How AI Streamlines Tasks in Business and Software Development DevOps with Ian Harris

By Test Guild
  • Share:
Join the Guild for FREE
Ian Harris TestGuild DevOps Toolchain

About this DevOps Toolchain Episode:

In this episode, we dive deep into the transformative power of artificial intelligence with our guest, Ian Harris. An experienced technology professional, Ian unpacks how AI revolutionizes customer service by understanding sentiment and handling frustrations, ultimately reshaping call center operations. He explores the game-changing impact of AI in software development in DevOps, emphasizing its role in enhancing code reviews and pull requests and automating mundane tasks, freeing up human creativity.

Joe and Ian also delve into the practical side of AI in business, discussing powerful tools like Google's Gemini models for data processing and OpenRouter for comparing AI model responses. The conversation doesn't shy away from the challenges, including data security concerns and the need to keep pace with rapid advancements in AI from major players like Google, Amazon, and OpenAI. Whether you’re keen on understanding AI's role in communication efficiency, or its potential in creating cost-effective, high-quality podcasts, this episode is packed with insights that you won’t want to miss!

Try out SmartBear's Bugsnag for free, today. No credit card required. https://links.testguild.com/bugsnag

TestGuild DevOps Toolchain Exclusive Sponsor

SmartBear’s BugSnag: Get real-time data on real-user experiences – really.

Latency is the silent killer of apps. It’s frustrating for the user, and under the radar for you. It’s easily overlooked by standard error monitoring. But now SmartBear's BugSnag, an all-in-one observability solution, has its own performance monitoring feature: Real User Monitoring.

It detects and reports real-user performance data – in real time – so you can rapidly identify lags. Plus gives you the context to fix them.

Try out SmartBear's Bugsnag for free, today. No credit card required. https://www.bugsnag.com/real-user-monitoring/?utm_source=testguild&utm_medium=sponsored_podcast&utm_campaign=testguild_podcast_smartbear_rum_bugsnag

About Ian Harris

Ian Harris

Ian has spent several decades building technology platforms, managing products. and helping customers pivot in disrupted industries. Of late he's been exploring the opportunities that AI brings to technology and marketing workflows.

Connect with Ian Harris

Rate and Review TestGuild DevOps Toolchain Podcast

Thanks again for listening to the show. If it has helped you in any way, shape or form, please share it using the social media buttons you see on the page. Additionally, reviews for the podcast on iTunes are extremely helpful and greatly appreciated! They do matter in the rankings of the show and I read each and every one of them.

[00:00:01] Get ready to discover some of the most actionable DevOps techniques and tooling, including performance and reliability for some of the world's smartest engineers. Hey, I'm Joe Colantonio, host of the DevOps Toolchain Podcast and my goal is to help you create DevOps toolchain awesomeness.

[00:00:18] Hey, it's been a crazy week, especially in AI, so I'm really excited to have our next guest joining us to talk all about AI, how it affects business, how it affects development, the whole shebang, with Ian Harris, if you don't know, Ian is a technology professional who spent decades building global platforms. He's been in the AI game for a while. He specializes in filling the gaps between tech and businesses, helping business leaders achieve their objectives and translate their desires into things that engineers can actually build. Which I think is even more important nowadays, especially with AI. And we'll see where that goes. But a lot of good information here. You don't want to miss it. Check it out.

[00:00:58] Hey, if you're app is slow, it could be worse than an error. It could be frustrating. And one thing I've learned over my 25 years in industry is that frustrated users don't last long. But since slow performance isn't sudden, it's hard for standard error monitoring tools to catch. That's why I think you should check out BugSnag, an all-in-one observability solution that has a way to automatically watch for these issues real user monitoring. It checks and reports real-user performance data in real time so you can quickly identify lags. Plus, you can get the context of where the lags are and how to fix them. Don't rely on frustrated user feedback. Find out for yourself. Go to bugsnag.com and try it for free. No credit card required. Check it out. Let me know what you think.

