Customer Roadshow

[TITLE: Your Business optimised. Telstra logo and Telstra Purple logos 

Your Business optimised Customer Roadshow 

Increase Productivity with Microsoft AI] 

JAMES NATION-INGLE: My name is James. 

I look after the go-to-market for the Microsoft practice here at Telstra and Telstra Purple. 

I'm joined today by Shane, who's one of the chief architects at Microsoft, Glenn, who's our Head of Workplace and Digital Technologies Strategy at Telstra and Telstra Purple, and Dom, who's the Head of Sales for our new Microsoft Practice. 

[SLIDE: Our panel] 

We're gonna have some fun today. As I said, we're gonna keep this relatively light. The session is about AI. Now, everyone's having AI hammered down on top of them right now. This is gonna be no different. We're gonna talk about a few different AI aspects. 

What I find about AI these days is when we're piloting things and when we're doing things at Telstra, I'm finding this AI stuff that we're doing that is present to me as a user all the time and stuff that's happening in the background that I'm like, 'Oh, actually there might be some AI in the back of that.' And these two sort of paradigms are really interesting as a user because I actively see some things that help me and then I passively see some things where I go, 'hmm, maybe something's happening in the background there that's helping me and I don't know it. And that's OK, and I'm starting to accept that as a user. 

What I'm gonna do though now is I'm gonna set up the demo later. Like every good magic trick, I'm gonna set up the demo later. So, where's my M365 admins? Got any techs here? I know you've been in that world. 

I'm gonna show you a couple of things. Just to set the scene, can I get the screen changed over here? Because you're gonna see some practical demos as we go at the end of the session. This is my PC. Hopefully, everyone's got one of these. 

Who's seen this picture before? Yeah, there's my admins. Awesome. 

So, what I wanna do is show you some of the things or set up some of the things that are AI in the back of some Microsoft technologies at the start. And I'll show you how I'm practically gonna use them later on. I want you to see me as a user. I want you to see that in my licenses and apps. So, for those of you that don't know, this is the M365 admin portal, and these are the licenses that I've got that allow me to do things. 

I want you to see a very specific and special license so that I can set this thing up properly, and that is the Teams premium license. So, this is one of the first licenses in Microsoft that had AI in it. So, we can accept that I'm a user, and I have some AI technologies in the back. 

Now, what I'm gonna do throughout this session is - because these sessions are not being recorded up here - I'm gonna use my little Polly mic. I'm gonna set up a Teams meeting to run in the background of this session while everyone's talking. I'm gonna use this mic and a Teams meeting, so I'm gonna grab. So, you can see my little Teams meeting. I'm gonna go meet now and start the meeting. I'm gonna hit record and get the transcript as these guys are starting to talk. And later on, I'm gonna use the transcript and AI to do some fun things, and you'll see a practical application of some of the bits that we're gonna talk about today. 

Now, without further ado, can I get you to switch that back over? Without further ado, I'm gonna hand over to Shane from Microsoft, and he's gonna talk to you about a bit of an update on what Microsoft's doing with AI. 

SHANE BALDACCHINO: Thank you, James. What an intro. Just by show of hands here, who uses Azure OpenAI? We've got a few people. Awesome. Who's had a bit of a play with the SDKs and things of that nature? Are there any builders in this room? Hands. There's another one there. Does anyone like my t-shirt here today? Does anyone know who this is? We're off to a start. So, a bit of an update. 

[SLIDE: Microsoft Cloud in the era of AI] 

So, look, as we enter this era of AI, there are three critical things in the Microsoft Cloud that enables, I guess, AI for any organization here. Today I want to step through these three aspects here. 

Firstly, look, AI, you've probably seen, as James just mentioned before, it is being sprinkled all over our product sets. You've probably seen the memes out there - AI on this, AI on that. But it is about accelerating productivity and creativity of people here. I'm moving across here to be on frame. There we go. Small stage. 

Second, look, we're delivering tools and services that you need to build your own intelligent apps. So, the ability to build your own copilots, the tooling, to not only leverage our copilot but to be able to build your own. 

