Navigating Life's Surprises and the Dawn of AI with Chris, Ana and Andrew: A Festive Fusion of Technology and Tenacity
Navigating Lifes Surprises
Ana Demeny
Andrew Welch
Chris Huntingford
William Dorrington
FULL SHOW NOTES
https://podcast.nz365guy.com/539
When life throws a curveball, like the mysterious absence of our guest Will Darrington, it's all about the pivot—a theme that resonates throughout our latest episode. We kick things off with a spontaneous round table, riffing on the unexpected nature of life's ups and downs. From Chris Huntingford cautiously navigating the 'trap' of discussing his recent joys to me opening up about a tough day's vital resilience lessons, this conversation peels back the curtain on the very real emotional landscape we all navigate. We then veer into lighter territory with Anna Demeny and Andrew Welch, who regale us with tales of over-the-top festive decor and the undeniable allure of reality TV as a mental vacation.
Ever pondered how AI could flip entire industries on their heads? We sink our teeth into that juicy topic, examining the profound influence AI could have on fields grounded in mathematics. Expectations meet reality as we scrutinize Microsoft's strategic AI manoeuvres and consider how these advancements are reshaping roles like executive assistants. Plus, we don't shy away from the bigger picture, tackling the exciting yet risky business of integrating AI into operations. This episode is a melting pot of tech insights, from the promise of AI customization to the parallels drawn with the historic adaptation of tools like Microsoft Office.
As our festive episode winds down, we don't just hang up our stockings and call it a night. Instead, we delve into the pressing issues of a two-tier economy emerging from disparate tech adoptions, the ethical crossroads of consulting, and the nostalgia of technology past. We even manage to sprinkle in our tech wish list and the profound potential of Azure Integration Services, all wrapped up with a bow of early morning Christmas cheer. So pour some eggnog, settle in with your family—or your cheese plate—and join us for a conversation that's as rich as fruitcake and as sparkling as the holiday lights.
AgileXRM
AgileXRm - The integrated BPM for Microsoft Power Platform
90 Day Mentoring Challenge April 1st 2024
https://ako.nz365guy.com
Use the code PODCAST at checkout for a 10% discount
If you want to get in touch with me, you can message me here on Linkedin.
Thanks for listening 🚀 - Mark Smith
00:00 - Power Platform Show Chat
10:18 - Future of AI in Technology
19:23 - Balancing Risk in AI Investments
29:44 - Emerging Two-Tier Economy and AI Opportunities
42:48 - Digital Equity and AI Implementation Risks
52:20 - Consulting for Customer-First Solutions
01:06:48 - Book Recommendation and Technology Nostalgia
01:16:20 - Exploring Azure Integration Environments
01:28:54 - Early Morning Christmas Greetings & Farewell
Mark Smith: 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. Welcome everybody to the end of the year of the ecosystem podcast. Unfortunately, will Darrington is stuck on a train between Scotland and his home, so it's not going to be able to join us today. Well, that was the excuse that he submitted. Being that Anna works with them, she was like that's funny. He wasn't going to Scotland today. So who knows, maybe he's just helping centre with his.
Mark Smith: Alps. But that aside, Chris, why don't you tell us why life is so fricking amazing at the moment?
Chris Huntingford: Oh man, there's a lot going on.
Mark Smith: That's a trap that's right in, I need to tell you for this one man.
Chris Huntingford: it's a trap. Life is so good right now, mark and Anna and Andrew. I'm very happy today. It's been wonderful.
Ana Demeny: Ha, ha, ha ha. You look like you've had a good day.
Chris Huntingford: I've had a shit day, but it's cool, we'll survive. No, it's been one of those days where it's like I've had to make some decisions, where other people have failed to make decisions. Yeah that's how we're, yeah, that's where we'll leave it.
Mark Smith: I remember having a boss years ago. This boss years ago said to me listen, if you didn't have shit days, you wouldn't know what really great days feel like. And so you know it's part of the cycle, right, we have great days and but it's hard when you're depending on people, right, and you get let down.
Ana Demeny: But we're always depending on people, right? Who's not depending on people?
Chris Huntingford: So I've always had this thing where I'm like, if I can't, like I'll always have a control freak, so I'll do it myself, because I know I'll do a good job, and like it's just this one thing where I don't give things away and I was like, ok, I'll give things away, like I'll let, I'll rely on people, and then when I do it, nope, nope, nope, nope. So yeah, I'm a little bit aggro right now. So if I get ranty more so than usual, it's because of a shit day.
Ana Demeny: That's OK. Andrew's really distressed. He had his second sip of wine.
Andrew Welch: Oh, that just means I'm behind.
Mark Smith: I am not sure, but on the last podcast you might have mentioned, there was some announcement that you're going to be making shortly. Yeah, is it? What are you up to?
Andrew Welch: What am I up to Like what? What am I doing with my time? Or what am I doing with my life?
Mark Smith: I think that day of the recording you were going to say something to your boss. I don't know, I don't know. Yeah, what was it, what was it?
Andrew Welch: So are we? Are we talking about? We're talking about my professional choices? I assume not my. Ok, not my choices in patriotic Christmas stockings Right.
Mark Smith: Honestly, you know when you put that sock up of the stocking sorry of the US flag, you know it's the time of year where you just do some trashy things right.
Chris Huntingford: And I'm saying that that's trashy. Of course that's trashy, that's trashy.
Mark Smith: Meg and I have been watching this. You know this series of I don't know even know what the name of it, but they go into these rooms and they talk with people without seeing them and decide if they're going to marry them, and then they get engaged and then and they haven't seen the person and then they meet and stuff. Oh my God, I love that.
Ana Demeny: It's an American love yeah.
Mark Smith: There you go, that's that's it.
Ana Demeny: No, no, no, but there was married at first sight, is it?
William Dorrington: Is it the rooms?
Mark Smith: I think it is lovers, oh, yeah, yeah the parts, exactly, we also love that. And there's one dude and he is American. Everything he has American socks on. He always has an American like American flag shirt on.
Andrew Welch: Yeah, he was a terrible kisser, it was. It gave those very hacking.
Ana Demeny: Yeah.
Andrew Welch: Yeah, I'm terrible.
Mark Smith: We're really fresh TV guys. So that's, that's what it reminded me of. Why would somebody watch that? Honestly, because you're fried at the end of the day, you've just got your kids to sleep after hours of a not bloody sleeping and you're just like I cannot think about anything and I can laugh at these freaks that go on TV and do the shit.
Chris Huntingford: Yeah. I can't look at with you, man. I downloaded an old 80s cartoon that I used to watch when I was like eight years old or six years old. I watched it.
Ana Demeny: The other day I like that's probably smarter yeah that's probably smarter.
Mark Smith: Yeah, and though how many sitcoms etc. Have not aged well right as in, like I think of Little Britain. Right, little Britain used to be my favorite. Go to now. You watch it and you're just like, oh my gosh, like the amount of racism, homophobia gender stuff, oh just extreme.
Andrew Welch: My all time favorite sitcom is how I met your mother, and every time I watch Anna and I over the last couple of years have rewatched all 10 seasons of how I met your mother and we say all the time, usually having to do with Barney, it's like, oh my God, that that's not a, that is not acceptable.
Ana Demeny: That's normal. You know, I don't know yeah.
Andrew Welch: Yeah, Misogynism would be is a kind description. So yeah, still a great show though.
Mark Smith: Surprisingly, we watched a movie the other day which I think was is it what's is an American pie, where he goes and puts his dick in an apple pie? Yeah Right.
Chris Huntingford: Well, so it's a Tuesday.
Mark Smith: And, and the thing is the thing is that movie we thought would not have aged well, yet it was actually it was OK.
Chris Huntingford: It was like nothing, Because everyone still eats pies these days. I mean it's so relevant.
Mark Smith: Chris Andrew is just telling us before how much he enjoys a pie. I mean, he's got no comment. No, like this. Oh, this is awesome, this is awesome, love it. Back.
Chris Huntingford: Back to you. Throw this out there. You just got taken to school by Madness Ball. Balls of earrings.
Andrew Welch: That is exactly that's the kind of man I want to get taken to school by, so nice, Tell me.
Mark Smith: tell me who's in the most Christmas spirit. That's your own magical shirt. It's first thing in the morning. I haven't even been drinking. My last gin was like nine hours ago Wow. Thirty nine hours ago. Yeah, well, it was the day before right.
Andrew Welch: Yeah, I quit, early yeah but, it was a problem.