[00:01:49] Joe Colantonio Hey, Ian. Welcome to The Guild.

[00:01:53] Ian Harris Joe, thanks so much for having me today. It's such a pleasure to be on the show.

[00:01:56] Joe Colantonio Awesome to have you. I guess before we get into it, is there anything in your bio that I missed that you want The Guild to know more about?

[00:02:02] Ian Harris Look, I think you've done me a great compliment, to be honest, it sounds pretty good. I've been programing computers in one way or another for a long time, and it's always been fascinating to me to be able to get these inanimate objects to do your bidding. I think it's really exciting. And for me, it's always had some form of media or communication. That's the kind of areas that I've been interested in exploring. And I think we start to talk about AI and its implications for the business world. I think we're right in smack bang in the middle of that. It's all about communication. It's all about being able to take one form of information and transcode it, transform it into something else and be able to do things more efficiently. And so it's a really exciting time to be in the software industry.

[00:02:43] Joe Colantonio Absolutely. There's been rumblings of AI for a while. And then, finally, the last maybe a few years ago, they rolled out a really good version of ChatGPT. And now just this week, they rolled out, I think, a new Google Gemini and a new chat GPT version that does multimodal type of AI. So just curious to get your take before we dive in, how do you explain what AI is? Let's set the bar and then dive into it from there.

[00:03:07] Ian Harris Well, we've been using AI in its many forms for a long time. So for example, whenever we've used if you have a camera at home that detects people as they walk past or identifies the fact that there's a dog that's walking past, or it can name the people as they come through the door, that's using some form of AI to do that. If you call in an Alexa or a Google Home equivalent, again, they're using AI to turn the vocals. The turns of words that you speak into actual words that the system can understand that's using AI, but it's always seemed a bit of a kind of utilitarian tool to do something, to transform something. Maybe a translation has been one of the good examples of where I had been doing some really good stuff, but what just kind of thrusts this whole thing into the mainstream is not just using it as a tool to serve some other purpose, but actually where we are starting to communicate with the AI directly and ask it questions almost in a personal manner. And so last year with ChatGPT, we had watched what they were doing for some time. They had a model that created some images that looked pretty cool, had to be on the waitlist for that. So it was kind of hard to get into. They had some models that did some with words and that was kind of interesting, but not really particularly useful until they actually transform that and turn it into a bit of a test case to say, look, let's see if we can talk to this and have a conversation with it. Maybe we'll get some uses around it. Maybe that'll be something interesting that people want to play with. Maybe we'll get some good feedback. It'll help with the training. And strangely enough, that's when ChatGPT turned from being just a tool that could complete some sentences basically into something that was far more than that. You could ask questions, it could respond to you, it could write simple poetry. It could answer questions. It had the history of the entire internet in its back pocket to be able to refer to and answer questions and string things together. And I kind of inspired people. Something about that. It had taken that leap from being some sort of tool, some sort of algorithm, and it kind of turned it into something personal, turned it into something that the humans felt like they were interacting with some sentient being almost. This is amazing, this transformation that AI had gone through. And so that was really the start of this AI revolution that we're part of, is that ability to communicate with an AI and have it kind of respond to you, like very much like a human would with actual words for the first time, and not just taking some words and turning into a different language, which is very useful and very cool, but actually rewording things and summarizing and giving you new ideas and creating things from scratch, which was really quite a new and different way of looking at what AI can do for us.

[00:05:48] Joe Colantonio Absolutely. So speaking of that, I like to dive into a little later on about the multimodal aspects of the latest releases and what that can actually do for you now, but you mentioned communication twice about multiple times, more than twice. I'm curious to know, as a business leaders and developers, do we need to change how we communicate with AI? Is there anything we need to know how to get the best out of AI the way we communicate. Any tips around that?

