The Impact of AI in Ecosystem Architecture with Andrew, Chris, William and Anna's Strategic Pyramid

Mark Smith: 0:30
Welcome to the Power Platform Show. Thanks for joining me today. I hope today's guest inspires and educates you on the possibilities of the Microsoft Power Platform. Now let's get on with the show. Okay, chris, why don't you kick off today?

Chris: 0:47
Yeah, I don't mind doing that. So, hey, everyone, welcome to the ecosystem podcast vlog cost whatever we're calling it this week, and thank you very much everyone for coming along. It's very exciting, Super happy to be here. I'm hanging out in a random Hilton hotel in Manchester, so if my signal goes, I'm sorry. But yeah, it's very exciting and I know this week we've got something very special planned. Andrew has designed something. Well, andrew and Anna have designed something quite cool, and it's around the strategic pyramid, which we're going to drill into a little bit more, and I know we're going to be discussing a number of layers, right, which is really, really exciting, and I think we will end up starting out with the implementing workloads layer, so I guess we should kick off there. How does that sound?

Mark Smith: 1:31
You know that onions have layers.

Chris: 1:34
Like ogres, ogres also have layers. I watched that in the Shrek.

Mark Smith: 1:39
That's what I was referencing. Is it a souffle as well? Is a souffle have layers?

Chris: 1:43
Yeah, souffle have layers. So ogres are equal, but not equal to souffle's are equal, but not equal to. Onions are equal, but not equal to Andrew and Anna's strategic pyramid.

Mark Smith: 1:52
Excellent, Andrew. What's this all about?

Andrew: 1:55
I actually had the idea. What we should do one of these days for this podcast is we should kind of force ourselves to each explain how you would tell your parents about ecosystem architecture and then have the rest of the podcast right using that analogy. Oh, I love it. I mean you're very enthusiastic.

Chris: 2:18
So the way that I think about it is, if you think about an ecosystem a bit like a fish tank okay, this is actually how I explain it to people An ecosystem is a bit like a fish tank and the workloads and the things you make on the fish. So if you have a really, really tiny little fish tank, you can't build that many workloads right, you can't make that many things, you can't have that many fish. But if you've got a much larger fish tank, you can have loads of fish, but you need all sorts of other things in there. You need oxygen tablets, you need granules of soil. There's lots of things that you need to make sure that that ecosystem thrives and survives. I don't know if you've ever thought about cleaning a giant fish tank. It's really hard, man Like. I actually owned one, but I've got a teeny, weeny little one at home that I can clean really quickly. So the more maintenance, the larger your fish tank, the larger your ecosystem, the more you have to clean it where, if it's smaller then you have to spend less time. So that is how I would explain how ecosystems work to people.

Mark Smith: 3:14
I love it. I love it. I love all the euthanisms part Just weaved in there, just perfect.

Chris: 3:22
Thank you. It was like watching one of those cartoons with a lot of suggestive content, but not a lot of suggestive content, right.

Ana: 3:30
I think actually that the fish tank or aquarium analogy goes very well with our strategic pyramid here, because you mentioned workloads. So you'll have a little goldfish ball with a bunch of goldfish, and that was really very good for a time, wasn't it?

Andrew: 3:52
Yeah, the strategic pyramid of fish tank architecture. That's what we're doing here, right.

William: 3:59
I'm trying to figure out what the filter would be. Would that be the sort of data cleansing side of this?

Ana: 4:04
Why don't you tell us what it is? Andrew, Come on, do the buildup slide for us.

Andrew: 4:10
Do I have to get in the spirit here and actually explain this in the context of a fish tank Up to you? So can we get? Can you guys see the screen here?

Ana: 4:20
Yeah.

Andrew: 4:21
Yeah, all right. All right, so it's a black slide. What we did is we came up with this idea. We were trying to explain to everyone how value works right Value of where we spend our time, value of where organizations invest their money and their time when it comes to ecosystem architecture. So what we did was we cooked up this idea of the strategic pyramid, and what you see here is that we have four layers and up at the top we start with implement workloads, and the reason we start there is because this is where most organizations, most Microsoft partners and most technologists, most consultants, most developers, most architects, most practitioners of the technology have been for a long time right. So when we talk about implementing workloads, we're thinking and I'll read this for those who do not have the benefit of the video here, though we'll put this in the show notes, I think, so folks can get to it but implementing workloads, we're talking about implementing the Dynamics 365 apps and modules. Or you might be implementing SAP or Salesforce or Workday or some sort of other custom application, developing complex custom applications. So that might be, say, a traditionally developed app, or that might be something that you build with Power Platform. Migrating and modernizing legacy apps, extending core business systems, building data and AI-driven end user components. So think anything from a simple chart dashboard report to pretty sophisticated AI-driven type workload and then finally, automating processes across the organization. So about a year ago, we were all talking about hyper automation all the time. But that's this idea, right, that we're going to go in. We're going to use automation tools and we're just going to automate everything that we can find. So, top of the pyramid, we have implementing workloads. Next layer down, we have creating the conditions for success. Now, a lot of organizations and a lot of people have sort of wrapped this in the cape of governance over the last couple of years, and I think we can talk more about this, right, how we've often, I think, conflated this idea of governance and enabling or creating conditions for success. But you're, basically, these are the services that make it possible to implement workloads. So platform management thinking, everything from the COE starter kit to Azure application insights, to Azure Monitor, to whatever it might be to manage the platform around you Enterprise architecture, data ecosystem, environmental architecture, et cetera, et cetera. So this really enterprise technical architecture, maybe ALM security, and then user and organizational empowerment. Next layer down, we get into what really I think this podcast is all about, which is building the platform ecosystem, and that includes deploying, building and deploying your landing zone, integrating your cloud landing zones, building an institutional data model so think the common data model, but specific to your organization a core platform data services, enterprise data governance, and then integrating workloads across the estate. And then, finally, our lowest level of the pyramid is what we call architecting these strategic foundations. Down at the bottom, we're into the realm that almost nobody touches, and that includes what's the executive vision. Too often, we're implementing technology or we're building an app without any sort of vision or direction as to how this is taking us forward as an organization Prioritizing workloads, mapping our ecosystem, creating reference architecture and then evolving that architecture over time, mapping the data ecosystem and, really importantly, strategic renewal and refresh. And just to round out this explanation of the pyramid, the problem that we have, I think, across the industry, is one that talent. There's a lot of talent up at the top of the pyramid. There's a lot of talent that is really good at building apps and implementing business applications, but as you go deeper into the pyramid, you have less and less talent and people in the market that are available and able and experienced in doing these things, which means that as you get higher in the pyramid, you have a situation where those sorts of tasks implementing workloads are much easier to insource and to reshore. So it's a much more commoditized kind of work. Right that, if you have price pressure at the top, you can just move where the work is being done really. And then, finally, those workload implementation is the stuff that is most accessible to either citizen developers building their own apps or to AI writing the code for you, building the app for you, whatever. So what we're really concerned with is taking people from this implementing workloads layer of the pyramid, where almost everyone has hung out for the last 20, 30 years, and driving the effort and driving the attention and the investment deeper into the pyramid around strategic foundations and building the platform ecosystem. Back to you, mark.

Mark Smith: 9:37
Okay. So the thing that struck me when I first came across this is that the top two layers, if we look at where technology is going are going to disappear in the next five years. So where everybody's hanging out, as you say, where you've got your citizen dev, you're operating. I think the citizen dev is going to be replaced by AI, and it's not saying the citizen dev is going away. In fact, I don't like the word citizen dev anyhow, but I think that what's going to happen right is that you know, I just did a presentation this week on AI, all the Microsoft AI, everything. Not just you know what sits in our normal warehouse, but you know security AI. And when you think about governance and stuff, I think it's going to be one of those key things that will be delegated to AI and it will do it with such precision and accuracy and unbiasedness. Particularly that green layer that you're talking about will potentially be outsourced to AI. And then the implementation of workloads I can't see, I can't imagine within five years that anyone's going to pop open a studio to create something. They're just going to give commands to the chatbot and it will probably validate. What type of interface do you think would be best and I'd go back to it. Well, what do you think would be best? Based on human behavior, based on what you know about organization, based on the data model that we have inside the organization, the organization data model, what's going to be the best way? And if everyone's living in teams, for example, I assume it's going to recommend an adaptive card type scenario feedback mechanism or assign it to my earbuds or whatever the output, and I will just use natural language query to build any type of interface. So I don't know if those two workloads at the top are going to exist in five years.

