those who are just relatively new to it and want to try different ways of experiencing it.

The other thing is more on topic, on brand for this episode, is the chatgpt cheat sheet, and I think you've heard of chatgpt used to write abstracts, and we also used it to do code snippets. I've used it for Excel formulas and creation, but you can also use it for like putting words into tables or doing sentiment analysis, so there's different ways that you can use chatgpt that may not be immediately apparent because you have just a prompt, and so this chatgpt cheat sheet is really great to just kind of explore the different ways that you could use it, so those are my two picks.

Awesome, great stuff, great stuff everybody, and yeah I got a few things I want to share too, but first let me just say, because I just want to, this has been so great to talk to both of you, Chris and Roselle, thank you for being here. Obviously this is a very topical current topic, I guess everybody seems to be talking about falling over chatgpt these days, but also GitHub co-pilot I think is really rad too, and we didn't spend nearly enough time talking about that. One of the links I'm going to share in the resources, you just go to the website to grab it, is this article I thought that was really good titled eight things you didn't know you could do with GitHub co-pilot.

I think the blog post I wrote, sorry, is it? Yeah, okay, I thought it might be. It was a couple weeks ago when I read through it, and I threw it in my notes, and then I haven't pulled it back up since then until today, but I remembered it being one of the best ones I had seen. So anyway, thank you for that. The other piece reading material I want to share is not related to AI, it's just something I've been reading on my own. It's called Platonic, how the science of attachment can help you make and keep friends. I think this book is just a great read in general for everybody. It's always important to understand our own attachment styles and how those play roles in in our relationships, including our work relationships and all of them.

So I'm a sucker for science explaining why I am the way I am in lots of ways, and so this is one of those that I wanted to put out there. The other thing, I've got a number of other articles, I won't go through them all, but they will be in the show notes or in our blog post that goes out with the recording, but I did also lastly want to just mention that there's an event coming up for anybody who might be in the Denver Boulder area or Front Range of Colorado, that's where I am, called the Learning from Incidents Conference. That'll take place, I believe it's the 15th and 16th of next week from when we're recording right now, so February 15th and 16th, Wednesday and Thursday. Tickets are still available. I am not sure if I will be there. I unfortunately suffered an ankle injury this week and will be undergoing surgery next week, so the timing is bad for me to be there, but a lot of my friends are going to be there, so I'm kind of torn on whether or not I can try to go right before surgery. We shall see. With that, Wesley, I want to hand it back to you to take us from here. Now I know we're missing PJ from this episode and of course Mary, but usually we end with a quote that PJ supplies and reads that is usually for a rap lyric, and I thought because of the subject matter of what we're talking about that, and of course the timing, we should keep in mind the ethics and the things that we're trying to do as we explore these new tools and how we can make them more accessible to all types of people. So my quote is we have before us the glorious opportunity to inject a new dimension of love into our veins of our civilization, which I think is something that we can think about when we're using these tools, and that quote is from Dr Martin Luther King Jr. It being Black History Month, I thought it would more be appropriate.

I'd like to thank our guests for being on the show. Thank you for sharing your knowledge and your experience, and I want to invite anyone who has feedback about ChatGPT and how it's changing your career to tell us and let us know and we'll incorporate it. Until next time, thank you for joining us here at the Community Pulse, and we'll see you next month. You've been listening to the Community Pulse. Find out more at communitypulse.io on Twitter at community underscore pulse or anywhere you get your favorite podcast. If you've enjoyed this podcast, check out our extra podcast, The After Pulse.

Resources

Highlights

No-code Automation for Azure Boards Using Azure Logic Apps, Power Automate, and Power Virtual Agents

Jason Hand — Mon, 29 Aug 2022 15:45:36 +0000

Recently, I debuted a new video project on the Microsoft Developer YouTube channel I call a "micro workshop".

What is a Micro Workshop?
In short, it's a video tutorial explicitly designed, recorded, and edited for multi-channel distribution in a variety of formats, including short form.

It is an instructor led technical walkthrough, that teaches the audience “how to” solve common real-world challenges while encouraging questions and interactions through bite-sized short form content.