[00:06:11] Ian Harris Look. It's interesting. I've been doing some work for a little while now in terms of trying to use AI to help bridge a gap in marketing, as it happens. So we've been creating podcasts using AI, so using these models to actually create the script of a podcast and then using a different AI model again to actually voice that. And we're now at a point where we can do that, where the quality is actually really quite good. It's really quite believable and engaging. And that idea of being able to use an existing marketing use case, for example, but applied in ways that are now much more doable at a different price point that would have been done before. I mean, lots of people listen to your podcast. It's a very, very successful show. But you of any people know the effort involved in recording and editing and mixing and getting it out there is it's a big effort. And so whilst it's a very valuable thing because you're reaching a particular audience to love to hear what you have to say, for many companies, that ability is really restricted because it's actually quite an expensive task to get a podcast out. What we wanted to explore with those podcasts, which is the company that we've been building this particular set of AI for, is can we create a podcast that is maybe not quite human, but both very believable and engaging, but at a price point that is not is if we have to have a voice talent and we have to edit it, and we have to create a script and the associated time and energy around that, could we do it as something that was, I don't know, a 10th of the price, a 20th of the price of what it would normally do to take humans and do the same sort of task. And that really is a great example of where AI is. In fact, the best way of describing it is a force multiplier. And we'll get to talk about software in a minute. But if you can take an ordinary bit of effort and get 10 or 20 times the energy, the capacity, the words, the whatever it is you're looking at from the same effort, then all of a sudden you've actually got this ability to do so many more things at a different price point, or a different niche or a different area. That was just not feasible before. But now all of a sudden, these opportunities in-house have now sprung into life.

[00:08:13] Joe Colantonio And this is a great, great just made me think of a point. I think the point you're building on top of an AI model, I assume, and I'm sure a lot of business users and developers are thinking of doing the same. How do you know? I don't know if this makes sense, but how do you know what to build without knowing what's coming down the pike? For example, I know like something called rabbit R1 came out I think a few months ago and it does translation and seem really cool. But then they just released ChatGPT for all that replaces bunch of the things that just going to be in your phone now. All that effort that went into that building on top of that, coming up with that idea, I mean, it must have been crazy amounts of money gone, because obviously it's going to be so available to everyone now just in their pocket. I think this model's free for people. How do you know, as someone that's actually done this, how do you know what to build and how to make it so or you won't be replaced? Or is that just something that's out of your hands?

[00:09:04] Ian Harris Oh, look, it's actually very hard right now because there's so much activity in this space across a whole bunch of different companies. You've got Anthropic that in making these amazing models now, you've got Google that are making amazing models now, Amazon is having a crack in at least in the code space as well. And they've got some business chat bots that they're working on as well. And of course OpenAI with ChatGPT is the kind of leader in this space. There's so much energy and activity that's happening that it's very difficult to say, I've got it now, and I've got this great idea and it's not going to be supplanted by something else. If you can imagine with the ChatGPT model, it's just come out, the open AI model, it's just come out in the last week where you're not only sending it text now, you're actually sending it audio and getting audio back directly. It's effectively replaced 2 or 3 other models in one go. Now, you don't need to actually record the audio, send it to an AI to actually turn that into text. Then take the text and send it to the AI to ask a question. Then get the text back and then turn that back into audio. All that stuff that you might have coded up and got going in. I know some organizations have had a go at that. It's all been completely replaced by just using a new model now. So you're right, it's a real challenge. What do you do? Where do you spend time? Is it worth doing anything at all? Should I just wait longer? But of course, business never stand still. We have a great opportunity with all sorts of tools in front of us to do new things and creative things, and generally speaking, everything will just get better and faster and more available. And so whatever you build, I think really what happens is that the opportunities from that will grow over time rather than actually being placed itself. Those ideas as humans that we need to actually apply these things too, are always going to be useful and interesting.

[00:10:42] Joe Colantonio Absolutely. With the newer models that came out is any ideas you give to business owners? I know like it does translations really well. You just can have it open, you can speak whatever. So English and speak to someone that's Italian and it's on all the time. It's able to listen. I saw someone who acts like you could trade in on your business solution software and it can then create tutorials on the fly for your web videos and it's crazy. Any ideas on how someone can approach us if they want to get into it and grow with it?