Andrew: 11:28
And this is where so I talk with a lot of we all talk with a lot of technologists, so individual folks out there in the industry who do this kind of work, and then also with Microsoft partners, and I think that over the next five years or so, and possibly even sooner than that, there's going to be there's a real crisis in the Microsoft business applications space. In those partners that are historically so focused on biz apps, right by definition, implementing workloads. Some of them are going to make a successful transition, I think, and start to look more and more like an Azure partner, right where they dive deeper into the pyramid. But the ones that don't make that transition, I think they're toast, and what alarms me when I hear people talk about this is not that there's alarm, but that there isn't alarm. If you are in the business of implementing workloads and creating the conditions for success, slash governance you should be freaking out right now, or you should be well on your way to making that transition to thinking more about ecosystems and platforms. That's the discriminator.

Ana: 12:47
So that's super interesting, but do you guys think that the reality is that many people are still obsessed with implementing workloads? And even though they feel, I feel like I talked to a lot of organizations who claim to do a lot more, because I actually think it would be super useful that you publish that, andrew, because I feel like people see implementing workloads as just a power app, whatever citizen developers build. But that's not true. I think we actually mean we talk about quite complex scenarios that are still implementing workloads, like the fact that you're implementing quarter million dynamics for sales. That can still be an implementation of a workload. So why are people still obsessed with it?

William: 13:52
My view on a lot of this is it's a bit like the invisible hand of the market going to a point so no partner is going to stop and go. I need to only focus on enablement. I need to focus only on the lower parts of the pyramid now Because, let's face it, it's still a gold rush. People are still going for it because it's there to go for, and my advice is absolutely go for it, because milk, milk, milk. What you can provide a fantastic service. But then have that strategic direction and intent to start having those more enablement and lower level pyramid type conversations, because otherwise you need to start going the customer towards that by actually talking about the workloads and going. But have you thought about the following? So we're all going to carry on with the gold rush because it's foolish not to. There's still hundreds of tenders coming out all the time where people do need these strategic tactical workloads. But then it's down to us to then start pointing towards the future as well and expanding that.

Andrew: 14:45
The smart money, and this applies, I think, to Microsoft customers. So, if you're an end customer organization looking to build a modern cloud ecosystem, if you are a Microsoft partner in the business of helping those customers build ecosystems even if that's not how you thought about it traditionally or the words that you've used, or if you are the end the technologist, the developer, the consultant, the data engineer, the architect, whatever the smart money, the smart course of action here is to diversify up and down the pyramid. I am in no way arguing nor have I ever that you should leave implementing workloads behind, but I think you really need to diversify your skill set, what you're offering in the market and what you're buying from the people, from those who are offering. Diversity here is the key.

Chris: 15:42
Yeah, it's a big point. So somebody in our business recently said to me we should be teaching people to do the basics very well, and I said I don't agree, and I'll tell you why. Because I think AI will do the basics very well. I don't think we will do. We need to do the basics very well. And I think that what is happening is that if we focus on just the basics and we don't learn how to leverage AI, that's a huge problem. When you look at GBT5, so what GBT5 will be offering in the next X years? You'll be able to type into a computer hey, build me an enterprise wide expense application and it will make it with related tables, with the right screens, the right actual logic in there, and it'll do it for you. You'll be able to write build me a claims management solution based on X. Not sure it's not going to know all the rules, but it's going to be able to create their relational structures for you. And as somebody who is a maker and has built my career on data modeling right, that goes out the window for me immediately. I need to learn how to prompt. So me being able to do the basics very well right now are fine, but in two years time it won't be. So my theory here is that learn how to use AI to do the basics very well is an assisted piece of work and also, yes, as Anna said, and I thought it was a great point and, by the way is focusing on yes, you have, and Andrew said as well diversification, but those complicated workloads will not go away. It's just using AI to do the basics and you use your actual brain to do the really tough stuff. That's just what I was thinking as you guys were both speaking.

Ana: 17:10
If you're looking at the pyramid as well, everything that all of the bottom layers are actually additive. They're not. So what we're saying is okay, fine, we know how to do workloads. We now have time, yeah, to do more, because we've done automation and we've done apps and so on, and now we have time to do ALM and governance, and then you know it's time for us to look further down. Right, that's what we're saying.

Mark Smith: 17:48
The thing is, with further down the pyramid it's quite a different skill set and it's hardcore consulting right when you need to be broad, you need to be knowledgeable and a lot of partners are fundamentally not like. They don't have a mindset of a management consultant. They don't have a mindset of Custer comes to them and says hey and I'll give you a situation. It's a bank. The request came to me via the UK, so it's a large international bank and they wanted a better way of managing email support tickets and they just wanted more resources to help build that out. I'm like okay, if we're doing email support in 2000, you know, in the 2000s, sure.

Ana: 18:33
But why don't?

Mark Smith: 18:33
you change the conversation, why are you still doing it via email? Why are you using a dumbed down email to manage a support process, even though you're one of the largest banks in the world? But the consultants, oh no, we can't challenge what the customer's asked for. They just want a better email system, Like your job's to consult, to bring to bear the best and breed an industry not to go. Oh, how many resources do you?

William: 19:01
want for this. That's a technologist's chat, isn't it? That's where the consultancy brain doesn't kick in, and we are surrounded by a lot of technologists. So absolutely, and that's tying what both yourself and Anna said together, which is only because the customer asked for it, you need to switch on and go. Is that best practice? What else have I seen across the industry, and it's our responsibility to bring that to fold.

Andrew: 19:21
I think it's interesting, will said the last time we recorded. Will went on a riff about how technologists have lost the ability, or maybe technologists need to get back to good, old-fashioned consulting and listening to the customer and these softer or more strategic skills. I think it's just very interesting. Will and I in our respective firms have the same job title. It has the word technology in the job title. It's these two guys who are here with the word technology in the job title telling people it's not about the technology, it's ironic.

William: 20:00
But it never was about the technology, yeah.

Andrew: 20:03
Yeah, mark, you were saying a minute ago I was just thinking about this was before the pandemic and I was in a room I was doing a workshop with some folks. This was not a bank, it was an insurance firm and it was part of a larger app rationalization. But this particular session was focused on what business problems or what day-to-day challenges do you have that we could bring to bear some sort of appified, digitized solution for? And someone in the room very seriously said well, we need. One of the most important priorities for us is to build an app to help people book conference rooms for executive meetings. And I said, yeah, I think there's an app already for that. And they were very perplexed by this. And after about 30 seconds of this back and forth, I realized this person really doesn't realize that Outlook does that. I had to let them down gently that this was probably not the number one priority for this Fortune 60 company to use.

Mark Smith: 21:03
There's the problem, right, people get into, and I found, with joining my current employer, this amazing thing in that I'd worked for nine on 30 years in the Microsoft ecosystem. And then you go into somebody else's ecosystem, in this case IBM, and they were using Lotus Notes when I joined two years ago still as their email platform, and what blew me away is the Zalits inside the organization that thought this was the best thing at all. Man, you can do anything with it. And I was just like are you serious? And what I realized is that people have no concept of what's outside their ecosystem because there's never been a reason to look. And so they look at what they have and they say this piece of tripe does taste amazing. And I'm like have you tasted caviar? Have you tasted a Swiss cheese? There's other things other than tripe, but they still believe it's the best thing because they've spent 20 years only on that. They've not had to learn yes, it's now muscle memory for them and using it. And then, like you say, with Outlook, what there's a way to book a room that doesn't mean you have to go to yet another bloody application to use. It's not just part of your standard day-to-day toolset and I'm just like, but a lot of people have no idea what's available.

William: 22:29
But then should they and I think this goes back to our previous conversation around. Actually, the education part's key, because if I was an accountant, I'm more excited about keeping track on what's happening with UK taxation laws, what's changing in the budget, what's happening that's going to affect my clients and their small businesses and what's going to impact them. I don't have time to focus on the technology, although I would take a step back and go wait a minute. Most accountants, as the example that was given, actually use the technology day-to-day to leverage in their job role. So they should be looking at it, but because it's not in their sphere of interest, it just goes past. It's a bit like if I asked you guys about taxation rules or maybe some new medical procedures have come out. The doctors would be able to tell you about it, but they'll be able to tell you about the technology behind it. So that goes back to the consultancy argument, which is we must start running familiarization education sessions around. Do you know this? Other stuff exists and it's really cool. I think it is super important as consultants to do that.