Chris Huntingford: It's what you like about.
Mark Smith: Yeah, anyhow, something I want to discuss with you all.
Ana Demeny: See, Mark, that's why. That's why you're always so fresh, so full of ideas.
Andrew Welch: He is the freshest man on this podcast.
Mark Smith: It's the morning. You guys are at the end of your day, right, oh?
Andrew Welch: my God, yes, I'm at the end of my tether.
Ana Demeny: Like we had. We already had a full day of work. We unpacked some because we were still moving in, right. You know, we've had an argument, we've had dinner, we, we've been through everything already.
Mark Smith: Is that why you're in separate rooms? No except for that point, it's got to play. Come on tell you.
Andrew Welch: Yeah right, exactly, exactly.
Ana Demeny: I don't get that, and I was, he was he and I?
Mark Smith: was he in the wrong? Was Andrew in the wrong? And the other was how? Of course he was. We knew the answer before we asked the question.
Ana Demeny: I mean, of course, what was the one with that question? Well see, we were talking about things that are never going to get old. This is one of the things that will never get old.
Andrew Welch: Me being wrong.
Ana Demeny: Yes.
Andrew Welch: Nice, I'm happy. I'm happy to know at least that that things not getting oldest within my power. So I will carry on being wrong. That's my commitment to everyone.
Mark Smith: And being young.
Ana Demeny: Being young, you said, not being old but we can get old.
Mark Smith: Yeah, right, right, what. What do you think is your favorite theory on what went down on open AI with Sam and his firing? Because there's so much speculation and no real transparency. Right, what's your favorite theory?
Chris Huntingford: Somebody panicked. Somebody, somebody panicked. I'm telling you, somebody was like nope, nope, nope, nope, nope. This man knows too much, he's too smart. Everyone's so clever. What do we do?
Mark Smith:
Chat, get rid of him, and yeah do you think that they got you know? There's the whole thing of Q star, q star being it could, it could handle math calculations brilliantly. In that they knew that the play forward was as if you can understand maths perfectly. All of a sudden, a lot of most science is built around maths. Right, most financial system is built around maths. All of a sudden, if you've got a AI that is a perfect mathematician did that and they say Eila, you know, the lead guy in the tech side went kind of oh fuck, he could see where this could go.
Chris Huntingford:
Yeah, they thought that we're driving towards sentience AI, right, which is we're very far away from that. I think I get why, right, but I don't. I think that's part of it. I always think that with these types of scenarios, there's a left and right hand thing that's happening around, and I think there's definitely got to be money that's being passed around and it's definitely going to be a power play as well. That, to me, was just an excuse that was like, oh Fine, we need to do something, so we'll do this.
Mark Smith:
Yeah.
Andrew Welch:
I wondered also the extent to which and I'm going to maybe disprove what my original theory was here right that the vision for AI inside of open AI, basically, is this groundbreaking technology that we have created? Are we squandering it by building a bunch of co-pilots to infuse inside of existing Microsoft products? Basically, has the vision for this magnificent technology become too narrow, and I could certainly see someone taking that position if you were going to be really aggressive with how the technology is used. What kind of refutes that idea, though, is that, immediately following, most of the company wanted to jump ship and go to the very company that was building this Microsoft?
Ana Demeny:
Yeah, and at this point in time we're building co-pilots to co-pilots. So even the statement of narrow, really I'm not sure can we even define it anymore.
Andrew Welch:
Yeah, I think Microsoft is doing this so interestingly and there are some things that I think will be revealed to the world from Microsoft perspective in the fullness of time, as they would say. But I do think that there's lots of evidence that Microsoft is really doing a good job of balancing the co-pilots and the co-pilots for co-pilots Sort of this non-sexy what I refer to as incremental AI that is way less sexy than you might imagine but still damn useful. But they're doing a really good job of balancing that investment with investment in some pretty amazing things that are not my story to tell Again fullness of time but I'm impressed with how Microsoft is handling this.
Mark Smith: The thing is it was interesting. I don't know if you saw the podcast that Steve Mordue did with Charles Lamana, but it's worth a watch because there's been some confusion around is it co-pilot for M365, is it co-pilot for Dynamics? Blah, blah, blah and the naming conventions. And I like how straight up Charles was. He was like mate, this is moving so fast, like yes, we've kind of this has been developed in flight type thing, and so what we said back in March, we've thought about it a bit more and we're iterating on it, and I think that's you've got to allow them the grace to iterate, right, not everyone's getting this. It's a moving target. I mean, I couldn't believe it. We just celebrated the first birthday of ChatGPT the other day. I shit down, this happened in this past year.
Mark Smith: I get people now reaching out to me and go how does it apply to my business? Like, how does it practically apply like to my business? And this one lady I spoke to yesterday in Australia she's an executive assistant and they just had a conference of executive assistants and or someone to that effect, and the discussion that came up is that some big research organization like I don't know who it was, I won't say a name because it won't confuse things had said that 54% of all executive assistants will be replaceable by AI and that you and she said it set a rocket through this community of. We need to get up to speed and become experts on using AI in every part of our life, whether it's scheduling, whether it's absorbing large amount of information and putting it into a way that's consumable for whoever they're in EA to like.
Mark Smith: What are all the different ways? And so you know, I actually did a side call and just showed all the ways that you know you can use it to create presentations to. When the presentation is like a one off on the fly and internal one, not a public facing one, how can you look like a real guru in it and how do you use AI and stuff. And I'm really big on on practical AI and at the moment I feel like a lot of co-pilot stuff still sits in the realms of wow, good demo candy, amazing. Some of the stuff, chris, is it coming out of your team? I think I'm got them on the podcast next week.
Mark Smith: It's massively practical writers and and I love what they're doing in AI. And then, and then I tell you another one If you look at the GPTS and open AI, I went and took and I created a GPT of. I took the 700 page success by design document, fed it to it right, and now you can just query it on anything. Give me an example of an ERD it builds it out on Python for you, right, and it's, and you know, showing relationships and explain to me. You know, and it does it and it's because it's trained on that.
Mark Smith: And then this week, microsoft well over the probably the last six weeks, they've dropped a shit ton of playbooks. I don't know if you've seen these. Yeah, yeah, heaps of playbooks for anything for Azure, architecture, for ALM. You know best practice for, and like I'm like all those you can now put into your own. You don't have to make it publicly available at GPT, just use it for your own team. Take all that body and knowledge, put it in there and keep feeding it over time. I think you'll create a valuable internal resource for you know, I'm just thinking.
Mark Smith: Take RFP responses, creating really articulate answers to the questions that are absolutely on point, that don't go into things like your staff. Have they had a background check and, you know, do you veer your staff before coming into the organization? This is the correct answer. This is not the correct answer. You know, I think it's really important to go yeah, how can we use technology? You like that injury? That's a private joke, by the way. Yeah, we're not going to go into details of it.
Andrew Welch: So trying to try to get myself serious about this again. So I told you guys earlier that I just finished yesterday the copy for a white paper that I've been working with some folks from Microsoft on, and the assumed title is going to be crafting your future ready enterprise AI strategy and the whole let's get. When we get back together in January we can dig into the paper itself, but basically what it's meant to do is to walk technology leaders, cios, cisos, cdos and, you know, for that matter, chief financial officers, their teams as well, you know to walk big organizations through how to build and then execute and action their AI strategy, rather than just like I want this and I want this and I want this, chris, can we get some raptor pants on that, thank you, so exactly. Anyway, the paper makes the distinction between incremental AI and differential AI, and the distinction really is that incremental AI is AI applied to helping a human become more precise, more efficient, more accurate, faster, whatever it might be, in a task that a human would have otherwise performed, and that's where a lot of the co-pilots fall in, whereas differential AI we could think of that as being aspirational or sort of moonshot AI. Those are the things like deep mines, alpha fold, that AI model that they built, or you know some other really interesting things that I've seen about classifying and interpreting, pouring through more medical data than anyone human possibly could take on in order to extract diagnoses and patterns based on where people live and sort of other factors like that.
Andrew Welch: And the reason that I think, at least right now, that this is an important distinction between incremental and differential is because organizations need to be looking at balancing their balancing risk across their portfolio. The differential AI are the bets that are less likely to work out, at least in these early days. So you need to kind of adjust the risk tolerance within the organization. For how much do we want to implement that is going to make a real impact for us, making us more precise, more efficient, faster, whatever, versus what kind of experimentation do we want to do? What kind of impossible problems do we want to take on? And I think that's to me that's a useful model for kind of looking across an organization's portfolio of AI investment.