JAMES NATION-INGLE: Whispers - what's a copilot? 

SHANE BALDACCHINO: What's a copilot? 

Look, a copilot, realistically, I think in layman's terms, James, is an AI helper to help you do more, to remove a lot of that heavy lifting in your day-to-day workflow here. So, I think I was having a chat with Glenn earlier before. I do a lot of programming in C++ of all these languages, and I think I'm pretty adept at that. 

[SLIDE: A copilot for every Microsoft Cloud experience] 

But by leveraging tools like GitHub Copilot, I'm often learning new tricks and better ways to help me to be more effective. 

So, a bit of a segue into GitHub Copilot, we announced this about a year ago, and the innovation has just been incredible. Last week, I was on stage somewhere in Sydney and was talking about GitHub Copilot of all things there, and over 40% of code checked in to GitHub in 2023 has been AI created, which is an amazing statistic here. So, it is about enabling you to ultimately change that value proposition from low to high to allow your people, your staff to be more productive. It's enabling people to unlock that creativity, productivity from things like GitHub Copilot through to the more office-based applications to be more productive and creative. These days there are copilots for things like dynamics, so enabling you to increase your productivity in your organization. 

[SLIDE: Empowering every organisation with Azure AI] 

Quickly, I wanna just talk about how these intelligent apps are enabling organizations to do more here. So, as you can see here, look, we've got Azure OpenAI, Cognitive Services, Machine Learning, Azure AI infrastructure. And I kind of like to think of this stuff as a bit of a layer cake. 

There is a lot of buzz today around things like OpenAI. Azure OpenAI is just one of the AI services that we offer. But look, there are other services like speech to text, vision services. OpenAI is amazing in so many ways. These large language models are not just great at doing one thing. They're great at doing lots of things. 

However, for things that are more specific here, for example, you might be doing image recognition. There are higher-order cognitive services that may be better suited for your needs here. And again when I say the words like TensorFlow or Caffe and PyTorch, etc, like that, then you've got offerings. Like Azure machine learning to be able to leverage those models directly. 

[SLIDE: Azure OpenAI service] 

So, look, in summary here, Azure is incredibly comprehensive in this space here. But look, in spending time with organizations around Australia and New Zealand, we typically see four distinct use cases. And I'll get into a few specifics. 

But look, content generation. I was gonna make this a little bit personal. I was sitting on a panel of a hackathon just recently, and I don't know about you, but if you go out for dinner or something like that, you're probably gonna look at the reviews of a restaurant. Well, if you get a bad review, particularly when there's no reply, it doesn't look great. So, the winners of this hackathon created AI solution that would generate content for social media responses, things on like Yelp and so on, Google, to automatically respond in the tonality of the organization against negative reviews. 

So, that's kind of like one example there. On the summarization piece here, you think about law firms, New Zealand based LawVu as an example, they create software for lawyers and in-house legal teams. 

You might have... A lot of us here have probably bought houses. Imagine if you've got a section 32. Do you actually read it? Do you wanna know what the summary is? Well, OpenAI is a great example of removing that heavy lifting to be able to get to that summary really quick. Cogeneration. 

As I mentioned before, I'm writing some C++, you might be doing something else. It could be creating that arm template to deploy your stack on Azure or coding in general here. And I think the last one is really interesting, that semantic search side of things. How do you extract that needle in a haystack? You may have things like Azure Cognitive Search. You may have an Elasticsearch stack out there, which is great for in-memory search, but how do you extract value in that? So, semantic search is fast becoming that new paradigm. 

We've got a logo here of CarMax. And I love this example here. And you can look at this on the Microsoft website. So, CarMax, if you're not familiar, they're a US-based secondhand car dealership where you can purchase a car online, and I think you can return it in 30 days or something like that. So, with thousands of reviews on their website, how do you get the summary of what people think of these cars here? So, they're leveraging Azure OpenAI to provide summarized views of specific models, what the pros, what the cons are and so on. Again, this is just one use case here. There's a few more logos on screen here too, which you can read off the Microsoft website. 