Mark Smith: 23:25
My wife has got a post-it note that she's put on my screen, which helps me with my arrogance and sometimes no one will type attitude, and it just says stay humble and curious. And I think it might have been Einstein that even came out with that, it was somebody famous differently, but it's such a powerful concept around that one curiosity to go into every day, going you know what? What am I going to learn today? What's going to be new? What don't I know? And you're only going to do that if you're a humble, right, because you need to step down that hey, I'm the person that knows everything and I'm going to be the voice and everyone thinks like, wow, that guy knows his shit To being. What can I learn? Because it ultimately, you know, through a curious type activity you're going to find and uncover more potential, more ways of working. But also it really drives a real listening to the customer, a listening to what's going on in their business, because oftentimes what you're given as a we need it to do X. That's a very tactical move. It's not strict, like they haven't gone further, they haven't unpacked that. And I think that consulting part of what we do needs to be to have the ability to pause and go. Can you explain that a bit more? Can you unpack that for me? Why are you thinking that way and use really good questioning techniques to drill in to what the customer's situation is, rather than just taking a verbatim? I just see too much of verbatim taking what's given and it's like oh, that's the fact, and it gets built and it's shit.

Ana: 25:04
I think that's a great technique to use within organizations as well, Because it's very it's very similar. We're talking about firms who implement at whatever layer of the pyramid. They will have smaller or bigger teams and they will have different interests. So unless they ask each other I'm going back to my reconciliation piece. Even after we learn about what the customer wants and we really learn what they want, I think it's still super tempting to fit whatever skills that we have. Like Chris was saying, we really need to learn the basics. I know the basics. Therefore, I'm going to fit my knowledge onto your problem and we're just going to make it work. At the end of the program, we're going to give you about 500 free hours so that you pay us, and it's just the way it works. So I think being humble and curious is actually a fantastic motto to learning how to use the strategic pyramid as well, in my opinion.

Andrew: 26:37
That's the title of this episode.

Mark Smith: 26:40
My knowledge over your problem. I like that my knowledge that you just said. Then apply my knowledge to your problem, yeah.

Andrew: 26:49
That's what Anna describes is the classic problem of if all you have is a hammer, everything will be a nail. And I see technologists, I see developers, consultants, architects, whatever do this all the time, where they look at a problem and they say the solution to this problem is surely the thing that I know. This is how I was with an organization earlier this week and they were talking about workload after workload that they said they had used dynamics to solve. Well, unfortunately, none of the workloads had anything to do with dynamics. What they did was they bought dynamics licenses for 4x the cost of power platform licenses and then they built power apps on top of dynamics. So they were sort of electively. They were electively paying four times more than they needed to be to solve their problem, because surely someone had told them oh, dynamics, because dynamics is the thing that the guy who told them they needed dynamics happened to know. And this happens all the time, I think, throughout tech, many people are guilty of it and as an individual technologist, you have to kind of force yourself to, as Mark says, stay curious and learn the new thing and be open to the possibility that the thing you know is not actually the solution to this particular problem.

William: 28:18
It's incredibly common. I was going to say one of the most common places I see this a lot and it's really quite awful is within ERP solutions.

Andrew: 28:30
Oh, the worst offenders.

William: 28:33
I see a lot of finance, supply chain management sold into companies that are tiny, that only need a BC, but because the part that went into consult only did finance the supply chain, or finance and ops for the older people in the room or AX, then that's what they get. I see it probably at least twice a month when we consult with a client and it's it's. It's verging upon some slightly criminal, to be honest.

Chris: 29:00
I actually so. When I just before I joined Microsoft some time back and I was at a partner X, I had a day between the two companies, and that day between the two companies I spent at a SharePoint conference with people like said Matthews and a bunch of others. Oh my God, I remember that I was an employee, so I've got to do some pretty, pretty weird stuff and I got at the stage in front of an entire audience I don't really care if you edit the sound or not but I said guys, sharepoint's not a fucking relational database, stop using this one. And yeah, I know it wasn't necessarily the right thing to do and I still believe it to an extent, but actually that's exact comment caused quite a bit of like controversy, I guess. But also I got to have my mind open to somewhat, so I started chatting to people like Paul comesy and you know a bunch of others and I really admire Paul. I think he's the guy is unbelievable. I mean, his grasp on like how actual ecosystems work is amazing and he's like dude. I get why you say that, but you can't be throwing it out like that. And here's why and I actually learned a lot about the SharePoint community and why people were building things the way they are and actually how even low code adoption inside power, platform specific and data versus being adopted the way it has. And it's because of the modern workplace community and I'm definitely one of the biggest culprits of trying to hammer in a nail with a giant hammer that didn't need it right. So I think that because of that, I've had the ability I've been allowed to learn and I think, like a lot of the time, you need to rely on the people around you to an extent to give you that knowledge, and I do think that a little bit of a step in the face sometimes doesn't hurt as much as people think it does.

Mark Smith: 30:47
Yeah, yeah, agreed, as in got a use case at the moment that is for a company that they want to migrate Nintex forms and workflows into the power platform. And the sales hat on me goes dataverse. You know it's premium, everything and every use case, of which there are about 17 of them just needed SharePoint as the back end. And you know, as somebody that's not come from SharePoint, and I find it like, hey, you've got to take the right road in those situations. And in this case it meant that the E5 calz were going to cover everything they needed. There wasn't a license, sale and play here, because nothing was confidential in nature. No relationship data was needed to, you know, be in that side of the system like dataverse and and yeah, it is horses for courses, right, it's knowing when to choose. And I don't know. One thing that's becoming heavily involved in all conversations with customers lately is decision trees. How do you decide in an ecosystem? What system gets built where you know when, if you go into an organization, let's say they've got SAP, they've got Pega, they've got Blue Prism, they've got ProductX and they've got the Power Platform right in place, and you're coming in to enable the organization to make the best decision. And one thing that I always laugh in the consulting arena is that when the pure consultants, in other words, the ones that don't have a speciality in the technology so we're product agnostic, we'll do whatever, whatever the customer chooses, and I'm like that's such bullshit. You, you recommend, you advise on what you're most knowledgeable about right. You never recommend something that might be the better product because Gartner said it never on our set it, but you don't have knowledge on it. You're not going to recommend that.

William: 32:51
It's knowledge bias. It's knowledge bias 100%.

Mark Smith: 32:54
And because what I sit in these rooms and there's like all the Pega devs and all the Blue Prism devs in the room and all the you name it you know and and they're all like their product is the best product. This should all go that way and to come up with an unbiased decision tree that says, in this situation, you're probably best to build a Fiori app, or, and you, you should use BTP on SAP to do this, because you're you're not going outside the SAP, the SAP ecosystem. What you're doing is super simple Like, stay in there. Yeah, it's probably going to take you three years to get a built, to find a resource, but that aside. But then in the minute you know the ecosystem touches oh sorry, the, the, the, the. What they need touches multiple ecosystems. So therefore, there's a document requirement. So there's a document repository somewhere, right, there's a tracking there. Let's say there's a. Let's use shipping as an example. There's a shipping code that's sitting in SAP. There's then conversational things that might have happened around that, more order sitting in a serum system. The minute you go broadly across system, you can't just say that only one tech is the way to do it right. You got to find the best mix to bring those, those ideas and concepts together. And so I think that it's very hard to to get that decisioning choice without bias right and and you know, without that tech bias, without knowing, hey, and that's why I think this discussion around ecosystem is so important and that, although we all come from a very Microsoft centric world, in the last two years I've got a massive exposure to SAP, because I am doing a lot around trying to understand and bridge that gap between the power platform and SAP and that probably the most powerful reason for having that relationship without using the word power is that the user experience you can provide on the power platform is superior to what one can experience on SAP. That's not a slight on SAP. Sap has data models and structures that are absolutely robust and will run for 20 years without breaking right, and that's why people choose a solid solution like that. But when you want user experiences over data, I think you've got some compelling reason why the power platform is in that mix. I go back to it, though. It's very hard when you've been trained only in Pegger or trained only in Blue Prism and that's what feeds your family to be unbiased about how you choose technology play inside the organization you work in.