Ana Demeny: I think that's really cool and very useful, because we've already been talking about so many types of AI and I don't know. This has been recording for like 15 minutes and maybe we talked about five minutes about real stuff and not, you know, merit or dispute. We've already mentioned so many types of artificial intelligence and it just makes me think like, do you guys think, yeah, that is my favorite wine.
Ana Demeny: Do you guys think that, like there's a real business to be built here, because I'm still dealing I don't know about you, but for people to look at differential AI, that's like super overwhelming. So what Mark was saying earlier and showing someone how to build, like a presentation or how to feed any sort of AI engine, a you know, a set of materials, can you imagine like to be able to ask a licensing question and someone to come up with the answer? Like no one can come up with answers about licensing and power platform right now, right, or no one I know at least, I definitely cannot. So that's really really important. This way, it's more tangible for people to go ahead and read Andrew's paper and think, oh, I can actually come up with some sort of AI strategy. I can actually do that.
Ana Demeny: And the other thing that's really really interesting is, like you were saying, Mark, the grace to give people the chance to iterate, and that's not just for artificial intelligence strategy and work, that's for everything. Do you guys still work with people who are like no, I need to see a plan and the plan needs to be, it needs to have certain steps and we need to have a high level of confidence with it, and you better take into account all risks and considerations and litigation scenarios, and only then are we going to be able to build on something.
Ana Demeny: So there's no iteration space there at all.
William Dorrington: Yeah.
Ana Demeny: That's just not reality anymore. Like, how do we deal with that?
Mark Smith: Or flexibility in thinking.
Ana Demeny: Or flexibility in thinking. That's right.
Andrew Welch: This, I think, is a really important principle that, again, we'll get I'm sure we'll get into more when we dig into the paper.
Andrew Welch: But particularly in differential AI, organizations are going to need and I'm not thinking Microsoft and Google and partners, I'm thinking the customer organizations they really are going to need to make some serious changes in how they budget and manage and measure the effectiveness of their IT investments.
Andrew Welch: Because if you treat development in AI the way you treat the implementation of a new ERP, you're cooked. You need to treat your investment in what is, at this point, still an experimental technology, not because it's not brilliant and not advanced, but, mark, you were recounting Charles and Steve's conversation about how we don't know exactly where it's going and it's a moving target. What you need to be doing is you need to be putting a few chips down, structuring the work in a way that, hey, I need to see some sort of result in the next two to four weeks and then we'll decide if we're going to put more chips down. You also need to have the discipline to walk away from something that isn't working out, either because the technology is not there yet or, I think, because the data that you have to train the model isn't there yet, but you need to have the discipline to walk away from a sunk cost as well.
Mark Smith: You know it was interesting. You said if you treat it like an ERP, it's not going to work, and I'm just wondering if a lot more companies should be taking on the idea of we need to do our own internal R&D on how this applies to our organization and think of it like that. We need to do some research, we need to develop our ideas. We shouldn't just be going, oh, let's jump on the bandwagon and implement AI. I think there needs to be a lot more room for research and research how it applies to your business. What are the use cases that are going to work? Because it's not like anybody's just woken up and got I've got a million use cases for this right. It needs the context of your data and the outcomes that the business wants to drive, and I think there needs to be this kind of research facility introduced back into businesses that haven't done it for so long to actually fund trial and error to find out.
Ana Demeny: But do you think that without that research capability is just easier for people to say, well, I gave you company X all this funding and we talked together and I didn't know anything Like you were supposed to be here and advise me and tell me what to do and you didn't tell me, or you didn't tell me well enough, and it's clearly your fault. So I feel like, to some extent, the lack of internal R&D does come from this opportunity to avoid accountability in a way.
Mark Smith: Yeah, valid.
Chris Huntingford: It's quite scary. Actually, I think it is really scary. We have this whole thing where I'll try and give you a scenario, right? Do you remember, man? This is years and years ago when Microsoft Office was released? Okay, like one of the versions, let's just pick 97. Okay, that's an easy one, right?
Andrew Welch: I was really sad that Microsoft Works was going away at the time, and it's just so hard.
Chris Huntingford: Yeah, damn. And when they released it, there were no R&D departments. People were like you know, like this is a business productivity tool we have to use. It's just something we know we have to use. And I don't think many companies well, I don't know of many companies who took, who dragged Office through an R&D cycle. It was just okay. And then I don't think there are many businesses that put ROI cases against every email they sent to every Excel they built. They were just like this is a tool we have to use, like it's something we have to use.
Chris Huntingford: Let me tell you, with AI, right now, we are lucky we have a prompting screen, because in a two years time, there will no be no prompting screen. This will happen to you, right? So if people, if R&D, r&d parts of the organization that exists now, like innovation and R&D which is why I am interested in innovation if people don't start taking this stuff up, like actually it's going to be far more detrimental than they even know, because this will happen to them in their business without any research or any plan. Okay, and I think like and you made a good point right Like it's up to us to tell people this. Unfortunately, the Ogden people that don't listen. But that's on them, right? Like we can only say as much as we can, right? And I think, like that R&D function has to be paired with a return on investment function. But it's not, it can't be on every single little thing you do.
William Dorrington: It has to be on the wider story and actually like productivity.
Chris Huntingford: What does that mean? So that's just what I've started to see, right? So when we go into companies, we actually tell them ROI per use case is bullshit, because it won't work.
William Dorrington: It's low code.
Chris Huntingford: Like, we will ROI every one of our 25,000 apps and flows that we make Crack on. Good luck.
Mark Smith: Yeah, yeah.
Andrew Welch: You'll destroy the R if you do that.
Ana Demeny: Yeah, yeah.
Andrew Welch: Yeah.
Ana Demeny: Right, right.
Andrew Welch: There's some statistics that I've shared in a number of presentations recently. Chris, I think you were there when we shared these at the emergency services tech show in.
Chris Huntingford: Yes, great presentation, by the way, dude, really strong presentation.
Andrew Welch: Thank you and I love, I think, that these are so interesting, so in and I have to read these specifically so I get my numbers right. But in the United Kingdom, between 2010 and 2019, there was an 11% average worker productivity rise in the most productive firms, whereas the least productive firms saw no rise. In Canada, productivity growth tripled in the most productive versus the least productive firms from 2000 to 2015. And in the US, the 75th percentile of firms. So if you look at the firms in the 75th percentile, in terms of their investment in R&D and modernization and technology, they enjoyed a return on investment capital 20 points higher than the median and that gap had doubled since 2000. Yeah, so the reason I love these three statistics is that it bears out that the organizations that are spending and spending wisely, not just spending randomly and haphazardly but the organizations that are really deliberately investing, that have a strategy around R&D and then taking that R&D and making use of it, are really starting to pull away from the median and certainly from the least productive firms.
Andrew Welch: And I think what is happening here, kind of behind the scenes, is that we're seeing this two-tier economy emerge, where you have the firms that are doing this wisely and are making smart investments and actually taking R&D seriously. And then everybody else. And the metaphor that I like here is the fable of the boiling frog. Right, you take a frog and you put the frog in boiling water. The frog recognizes that the water is boiling and the frog jumps out. But if you put the frog in room temperature water and then gradually turn up the temperature, the frog doesn't realize until it's too late. And what I think is happening here is that many, many firms are there that frog, the water is starting to boil around them, but they don't realize it yet and they don't realize that they're sleepwalking into the wrong side of the two-tier economy.
Mark Smith: And you're talking about large organizations. What about everything that sits in the SMC space that's not even thinking like this at the moment, or anybody that's sitting in? What's the space below SMC? Smb, SMB, yeah, yeah, yeah.
Ana Demeny: But you know what? It's really easy now. It's really easy now. Actually, we've run for SMC, smb and without a lot of research capability or energy or whatever. And I'm not saying that there isn't going to be that huge difference between firms who do do their research and like, plan and have a strategy and vision and organizations who do not. What I'm saying is just so easy to adopt your technology now, because the new technology is very adaptable to the old stuff.
Ana Demeny: So let's just say you've been working on a project for like the past five years.