[00:11:11] Ian Harris Okay, I think one of the most exciting parts of that is that because you can now send the audio directly to the model and get audio back directly, the latency has improved so much. You've probably seen a few of these demos in the past where cool, you recorded some audio and you send it to the chat thing and it comes back to you and it's got like 2 or 3 seconds of lag behind it and you're like, yeah, I mean, it's cool, but it's kind of painful at the same time. Now it's pretty close to human level responses to things. It's that quick where you feel like this smidge of the lag, but even that will go over time. It's just a smidge. But people kind of pause and I'm thinking as I'm talking and everyone takes a moment to respond. And I don't know if you've ever talked to a call center and often found that, yeah, they're not really paying attention to me. They're paying attention to something else. And if you can imagine that even as humans, we're not that great of it. But now we're at a point where we can actually have a real time voice conversation with an AI that knows everything that you want to know. I think that's one of the biggest opportunity that's coming up now, is that for you to make some information tutorials, a help desk, some information about a whole bunch of text you have within your organization, or some knowledge that's been built in your organization and make that available for query over a very human interface. Talking is just really quite phenomenal because as you probably work in lots of organizations, Joe, where if you want to find something out, you got to find the right person to talk to. And you find John and he says, you should talk to Jenny. She knows exactly about exactly that thing. You find Jenny and you have it. You set up a meeting with her, and you chat to her about the things she talks you through, how something works. Right now I know this thing. Imagine if all that corporate knowledge was now condensed into an AI, that you didn't have to track down the right person or find the right information in a file somewhere. But you just talked and said, oh, can you tell me about the latest way of doing QA in our company? How can we do that? How do we do that at the moment? What's the right server that I need to talk to? What's the software? Can you talk me through what we're doing at the moment and or even what if you had a production system and you wanted to say, tell me about what the load is on the service at the moment. Is there anything been happening over the last day? Do I need to know anything that's happened over the weekend? Being able to query things by voice, it's really quite an astonishing change in the way we interacting with machines now.

[00:13:32] Joe Colantonio Yeah, that's a great point about the performance. And if you think about that, that's a great use case for developers listening. And also what I thought was interesting I don't know if you caught this, maybe I didn't hear correctly is that, it can understand sentiment. So if you have a tech support and someone's getting irate I mean it could probably tell that and then maybe switch you to a live person or something like that. It seems like a lot of use cases you can get out of it.

[00:13:54] Ian Harris Yeah. Not just understanding things and then being able to respond, but going, hang on. This person's getting a bit frustrated, I might take a different course of action that they wouldn't have done otherwise, which is really great because and again, we often want to talk with expertise, real humans that know a particular topic and be able to help us out. But really, we've all been on text chats with service lines to help out with our laptop or our mobile phone, or a bill for our phone company. And the time we spend waiting to talk to someone. And we finally get to talk to someone. And either they were all kind of chatting with different people at the same time, and they respond to that chat. Or even if we're talking to them as a human, it's taken so long to have that conversation. I would be quite happy if I could just talk straight away to an AI that had all the answers to me, and I didn't have to wait, and I knew everything, and it was empowered to help me, and it was protected from doing silly things, but it was genuinely useful to me. I think that would be a big change. I think the way we do call centers is going to change fundamentally. I'm sure we'll still have really expert humans that can solve really complex problems that are unique. But for the majority of the cases, when we interact with companies, we really just want the same answers everybody else is asking. It's just the details might be specific to us, but really, we're not special. And so we can get that information that answers much more quickly, much more conveniently than we have been up until this point. So I think that's a really exciting change. And of course, the change from a businesses, humans are really expensive. And if you are a small business having a call center, you often have to outsource that. It actually is very expensive. If you don't have a large call center and then you have to do training and set up what? All that setup and configuration and details. It's really hard for small businesses, especially to have a call center all of their own. They have to rely on other people to do that for them. We'll see. And I think now the startup of these, I want to switch on a call center. Here's the questions I want to able to answer. And he's all my information. And we can spin up our call center for very little money and a very different use case for a very small niche audience that is being able to respond in ways that, up until now, would have taken hundreds of thousands, millions of dollars of set up and people and communications gear and all sorts of other infrastructure in order to be able to do that. I think the opportunities as well will grow from that.