[SLIDE: Microsoft Cloud runs on trust] 

So, look, a lot of the questions I get asked often is, hey, what's the difference between Azure OpenAI and OpenAI? I can leverage GPT-3, 5, Turbo or GPT-4 or other models. What's the difference? And I think it comes back to - we've got this thing here, Microsoft Cloud runs on trust. And that is the truth. Everything we do is built on trust. Security is job zero here, and responsible AI is a core ethic for us here at Microsoft. 

Pop the term 'responsible AI' into your favorite search engine, and you'll see Microsoft will appear right up at the top there. You can read our responsible AI principles. But I think the summary, the main difference between OpenAI, the native versus Azure OpenAI, it's the same large language models running on our end. It's not altered in any way. However, we're running a copy of OpenAI's model on your behalf in a private way. You can access it via a private endpoint, but most importantly, we will not leverage or store the prompts - so the questions that you ask - to further train OpenAI's model. So, if you pop in, maybe, Samsung and OpenAI into your favorite search engine, you'll probably see some examples of where things have kind of gone wrong. So, look, protecting your data is top priority for us. 

Now, I did a senior demo just before James. You had the OpenAI portal up. And, look, you've got the ability with Azure OpenAI to easily ingest your data. So, not only can you leverage these large language models, you can leverage these large language models with the context of your data in your organization, be it from flat files, in-memory search engines, and so on. 

So, look, this has just been a bit of a whirlwind here.  

[SLIDE: AI Innovation] 

But look, there is a lot happening in the Microsoft Cloud regarding AI here. So, we've got three months on screen here. Obviously, we don't have the rest of the year. But, look, we're just at the beginning here, the infancy. The Microsoft Cloud is releasing on average five updates, five updates feature changes per day. 

Now, one of the things I would say, if your world is Azure, and I hope it is, pop into your favorite search engine 'Azure Heat Map', and what you're gonna get is a page that's gonna come back and it will rather than parsing our blogs, our RSS feeds of what's happening on our platform, it's a Heat Map based on certain topics. And the topic that shines brightest at the moment is Azure AI. So, we are releasing the greatest propensity of updates on our platform there. So, look, we're incredibly excited about this. I hope you are as well. 

And look, with that, I'm gonna pass the baton over to Glenn. 

Thanks, Glenn. 

GLENN CARMICHAEL: Thanks, mate. 

That's a one for you. 

Hi, my name is Glenn. I'm from Telstra Purple. So, Telstra Purple is a part of Telstra. 

[SLIDE: What does AI mean for customers?] 

We're professional services, part of the business. So, we've got people that are designers, developers, data scientists, people that do implementation of contact centre and things like that as well. 

So, what I want to talk about today is what exactly does AI mean for customers.  

[SLIDE: The opportunity of AI. How Telstra Purple can help] 

Now, we've been talking about AI a lot over the past sort of six to 12 months. There's a lot of tools out there. They're all really, really cool, really, really exciting. But they are just that. They're just tools. They're something that enable people to do something that previously they weren't able to do, or they make it easier for them to be able to do it. 

I've got an eight-year-old that's using ChatGPT to help him write an app that keeps track of the Pokémon cards that he's got and the Pokémon cards that his friends got that he would like to get from his friends, which is genuinely impressive because being able to have a conversation and get that code back and debug conversationally, going, 'Oh, that didn't work.' And it apologizes. It's like, 'Oh, I'm very sorry. How about you cut and paste this in? And that should work just fine.' 

So, some of the stuff that we've been talking about, we're talking a lot around generative AIs, and that's quite a narrow band of the full spectrum of what potentially AI could talk to. I'm gonna talk a bit about generative AI, but one point I do wanna make is that there's this stacking effect that we were talking about earlier. 

There's other variations and formats of AI. You can stack generative AI tools, but there's other productized AI services that you can put on top to extract additional value, particularly as a business customer. 