Ana Demeny: Yeah, let's just say you're doing that and you're using some old tech. Maybe you're still using some on-prem even, and you know some of the functionality may be very manual, etc. Etc. You really just need a few folks to pay attention to Ignite and then Inspire and all these big conferences and take some notes away and with the capability that technology has right now, you are able to start including some of that new stuff, such as I don't know fabric or some of the artificial intelligence capability, even if it is not the co-pilot that you install on your new organization. And then you need to move your data to North America, etc. Etc. Etc. It's still like the opportunity now to adopt technology and to be able to start doing a bit of research is enormous. You no longer have to plan and put money aside for the last quarter of next year. That's when we're going to start doing it. You can do it right now without a huge amount of investment, and that's what's fabulous from my perspective.
Andrew Welch: As of May. So I agree with you, anna, and I think that one of the things that you should always be doing in business is looking at what your competition is doing or, in some cases, what they're not doing. So I'm sure this statistic is somewhat outdated, but, as of May, 40% of American small businesses reported being uninterested in AI tools. What that does is that creates a mind-blowing opportunity for the 60% of small businesses in the United States that are interested and maybe are going to seize the initiative to act. And I think an interesting companion study to that.
Andrew Welch: Donna Sarkar got me turned on to this study from the Boston Consulting Group and I'm not as conversant yet in those statistics because I haven't shared them in multiple conferences at this point. But basically, one of the findings in this study is that AI is actually tending to benefit the people, kind of the median performers. Right, ai actually has statistically bearing out right now that if you're an average performer, ai can help bring you up to the level of productivity and the level of performance of the top performers. So there's, oddly, a real equalizing opportunity here that probably will be, on the whole, pretty good for humankind and for, like, the working professionals. I don't agree.
Ana Demeny: That's exactly what I was trying to say.
Andrew Welch: Oh, I like that. I like when we don't agree. Chris, convince me. No, no, this is dangerous.
Chris Huntingford: This is a lawsuit in the making.
Andrew Welch: That's fair. I know where you're going. I know where you're going, and that's fair.
Ana Demeny: I think that Donna is very hard right now and for everyone who wants to learn a little bit of AI, all they need to do is follow Donna.
Chris Huntingford: Yeah, just follow Donna.
Ana Demeny: Just read everything she posts. Yeah, it's so funny because when Andrew.
Andrew Welch: She's so hard right now.
Mark Smith: When I went to oh God, when you were talking to Andrew and you mentioned Donna in my mind I was actually going to mention Donna because of one slide I saw she had up somewhere over the last week where it had. If you're looking at the AI landscape, 40% of it is around data.
Andrew Welch: I love when you say data.
Mark Smith: And you might remember that graphic and I have the slide.
Chris Huntingford: I'll pull it up for you.
Andrew Welch: Oh, I love a good slide. Bring me the slides.
Mark Smith: I just thought it was that one kind of I don't know Perhaps you shared it, chris, it might have been on LinkedIn and I was like man, people really need to understand what that one small graphic is showing, because so many organizations don't get that 40%. And I want to give you a practical example of small business. Yesterday my wife met with our accountant, who's been my accountant for maybe 13 plus years, and she was discussing our new business activities and what we're doing in our own business and he didn't realize the stuff that I'm doing in AI and stuff. And he was like blown away by the lack that they have in the accounting. And he's got an accounting firm of about 30-odd staff and massively concerned that they're going to get left behind. And when Meg was telling me this, I was just riffing in my mind what would I do in that scenario?
Mark Smith: Going in and if I was to do a talk to a small accounting firm and straight away they do a survey at the end of a financial year to every one of their clients that says a whole bunch of questions qualifying questions about how their tax is going to be sorted out and stuff, and that's all a manual process for them. They capture it electronically and then they answer one, answer two. They've got to go through it and nobody got time for it, right? That is like we need to extract a few of the numbers otherwise, and I'm like, listen, you could run AI across one, all of that data, and how did everybody answer question one? You could then plot that and go okay, what were the outliers and are the outliers concerning Because there we're going to actually have to run a follow-up call with that customer and address this or would it allow them to expose whole new tax minimization opportunities for their customers?
Mark Smith: Because of certain things they could give a lot more robust, because I don't know if this is the same overseas, but every accountant in New Zealand has now become a business advisor. It's kind of like, as they've expanded the accountancy game, they've gotten to business advising and that's a really big part of their portfolio, of what they offer. Now, that was just me thinking in five minutes. There's so many ways I would run AI in the accounting profession as someone from the outside that I think would create massive improvements.
Ana Demeny: So that's exactly what I was saying, mark. So before, what would we before last year, at the beginning of last year, let's just say, how would we tackle this problem? Well, we would go there and we would help them re-engineer the way they do surveys right, we would try to implement new technology on how they do surveys, and then we would capture that data into a certain type of way and you know what I'm talking about and then we would do graphs and analytics and so on and so forth, whereas now you were able to just go in there and say, hey, dude, I can actually work with your answers right now as you have them right now. This is exactly what I'm trying to say. It's just so easy when, if you are behind, to start making some steps, to gather some information and then start doing some revolutionary changes, some true digital transformation. And I remember this summer and we had this uncle he sells spices, so naturally he's a multimillionaire, I'm not even joking. So the guy sells rosemary.
Andrew Welch: And he once owned the largest supply of catnip in the world. So that's nice.
Chris Huntingford: Wow, Thanks for letting me know that man Wonderful yeah.
Ana Demeny: The dude is is fascinating. He's just fascinating. He's like in his 80s, I guess nearly as right. He fixes boy boats. He's out on a boat all the time, but he's really brilliant. So he and he comes by and we're like working, but he feels like having a chat. So he sits down in front of you and you're like on a call and you're like okay, it just sort of looks at you until you realize you have a house guest man.
Andrew Welch: Come on, Get off the phone.
Ana Demeny: So all of a sudden he starts talking to me and he's like so what do you think about this AI business? Like like super startled. Like the guy sells rosemary on the internet Not just that, but anyway.
Ana Demeny: I was like oh Neil, what? How do you mean? Like, what side of AI he's like? I have decided it is way too complicated. So what I'm going to do and what I do right now is I look at my, my next competitor, who are big, big business, and I just tell whatever AI technologies they're doing, because they like to brag. So Andrew Zanko revolutionized his entire marketing strategy using artificial intelligence in his eighties to make his business flourish, because he never. He doesn't want to be as big as the next guy, because, first of all, he makes way more money by being a smaller business. You know, but he's just implementing the same things. These are some of the methods SMB, smc could follow as well, and that's not a that's not a bad place to be Like. You're just not reinventing the wheel and with artificial intelligence is just so easy to do it right now.
Mark Smith: I I just to pivot slightly or to recap or re re go back to a point that we didn't explore. Chris, you saw that there was going to be a legal case or or a disagreement with what Andrew was saying earlier. Andrew, can you, can you put your point forward again, just so the list is don't have to go back and rewind. And then, chris, I'm keen to hear your thoughts.
Andrew Welch: Basically, what I was saying is that there's this, there's a study out there that has found that AI really benefits sort of median performers and helps them, helps those performers to to perform and we're talking about the individuals here at a at a kind of a higher level, and that that benefit seems to be outsized kind of around the average right. And what I was saying is that that in so far as that phenomenon tends to bring up the productivity and the performance of, of of those kind of in the median, that that's probably really good for the economy and and for kind of the, the, the state of all of these humans in in the minute. And I know where Chris is going to go and I don't disagree with you, but I but I don't know where Chris is going to go, so so where are you going, Chris?
Chris Huntingford: I will show you where I'm going.
Andrew Welch: There's a lot of risk there, I think.
Chris Huntingford: Oh, it's something, something way more dangerous than that man Digital equity, yeah, so what's going to happen is that when you look at the way in which AIs released across companies, there's a problem where you're going to have, let's say, andrew and Anna you're both in recruitment, right, and Anna, you get given M365 co-pilot. Andrew, you don't, okay. So you're both reviewing 1000 CVs. Anna, you're able to reason over the data in those CVs really quickly, bring 50 candidates on board and whack these your paycheck. Andrew, you're still reviewing the 1000 CVs and get two candidates on board Now. Anna, you might be a subpar recruiter and Andrew, you might be a great recruiter. You just got unlucky. What's happened is that you don't have the tools to do your job.
Chris Huntingford: All right, and if you look at the way in which tech is being released in these companies, they call it rings of release. So a ring of release works exactly the same way as an update works, where you'll drop it in rings, right, and what you'll start finding is that when people start trying to adopt the stuff, if the adoption plans aren't done correctly and the right training I'm using the word training very loosely, we call it enablement isn't given and you start watching the monitoring. People that aren't using it but aren't able to do their jobs correctly are going to really struggle and they're going to start comparing themselves to other folks in the company who are doing really well those subpar potentially performers. And I think there will be a bunch of legal cases that come out that say you know what like this isn't fair. We've been unfairly treated, especially in the UK. Interesting yeah, so digital equity in this whole thing is going to be huge Now this is so interesting.