This specific "No Code"Micro workshop teaches you how to build real-life solutions, from start to finish, using Azure Logic Apps, Power Automate, Power Virtual Agents, and Azure DevOps Boards.

If you would like to learn how to create real-world automation to solve common scenarios with Azure Boards, this micro workshop will not only show you what you can build in just a few minutes but inspire you to continue your own "no code" development journey.

Watched beginning to end, the entire micro workshop is just under 45 minutes in length but can be watched in 4 individual parts - each 15 minutes or less, complete with step by step chapter markers and instructions in the video description.

Things You'll Learn Upon completion of this micro-workshop:

Part 1: How to Create a Pre-Populated Azure DevOps Project

(Watch part 1 - 6 minutes)

We'll begin our workshop by signing into (or signing up for) a couple of accounts we'll need.

To follow along with this micro workshop, you'll need the following accounts.

We'll also use the Azure DevOps Project Generator to create sample data that we'll use in parts 2-4. We'll also take a look at what we will automate in the following steps.

🔗 Azure Signup
🔗 Azure DevOps Signup
🔗 Power Virtual Agents

Part 2: How to Create Azure Logic Apps to Auto-create Multiple Work Items

(Watch part 2 - 15 minutes)

Automatically creating multiple Task work items at once when someone creates a new User Story is the goal of Part 2. Logic Apps are where we'll create the "no code" automation to do all the work.

I'll walk you through step by step as we create an Azure Logic App from scratch.

Part 3: How to Create a Scheduled Azure Boards Status Email w/ Azure Logic Apps

(Watch part 3 - 9 minutes)

Azure DevOps has built in functionality to provide a daily email report. However, it's not all that customizable. The good news is that you can create a daily summary email from scratch to provide information out of Azure DevOps exactly the way you want to see it.

Part 4: How to create an Azure Boards Chatbot with Power Virtual Agents

(Watch part 4 - 10 minutes)
The final part of this micro workshop leverages Power Virtual Agents and Power Automate to create a chatbot that interfaces with Azure DevOps. We'll have our bot retrieve all of the work items that are "Assigned to Me" and put a link in chat that takes me straight to the work item.

That's it!

I hope you enjoyed this "micro" workshop. If you have any questions, be sure to drop a comment.

Related Resources

🔗 Get Started with Azure DevOps (Docs)
🔗 Introduction to Azure DevOps (Learn)
🔗 Get Started with Azure DevOps (Learn)
🔗 Overview for Azure Logic Apps (Docs)
🔗 Introduction to Azure Logic Apps (Learn)
🔗 Power Automate vs Logic Apps
🔗 Built-in Connectors in Azure Logic Apps (Docs)
🔗 Overview - Power Virtual Agents (Docs)
🔗 Introduction to Power Virtual Agents (Learn)
🔗 Get Started with Power Automate (Learn)
🔗 Azure Signup
🔗 Azure DevOps Signup
🔗 Power Virtual Agents
🔗 Azure DevOps Project Generator

What's New with Static Web Apps and Durable Functions

Jason Hand — Mon, 14 Dec 2020 22:59:13 +0000

In this episode, Senior Program Manager, (and Former Azure Cloud Advocate) Anthony Chu talks to us about a variety of topics in including Azure Static Web Apps and shows off a couple of fresh demos around Durable Functions, Codespaces, and Custom Handlers.

We'll Also take a peak a Learn module introducing us to Azure Durable Functions and see if we can uncover an even better understanding of the topic covered in the show.

To get things started, Anthony gives us a real quick overview of what Azure Static Web Apps are followed by what the process looks like to deploy them.

Before we wrap up, Anthony and I take a look at a Learn module about Durable Functions. This is a great place to start if you'd like to learn more on the subject.