So, things that I'm gonna talk about. The first one is the democratization of AI. This is largely been brought about by Microsoft's significant investment, particularly in the business space in OpenAI. But I'm gonna talk a little bit about sort of where things have come from and where we think they're going. 

Then we're gonna talk about how we see that AI is gonna change work. And then we're gonna talk about how businesses can take advantage of some of the changes that we see coming. 

Talking about the democratization of AI. Everyone here - like quick show of hands - everyone's played with ChatGPT? Yeah. Fantastic. It's amazing. People freak out when the computer can do things that you can do. And I get that. But now we've kind of moved past that initial surprise phase and we're into, well, how can I get this to do more for me? And it's not just in generating text. Has anyone had to play around with like DALL-E or Stable Diffusion or Midjourney, these image generation tools? 

So, genuinely, genuinely cool and really impressive. And the thing that's blowing my mind is how quick they are improving. Like I've been in the technology industry for a really long time and the speed at which these things are getting from frankly a bit shit to incredible. You look at DALL-E from 12, 18 months ago and they were pictures, but compare it to the photorealistic and gorgeous artwork that it's able to churn out today. It's a lot of fun actually. If you go back and you search 'DALL-E improvement over time'. Lots of people have kind of done experiments where they've done the same prompt over time and sort of laid it out as a comic strip. And you can kind of see just like melted weird hands, weird pupils through to like gorgeous images. And it's kind of fun to see like where additional training and everything went into those models. It's really, really cool. 

And we're at now with video where we were at 18 months ago with photos. And so, you can sort of see that acceleration curve. The video, though, is still... If you'd like some nightmare fuel, if you would like to not sleep tonight, can I recommend just go to your favorite search engine and look up 'Will Smith eating spaghetti AI video'. It is horrifying. Genuinely horrifying. 

So, these things are getting really good, really, really fast. And it's super exciting. Hearing about the copilots and bringing them into more than just GitHub. Bringing copilots into Office 365. This is what we're talking about when we're talking about the democratization of these tools, particularly in a business context. 

Because copilot for GitHub, it's been around for just over 12 months now. A little while. It's not new and it's an amazingly powerful tool. It allows our developers and our developers at Telstra Purple use it all the time. It's a way for them to optimize and automate what they do. They are more efficient, they are more effective, and they can do less time actually working, which everybody wins. So, they're a really powerful tool. 

But the reason that they've been able to do that so well for so long is because developers have got this superpower that normal people haven't historically had. They can talk to the computer in a language that the computer understands. They can communicate their intent to the computer very, very clearly. 

And so, what these new tools in the copilots that are coming allow us to do is it allows us to communicate our intent to the computer in our chosen language, whether that be English or something else. And so, it's effectively taking human languages and turning them into programming languages. 

So, what this means is that developers are able to do that with GitHub. But I'm a designer by trade, so I'm obligated to hate and be bad at Excel and spreadsheets, and I live up to that. But what that means is that now I just have to tell it what I want the spreadsheet to do because I've always had that. I've always been like, I know what I want it to do, but if I go and search what formula to use, my brain will explode. But now, if I can describe what I'd like Excel to do to the spreadsheet, it can do that. And that's the thing that's quite different. 

So, it's taking that ability to articulate your intent to the machine and driving that efficiency and automation gains that we get and giving it to everybody because everyone can punch in those prompts. So, more people with those skills, more people with that power, that's gonna make a huge difference to the productivity of our workforce. 

[SLIDE: The opportunity of AI. How Telstra Purple can help] 

So, talking about how AI will change work. To be honest, I don't think that it will. I think that it has. If you look, there's a number of different surveys that have been done. IT Brief Australia is the most recent one, 83 now. 83% of people have used AI at work in some way, shape, or form in the last six months. That's more than four out of five people that have had a crack at this. And of those 83%, 81% believe that it saved them and their company time. They believed that it was beneficial in that way. In about the same amount, believed that it helped their company develop more innovative and creative work. So, that's really cool. 