[00:16:11] Joe Colantonio Yeah, I agree. I mean, I get excited hearing that. However, other people hearing it may be like, oh, wait a minute, that doesn't sound too good. I work at a call center like my whole family works in call centers. I don't want to get philosophical here, but like, where do you see it going? I know before people say all this happens all the time, you get a new technology, replace it, and then we have newer jobs that happen. Have you thought about how much is AI going to disrupt things? And if where does humans stand? Do you actually see replacing humans completely? And then we live in a utopia. Like how do you see it?

[00:16:42] Ian Harris That's a really good question, Joe. I think everything is going to change. I think we're going to look back in two years and the changes will be unimaginable, that we're looking back to see where things have gone. Technology's funny. Things seem to change little by little. And then all at once. And it's a great phrase, right? Because they sort of things seem to creep up on you, and then everything is different as you look around. But for us humans, as we're looking at all these fundamental changes in the way we interact with machines and what we can do with AI, we always still have complex human level problems that only humans are really good at solving. We will never really get rid of the humans in all these sorts of problems, and we'll always need expertise in all these different areas. And someone's going to create the questions and the content and the stuff in the first place, and the business ideas. So humans will always have a part in these sort of enterprises. But what will happen, I think, with AI is that a lot of the mundane stuff, a lot of the stuff that is not creative, that is repetitive and is boring, will go away and free us to do more interesting tasks. And this is always the way with technology. We have these opportunities that supercharge us doing something, and then it frees us from doing stuff that we don't really, really want to do, and that allows us to take on more interesting tasks. I think it's a positive way of looking at the way AI is going to revolutionize business.

[00:18:02] Joe Colantonio Absolutely. So we spoke a lot about from maybe a customer's perspective or a business perspective, just to know as you're developing software on your own. Do you leverage AI and with development, for example, like these newer ChatGPT models? Once again, I think someone was using it for code reviews. You can just read your code review it, can look at it and tell you what's wrong with it. Are there any use cases you've seen from a developer's perspective you think could help folks.