Chris Huntingford: Yeah, right, so you wouldn't. You wouldn't have seen this come out in in the low code stuff. And the reason is because the low code stuff you still had to be kind of driven, so this would have come out in office to an extent. But people adopt the way they adopt, right. But the gap that Anna was talking around between subpar performers versus par performers, all of a sudden a subpar performer is now really strong and this whole concept of digital equity versus digital equality is going to come out. And I'm telling you, in the next few years this is going to surface and it's going to be a problem.
Mark Smith: I think it's going to be awesome.
Ana Demeny: I believe that a subpar performer will still be a subpar performer with with AI as well.
Chris Huntingford: With AI, but if the AI isn't released correctly out into the ring, the rings of release, what you're doing is you're actually stunting other people. There's a whole big study on this, yeah.
Mark Smith: Yeah, yeah, no, I can definitely see a couple of things that spring to mind out of this One co-pilot, which was, you know, I think, loosely first implemented by Microsoft in GitHub. Right, was it GitHub?
William Dorrington: Yeah or was?
Andrew Welch: it yeah, or was it? I think it was GitHub co-pilot.
Mark Smith:
It was on the developer side, right? Yeah, what they've shown, what data is shown, is that it enables a good developer to become an amazing good developer, right?
William Dorrington:
Yeah.
Mark Smith:
But a bad developer or a learning developer. It is a disadvantage to them because they try and use it as a they don't know what they're talking about type thing. And so therefore, like I was talking to a bloke yesterday around AI and how he'd use it for a customer, and it lied, right, it gave a hallucination because he didn't know his data. He didn't know his hallucination, right.
Chris Huntingford: Yeah.
Mark Smith: And the customer came back and said what the freak is this, you know, and I think that it could be a dangerous tool in the hand of somebody that doesn't know what they're doing. But it doesn't make them better, it just makes them. It becomes luringly obvious, you know, if there's a mistake.
Ana Demeny: Yeah, well, it amplifies everything, doesn't it?
Mark Smith: Yeah, one last thing Think bigger. I saw it on your slide there, chris. It prompted me that I've seen a bunch of your team on LinkedIn with the think bigger jerseys on, or hoodies, or whatever, yeah it's obviously an internal thing. What's it all about?
Chris Huntingford: I think we could easily negotiate this, Really really easy.
Mark Smith: So what's think bigger about?
Chris Huntingford: Tell me there's a bunch of different meanings, man, but I think the main one is that the corporate culture of this company is that of a trusted maverick Right. So what we try and do is we try and do stuff like really differently. The whole idea is that we want to disrupt, but positively, and we want to do things a little out of the box. So when people are driving down the left lane of the highway into the line of traffic, we're like, well, why go the other route? Like do something different, do something bigger. Yeah, it's been interesting, man.
Chris Huntingford: I think the culture, the culture is really awesome and I think that's why we've been allowed to go down the route of actually productionizing things in AI. I mean, you're going to talk to John and Mike, I believe. Yeah, so those dudes built. Those dudes had time to build out their own prompting engine that creates entity relationship diagrams. Like it's taken eight to 16 hours of my work when building our user stories and it's made it one hour. Yeah, they made their own model driven app, co-pilot. So that's what it's all about. It's really just thinking really differently.
Chris Huntingford: I love it who do you saw all of you.
Mark Smith: Yes, thank you. I just want to share my screen here quickly and play a video to you and listen if you're watching the podcast or listening to it. Remember we do the podcast. The audio version comes out a little later than the video version. The video version is. I'll get Andrew. Andrew, where can they find the video version? What's the name of the site?
Andrew Welch: Cloud Lighthouse, so it's cloudlighthouseecosystems.
Mark Smith: Awesome. Go there, watch the video I'm going to share one of our avid viewers of the show. He submitted a question in video format, so what we'll do is we'll watch that and then we'll go through and let everybody give their answer. Opinion based on what has been requested.
Andrew Welch: Nathan Rose is my spirit animal right now.
Mark Smith: Here he is Nathan Rose.
William Dorrington: How do you respond to someone standing in front of a burning house asking if they should put a fresh coat of paint on it? Hello everyone, I'm Nathan Rose. I'm a lead consultant at Microsoft MVP, based here in beautiful Auckland, new Zealand. Over the past few months, I've become a massive fan of the ecosystems podcast put out by Mark Smith and Cloud Lighthouse. If you haven't had a chance to check it out, they have some incredibly smart things to say. I highly recommend it. So the most recent episode of the ecosystems talks about the centrality of data in the ecosystem and the fact that many organizations and consultants erroneously put the app, the workload that we're implementing, at the centre of the ecosystem, and we should be focused on the data and getting that right before we go working on the app and whatever it is that we're building. So the episode reminded me of a recent customer interaction that I had, and I'd love to share that story and pose a question back to the ecosystems team for feedback. So I was recently asked to help out with a pre-sales opportunity.
William Dorrington: An organization here in New Zealand wanted a customer portal pretty standard. I was asked to attend a late stage meeting with the customer. They're most of the way through the sales process and they just need someone there to answer any potential technical questions that pop up. You know what this looks like. You're not there to challenge, probe or otherwise rock the boat. You're just there to answer specific questions that they have. So over the course of this meeting, the customer's lead architect says you know, we're actually a really good position here. The core system you're going to be hooking up to is custom built on premise and the one person who understands it has been with this organization for over three decades. Yeah, that was said in 2023. So alarm bells are flashing at this point because, while the customer feels their greatest need is for a portal, the reality is they have this antiquated core business system that is presenting a massive amount of risk to their organization, and the purist in me would love to do nothing more than stop the meeting and take time out. We can't talk about anything else until we've addressed this core business platform. You have no business talking about a portal until we've addressed and fixed this core business platform. Now, anyone who's working consulting for more than five minutes knows why this is a terrible idea.
William Dorrington: The customers stated felt need is for a portal. There's budget for a portal. There's attention for a portal. They don't want to talk about anything else. The data, the core data system, is out of scope. It may be on a technology roadmap somewhere in the organization, but it is not a priority today. There is no attention, there is no budget. So my question back to the ecosystems team is this as a consultant, how do I navigate that natural tension between where the customer is focused today, where there is budget today, where they will do things today, versus where they should be focused, and how do I ensure that I don't lose those opportunities that I have today to work with the customer while ensuring that I am taking them towards where they need to be, the things they need to be focused on, in this case, this core business system that is presenting a lot of wisdom in the organization. Look forward to hearing your thoughts. Can't wait for the next episode. Thanks so much for watching.
Chris Huntingford: Wow, what a great, what a great story and what a great question.
Ana Demeny: I love it.
Chris Huntingford: Oh my.
Andrew Welch: God yeah.
Mark Smith: So why don't we just go around the room? I was like I'll say mine first because it's so somebody doesn't now say it. But let's not do that. Let's, chris, let's take your view on it. You've just seen that fresh. What are your thoughts? How, what would you? What would your guide be? And, of course, he explained something very specific in that the challenge you have is a. You know he's been poured in and you know the sales person has told him.
Chris Huntingford: Oh, my God.
Mark Smith: They've read the right act right. Don't fuck the steel up for me.
Chris Huntingford: Right or do you need?
William Dorrington: to do it.
Mark Smith: If they have a question, answer that because I've been in the situation where I've gone this is wrong, and I've been came afterwards by the sales. Right, you don't need to go. I'm very big at calling out elephants in the room and I'm probably become more aggressive the longer my career is gone, because it's created good outcomes, right? Anyhow, before I go down there, chris, tell me your thoughts.
Chris Huntingford: I actually got into trouble for telling a customer to not buy PSA Years ago. I told them I sat in the room. I got absolutely spanked afterwards but I'm like why I don't believe in the products. It's shit. I don't buy it, sorry.
Ana Demeny: I mean, that was good advice.
Chris Huntingford: Yeah, right, thank you.
Andrew Welch: I've had some recent experiences, but it was dynamics for HR. We'll just yeah.
Ana Demeny: But the point was and he's like, okay, and I don't. I do want to do the best thing here for the customer. However, how do I not miss out on this sales opportunity right now?