But only a little over half thought that it was on par with human outputs. And this comes back to that original analogy of these are tools. You as a human can have different levels of mastery of using those tools. You can give it a pretty rubbish prompt, and you'll get some pretty rubbish outputs, and then you're gonna have to finesse those outputs. You're gonna have to use your human ability to massage that and turn it into something good. 

You can spend the time and have a bit more mastery of the tool and generate a much better prompt. Maybe you have to do a little less massaging at the end, but that human element is still critically important. Is that why there's jobs for prompt engineers these days? Well, yeah. The prompt engineer at SXSW was the fastest-growing job title that they had. So, from year on year, the largest leap from people not existing to, hey, this is a job type now. 

Whether or not prompt engineer is a job in the future, there's no prompt engineers in the room, is there? No. Good. Look, the way that you type search terms into your search engine, the way that you type that is not the way you would normally write. Like you know how to enter keywords and work the machine to give you the results that you want. That is something that is not natural. It's something that you've learned over time. It's a learned behavior. I think that that will be true of the way that we write prompts. 

People will learn what priming the AI is, they will learn how prompting works, and that will become kind of a baseline skill. It will become an expected skill, which brings me to my next point. So, adapting and adopting AI in how you work that is going to be essential. It's gonna be table stakes. It's not gonna be like, look, we're doing AI, we're different, we're innovative. It's gonna be like, we're doing AI. We haven't fallen behind. And that's gonna happen really, really quickly. 

So, the example that I can give you is people understand what AI is capable of. And if you look at any industry, people even outside of that industry have a pretty good idea of what tools are standard and best practice for that industry. So, like you're sitting down with your accountant and your accountant is taking notes on pen and paper and you ask them openly just out of interest, what accounting platform are you using? Do you use MYOB? Do you use Xero? Like, what do you use? If your accountant turns around and goes, no, I don't like any of that. I like to print out the forms and do it in my head. I do it with pen and paper. I'd be concerned. I'd be concerned about the accuracy and effectiveness and probably the efficiency of his work because I know that the best tool for the job is available to him and he's choosing not to use it. So, I'm probably not gonna use that accountant. And it's not limited to knowledge workers. You employ a builder to come and build your house. You've got your slab poured, you're building out the framework. Now, none of us are probably builders, but you probably know that the best or most efficient way to nail a frame together is to use a nail gun really quick. Really easy. If you walk out and he's hammering nails in, you ask him, hey, why aren't you... Why aren't you using the latest, most modern tool for the job? And he goes, I just like doing it this way. It's like, well, I'm paying for your time and that takes ten times longer, so I'm gonna find a new builder. So, people's expectation of what tools you are using in your business matters and people's expectation that you are using AI in your business is coming really, really quickly as table stakes. It's gonna be standard. And so, this isn't something where you can pull everybody into the boardroom and go, cool, we're gonna have an AI training session and we're gonna be an AI company. This is moving so quickly that you need an approach to adopt this and then you need to be assessing and understanding what are the advancements in the technology and how can I roll that into my business. Which brings me to how businesses can take advantage. 

[SLIDE: The opportunity of AI. How Telstra Purple can help.]  

So, I think that using things like copilots, using stuff that's off the shelf, really, really powerful and an amazing first step. 

It's that vertical piece, that next level where you're starting to think about building things that are specific to your organisation that require a higher fidelity or accuracy or greater precision about the information that you're presenting it with and that you're asking it to feedback to you. 

So, lots of organisations have got different ways of doing this, but a bunch of our Telstra executives recently had the opportunity to go to America in a study tour and they spent some time at Microsoft. I got to read the outputs of that study tour. Actually, I fed it into ChatGPT and said, give me a summary, but I read the summary of the outputs of the study tour. 

And I loved the way that Microsoft was encouraging their teams around coming up with ideas of how to use AI in their business. So, they've got their guardrails, their frameworks that come down from the top level, and then they're encouraging these green shoots within their business. And as ideas come up that are gonna generate significant value in their organisation, they can choose to nurture and fertilise and grow those up. 