📺 All Around Azure no.045 - Anthony Chu & Jason Hand

Useful Links

🔗 What is Azure Static Web Apps?
🔗 Quickstart: Building your first static web app
🔗 Fan-out/fan-in scenario in Durable Functions - Cloud backup example
🔗 What are Durable Functions?
🔗 Azure Durable Functions documentation
🔗 [Durable Orchestrations](https://07f.ca/45-url6
🔗 Azure Functions custom handlers (preview)

All Video segment of the Episode

📺 Episode Summary
📺 Episode Demo
📺 Learn Module Overview
📺 Full length unedited video

Learn more about Anthony Chu

🔗 Twitter: @nthychu

Find Us On

📺 Channel 9
📺 Twitch
🔗 Twitter
🔗 YouTube

How to Kick Off Your Career with Custom Vision

Jason Hand — Wed, 11 Nov 2020 16:14:30 +0000

📺 043 - Mikolaj Sienkiewicz 🎙️ Jason Hand

Episode 43 introduces Mikolaj Sienkiewicz, a Microsoft Learn Student Ambassador to the show to discuss Azure's Custom Vision. Mikolaj shares his experience in the Imagine Cup Competition as well as his project Asisto project.

Follow along as Jason and Mikolaj go through Azure's Cognitive Services, specifically the Custom Vision portal (https://www.customvision.ai/). See how easy it is to build powerful AI applications using a variety of cognitive services offered on Azure (https://azure.microsoft.com/en-us/free/students/?WT.mc_id=allaroundazure-devto-cxa).

Resources and links are found below for datasets and tools to follow along with the live demo and perform your very own image recognition.

All Video segments of the Episode:

📺 Episode Summary
📺 Mikolaj Sienkiewicz Interview
📺 Episode Demo
📺 Full length unedited video

Useful Links:

🔗 Microsoft Learn Student Ambassadors Program
🔗 Asisto project
🔗 Imagine Cup Competition
🔗 Azure For Students
🔗 Cognitive Services Showcase Page
🔗 Custom Vision Portal
🔗 Platform to find multiple interesting datasets and AI problems
🔗 Dataset that we use in the demo
🔗 Postman tool
🔗 Cognitive Services
🔗 Learn Module: Classify Images with Custom Vision

Find Us On:

📺 Channel 9
📺 Twitch
🐦 Twitter
🔗 Dev.to
☁️ YouTube

Serverless Azure Functions in .NET

Jason Hand — Mon, 19 Oct 2020 16:54:00 +0000

In this follow-up to episode 008, Jason and Brandon pick up where they left off with a beginner's understanding of serverless, Azure Functions, and .NET.

Now we'll explore alternative ways to trigger our functions.

Useful Links

🔗 Visual Studio Code

🔗 Azure Functions Overview

🔗 Azure Functions Best Practices

🔗 Learn Module: Execute an Azure Function with triggers

Learn more about Brandon Minnick

🔗 @_TheCodeTraveler

🔗 Bio

Find Us On

📺 Twitch

🔗 Twitter

☁️

Microsoft Ignite 2020: All Around Azure Re-cap

Jason Hand — Thu, 15 Oct 2020 22:51:44 +0000

📺 000 - Melody Zacharias 🎙️ Frank Boucher & Jason Hand

In this special episode Frank invites Microsoft RD, Melody Zacharias to the show to discuss higlights of from this year's Microsoft Ignite. Melody spends a little time showing off some of the new functionality in Azure Synapse Analytics and answers questions from the live audience on Twitch.

Later in the episode, Jason Hand joins to host a live Kahoot game pitting Frank and Melody against the audience in a game of Azure trivia.

☁️

All Around Azure Unplugged (Special Episode)

Jason Hand — Thu, 15 Oct 2020 22:51:22 +0000

📺 040 - Jason Hand 🎙️ Frank Boucher

In the 40th episode of All Around Azure, Frank and Jason pause to discuss previous episodes as well as answer questions with the audience.