[00:18:24] Ian Harris This one create opportunities there with AI again, humans are terrible at detail of large quantities of text as you're reading through code and you're trying to debug it and understand what's going on, you're trying to understand the meaning in the detail or someone else has put into this code. Having an AI be able to look at software and explain it to you is a fundamental shift in the way we interact with software. It's often the case that the person that is originally wrote the code that you're looking at is not there anymore, for whatever reason, and have something else explain it to you in a way that makes sense, that it's relevant to the topic that you're talking about, understands the context of the code as a whole, understands the particular language, and then can say, but by the way, there's a few errors in the code, which under these circumstances, you'll have these bugs that you might not have realized before, be able to spot those edge cases that humans are we really struggle with discovering these things ourselves. They'll be able to spot things that we can't see, and they'll also be able to refactor and be able to reorganize the code, make it more obvious to see what's going on, even from a human perspective. Being able to review the code like that. I've heard a lot of good work going on at the moment with AI helping out with pull requests. So being able to find the changes of software that's going into an organization and and being merged into it, into the main branch, being able to describe that in a way that is clear and understandable and repeatable and contains all the information you need. Again, humans are terrible. I have to add all this stuff in, I have to add this detail, and I have to describe something I've forgotten to even reviewing pull requests I've seen AI do fantastic. Oh, you forgot to describe exactly what the use cases is then. What description doesn't actually match that code change in fact, or your code change does more than what you've described in the pull request. Fantastic. Well, this level of detail and being able to do basically do a better job at software because we've now got an AI that really deeply understands through millions of lines of code. The software itself, probably better than any human would be able to do in terms of that level of speed and accuracy and being able to get that information out of that code itself. So there are a couple of examples. But I'm also saying now we're starting to look at ways of describing software in not so much architecture diagrams but documents. So taking product documents that describe how software should function and what its limits do, and then being able to read the document and say, okay, well, that's what you want to do, then you're going to need these sort of software modules, and they're going to need to interact in these sort of ways. You're going to need a whole bunch of APIs that you can go and get it to write the APIs for you. Fantastic. And then you can say, okay, I've got these APIs, but we need to store the data, obviously, and it can create the database schema for you. Great. Okay. That's another mundane task that slog through it and work out what the different fields are going to be or AI can do it for you and do it much more accurately than I would, and I'm going to get there, but I'm going to make mistakes along the way. And again, it's back to that force multiplier. Humans have created all this stuff on average overall eventually, but if I can do it in 30% of the time and I haven't had to sit there, and even if I've got the idea and I know the fields and I know what to do, I've got to sit there and type it. And no matter how good a typist you are, if you're thinking and typing at the same time, it's slow. It just takes time. And then you got to review it and you got someone else to check it. Meanwhile, the AI just done it. You're done and you move on to the next thing. And so you see that developers will get faster. And this is great, because the fun part of software is the thinking, the designing, the overarching architecture. It hangs together, the typing it up, the code, the writing, the code bit. That's actually not that fun. The fun bit is getting it, working it, debugging it, and seeing it all connect together and seeing the whole thing function as a whole. That's the exciting part of software. And so I think it's going to make software more fun, and it's going to make it faster. So you'll be able to get much more done with less people.

[00:22:11] Joe Colantonio That's a great point. I used to work on an automation codebase, and 8 sprint teams would use it, and they put me on every code review that they checked in, and then they put me on a blocker because I was able to get to all day, I missed something. Sometimes teams were afraid to make changes because they didn't want to affect another code, so they create a new method that's basically the same method that was already there. So like you said, it's not replacing anyone. It's making things. It's taking where almost the minutia and freeing you up to do, like you said, the fun thing. I love that perspective. Do you see a lot of developers once they get into it that they start embracing that and not be afraid?

[00:22:44] Ian Harris Yeah, I think most developers. I've seen work for some pretty big teams, and I think there's a few people that still the do genuinely like that the act of typing and coding and thinking through it step by step and thinking slowly, which is good. That's good. It's good to have people like that. But most people like I'd say probably 80, 90% go, oh my goodness. Like, I didn't have to write out this full loop anymore. I could just say boof and it's done and I mean, we all know it has to get done. It has to cycle through this thing. I just describe it in a few words. It typed it up for me. Now the using GitHub Copilot or something like that, they can just describe what they want. It dumps 20 lines of code. I'm looking at it. That seems to work. That seems to me right sort of thing. That's good. Okay, next I'm onto the next thing and I'm really chunky through. And I'm actually really feeling like I'm actually moving through the software and actually getting things done, which that velocity is relatively unheard of in terms of production. That sort of velocity is relatively unheard of in terms of performance improvement in the workplace. With this, with AI and software development, we're seeing 20%, 30% improvement in the amount of stuff that gets done, in the amount of time that humans normally take. And that's an incredible improvement. Imagine that. Imagine that if we were really looking at any other industry where you went, oh, we've got a 30% improvement today just by switching this thing on. And you don't have to do as much boring things. Oh my goodness. Like you just fall over from an economic perspective from any. I can't even imagine the any other area where any of the manufacturing or any other business where you'd find that level of improvement with such just switching on and off on these tools at $20 a month. Phenomenal. What value is that? That's amazing.

[00:24:26] Joe Colantonio Absolutely. When you work with businesses, how does one know which model to use? It's almost like ChatGPTs like a platform, not like AWS. You just build on top of it and as it updates, you get all those benefits. The cost goes down, the whole shebang. Or how do you see development going forward then with these models?