So, these kinds of implementations, these kinds of custom versions of AI, this is something that Telstra Purple can help with. 

So, are we gonna be low on time?  

JAMES NATION-INGLE: Whispers- Keep going.  

GLENN CARMICHAEL: How long have I got?  

JAMES NATION-INGLE: Whispers- 2 minutes 

GLENN CARMICHAEL:  Okay cool, you did this (shows 6 fingers in the air). And I was like, I have six minutes. Like, its terrible communication. [Audience laughs] 

So, Microsoft's approach to this is really, really cool because they've got those guardrails, that framework, and they've published that framework of like ethical AI use and implementation. It's really good. I know that the Microsoft guys said, go look at it, but I'm telling you to go look at it too, and I don't work for it. So, it's genuinely good. 

So, Telstra Purple can help you with the implementation and the development of those things. 

So, I wanna finish up on just like a story about a customer interaction that I had recently. I had the opportunity recently to have a chat with the CTO of my local Council, which was really interesting. It was actually we were playing golf. But just a few weeks beforehand I'd been searching on their website. 

I'd been trying to work out - we want to do some landscaping out the front, and I wanted to find out if I was allowed to put my letterbox on the nature strip. Now, I notice in some parts of the suburb that seems to be a thing that people have done. And in some parts of the suburb it doesn't seem to have happened. And I was like, well, what's the rules? Like, how do I find that out? So, I went searching on the website and couldn't find anything. It was not particularly well structured. Terrible AI. I'm not gonna tell you which Council, by the way. I know. Terrible information architecture, but it took me down this rabbit hole. 

So, I found the Zendesk instance that had some stuff, but not the stuff that I was looking for. Then I found like a public-facing SharePoint instance that also had some stuff, but not the stuff that I was looking for. Eventually, I downloaded a PDF that allowed me to search within it to find the answer to my question, which was no - no, you can't. So, it took me a long time. 

But I was playing golf with the CTO of my local Council and we got to the third hole. We kind of chatted about what we do and everything, and the conversation turned to AI, and he's like, yeah, there's just not a use case. I was like, are you... I told him the story about the letterbox. He was like, yeah, it's pretty rough. Not much we can do about that. What are you talking about? Like, we can train or we can use all of this unstructured data and we can help you to unearth information that is stored within all of these different places in a much better way. In fact, we could set it up so that somebody could conversationally interact with it. And he thought that was pretty cool, but pretty uncomfortable with putting that in front of his constituents, in front of people like me that are asking questions. But the idea of somebody that works for the Council being able to take a call and understand my question, and they're two days into the job, they're just as good at finding the information that I need as somebody that's ten years into the job, that's really, really powerful. 

Now, that, until I had a really great conversation is where the story ends. I was like, cool. AI did a thing. We're having that conversation with the customer. 

We're gonna move that forward. 

But then I had a chat with Emmett, who works with me at Telstra Purple. He's our Head of Contact Centre. 

And he's like, yeah, but they're still typing the query in. Like, you know that we have these other products and tools that can sit on top of that. 

So, if we're doing sentiment analysis and we're listening to what's being said, why does the human need to generate the prompt? 

So, you can have an AI generate the prompt that searches the database and surfaces the information, and so the two humans can have a conversation and the information that is pertinent to that conversation. 

So, I just spat at you. The information that is pertinent... there wouldn't be any water for the front row. I'm sorry. 

So, the AI is generating a prompt for the LLM to surface that information so the two humans can have a really natural conversation with the relevant detail being presented to that person that's working for the Council. 

So, that is an example of some of the really cool and amazing stuff that is possible when you take just those productised things and you just start to stack on some of that more precise stuff that is specific and unique to your business, your use case, your data, and it's all stored securely using Azure. 

With that, I'm gonna hand over to Dom. 

DOM JOHNSON: Thank you. 

GLENN CARMICHAEL: Thanks, mate. 