Chris Huntingford: Yeah, this is hard. So, in my role, right, I'm effectively a pre-sales person, okay. So, like, I'm responsible for both selling and consulting, that is legitimately what I've been doing since 2010. Okay, so I get where he is, right Now, understand why he's in the position. So I do something. And I can only speak from my own experience. Right, because I'm like you, mark.
Chris Huntingford: I have zero problem telling people that this is a shit idea. One example I want to give you, and then I'll tell you what I think about this, is that recently, I was working with a firm in the UK and they told me that somebody had architected a solution on top of SharePoint, and it was private information, pii. And I said to them it's SharePoint for a reason. And I said absolutely not. And what was funny was that I was watching the bottom left-hand corner of the screen, one of the solution architects. She was shit, hot man, so good at this stuff. She was like clapping Okay, in this scenario, right, this is a double-edged sword.
Chris Huntingford: You have to consult. If you want to be a trusted advisor, you have to consult. You have to tell them that they have a much bigger problem than X, right? So let's just get that out the way. I don't believe that a sticking plaster is the right thing here. You have to tell them that actually you have a much bigger problem than this thing. A temporary thing or a tactical solution to help them, okay. But you say, look, we will do a tactical thing that will help you here to do X, y and Z. However, your biggest problem is this, and if you don't do this, that solution is null and void anyway. So I would hit it from both angles by being a consultant and telling him the truth, but also saying look, from a tactical point of view, we can solve the immediate problem with something like X, okay, but you would have to bring them together and do not lie Like. That is the one main thing. If you see a problem, you have to point it out as an advisor.
Mark Smith: Interesting, interesting, and I think how you point that out is going to be critical. Andrew, what are your thoughts?
Andrew Welch: One thing that I want to point out here first is that I appreciate the difficulty of the situation, and not because, listen, you have a technically smart person who wants to offer the best advice and you may have a conflict with the sales motion that is happening. But I mean, that is a difficult situation. But I think that that kind of is one of those things where you can boil that down to what's the right thing to do for the customer. And I think that if you hang that idea what is the right thing to do for the customer in the sky as your North Star, that's going to always guide you. But one of the difficulties there is that is, actually you can be perceived by the customer as trying to actually sell them more things. So that's what makes this a sticky situation. I have no doubt that Nathan in this situation is wanting to do what's best for the customer, but I also have no doubt that there's a naysayer in the room from the customer's perspective saying no, no, no. This guy is just kind of a wolf in sheep's clothing trying to sell us more things. So I just want to get that out of the way out of the way first. I have some high level strategic thoughts on this, and then I have some practical guidance. I'm going to start with the practical guidance.
Andrew Welch: I think that if you have as your North Star that you always are trying to do what is best for the customer, even if that is at the expense of a short term sale, I think that the way that you need to present this is that you need to take on the persona of an educator, not a salesperson and not even an architect. I think that you need to say listen, I appreciate what you're trying to do. From a short term perspective. You have an immediate problem. We need to solve that. We can solve it. It is technically possible to implement a bandaid or a plaster solution here. Here are the benefits of it, and I think you need to be sincere in acknowledging that. The benefit is we're going to solve a very immediate problem for you, but here are the downsides. The downsides are that we are putting a bandaid over a much bigger problem. We should be in an ideal world where time and money are no object. This is how we should solve the problem. I think that over time, that is the way to earn the trust of the organization that you're working with. They might be internal to you. You might be part of an internal IT team giving guidance to your business colleagues, like your business or your operations colleagues. I always think that when you have to present bad news like that, the best way to do it is to do your very best to take on the persona of an educator, of a mentor, of a coach, like I'm here to coach you through what I acknowledge and understand is a very difficult decision where you have to weigh immediate need and cost and time and all of these sorts of things. That's my very practical answer as a companion to that. I think that this is why ecosystem-oriented architecture and taking a more strategic approach to your cloud investments is so important.
Andrew Welch: We've been talking a lot. This podcast is called. This is the Ecosystems Podcast. We talk all the time about ecosystem-oriented architecture. What is happening, it sounds like, in the scenario that Nathan presented is a very point solution. Old-fashioned way of looking at things I would try to do if I were him and if I were in that situation, is I would try to guide that customer towards saying hey, listen, I totally get why you're trying to do what you're trying to do. There is this emerging and quite successful at this point. School of architecture. We call it ecosystem-oriented architecture. Let me explain to you how this works. Let me explain to you what the principles are. Let me explain this Again. We come back to this being a guide, being a coach. A lot of technologists, I think, always want to be the magician. I think that the persona to take on is not of the wizard who does mysterious things in the back room and then brings you magical objects, but try to guide your customer. I think that this is the crux of what ecosystem-oriented architecture is all about.
Mark Smith: I like it Anna.
Ana Demeny: I'm a bit in between what Chris's approach would be and what Andrew's approach would be. I cannot forget the fact that my consultant is there to just answer questions. He's going to offer answers to whatever questions he's been asked to give, but as that consultant, I would actually focus on where I see the problem is. The problem is a people's problem there. In my opinion, pure and simple. That was a system created by someone over three decades ago. I believe that, and I would actually try and understand whether they want to create a portal, because actually they want to get rid of that huge risk of that person who creates the system over three decades ago and they don't really know what's there.
Ana Demeny: I would start with that and then I would go into ecosystem architecture, because ecosystem architecture does mean that we will connect everything together and we will use a lot of aspects of Microsoft technology, but not necessarily at the same time. Hybrid solutions are actually perfectly valuable. To begin with, I would try and prove value like that and bring out the fact that really, what you're looking for is not necessarily the best solution, but we could get there by using all of this rich technology that the Microsoft ecosystem offers us. While doing that, we would alleviate some of the biggest risks that they have, which is again a people's problem. That's how I would go about it. Actually, your turn, mark. I'm so curious, everything you mentioned. It was so hard for me to focus on anything else.
Mark Smith: There's this guy called Chris Deadman and he's a director of business strategy in Microsoft. I remember I was doing a code training session with him, with GBBs. He said you've got to understand. When people come to us and they buy our technology, this might be something they've done once before in their lifetime or in their career, maybe twice. Right, we do it day in, day out. We have expertise. We have years of experience of doing what we do. Sometimes you need to explain it to a customer. This is your first rodeo. This is my thousandth rodeo. There's things that I have seen that I'd like to bring to the table here so you don't make the mistakes that we see. You got to do it in a very humble way. It's not about all of me. It's about going. Listen, you're in good hands because I've got experience.
Mark Smith: I love what you said, andrew, about come as an educator, not as the wizard that can pull things out of a hat. Too much in IT has been sleight of hand, wizardry in people's experience, in that, oh, my password's not working. When you're on the help desk and the guys like tap, tap, oh shit, the system, the service needs to be rebooted, and go. Can you try that again? And they enter it. Oh, you must have made a mistake when they just know they started a service Slight of hand shit that IT has been doing for years to make people feel that they're dumb when it comes to understanding IT, because the IT person doesn't want to lose face. So you've got to be careful, right. That's why you've got to come as the educator, you've got to come as humble. Now, I think, also tying in what Anna has said there I would have gone, go through a meeting like that and show that we can absolutely deliver what they're looking for.
Mark Smith: And then I just had it with a car manufacturer recently on an RFP. They had asked for all the stuff in an RFP. They in this case they wanted a Dynamics 365 sales implementation right as in. So they wanted to see a room and basically all I'm saying behind the scenes here. They went and asked all the stuff, blah, blah, blah, tick, tick, tick all the boxes. At RFP we go to Orals, we present and we show that we met everything. And then I said so there you go, 100% meet your specification. I said but can I just bring up some stuff that you didn't ask about? That I think is really important to your organization.
Mark Smith: And then I went into another five to 10 slides around all the stuff that they didn't know around advancements and technologies that applied to their business. That would have massive impact. So one of the examples were there was going to be a call center involved and I said, listen, if we put biometrics on the qualifying of somebody calling in, we will shave a minimum of three minutes off every operator per call. Let's look at the cost impact of that. Shit. Didn't even know this technology existed, right. And so by the end of it they were like, oh, actually we need to adjust the RFP Because. So, in other words, we said we could deliver. And then we said, hey, but here's, like Andrew, what you said. But here's the power, here's how we think about platforms and ecosystems and things like that.