Ana: 35:31
I think, looking at SAP specifically, in my experience as well, I think you're totally right, mark. But there is another angle. It's not just the user experience, it's also when you want, when you need, when you need change quickly Sometimes you just need change quickly You're going to have to somehow plug in that information into SAP and work with it, and you're going to have to do it quickly. So it is very likely that you're going to create a you know, some sort of power platform, low code, an Azure solution and then just plug the data back into SAP, because you know a lot of the time data still needs to leave in SAP and that's perfectly fine. But I think what's important here is that you do take into account, like, even when you go in and your bread and butter is Low code or power platform or dynamics and and you do decide this is the best way to go, and You're plugging it into a bigger system, ecosystem that may be of a completely different technology. I think what's important there is to still talk within your teams and don't just cowboy your way there and Quickly build you know a bunch of apps to just show off but rather think about okay, how are we gonna deploy? How do we actually integrate all of these data together? How are people gonna log in? Who's got access to what you know? How do we make this, as Andrew says, future ready? Are we gonna be able to, you know, use this for AI in six months time? Like that's what we mean here. In a way, we're not trying to slag off or like tell off any ERP developer or Dynamics or power platform or anything like that. What we're seeing is Okay, so you have a fishbowl and you've been selling the fishbowl as a Workload to all sorts of people, and finally you get to a point where you you realize that you know all of this is gonna be automated and people are just gonna work on them off of Amazon and All of the sudden, you want your little goldfish to actually live in a bigger fish tank. Chris is gonna be like you.

Andrew: 38:13
You got my analogy completely wrong right, I mean wait, amazon is all in on this analogy.

Ana: 38:19
Absolutely the moment. You want that your little fish to to live in the, to be able to survive in the bigger. You know fish tank you, you want to make sure that the other little fishes aren't gonna eat it. You know that it can breathe, that it can, etc. So, and that's what we're talking about, what this is one thing, to implement your workloads, and no matter what they are, what sort of fish, your ball fish, your year actually selling, and Then now you're able to Integrate your little ecosystem of fish into bigger ecosystems because you're protecting them from the other little fishes and Everyone knows about them, and so on, so forth. So you've created the conditions for success. But what we're saying is oh my god, it's very likely that in like six months time, amazon or eBay or your provider of choice will do that automatically for you and they won't let you buy a fish that's gonna get eaten by the other fish. And what's next? And I think that's what we're trying to do with Strategy, like we were saying earlier, not not on tape, unfortunately, a lot of people come and say I want to do strategy and he's like okay, so what do you mean? Oh, you know.

William: 39:44
We've got. We've got 100 strategists. Now nothing's moving forward.

Chris: 39:50
Come on, baby, maybe a baby a hundred grand soon, not make decisions.

William: 39:55
But I'll think, and I'll think hard.

Andrew: 39:58
So I think that there's a just to put a bow on this, this workload, implementation layer of the pyramid, because I think that Anna is rightly moving us into the create the conditions for success. But a few years back, chris and I were with a big group of people. We're at a Microsoft conference in Kiev, ukraine. This was the the conference at which Chris infamously, he did not jump into a fountain in the center of downtown Kiev because there was a sign that said no jumping in the fountain. Chris rolled into the fountain because he felt that this was that, that that would was was okay.

Chris: 40:41
Come on, bro, this was off to the million.

William: 40:43
That's it. Yeah, you're following the rules. He didn't jump. I'll be your lawyer.

Andrew: 40:51
So after, at one point, there was a break in in the conference and, by the way, this was one of the best conferences in one of just the most absolutely beautiful cities that that I've been to but in any case, the so someone came up to me and they asked they asked me if what I thought power platform wasn't good for, and I struggled to answer the question, right as if. Oh yeah, power platform is good for everything. You could use them for literally anything. And I struggled to answer that question. And as I look back on that now, I think to myself that there's a real difference between expertise and wisdom. Right, expertise is knowing a lot of things about a thing. Wisdom is knowing when that thing isn't right. Right, wisdom is about knowing when what you know is not applicable or is not the right choice here. So I would kind of challenge everyone as you think about the technology that you know very well or, if you're a partner, you think about the technology that you implement very well Ask yourself what are the situations when the thing that I am so attached to is wrong? What are the situations when I would recommend to my customer or recommend to someone else? Don't use the thing that I personally know very well, even if it means that you don't win the deal or that someone else has to come to come do the work. And if you can do that, I think you you've taken a really big step into one a deeper level of the pyramid, but also into this realm of of graduating from expertise to wisdom. And that's the thing I couldn't do on that particular topic a few years ago, and you know I I could sit here and make you a list now.

William: 42:43
So, whether this stays or not, I just thought of something that just ties up the bone. I see that a little saying that I that I heard, which was, you know, expertise is knowing a tomato as a fruit. Wisdom and experience is knowing not to put it In a fruit salad is a bit like our love for you know, we all love the power platform. We all really enjoyed that. But you know, we know it can. It can do a lot of stuff, but knowing when to use it it's really appropriate. So, yeah, so I just wanted to a to chip that in. I like it that also goes.

Ana: 43:11
That also goes with. I have a little bit of a Beef with with people who are maybe building new teams or they're trying to do new, new stuff and they're like we're gonna do this kill can be learned. Expertise, like it's super easy to be learned, and I like Is it, though. Is it Because maybe, so, maybe we should add you know, the consultancy layer on top of it. You know that that you that you're talking about, because I feel like many times yeah, you're right, you're both right people are confusing expertise with with wisdom, and they're not allotting a lot of Value to that wisdom. You know, and what you and what you've discovered before and how you, you stayed humbling and curious. I really like that. I, I need, I need that posted in my life. Yeah.

William: 44:15
To add to that, there's a pet peeve I have and I wonder if you guys have this which is when a client brings up a problem they're having and it's normally the quite junior architects jump straight in with what they think the solution is, after the client's only expressed it in one sentence. It's very similar to the, to the, the expertise and wisdom, which is, oh, I think I know the answer, but the wisdom actually comes in to say, hey, let's ask a few more questions and really explore it. I think that that's another, you know, as, as Anna would say, it's something that I have beef with and it always cries my gears. Let's get all the analogies out there.

Andrew: 44:45

When, when that happens so, as we, as we sort of break over, to Create the conditions for success, would would anyone like, now that I know I can share my screen for those who are watching on youtube, but anyone like, a very, very interesting live demo of something? To harken back to something chris said a few moments ago, Is it gonna be funny? It's gonna be funny, it was gonna be great.

Mark Smith: 45:11
Okay, mr.

Andrew: 45:13
All right, so here's my live demo, so I'm gonna share my screen again. Sorry for those of you On the podcast version and we've got you. Guys can see an empty browser tab here. So, what we're gonna go. What we're gonna do is we're going to go to the site. Sharepoint is dead. Dot af. Fun fact dot af is an afghanistan registered domain and and does anyone want to guess it? At what you're gonna see.

William: 45:44
SharePoint dot af, and what do we have?

Chris: 45:55
Lord hunts and fakes. I can't believe that's still live, wow.

Andrew: 46:03
This is my far, the best thing I've ever purchased is that domain. It's lapsing soon, so when it's gone, it's gone. Sharepoint is dead. Dot af will be available.

William: 46:12
Shall we all share that after this on our, randomly on our twitter accounts. Yeah, yeah, there we go.

Mark Smith: 46:18
Out of nowhere. Tell me, what do you, what? What are you thinking about Some of the latest news and events? I find this one here jumping to my feed three days ago and past company front page magazine Microsoft's moment and If you drill into the article it says that Microsoft's such an indella is winning the big tech AI awards. And here's how and it and it breaks that down. I saw a graph the other day showing it kind of where open AI chat, gpt is that and where a bar does that which is its closest Kind of parallel competitor and market, and they're light years apart around the monthly active users that type of thing. It's interesting. In this article Satya talks about this was 10 years in the making. He goes in general. I get excited by new things, but the story of our transformation, transformative moment has been a decade in the making and I think a lot of people thought that Microsoft like how were they so prepared for this wave? And, as he said, it was a decade in the making. It wasn't that. They just, you know, open AI landed in market in november with their publicly facing large language model and all of a sudden, microsoft was like able to pivot and and react. I think Amazingly and and I feel that, with his, his tenure of, or encouragement of his staff of learn it all, yeah, be a learn it all, not a know it all, I like that. Well, thank you for correcting me with that Is such a powerful thing. And I end, do we think what's your opinion on this around the moment that Microsoft is in in this AI play? What's it gonna mean? What's the impact gonna be?