Useful Links

🔗 Free Azure Account

🔗 General Learn

🔗 Exam

🔗 Sonia's blog

🔗 Learn Module

🔗 Cloud Skills Challenge (until Oct 15)

🔗 Franky's notes

🔗 AAA w/ Erik

🔗 AAA w/ Jay

🔗 Bicep

🔗 Migration tool

🔗 API Mgmt

Learn more about Jason Hand and Frank Boucher

🔗 @jasonhand

🔗 @fboucheros

Find Us On

📺 Twitch

🔗 Twitter

☁️

Demystify Apache Spark with Azure Synapse Analytics

Jason Hand — Mon, 05 Oct 2020 15:18:36 +0000

📺 034 - Adi Polak 🎙️ Jason Hand

Spark is the leading Big Data analytics framework. However, it also has excellent features for Stream processing and machine learning development. Big Data systems are a new reality. Whether we're creating microservices-based solutions or machine learning-based products, we work with data, most often, big data. Apache Spark is your friend if you want to position yourself on the road for success as a Big Data Developer, Data Engineer, or Data Science with hands-on production experience. Come to this session to learn the basics of Apache Spark and how YOU can get started with it.

By attending this session, you will learn how to get started with Apache Spark, it’s capabilities, what Spark pools are and how to take advantage of Spark pools in Synapse Analytics.

Useful Links

🔗 Azure Synapse Analytics

🔗 Learn Module: Spark architecture fundamentals

🔗 Learn Module: Introduction to Python

Learn more about Adi Polak

🔗 @adipolak

Find Us On

📺 Twitch

🔗 Twitter

☁️

Azure Migration Options: Enterprise Tooling in a Cloud World

Jason Hand — Wed, 30 Sep 2020 15:06:52 +0000

📺 032 -Jeramiah Dooley 🎙️ Jason Hand

As a rule, much of the migration efforts that cloud providers focus on are around individual workloads: How do we move that application? How do we repackage that app? How do we move that database? By design, this places a lot of burden on the customer to do whatever work needs to be done in order to get the workload moved. For large and medium enterprises, where there's a significant amount of investment in tools and operational practices, this cost can be too high, and can lead to slowed or even abandoned plans to migrate. What if we could facilitate the migration of the entire platform itself?

Useful Links

🔗 VMware on Azure

🔗 .NET Architecture

Learn more about Jeramiah Dooley

🔗 @jdooley_clt

🔗 Bio

Find Us On

📺 Twitch

🔗 Twitter

☁️

Accessing Key Vault secrets from an Azure Kubernetes Cluster

Jason Hand — Tue, 29 Sep 2020 21:23:16 +0000

📺 031 - Neil Peterson 🎙️ Jason Hand

In this discussion we will talk through and demonstrate a few different ways to access Azure Key Vault secrets from and Azure Kubernetes Service cluster.

Useful Links

🔗 Tutorial: Configure and run the Azure Key Vault provider for the Secrets Store CSI driver on Kubernetes

Docs

🔗 Azure Kubernetes Service (AKS)

🔗 Tutorial: Run applications in Azure Kubernetes Service (AKS)

🔗 Tutorial: Deploy an Azure Kubernetes Service (AKS) cluster

🔗 Kubernetes core concepts for Azure Kubernetes Service (AKS)

Learn

🔗 Introduction to Kubernetes

🔗 Introduction to Kubernetes on Azure

🔗 Introduction to Azure Kubernetes Service

🔗 Azure Kubernetes Service Workshop

🔗 Key Vault Configuration

🔗 AAD Pod Identity

🔗 CSI Storage Driver Support in Azure Kubernetes Service

Learn more about Neil Peterson

🔗 Twitter: @nepeters

Find Us On

📺 Twitch

🔗 Twitter

☁️

Grab Your Wine, Lets Demystify ML And NLP 🍷

Jason Hand — Tue, 29 Sep 2020 21:17:28 +0000

📺 030 - Cassie Breviu 🎙️ Jason Hand

In this episode we will learn how to use Machine Learning to predict wine prices, points, and variety from the text description.

We will process the wine description text with the library "SciKit Learn" to create a Bag-of-Words Logistic Regression Multi-Classification model.

If that didn't make sense then you are in the right place!

If it did make sense join anyway because wine.

After we build our models I will show how we deployed them using Azure functions and Static Web Apps.