Mark Smith: Because and then you know, the other story is, I would add, in the bus effect, right, if that dude with his three decades of experience and he's the only one that's the guru of the system what happens if he gets taken out? Do you have solid documentation, et cetera, around that? I would, you know, I would decide whether that was going to be part of my story. But one of the things and on my 19 entering challenge. I really encourage people to read this book called Getting Naked and and the whole book is a fable set in San Francisco around a massive consulting firm and a little medium sized consulting firm that keeps beating the big consulting firm and one of the things they have these kind of they face these fears of consultants and one of them is of calling stuff out that is going to kind of drop a bomb in wherever the discussion is going and knowing how to do it and doing it in a humble and curious way that really that doesn't come across as, yes, that's going to be. Now we're going to sell you X, y and Z behind it.
Mark Smith: And that's why I think, come back and back to what you said, andrew, the edge. You're there as an educator. You're not there, you're not coming across to. You know how many people can we get on it.
Mark Smith: But you know, the other thing is that I often used as a because I live in New Zealand and if I'm on a New Zealand based project which is very rare that I am, because I'm generally outside of New Zealand the projects but we had an earthquake here and in that earthquake companies disappeared because their data got totally piped outright on premise because their buildings got crushed and destroyed, and you know, let alone the human toll.
Mark Smith: And then I use 9 11 as another reference how many businesses had their data sitting inside one of the world towers and that got wiped out. And this is and this is how I go into why cloud is important, why redundancy, why geographical redundancy is so important, and when you're in a country like ours that has a high rate of seismic activity, you're only too aware of a news or, you know, article of where why a proper redundant strategy is important and anything on prem that's important to the lifeblood of a business. These days, I'm like, yeah, I'm super. So those are kind of the different angles. I think that the key, though, is understanding that people conversation, understanding that you're there to educate, but I would do that after we were quite clear that we could deliver on. Why ever I was bought into the room to deliver on.
Ana Demeny: Yeah, I think that's really really good. Also, I really love your comparisons. I'm not sure they work very well here in the UK because those are all acts of God and everyone knows that God will just avoid.
Mark Smith: I thought you were going to say God doesn't exist and I was like, well, that's a whole different subject.
Ana Demeny: No, but like this is like yeah, that that stuff would never happen in Great Britain.
Mark Smith: So you'd never get floods and great but and you'd never get a snowstorm that cuts off power for days or I mean, yeah, we do, but even those are not like people.
Ana Demeny: People don't bother themselves with such thoughts. I feel, yeah, valid.
Andrew Welch: I'm surprised to play a certain flood more frequently Me too, actually me too, but that's because even the water doesn't want to be here.
Chris Huntingford: I want to be here.
Andrew Welch: This is as your American, South African and Romanian.
Mark Smith: Yeah, I know.
Ana Demeny: Living in all, living in Britain, or right, exactly, no, honestly, like I used to work, I was to work with this, with this company, and they used to sell to send all of their messages through a server that lived underneath the desk and that server obviously just kept going like I'm plugged or it would freeze and stuff like that, and I kept and I kept saying, guys, what if there's a flood or an earthquake?
Andrew Welch: Or they were like, yeah, Can I share just sort of a funny story from my earlier professional days that you just reminded me of? So so I worked for a work for a company in the Washington DC area and a lot of the work that they did was for US government clients, the US government agencies. And one of the one of the practices that's very common there is that when someone reaches someone who's been a government employee, they've been a senior employee, and they kind of reach the end of there and I was quite junior, I was young, but I anyway one of the practices is very common is that they'll hire a senior person away when they retire from the government. They'll go to work for private sector and the whole reason is that they have a lot of knowledge and you know, they have a lot of connections and whatever.
Mark Smith: One on one consulting, as in big consulting companies, right yeah exactly Hire someone with the relationships. That knows where all the bodies are buried? This?
Andrew Welch: guy was named Wayne and Wayne Wayne Wayne.
Mark Smith: What was his last name?
Andrew Welch: I do Wayne as well actually Wayne Wayne.
Andrew Welch: I did have a colleague named Wade as well at another point, but this guy was named Wayne and Wayne. One day he calls me over and he says hey, I'm really struggling, can you help me with something? And I say, yeah, sure, what is it? And he said well, I've noticed that you send emails very quickly. And I kind of.
Andrew Welch: I had asked him a few follow up questions and he said Well, what I really would love to know is can you send an email, I'm sorry, can you send an email to more than one person at a time? Wow, and I was like, wayne, what, what are you? How are you doing this now? And he said well, when I want to send the same message to multiple people, I type the message and I send it to someone, and then I copy and paste that message into a new email and I send it to the second person. And I said and you do this as many times as needed in order to send it to the group that you want to send it to? And he said yeah, and I said, okay, so I'm going to show you.
Andrew Welch: So we have you just type the email in and you put a comma in or a semicolon in and you can put another person. And he said and can I put three? I said, yeah, yeah, you can put three. Wow, it is well, there's okay, this is, this was. But the point is that this was a very formative experience for me, in, in, in, in my, you know, in the younger part of my career, in that I think that it, it the, what it taught me is to not take for granted yeah, that everybody knows, even the things that you think are the most simple things, and to kind of approach that with a degree of humility, which is difficult, right, because what I wanted to do is I wanted to, I wanted to die laughing and call every single person I could call to tell them the story.
Ana Demeny: Instead, you waited for 20 years and you told the internet yes, instead.
Andrew Welch: I have waited for. I've waited for almost 20 years to tell this story.
Chris Huntingford: That's true. That's solid gold, absolute solid gold.
Mark Smith: Yeah.
Ana Demeny: Yeah, yeah.
Mark Smith: And you know. That just got me thinking the CC field. Right it stands for carbon copy. Right In a BCC, blind carbon copy To most people in the modern world. If I asked my kid what a carbon copy was, what's a, what's a carbon cot like? Isn't it funny that those terms have come from 20, 30 years ago of technology that you'd never used today?
Andrew Welch: Do you remember the movie Big Tom Hanks Big, where he sort of?
Mark Smith: Swaps of this kid or something, yeah.
Andrew Welch: Yeah, he's a kid and he becomes an adult overnight physically, but then he's got, yeah, so and I remember he goes to I haven't thought of this in decades. He goes to this, you know this employment shop or whatever you know kind of placement and or file for unemployment or whatever, and the woman behind the desk says, yeah, fill this form out in triplicate, give me the pink copy back, you'll hear from us and you'll get a decision in six weeks yeah. And now I'm just thinking, who even knows what triplicate is?
Mark Smith: Yeah, yeah, yeah.
Andrew Welch: It's like the disk icon as well. Like what? Why?
Chris Huntingford: did you say I was just going to bring it up and look at that?
Mark Smith: Yeah, yeah, yeah, oh wow. That's cool, that's so cool, you do you do have an interesting set of statues. Chris, thank you as we draw to a close and wrap up, a couple of things. Christmas gifts from Microsoft. What are you getting? What do you got? What are you thinking?
Chris Huntingford: What? What? That's a question.
Ana Demeny: For me it remains fabric. That's the. That's the Christmas gift.
Andrew Welch: The.
Chris Huntingford: Christmas gift.
Andrew Welch: I think fabric being hitting general availability. That's like the you know the bicycle or the set of skis under the Christmas tree, that is, that's the big ticket, the big ticket item. I have a stocking I'm going to call it a stocking stuffer from Microsoft that I will share that is, has gone really under the radar and it's really, it's obscure, but I'm really geeked out about it. Has anyone heard of integration environments? Have you guys ever seen this?
Chris Huntingford: I have not. That's used to me.
Andrew Welch: Okay. So integration environments and you'll hear why I am so geeked out about this is that it's part of Azure Integration Services. The idea is that very similar. So if you're in the Power Platform world, you're probably very familiar with environments and solutions and how environments really allow you to group, to logically group components so apps and the data model and tables and everything and then to build ALM on top of that and to build security on top of that and so many things.
Andrew Welch: So I think the same principle with integration environments for Azure Integration Services. So what integration environments allow you to do is you can establish a logical boundary and I think it might even be in private preview right now but the supported components are, I think, service bus cues, logic, apps and functions. Maybe those are the supported components and you build these and you load them, you group them logically in your integration environments and I think that there that's a really subtle but really important capability to building the integration neighborhoods that we've been talking about in ecosystem oriented architecture, where, rather than building point-to-point integrations, you're building reusable, repeatable components and standards and processes from an integration perspective. The other thing that's cool, that's the companion to integration environments, are business process capabilities.