[00:24:44] Ian Harris Look, that's a really good question, Joe. Like when you're getting into AI, it's hard to know who do I trust? What's good at what. There's all these open source model things out there that you can do stuff with. You've had all these different companies we've talked about three. There's a whole bunch of others that no one's ever heard of. I'm sure, peddling their particular models as well. We've got new companies that are coming into it, like IBM is looking at it as well. Where do I go? How do I choose? What do I do? I mean, you can't go wrong with some of the big names companies, of course. But in terms of those models, it really is a bit of a trial and error in terms of working at what model is good at the use case that you want to do things with. There's some languages, for example, that GitHub Copilot doesn't handle very well, like its code or some of the mobile languages it's not so good at. So you have to make a choice based on languages for that sort of thing for creative writing. For one way, we do a lot of experimentation with creating scripts and editing words for the podcast that we're creating. We found that still GPT 4 , produces the best results for us for the sort of thing we're doing. But we also found that for particular use cases Anthropic Claude version three that they've that it now has it's very good on the creative side. It's much more kind of like, telling a fantasy story where you're trying to be a little more creative or a little more flamboyant. That's very good at that kind of thing. And now we've got Google's Gemini models, which again, the particular use case for them is being able to throw millions of tokens at them. So the tokens are in the way that these AI engines work is they really just work with numbers. So we take our words, we chop them up into little pieces, turn them into numbers, send them to the model and send these numbers back, which it turns back into words. Those numbers, the tokens that we've broken these words up into. Now, the Gemini models from Google can handle millions of tokens. And this has been one of the frustrating things about AI up to this point is that you can hand it some thousands hundreds of tokens and it would be able to do something useful with that, but it certainly can't write more than a few paragraphs and handle a whole ton of information. Now we can throw an entire book, not even a book. You can write a whole bunch of books, in fact, we can throw an entire video of data and then get it to work out what's going on in the video. We are now starting to get not just being able to interact with these AIs, but also for them to be able to process large amounts of information as well. How do you choose? There's a lot of aggregators out there. One of them is called Open Router. So Open Router is one of these ones where you connect it with using an API or it's got a playground, a chat interface as well. And you can choose a whole bunch of different models. You can have a play and see how they respond. And you can actually send it to several different models at once. You go, okay, I want to write a podcast script about, AI engines, for example, and you can send it to different engines and it would give you a response back and you can start to compare them and see what is best for your particular use case. So trying those different things, finding services that give you the opportunity of being able to explore those different engines. It's a case of having a look and seeing what works for you in your particular use case.

[00:27:39] Joe Colantonio Absolutely. We talked a lot about the positive side of AI, the very AI positive. But is there an ugly side? You're a developer, is there anything you need to know, like, well, obviously security probably comes to mind. Is anything like any ugly sides of AI you want to reveal?

[00:27:54] Ian Harris Yeah, there's a few areas you need to be careful of. So in ChatGPT, for example, if you're using the standard ChatGPT interface, then the things you shove in there will be used by OpenAI to train their models to get better at what they're doing, because they look at your response and you go, oh, no, but I want this instead. But I want this when you look at that and go, okay, well, what you really made at the beginning, there was something different. And so we'll help the model get better at things over time. Now, that's good for OpenAI and for the wider community. But it doesn't mean that the stuff you put in there becomes part of what that model is. It really does become part of the training set. And so for organizations that have a concern about the security of their code, the confidentiality of their code or the information that they're processing that particular interface is not appropriate for them. And so you need to make sure that the API you're using handles the data that you're putting into it in a secure way. So that's one consideration. And the other is from a more human perspective, we are using a lot of humans work in order to be able to train these models to do what they can do. And OpenAI is now going to deal with Stack Overflow, for example. They've been taking all the information of Stack Overflow, if you don't know, it's a website where you can put a question and people will help you with an answer to a typically a programing question, how do I do this in this language? Or how do I do this in this particular database, for example. So now Open AI has a deal with StackOverflow to take all that information and make it part of the model. But there's a number of people that are frustrated by that because they've contributed to this website over a number of years. They've done it as a service to the other developer community. And I kind of feel that's being used for a purpose that they really hadn't intended. Joe, if I was helping you out with a, I don't think I'd be helping out with your QA and test automation. You obviously know all about that. But if I was and then suddenly that information was then being provided to an AI model, and that really wasn't my intention. I could see how that could be kind of frustrating for me as well.