Chris: 48:04
I mean huge Dude. I mean think about, think about what's happening. Microsoft are not very good at making new things. They're very good at taking stuff that it exists and democratizing it to make sure people can use it Already quickly, like all the llama stuff that's come out. Now. That was not a, that's not a Microsoft thing. That's coming in. The large language model asked me anything stuff. I mean that's that's getting dragged in. I mean Look at what's happening with gbt 3.5. It's crazy. It's crazy. I. I think that this is gonna change the way in which people work completely over the next few months. Even so, I'll give an example. Right, andrew and I are both doing a presentation at this one thing at Birmingham in a couple of months. So Andrew is doing Something what's the thing for us?

Andrew: 48:53
invite, invite your friends, invite your friends.

Chris: 48:55
Oh, okay, yeah, so it's the emergency services, um thing. I don't actually know that name, but you want us to call it again.

Andrew: 49:01
I think it's the emergency services tech show.

Chris: 49:04
That's the one, yeah, yeah. And um, somebody said, hey, chris, can you put together a presentation on on co-pilot and AI? And I started laughing and they said why do you think that's so funny? I'm like dude, that's in a few weeks. This whole thing will have changed by then and it is making me laugh a lot because, like, microsoft are moving that quickly and you know, we said it in the last podcast like all um vlog cost, whatever we're doing these days, now that the waves are so fast, it's impossible to just keep up with everything. Like somebody just posted on twitter, they said, starting today, developers create and find, can create fine tuned models for the use cases. Some important details this is using chat. This is using gbt 3.5, as far as I understand. Like, this is insanely quick. I have no way of keeping up with it. And, um, I think, microsoft being involved, they've got the resources, they've got the people and they've got the skills to drive this a lot faster than, in my opinion, anyone else.

Andrew: 50:04
Agreed one of the things that Anna and I have talked about a lot is that this is, this gets back to kind of what someone's individual skill set is and where someone chooses to spend their time, the. The technology is progressing and evolving too quickly at this point for any one person to be An expert in everything, right, and I think that we're, we're almost and and to keep me honest on on on this, on the conversations We've had but we're almost to the point where you, you, you have a distinct specialty, which is knowing Enough about how all of the pieces fit together so that you can assemble them, and then you kind of have to turn that over To someone who knows a specific part of that very deeply, right? So, and to me that's what people ask me what, what, what is an ecosystem architect? An ecosystem architect, for me, is someone who can put all the pieces together but doesn't necessarily need to know, nor should they know, because they couldn't possibly have the time to know. Uh, the latest and greatest. They're going to then turn the, the data storage piece over to someone who deeply knows data storage. They're going to then turn the integration piece over to someone who deeply knows integration, and I think that there's going to be a little bit of friction in the talent market for a while, where there's this Stigma either that you don't know something deeply enough right, or that you don't know enough things, and we're going to need to sort that out right as an industry, because I think it really is going to be a distinct specialty to simply know how all the puzzle pieces fit together, even if you don't exactly know how to make the puzzle piece.

Chris: 51:49
Dude, there's something that that needs to be brought up here very quickly around the concept of an ecosystem architect, and that it's not a tech role and it's not even it's not even remotely pointed at technology. Let me give you an example. So, integration, fine, but one of the things we ask about ecosystem architect layers Can you design a support plan for someone? Can you build a governance framework for someone? Do you know what a governance board is? Do you understand her roles and responsibilities work? So, actually, something Anna brought up pretty much earlier was around. Like you know, stuff can be learned. I don't think this is a expertise role. I think this is a wisdom role For real that's interesting, that that's interesting. I struggle to even go to Microsoft and say to hey, microsoft people, please get you to find a support plan for organization with X amount of users that have X amount of things in play and using a slash data for X, y and Z. It's really hard, man and even change managers like customer service managers and things like that, they don't know. So we've invented something new and I think we can have to train the market.

Andrew: 52:53
Mark, to get back to your question about what you're asking about the, the, the satya piece and the statement that this has been ten years in the making. What are the things that if you, if you look at where Microsoft is right now from an AI perspective, there either very lucky in that they happen to have all of the right pieces lying around the shop room floor At the time that the open investment occurred, or they're very smart, right, and just to give you an example of this, if you look at, if you look at Shadow IT and where shadow IT has been, the idea that the, the idea that these quasi apps that have been built for a very long time by folks could have anything to do with it, could feed the AI beast today, less than a year after I got this wave began crashing on the shore, is insane. But here's how this works, right? So you build an app or you bring an app in from the cold Right and you migrated into power platform. Well, that data that you can house that data inside of data verse. Well, data verse can push and now short cut its own data into a data lake. Well, cognitive Azure, cognitive search indexes that data in the data lake, and then you point the AI model at the cognitive search index. This is how, how. This is the basics of how these models work. Right, you have this index data and you retrieve that data from the index and you run it through Azure open AI. Well, we built this sort of daisy chain from productivity app that keeps its Data in data verse, which pushes its data into the lake, which is index by cognitive search, which is used by AI workloads. Okay, that is either a very, very lucky coincidence or some very, very smart ten moves ahead chess playing, and that's the stuff that I don't hear people talking about enough, because they're just talking about the end of the story, which is the AI workload.

Mark Smith: 55:12
Yeah, that's interesting thing is a Microsoft for a very long time has been there are indeed budget, I think eclipses every other tech company in the world. Right, I remember Ten years ago Microsoft events. I remember in the Gold Coast of Australia sitting there and one of the things that these would be partner events and the Microsoft exec would get on stage and talk about how many billion Ten years ago they were investing in the R&D and I mean a lot of that. Never, ever comes is a lighter day, but the stuff that does Is well, research. They've been doing a lot of work. It just didn't, as you say. They found a few parts on the shop floor and they all came together accidentally. These are some very, very smart fellows that have been on Microsoft's payroll for a very long time. They're very bright people I know. Even I remember being in Was it Cambridge?

William: 56:06
except you guys live Cambridge yeah, yeah, the research center here for Microsoft, yeah yeah, I was gonna say, is there a fast flash university around there?

Mark Smith: 56:14
yeah, they. I mean, they got the top, the best in the brightest, right from Cambridge University. They probably got the same thing going on Oxford there. That they are getting these very, very, very smart people to Research things that are ultimately is going to be monetized and bought to bear, and I just think this is a case in point. Yes, ml has been around for a long time and my wife and I had an argument of realm, where does AI? Is automation AI? And it's kind of like very hard to go. You know the difference between what is AI and what's not AI, because if you simplify it down, you could say a lot of stuff that we've been doing for 15 years is AI. But I think if we're looking at large language models and the way that broke on the beach in November, that's taken the world by storm. I think it's a different level. But All that research had been going on for some time to get to this point and it's not.

Ana: 57:10
It's not just research, it's a. It's a, it's a whole Journey, like you guys were saying. You know, before we, before we're able to do this, we Like we did Azure data factory for a long, long time, so we prove that you can actually move data before, way before we have fabric right. So we're not just jumping into oh, let's just use AI and we have, together with all of that research, we we've had huge teams that democratize stuff, like you guys were saying earlier. Right, they bought things and then they they made it available for everyone. And the moment you make it available for everyone is not just the fact that new solutions are built, is the fact that new ideas are being born. And then you build on top of those and then you get, like All of the smarties, big heads, who are creating the very first Low code AI program, and I'm talking about the AI builder. That people you know largely ignored, I'd say. But that created a basis. You know it did create a basis of when that Richard, what, where that research was gonna land next.

William: 58:36
And also pre trained democratized models and how you can do that on mass across a large estate and we're not talking local state, we're talking a Microsoft technology estate rolled out to their clients. And one thing that I think we would be really amiss not to point out time back to everything Anna has just said, plus the the Sachin Adela aspect is his whole concept of hit refresh. You know, 10 years ago he became the CEO and he went from a sort of no, no, no, microsoft can only connect to Microsoft and if you're not going to connect to Microsoft, we're going to make it bloody hard for you to use it to. Suddenly, actually, we should allow Salesforce SAP. You know you should be allowed to put your data where you wish to put your data. But the key point behind that, in the driving factor which I think a lot of leaders can learn from, is he suddenly went what's the what's the what's the Azure team? Was the Azure team not sit with the dynamics team for a bit? Why is the dynamics team not sitting with the office 365 team for a bit? Let's mix these teams together and see what happens if we start saying wait a minute, you're not in your own departments, stuck over in Cambridge or in London or in Paris or in the US. Let's actually start merging some of these different skill sets, keep them together for six months and see what they come up with. And all of a sudden you got this, you know. And let's look at the story of power automate. You know, a hackathon in inside Microsoft with a when wait a minute logic apps is really complex and actually, if we can find a way of making the UI more friendly, we can democratize is further and enable better integration. Let's put a better UI on top of that. And all of a sudden we got power automate digital process automation off the back of allowing that hit refresh concept and trying new innovative approaches. So yes, it's a story I love to tell, but I think you know that is one of the driving factors of what's actually came in, which wasn't always that technology vision but that leadership of let's start this true melting pot of innovation.