(AUDIENCE APPLAUSE) 

JAMES NATION-INGLE: While Dom is doing this, what I'm gonna do is I'm gonna stop the meeting that was said before so that the transcript has a time to for me to grab it and use it. 

Dom is gonna talk to you about our new Microsoft practice. 

Cool. 

DOM JOHNSON: Hi, everyone. 

So, Glenn, earlier just talked to you a little bit about what Telstra Purple is. 

Who's here has worked with Telstra Purple? 

Show of hands? 

Few of us. 

Who here knows what Telstra Purple do end to end? 

Few of us. 

So, maybe I can break that down from a Microsoft standpoint. 

Ten minutes. 

Thanks, Adam. 

I'll try and make it five. 

GLENN CARMICHAEL: Did you do like this or something? I was like, yeah. Clarity, gestures. 

No, it was good. 

DOM JOHNSON: Thanks. 

So, our Telstra Purple Business. We're a professional services company of Telstra and we are there to, in our practice, bring our customers complete clarity in adopting Microsoft Technologies across the entire stack end to end. 

So, I have a few of my teammates standing right behind the room there, including my boss. So, hey, Mary. 

And so, our role is to help you adopt the Microsoft technology. So, case in point, Microsoft Azure. Who here leverages Cloud today? I'm assuming I see a few hands there. 

Absolutely. 

So, taking applications that were legacy or building new applications and deploying them at scale, deploying them at speed. Telstra Purple has done that for numerous customers around the country with some terrific case studies. Particularly one that I'm absolutely proud of is the work that we've done inside of Ambulance Victoria in the past, where in the midst of the pandemic, the government came to us and said, hey, we need to track all of the ICU beds across the country in real-time. Within one week, our digital teams and our infrastructure teams got together and launched that. So, every first responder, every person that was required to have a deep understanding of what's happening around the country when it comes to hospital beds was able to ensure that they knew that information. 

So, we then have Azure infrastructure as a big, big play for us as well. So, whether it's taking workloads... Sorry, there was a bit of guitar playing behind me - taking workloads that was legacy in an on prem data centre into the public cloud, whether it's Azure Native, Azure infrastructure as a service, or on Azure VMware Solution, we play in that space also. 

Who here leverages the Power Platform stack in their organisation? If you actually go to Bing and type in 'Microsoft Power Platform', you'll see a case study, a video that comes up where Microsoft did a case study on how we use it internally. Does anyone here know that story? No. 

So, we are one of Microsoft's biggest consumers of power platform on the planet. And it's a great story. 

When you have a moment, go check it out through your favorite web browser. 

But effectively what happened was there was a situation where our engineers and our people that were servicing all of our towers, exchanges, our pits, they had... 

We have about 2,000 of those people around the country. 

And what they had to do was leverage about three or four different applications, ten different spreadsheets to surface information together. 

I'll fast forward a little bit and just say we built a Power Platform application using our power apps. We surfaced that into Microsoft Teams. All of a sudden, we ended up saving millions of minutes within the business, improved employee sentiment, but also gave executives the visibility that they needed to... 

Am I going robot? 

No. 

Gave executives the visibility that they needed to get the information that they wanted under their fingertips as well. 

So, I'm getting the wind-up, which I thought I had ten minutes. 

Play me off, Johnny. 

Look, to fast forward, we have tremendous capabilities in Azure Power Platform, security and frontline workers in our practice. We have people like Glenn, people like James, my wonderful staff at the back of the room, that can sit down with you and understand what the business problems that you have are, what are the ideas that you are trying to build, trying to take out to market? If you've got that million-dollar idea to save $1 million, to make $1 million, or to change the lives of people in your organization or outside of them in the world, we take a design thinking approach into everything that we do all the way from infrastructure through security, through to frontline work. 

And so, reach out to your client partner in Telstra or to one of your digital BDMs and we will get that conversation started and help you with the adoption of AI by putting brilliant people like Glenn in front of you to have that conversation. 

So, thank you so much for your time. 