Useful Links

Learn

🔗 Build AI solutions with Azure Machine Learning

🔗 Create no-code predictive models with Azure Machine Learning

🔗 Create machine learning models

🔗 Introduction to Azure Machine Learning

Docs:

🔗 What is Azure Machine Learning designer

🔗 Azure Machine Learning documentation

🔗 What is Azure Machine Learning?

🔗 Azure Functions

🔗 Github Repo

🔗 Grab Your Wine. It's Time to Demystify ML and NLP 🍷

🔗 ONNX: No, it's not a Pokemon! Deploy your ONNX model with C# and Azure Functions

Learn more about Cassie Breviu

🔗 @CassieView

🔗 Cassie's Bio

Find Us On

📺 Twitch

🔗 Twitter

☁️

Securing our Serverless Apps on Azure

Jason Hand — Fri, 25 Sep 2020 12:23:08 +0000

📺 029 - Dean Bryen 🎙️ Jason Hand

Let's take a look at how to secure up our Serverless Applications on Azure. We'll deploy a purposely un-secure app, then we'll wear two hats. First we'll put on our black hat and attack the system to see how an attacker may 'hack' in-secure serverless apps. Then we'll flip over tot he blue team and look at fixing the issues and make our app more robust!

Useful Links

🔗 Serverless Security Azure Functions
🔗 WAF on FrontDoor
🔗 Key Vault References in App Settings
🔗 Azure Functions
🔗 GitHub Repo
🔗 Serverless web application on Azure

Learn more about Dean Bryen

🔗 Twitter: @deanbryen

🔗 Bio

Find Us On

📺 Twitch

🔗 Twitter: @jasonhand

☁️

DEV Community: Jason Hand

How will AI tools like ChatGPT and GitHub Copilot affect Developer Relations?

Resources

Highlights

No-code Automation for Azure Boards Using Azure Logic Apps, Power Automate, and Power Virtual Agents

Part 1: How to Create a Pre-Populated Azure DevOps Project

Part 2: How to Create Azure Logic Apps to Auto-create Multiple Work Items

Part 3: How to Create a Scheduled Azure Boards Status Email w/ Azure Logic Apps

Part 4: How to create an Azure Boards Chatbot with Power Virtual Agents

Related Resources

What's New with Static Web Apps and Durable Functions

📺 All Around Azure no.045 - Anthony Chu & Jason Hand

Useful Links

All Video segment of the Episode

Learn more about Anthony Chu

Find Us On

How to Kick Off Your Career with Custom Vision

📺 043 - Mikolaj Sienkiewicz 🎙️ Jason Hand

All Video segments of the Episode:

Useful Links:

Find Us On:

Serverless Azure Functions in .NET

Useful Links

Learn more about Brandon Minnick

Find Us On

Microsoft Ignite 2020: All Around Azure Re-cap

Useful Links

MS Learn links

Learn more about Melody Zacharias

Find Us On

All Around Azure Unplugged (Special Episode)

📺 040 - Jason Hand 🎙️ Frank Boucher

Useful Links

Learn more about Jason Hand and Frank Boucher

Find Us On

Demystify Apache Spark with Azure Synapse Analytics

📺 034 - Adi Polak 🎙️ Jason Hand

Useful Links

Learn more about Adi Polak

Find Us On

Azure Migration Options: Enterprise Tooling in a Cloud World

📺 032 -Jeramiah Dooley 🎙️ Jason Hand

Useful Links

Learn more about Jeramiah Dooley

Find Us On

Accessing Key Vault secrets from an Azure Kubernetes Cluster

📺 031 - Neil Peterson 🎙️ Jason Hand

Useful Links

Docs

Learn

More

Learn more about Neil Peterson

Find Us On

Grab Your Wine, Lets Demystify ML And NLP 🍷

📺 030 - Cassie Breviu 🎙️ Jason Hand

Useful Links

Learn

Docs:

Learn more about Cassie Breviu

Find Us On

Securing our Serverless Apps on Azure

📺 029 - Dean Bryen 🎙️ Jason Hand

Useful Links

Learn more about Dean Bryen

Find Us On