Andrew: 1:00:26
Will's rattling off Cambridge or London or Paris or you know the US somewhere.

William: 1:00:31
I've never heard of it before. You brought it off the first time today. Apparently it's quite a big place. I don't know. It's really.

Mark Smith: 1:00:42
Yeah, tell me. I've got a question for you all and see if you can answer this one, and I'm going to go around everybody to get an answer. What have you learned in the last week that you didn't know a week ago? And if it's nothing, why?

Chris: 1:00:57
I think I've learned that there is a lot more, a lot more curiosity about around Things like AI and co pilot than I thought and not not BS curiosity. I mean like genuine and not fear based curiosity either. I think what's happened is that there's been a door that's been opened by all this. Like GBT bought stuff that's come out and, yes, I think there's a fear based thing like your chimp brain automatically fries out and goes, oh my gosh, if we're not going to do this because of screwed. But there are been changing to a number of people in certain organizations where like, yes, you know they're not stupid people like they've seen this kind of stuff coming on the horizon and they actually genuinely curious, curious and they're asking some really amazing questions around like we don't, and being open and saying, look, we don't necessarily understand the use cases, but we did keen on understanding more about really what this actually does in the real world. And I think, like I've had a bit of a perception change from the fierce the fear side of things to more the curiosity side. So I don't think it's happened in the last week, to be honest. I think the last week was the realization. It was more the build up to that. So that's quite cool is that organizations are actually genuinely curious and genuinely wanting to do something awesome, not just freaking out and panicking like they did in the life is over digital transformation era.

Mark Smith: 1:02:19
Nice is next.

Ana: 1:02:21
I learned something completely unrelated to our jobs. I learned two things. Number one I learned how to do a mini fan. I have a friend who is meeting these little fans. You know she's like super successful actually. She's creating the she's creating this arts and crafts books and she started doing them, you know, during COVID and they got out of hand like wildly successful, translated in several languages. When you're an artist and doing creating creative things, often people just like want you to do all of your work for free. So unfortunately she's not getting the success that she, that she should. But I learned how to make one of those fans and that was that was actually really cool and a lot of work. And number number two, I learned that competition backpacks, actual bags, do contain everything. So we just bought like this huge try out one.

Mark Smith: 1:03:37
We need photos you need to share. You need to share a picture, so so so we know what you're talking about. I like it.

Ana: 1:03:45
So it's guiding, it's super smart. They thought about everything. One of the the most important, the most interesting things it's got. It's got a plastic compartment or a rubber compartment where you put all of your swimming stuff and your wet things and all of that and that's completely separate and it comes out but still remains connected. Shut up, it still remains connected. That's where, when you swim, Jesus.

Andrew: 1:04:16
I feel like everyone, everyone can hold it together until Mark starts laughing and then we all laugh at Mark laughing at himself.

Ana: 1:04:25
And that's what I learned nothing, nothing to do.

Mark Smith: 1:04:28
That's good. It's good. It shouldn't be about just what will, will, will, will.

William: 1:04:38
So a few things. One I'll just keep very brief, which is there's a ton of really awesome stuff that's coming out for co-pilot sales in particular, which sales is always. People just assume it's such an easy first party app. I would have to do it, you know leads, opportunities and quotes, blah, blah done, but no, there's some really cool stuff coming out before co-pilot sales and a mirror bell. Gilali has done a phenomenal video on that, so maybe we're, we're, we're tacked that in better. You know she's one to watch if you want to stay up to date with especially sales co-pilot. The other thing was I was a, I was with a client this week and I was with quite senior members of their team and the takeaway from this is never underestimate the basic, the regard, the simple advice you can pass on, because I kept noticing that all their model driven apps they used all the, all the sort of sort of more modern controls, but every time they either as a billion, there was still the yes, no text field. I said you do rise, there is a and out the box or flip switch, for this is it, and they've just always overlooked it. So always, if you think you're going to pass on a piece of information that you think stupid, because I already know, because it's so so, so basic. Just say it anyway, because sometimes you can really get a surprising reaction from it. And then the last thing going with the same sort of vein as Anna, which is nothing to do with what we do is I learn a new level of compassion for wasps, and why that is is everybody hates what's right. They just the annoying things out. You know that attack you in summer, but this time of year they actually get kicked out their nests. The Queen release the pheromone, gets rid of all the wasps, all the worker, worker wasps. They they're not allowed back. She lays one more load of eggs they're all Queen eggs. They get buried away and they go out and they have to fight for their own. They get no more food because it's what they want. The wasps and the Queens mate together. So they're actually, when they come into you, they're just trying to look for sugar. So if you leave a little bit of jam to the side, they'll be happy. Don't wave your arms around, because they release pheromones that then other wasps come over and think they're being attacked. So if you just stay, still, give them a little bit of sugar. They'll get their food and they go off, and I feel sorry for them because I have nowhere to go. So that's what I learned today.

Mark Smith: 1:06:43
I watched that same video, that's what I did on TikTok. Because I was like, if you don't say the pheromones bit, I'm going to say the pheromones bit at the end.

William: 1:06:55
I have a good friend who's an entomologist. I only knew it, but it was the compassion and the storytelling that he delivered it, maybe we should share it, mark, because it was so nice.

Mark Smith: 1:07:04
And this is what's interesting, right is that for everybody hating on TikTok, if you train the algorithm right on TikTok in other words, based on what you interact with you get some amazing insights. And then I also love the random nature of shit that jumps into your feed.

Andrew: 1:07:22
Weird stuff.

Mark Smith: 1:07:24
I don't know what you're talking about, but it takes you down a pathway of and I like it, that that's. You know. So many other social media platforms are echo chambers and I find that TikTok just has a habit of disrupting that echo chamber with new content. That is bloody interesting. I find a lot of stuff around, you know, fasting, for example, and and how people get on diets and things like this that just go nowhere and ineffective, and it's like all of a sudden you're hearing these non-commercial voices in the mix. You know the dietary industry is a trillion dollar business. It's just phenomenal that it's such such a big business. My my learn because and you notice, andrew, I'm leaving you last so you don't say something like fuck or because you get put on the spot at the last minute, so I'm leaving you to the end.

Andrew: 1:08:18
Are you saying that I'm slow? Are you saying that I need? I need time to process my thoughts.

Mark Smith: 1:08:23
I'm not saying you're slow, I'm just trying to give you great mental health for the week ahead by not throwing an expletive into the mix too early.

William: 1:08:33
You've got to do it now, andrew.

Mark Smith: 1:08:36
So here's the thing. Has anyone heard of the spiders web, britain's second empire? Look it up on YouTube. And this British guy team, for 4,000 quid, built this documentary that's I don't know an hour and a half long, and it shows that when Britain was losing its hold on the British Empire, they pivoted and established within the city of London an economic empire, and what blew my mind was a stat. I always thought the US dollar was the currency, was the, if you like, had control over the last, over the largest economy in the world. Right, it actually has about 17% influence over the global economy. The city of London has around 40% influence over the global economy 40%. And do you know what they love? They love that you think that Switzerland safe as a Swiss bank. They love that. You think that because it's kind of like a rate like you're all focused over there where this is where the shit's going on here. The city of London created protections under outside the UK legal system that allows them to basically run as though they're in a totally different geography. The laws of the UK don't apply in that jurisdiction and there's two other city states in the world, right, which is a state that operates within a city. There's a city of London. Does anyone know the other two?

Chris: 1:10:18
Is it New York?

Mark Smith: 1:10:20
It's within Washington. There's another city state within Washington and I don't know. Is it Columbia or something like that? I'm not sure. I don't know. It's the one I'm least knowledgeable on.

Andrew: 1:10:30
I suspect what you're talking about is. You're talking about the. I think it's called like the national capital district. It essentially is the federal land, something like the national capital area or the national capital district, or it's the federal land within Washington DC. That is in theory or not in theory, but is that directly controlled by the US federal government?