[00:29:51] Joe Colantonio Absolutely. Good points there. Okay, Ian. Before we go, is there one piece of actionable advice you can give to someone to help them with their AI DevOps efforts, and what's the best way to find or contact you?

[00:30:01] Ian Harris I think one of the best things you could be doing is look at the areas in your organization that contain mundane, repetitive, time consuming tasks. Look at the opportunity of using AI to automate that part of your process. It doesn't have to be the whole process. It doesn't have to be creating new content or scripts or even mocking materials, even in programing, or if there's reviews or things that are getting things to getting stuck in your organization that are just slow and inefficient, it's quite likely there's a way you could use AI to smooth that over and make that faster and more efficient for the rest of the organization, and take those blockages out of your development process or, all the processes that you have in your business. I think looking around yourself and seeing what people are just not excited about doing and things that are slowing things down, there's probably a use case there for AI and exploring that I think would be a great idea. Trying to get hold of me. I'm looking out on a website www.PulsePodcast.com. That's the project I'm working on at the moment, doing lots of AI stuff, and you can see some great examples of AI voicing and AI script generation. So we have some of these models are actually taking entire keyword research of news and then filtering that, basically using AI to actually filter those articles and find the ones that are actually really about the topic that we're interested in. Because keyword research can turn up a whole bunch of stuff out of that, and then actually generating scripts and generating voices and generating music. So if you want to see some of that stuff, have a look at our website. And it's a great example of what AI can do for you in the business.

[00:31:31] Remember, latency is the silent killer of your app. Don't rely on frustrated user feedback. You can know exactly what's happening and how to fix it with BugSnag from SmartBear. See it for yourself. Go to BugSnag.com and try it for free. No credit card is required. Check it out. Let me know what you think.

[00:31:52] And for links of everything of value we covered in this DevOps Toolchain Show. Head on over to Testguild.com/p147. And while you're there make sure to click on the Smart Bear link and learn all about Smart Bear's awesome solutions to give you the visibility you need to deliver great software that's Smartbear.com. That's it for this episode of the DevOps Toolchain Show. I'm Joe. My mission is to help you succeed in creating end-to-end full-stack DevOps Toolchain Awesomeness. As always, test everything and keep the good. Cheers.

[00:32:26] Hey, thanks again for listening. If you're not already part of our awesome community of 27,000 of the smartest testers, DevOps, and automation professionals in the world, we'd love to have you join the FAM at Testguild.com and if you're in the DevOps automation software testing space or you're a test tool provider and want to offer real-world value that can improve the skills or solve a problem for the Guild community. I love to hear from you head on over to testguild.info And let's make it happen.

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}
Harpreet Singh-TestGuild_DevOps-Toolchain

DevOps Crime Scenes: Using AI-Driven Failure Diagnostics with Harpreet Singh

Posted on 06/12/2024

About this DevOps Toolchain Episode: Today, we have a special guest, Harpreet Singh, ...

A podcast banner featuring a host for the "testguild devops news show" discussing weekly topics on devops, automation, performance, security, and testing.

AI-Powered Salesforce Testing, Shocking Agile Failure Rates, and More TGNS124

Posted on 06/10/2024

About This Episode: What automation tool just announced a new AI-driven solution for ...

Paul Grizzaffi TestGuild Automation Feature

Degrees of Automation with Paul Grizzaffi

Posted on 06/09/2024

About This Episode:In this episode, Paul Grizzaffi, a senior principal automation architect at ...