JAMES NATION-INGLE: Cheers, Don. 

Thanks, man. 

Can I get you to switch the thing back over? 

Now, one of the things I get asked a lot is, will AI kill creativity? 

And so, what I wanted you to do... 

So, you've seen today present AI, so copilot, something that we know about. 

That's something that's gonna help us a lot actively. 

And then the OpenAI stuff that uses your data in the background securely that we talked a little bit about ChatGPT. 

This is the engine that we're playing with now that allows us to get the ChatGPT look and feel. 

I'm gonna show you under the covers of the OpenAI studio, which is this one up here. 

Now, what everyone... Everyone's probably seen a few of these in their time. 

Meeting with James Nation-Ingle. 

These things. 

This is my meeting prompt. 

Cool. 

So, I've got that. 

I'm gonna drop... 

I'm gonna use the text. 

I'm gonna drop this into the AI so that it knows it. 

What I'm gonna do to show that it can help with creativity is that this meeting summary, 

I'm gonna ask it to write us a song and then I'm gonna play this song for you. 

So, I will go, "write me a song using the following text summary." Cool. 

So, it's gonna start writing me a song. 

There we go. 

So, we're writing a song now. 

Oh, everyone's had a go at this. Anyone that ever played an instrument? 

As soon as they saw ChatGPT, they're like, yeah, I'm gonna get it too. 

I'm gonna be a star. 

So, what does a song need next? 

I need some chords. 

Glenn, what I might get you to do is start chucking some prompts into me while we do this, and for me while we do this. 

So, we need some chords. 

So, can you ask it to... 

Cool. 

It's written a song. 

Can you go back to the top of the song? 

Let's have a quick look. 

It's cool. 

You're in the text box anyway, so if you just write, "make the song two verses in the key of G major with a chorus." 

[GLENN CARMICHAEL] For the chorus? 

[JAMES NATION-INGLE] With a chorus. 

[GLENN CARMICHAEL] With a chorus. 

[JAMES NATION-INGLE] Not for the chorus. 

Oh, let's do this. 

I'll kick the thing. 

Cool. 

Everyone hear that? 

(JAMES PLAYING ACOUSTIC GUITAR CHORDS). 

Cool. 

So, this is how creativity can work with AI. 

So, quickly I can go. 

So, we've got G, C, and back to G. 

And we can go, In the era of AI, Microsoft Cloud takes flight, boosting productivity and shining a new light. 

So, this is cool. 

We got some basics yet. 

Now, I'm feeling my creative juices. 

Can you ask AI to turn it into... 

Use the same song, turn it into a 12-bar blues. 

Put it in the key of E. 

Put it in the key of E. 

So, we can go. 

12-bar blues, baby. 

Real time now. 

In the era of AI, Microsoft Cloud shines bright, Boosting productivity got in through the night, Built-in AI for each solution, A co-pilot in flight, Github Dynamics, Azure AI, Our Guiding Light. 

I got one more.  

[AUDIENANCE APPLUSE]  

Thank you. 

Very kind. 

Now, put it in there. 

Make the last verse a nice joke about Microsoft and Telstra. 

Go back to the 12-bar blues. 

Everyone likes blues? 

There we go. 

(JAMES SINGING), In the era of AI and Microsoft Cloud, we hail. 

With Telstra by our side, we're all set in sale. 

A partnership so strong, we are sure to prevail. 

But remember, even in the cloud, don't forget to check your email. 

Nah, it wasn't that funny, but it was alright. 

So, hopefully you can see that AI is not gonna kill creativity. 

It's not gonna stop people having a job. 

It's gonna help. 

We're already using it to help. 

It's important to do it securely. 

It's important to use private data. 

It's important to plan for these things. 

Hope that you've all taken away something good. 

Thanks very much for joining us. 

Thanks to the team for helping out. 

I wish we had more time, but you all have to move to the next session, which is in the big room. 

Thanks very much, everyone. 

[Your Business Optimised. Telstra and Telstra Purple logo]