Mark Smith: 1:10:54
Yeah, but it's not necessarily anyhow, I'll leave that aside because it's my least knowledgeable area. The other one, of course, is the Vatican City. Right, it's another city state operating within Rome. But yeah, just what I love is that documentaries have been produced like this that are not mainstream media, and the facts and stuff, then the references and the you know, the very researched people that they interview on it is just mind blowing. And they did this whole doco for 4,000 quid.

Andrew: 1:11:28
Wow.

Mark Smith: 1:11:28
And.

Ana: 1:11:28
I tell you, go, have a look at it.

Mark Smith: 1:11:30
It's had something like oh no, millions and millions of views, but it's just, it's a massive eye opener to what you think is going on in the world and it's not what it is.

Ana: 1:11:42
Definitely check it out, man yeah.

Mark Smith: 1:11:45
Okay, Andrew, I hope we haven't sprung anything on you, but what have you learned it?

William: 1:11:52
looks distressed already that you didn't know last week.

Andrew: 1:11:55
I've got three. One of them I thought was funny because Anna was talking about backpacks earlier. One of the things I've learned in the last week is that I am that Anna has banned me from purchasing any new backpacks. I love backpacks and I was talking about buying one the other day and Anna said absolutely not, you are banned. No more new backpacks, you can only buy like six of them a year. So I apparently am banned from purchasing new backpacks. More consequentially, I also learned that in the UK there is a difference between a barrister and a solicitor, which would have been very helpful information for me when. I was flipping out on our residential property solicitor, wanting to know I mean, I'm the American in the equation who just wants my lawyer to go destroy someone in my name. I'm flipping out on my solicitor, on our solicitor, because I think they're being insufficiently aggressive, representing our interests. And then, several weeks later, I found out that barristers and solicitors are very, very different occupations and licensing schemes, et cetera. And then, finally, the other thing that I've learned within the last week and this happens, by the way, every now and then when you live in a country that is not your own is that you go on and you finally realize after a time oh my God, that would have been really useful information a month ago. But my final thing that I have learned because I've sat down this is relevant to the podcast and I've really started to put a lot of time into writing this enterprise architecting an ecosystem white paper that the five of us have been talking about and it turns out that explaining ecosystem architecture is really, really hard. So this is a writing assignment that I'm not struggling with. I'm just finding that it is much harder to explain the kind of concepts that we're talking about here than to explain them in writing than I think I expected. So the paper is coming along nicely, but it is quite a challenge. So that has been definitely a key learning of the last week for me.

Mark Smith: 1:14:25
Fantastic. I want to share another little. I don't know in the news type item and I don't know if any of you are any else any, you fellas? I don't know if any of you have something else you want to share on your screen, but let me share this. It came out two days ago, which is pertinent to most of the people probably watching this, which is a post from techcommunitiesmicrosoftcom which is introducing new results for all role-based Microsoft certification exams. In the exam you are now allowed full access to Microsoft Learn while sitting a Microsoft certification Open book Within the interface. There's a little button here that allows you to click and open Microsoft Learn. You can't log in so it won't let you go to anything like all your favorite, but you can then search anything on Learn apart from the Q&A section and support in your exam. And they kind of liken it to you know, years ago the calculator come out and all the maths exams are like you know now calculate us. We don't want people cheating and of course now if you've walked into an exam without a scientific calculator on math, you'd be stupid and I just think this is a good, practical way, like a lot of people know their shit but they freak out at doing exams. They can't hold a lot of stuff in, you know, in memory, if you like to answer exam questions, and so it's going to allow you to use that resource because, like in the real world, if you run into a problem, you're going to look it up right, you're going to do some research, and I just think that it's a bit of an evolution on the exam front to allow, you know, full access to Microsoft Learn. You're not allowed to link anything off there, they're blocked. It's only within the Learn domain that you have access to. But interesting changes that will be rolled out by the end of September. My understanding is to all roll based certifications. So that's not fundamentals exams, it's roll based certifications explicitly. So just a little something, something for you to bite down on.

Chris: 1:16:30
Useful. That is a poison. That is a poison. That is a poison chalice.

Mark Smith: 1:16:34
Yeah, what do you think? What do you think?

Chris: 1:16:37
It's a poison chalice. I currently write university exams right now. Yeah, I write university exams right now on a time limit basis and you can end up in a whirlpool of research If you don't know your fundamentals and you don't know your stuff. I can categorically tell you that's a poison chalice For me in an exam. It should be a last resort. Honestly, I failed Igneous Petrology because of exactly that.

Mark Smith: 1:17:06
I failed to know what Igneous even stands for.

Chris: 1:17:08
Yeah, it means the transformation of rocks under the surface of the earth. But basically what can happen is that in a multiple choice exam there are answers that are you know. You're basically going to struggle from a selection perspective. If you go and read it up, microsoft documentation is not exactly uber clear. So you can spend about 10 minutes per question if you don't know your beans and what will end up happening is that you'll end up spending hours on one question. Well, minutes on one question and not enough time actually running through the exam. So I love it, it's great. I think it's good if you really, really don't know something, but to me that's a poison chalice.

Mark Smith: 1:17:51
But everyone has their exam techniques right, which is you go through, you read the question, you click the obvious, answer anything you can't like light bulb on, you quickly market, skip over it and you get to the end, unless there's some of those questions. Then you go redo it again, and then you're only dealing with the delta right, the stuff that you're not 100% indexed on.

Chris: 1:18:10
So, my dude, this is dangerous, because what's going to happen is there will be far more people paying for exams that don't know their carrots, and then they'll take the exam, they'll try and research during the exam and fail, and then they'll have to do it again.

William: 1:18:24
So it creates almost a forced inertia for those people. But I do agree with Mark and I know Chris, you agree with him as well which is actually for the people who use it appropriately, that's a very useful tool. But for those that think that that's they can pass the exam because they don't need to know it, then that you know they. Or if they think Microsoft because they're democratising everything else, they've democratised exams for them.

Mark Smith: 1:18:45
It's not going to help. On the situational question front, right, You're in the situation. You have this. Good luck looking up that explicit situation on Microsoft Learn.

Chris: 1:18:53
Yeah, like I said, it is a poison chalice.

Andrew: 1:18:56
I make feelings about this, but I think it. It it returns back to my original thesis on exams, in that I think that these exams and these certifications, I think that they're they're terribly overrated. I think that exams and certifications are a great way, particularly if you're looking at a topic or you're at a point in your career where you haven't had the opportunity, through experience, to demonstrate your knowledge, to demonstrate, you know, the experience that you have, etc. Etc. But I think that a lot of people waste a lot of time taking these exams at all. Right, and that's not a criticism of the exam. That is a criticism of kind of this obsession like how many badges can I put on my profile? Now, microsoft kind of creates the conditions where this is, where this is in some ways necessary. Right, because a lot of partners, in order to have a certain specialization or whatever, you have to have taken the exam. But I think that probably, these exams and these certifications are 50% nonsense. So it's sort of like we've democratized nonsense.

Mark Smith: 1:20:06
Definitely sound like a criticism to me. I heard a criticism right there that was definitely a criticism.

William: 1:20:10
You should name a podcast which is Microsoft certs are useless.

Mark Smith: 1:20:15
You know what shocks me? I see a lot of CVs. Right, we're always hiring, we're always and I see people that have been in the game that we're in for five minutes and how their CVs look like they've been in it for 15 years. And you see others that have been in 15 years and it looks like they haven't done anything in those 15 years. And I am shocked and appalled about how so many people in our career space do not take their career seriously or, if you like, they don't chart their own course. Their knowledge is at the limit of their last project and we were asked to do X, so I research X. Like there's no kind of commitment to you know what in the next three years, my career is going to go here. So these are the things that I need to do and these are the things I need to show and these are the things I need to demonstrate. Everyone is like a ship on the ocean with the rudder cut off and they're just doing whatever the client asked. Next, they researched just one step in front of their nose and I'm like no wonder we're not getting really amazing consulting or anything, because people just think you only have to stay one step ahead because you've been told just stay one step ahead of the customer. That's all it's needed. What about you know what? If your doctor was like shit? I just watched the YouTube on how to do this brain surgery just before I'm ready. Yes, you'd be like negative mate, you're not doing my brain surgery yet. So many people I feel. I look at these CVs and I'm like and what are they asking for? $10,000 more than their last gig, right, and they're and I see this 18 month two year cycle showing in their history right.

Andrew: 1:21:57
Mark of those New Zealand dollars or those real dollars.

Mark Smith: 1:22:00
Oh, so well, it's monopoly money. It's been real monopoly money, right, and it blows my mind at just the lack of people thinking about their career progression outside of. You know getting a pay increase, but you look at what's there and you're like ugh.

William: 1:22:21
But it's pay and roll, because they're chasing to get to architects as well. It's quick as they can. Oh, everyone wants architect. Yeah, I've done it. I've done it for two years now.

Andrew: 1:22:29
Everyone wants architect, except for the people who don't. I don't want to be just an architect. So it's funny You're either dying to be an architect or dying to not be an architect.

William: 1:22:39
I want to work in the strategic position that I pulled up on.

Mark Smith: 1:22:43
Yeah, Another one I pull up on is techno functional. I'm a techno functional person and I'm like techno functional. I mean, what is this invented title for yourself?

Chris: 1:22:59
It sounds like a shitty dodge that you find in a crappy EBM club.

Mark Smith: 1:23:03
You know they're saying they're functional but they can't develop. They don't know any. But I'm about to be an architect. I'm like one project away and I'm like you know, you're 15 projects away from architect. You don't have enough experience.

William: 1:23:15
Mark, we've got the next podcast, haven't we Recruitment nightmares? Yeah, we should. It'd be a full podcast and that'd be amazing, yeah.

Andrew: 1:23:23
I'd put it on record in front of the whole Microsoft community or anyone who bothers to listen to this podcast and say that I happily right now have no control and no involvement in hiring or personnel of any kind and that is a happy thing for me. But if I ever do in the future and I see a CV that comes across my desk and uses the phrase techno functional, you're done Like an automatic DQ for me for use of the word techno functional.

Ana: 1:23:54
We were talking about this because Andrew was rather obsessed with managing as many people as possible, and I'm rather the opposite. I much prefer a dotted line and to me, leadership is very different than people reporting into you anyway. But what I will say and I will take us back to exams is that I actually don't think that exams are that bad. I've been taking Microsoft exams for a long, long, long time and they used to be absolutely horrible. No, you're amazing. I feel like I should get on a glass of wine right now. But now they're actually pretty good, and I think that the reason why they can give you the manual right now and they couldn't have definitely five years ago is because right now, a lot of the exams make sense and fine, you don't take exams, and maybe we don't take exams anymore, maybe because we focus more on the consultancy and on other skills and so on and so forth. But we still need the experts. There still are people who are okay, I'm focused on X and this is what I do. I know incredible Power Apps developers and I know amazing automation developers and dot net developers and so on and so forth, and I think exams are pretty useful because when you're embedded in, like you were saying, mark. When you're embedded in a project and that's what you do, and your project has been going on for like three years the only thing to do in order to focus your career, on your career, even a little bit, is to take an exam.

Mark Smith: 1:26:01
Exactly. It's a form of learning, and even you know there's, of course, the dark side of exams, which no one ever talks about publicly, which is the cheat sheets and stuff that are out there that people absolutely tap into. I'd love to see the stat of how well I knew. I saw something a while ago where, back in Windows NT days, some fella created a cheat sheet and he made over a million dollars in selling that and Microsoft sued him back in the day, and so if you think that was like 15 to 20 years ago, how massive is it now? Right, like that little underground ecosystem of cheat sheets and stuff. But at the end of the day, do you know what? If someone's committing all that stuff to memory, it means they've got a reference point in their brain that when something comes up in a project, they're like yeah, I remember this, I remember cheating and learning this, and they can go find the actual bit that's going to help them today, right, so I don't think it's necessarily all bad.

Ana: 1:27:11
Yeah, because you get, because you've got like 45 questions Used to have 90 questions, do you remember?

Mark Smith: 1:27:17
that.

Ana: 1:27:18
There used to be 90 questions and the score to pass is 700 points. So for 90 questions you can be wrong quite a bit and still pass your exam. You know, because there are many questions Right now there are 40 to 45. When you've got 45 questions all of the sudden, you can't really get that many wrong, and a lot of them are quite complex and a lot of them really do make you think you know about the answer. So yeah, I think they're pretty useful.

Mark Smith: 1:27:57
It's good they have to draw a line under this. We've been cranking on for a fair wee while, so let's wrap it up with a quick little. Let's test our knowledge, and I've just done this on the fly, and so, andrew, I'm going to give you the last question. We all know why we need to do that. But we'll go with you at the end, right? Well, I'm going to give you this question, although perhaps I should give it to Anna. In fact, I'll, either of you, jump in if you're not. Articulate Vector database, what is it?

William: 1:28:31
I've just read a huge article on this because Kyle Hill brought it up to me and, for the love of God, I wouldn't be able to be eloquent enough to place it right now. Anna, did you want to give it a go, or is Mark just tripped us up?

Mark Smith: 1:28:47
It is new right, it all has to do with this area of AI. According to Microsoft Learn, it says a vector database is a type of database that stores data as highly dimensional vectors, which are mathematical representations of features or attributes, and one of the capabilities of it is speed around when you catalog and index data and make that highly available to things like large-scale it can jump across reference points, can't it?

William: 1:29:16
So, instead of actually having to do a much more of a linear or horizontal, it allows you to have lots of categories and search faster. But that is a really underwhelming explanation of what it really is, because if you look at a lot of the investment rounds, a lot of businesses that go out there, vector databases at this point is huge. I think it would be good to push some articles out from us under this podcast.

Mark Smith: 1:29:41
So people can actually learn better than we have. Okay, here's the next question. This one's going to be for Chris Andrew. Yours will be the last question.

Andrew: 1:29:52
We all know why You're going to remind me of this every time.

Mark Smith: 1:29:58
Chris, what does GPT stand for?

Chris: 1:30:01
That is a great question. Give me a second. Gpt stands Global positioning technology.

Mark Smith: 1:30:08
I didn't ask Will because Will actually mentioned it before.

Chris: 1:30:12
Jitter, it's a pre-trade transphaleman.

Mark Smith: 1:30:15
Fantastic Prometheus, prometheus. Well done, well done, anna, here's yours. What does LLM stand for?

Ana: 1:30:24
Really yeah. Low language modeling.

Mark Smith: 1:30:28
Yeah, fantastic, that's perfect. That's perfect. I wasn't. Every question doesn't need to be a hard one. I just got my order mixed up because that was actually Andrew's question.

Chris: 1:30:41
I gave it to you earlier, Mark. What does LLMA stand for?

Mark Smith: 1:30:44
Oh shit, I don't know.

Chris: 1:30:47
LLMA asks me anything.

Mark Smith: 1:30:53
What does LLMA stand for?

Chris: 1:30:56
Yeah, so a LLMA model is effectively large language model and it's AMA.

Mark Smith: 1:31:01
Got you Ask me anything? Yeah, got you, andrew. Yes, describe to me generative AI.

Andrew: 1:31:10
Oh God, generative AI See.

Mark Smith: 1:31:14
I didn't swear. How would you explain that Some people would say bless beaming is a swear word.

William: 1:31:23
Great yeah, I almost said the same.

Andrew: 1:31:28
So the way that I would describe generative AI is by distinguishing it from the previous generation of AI, which was very much sort of like it was logic-based AI right, we're going to train the AI model in something very specific and it's going to essentially access or select whatever pattern we have intentionally trained into it, right? And kind of run it goes back to the decision tree. Essentially, the previous generation logic-based AI was a giant decision tree and we trained the AI on every possible outcome, right? So generative AI. What we're doing is we're feeding the model the entirety of all of the data and the knowledge that it can access and we're essentially allowing it to train itself on a vast trove of information, and what has made this possible nowadays is that the model can process essentially the entirety of human knowledge out there on the internet.

William: 1:32:38
So I would describe it as but the key thing is it can then generate the text, the image, the video off of that. Then right, that's in its simplest terms.

Mark Smith: 1:32:46
Okay, who wants the last word for today? Who's taking it? Last word for the day.

Chris: 1:32:51
Lomit does not stand for large language. Ask me anything. It stands for large language model, meta AI.

Mark Smith: 1:33:01
I've never looked at that and we're done, we're all set because we're ending on that. Hey, thanks for listening. I'm your host business application MVP Mark Smith, otherwise known as the NZ365 guy. If there's a guest you'd like to see on the show, please message me on LinkedIn. If you want to be a supporter of the show, please check out buymeacoffeecom. Stay safe out there and shoot for the stars.