Andrew Welch: So, now, rather than just building out your service bus topics and cues and your logic apps, you can actually map them to steps in a business process and where these integration motions align to something that's actually happening from a functional perspective. I'm just connecting some dots here. This is I have not seen this on the roadmap, but to me this seems like this seems ripe for taking AI capability and marrying it with both the integration technologies, but also with what's actually happening in the business, and building some model to basically run with this for a process optimization perspective. So that's what's in my stocking and I was super excited when I saw it.
Mark Smith: I love it Huntington.
Chris Huntingford: Dude, I don't care, I just want a freaking M365 co-pilot per user license, like that's all I want.
Mark Smith: I don't care about anything else, bring me.
Andrew Welch: Santa, bring it to me.
Mark Smith:
I love it. I love it. I would have to echo what Anna said and that I think, if, and then probably more if you're a consultant listening to this, I think one of the biggest areas of investment in your career in the next 12 months would learn everything you can around fabric. Learn it, learn it, learn it, learn it. It'll touch every part of what you do in Microsoft business applications massively. What would I like, what would I wish for One? That AI builder would go away and become part of co-pilot studio.
Mark Smith: And the reason is I think it's confusing people.
Mark Smith: I think it confuses people to have inside the BizApps thing Like come on, put it, there have one tool.
Mark Smith: The other thing is, with Google's announcement, with Gemini, is they show amazingly, the concept of solving a problem and how, behind the scenes, it would build out a liquid interface. In other words, the interface changes based on the topic, the data, everything you know whether, and in the concept of a chat virtual agent. There's sometimes in the virtual agent that a text-based response is not the right way to represent the data. It should be a visual response, it should be an auditory response, et cetera, et cetera. What they demonstrated in the interface design that the interface not only changes based on the first kind of inquiries but as it gets refined, it'll keep modifying the interface based on how to present data in the most optimal way for the backward and forth happening with the AI, which I just think is sick. So this kind of like the interface is up for grabs and it showed how it actually defines all that behind the scenes and redevelop someone's question to make it more clear, et cetera, but really using the full fidelity of what future interfaces could look like.
Ana Demeny: And how much into detail they should go as well. Yeah, amazing, At some point it stops. It's like oh okay, that was too much.
Mark Smith: Back on the open AI thing and just rounding back to where we started, I can remember why the math issue is a big problem. Why Ilya apparently shat his pants on what it meant is that it meant with the right amount of compute, every encryption could be broken. That's right.
Mark Smith: That's right. Every form of encryption could be broken very rapidly, and that's before we get to quantum computing, right. And so I think that there was a oh fuck moment. Every banking system, protocol, et cetera, destroyed instantly every passwording, hashing, blah, blah, blah problems, and so that was. It's not necessary that it's become sentient, it's become so smart. Another little thing I saw the other day is that I think was Einstein had an IQ of 160, I think.
Chris Huntingford: Around there. No, I know it's post 150. Yeah, so 160.
Mark Smith: I think that GPT-4 is currently at 155 when it does an IQ test and they reckon within five years it'll be at around 6,000, right off the scale type thing. What will that mean for humanity?
Ana Demeny: Huh, do you mostly use GPT-4?
Mark Smith: Yeah, yeah. And Moen GPTs. I've started developing Moen GPTs just for personal right, so personal use stuff Been able to. You know I'm doing some training at the moment for some Microsoft staff and I do a big case study. What I was impressed is that the speed that the Microsoft folks just took my PowerPoint, which, let's say, 10 slides of facts, size of organization turnover, all this kind of data I have in there of my fictitious case study and of course, it leads into all the various sales plays. So, based on your role doing my training, you could see where you're going to take the story.
Mark Smith: The speed that they just dropped that straight into AI and had it summarized and be absolutely concise I'm like that's smart, right If you're working in agile supply chain or something like that. The speed that you can ask questions to what are the top five challenges in this industry right now and it goes out and gives you reference to the challenges. We needed a demo recently for an organization, a government organization, and you know it's a big organization Delt with the military. And the speed that I was able to find out their challenges out of news items and stuff in the public domain in the last 12 months and then to play that back. I understand these are probably some areas that you're facing challenges with inside the organization. Bang bang, bang and everyone's just like nodding and I'm like I would have never had time to research all that shit, right?
William Dorrington: I just sent it out, I got it back.
Mark Smith: I validated the references, like read enough to go yeah, okay, it's not bullshitting. And then played that back and then going hey, if I was to build into a you know a solution, what would be their product? And I was just able to create, you know, going to the old ERD way, I'm like I start with data and model out what I think the app should be. The speed that AI accelerates all that and allows you to turn around and do a demonstration to a custom where they're like fuck, you've already built this.
Ana Demeny: Yeah, just amazing, but you need to know about it as well. I just did just a little thing and I asked you some questions and I just wanted to to create a little project to see if something would work or not and, to be fair, the results I got were 70, 30. So 30% of the stuff that I got back was not accurate. It was, in fact, incorrect either because it was old information or it because or because it was information from AWS or Google Cloud. What model?
Mark Smith: were you running? Were you running GPT-4? I was not. Yeah, I find so many people are farting around 3.5. It's three year old. Like don't touch that Pay for the proper version. Get on the latest stuff Because honestly, I can tell the minute I haven't like the settings are switched back because the answer is such bullshit I'm just like fuck off, I'll re-prompt it. Here's the other thing I think everybody should be doing at least five prompts a day.
Mark Smith:
That could be around image creation. It could be anything. You need to force yourself and go. What are my five prompts today? Because I think if you want to learn AI beyond doing courses and stuff, you need to actually fucking do it every day. So it becomes like muscle memory because your prompts get better and better. You're like I know how to position this. Often, like my wife will look at my I'll come up with this. She's like wow, that's an amazing output. But look at the prompt, look how I readjusted it. That's how I tuned it. If you like to get to that point, so you do it. As you say, you have to know stuff, but I think you got to get really good at going. You know sometimes and there's so much data in the political arena around this the way you ask a question.
Mark Smith: Totally in moments is how it gets answered Totally Right, and I think what AI does now affects that massively.
Ana Demeny: And then you can forget what it told you before, or at least, okay, not GPT-4, but it can actually forget, like you.
Mark Smith: The windows a lot, the memory pools a lot larger. Now Right.
Ana Demeny: If most people are using that and are be disappointed and are like thinking what are these people talking about? Ai is pretty shit still.
Mark Smith: Yeah.
Ana Demeny: That's because.
Mark Smith: You're on the old version.
Ana Demeny: You're on the old version, yeah.
Mark Smith: Yeah, yeah, you're in the embryo version. It wasn't even born.
Ana Demeny: Embryo version. That's right.
Mark Smith: Right, it hasn't even been fertilized. Oh sorry, it's actually only the egg and sperm phase. You're not even at embryo yet. Yeah.
Ana Demeny: Oh, I mean it's not that bad, but yeah.
Andrew Welch: I had an experience recently where and it had been a while since I had a the need and the time to sit down and build an app. And recently I had a need and I made some time to do this and I sat down with PowerApps Copilot and I just started to converse with it and it got some things right in terms of the data modeling. And then I got to I needed to create some status reasons in this table and I said, listen, I need some status reasons and I need my inactive status reasons. My active status reasons to be A, b, c and my inactive status reasons to be X, y, z, slash Z. And what PowerApps Copilot came up with was an active status reason that was A, b, c and then one other inactive status reason that was X, y, z, and I was just like to hell with it. I deleted the whole thing and I built my. I built it through scratch.
Mark Smith: I was like you're not ready.
Andrew Welch: You're not ready here. I was very disappointed, but I do think that there is a degree that everyone kind of from an end user perspective and that's what I was in this moment was an end user or a developer end user you do need to be a little bit skeptical of the results that you pulled back. I don't know, maybe it was running on 3.5. Maybe that was my mistake.
Mark Smith: Now, with that close on a, I couldn't even understand the X, Y and stuff. I was like I'm not even holding that in memory.
William Dorrington: It's early morning for me, but I know for you it's.
Mark Smith: You're all at the wee hours of the morning now, so with that, thank you for joining us. Have a merry, merry Christmas, and merry means it needs to be lots of wine, eggnog, eggnog holiday here, family Cheese, food, Cheese food Life family life.
Andrew Welch: Patriotic stockings, patriotic stockings.
Mark Smith: Ciao, ciao, love you guys. Bye everyone. 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 would 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. Forward slash. Nz365 guy. Stay safe out there and shoot for the stars, thanks.
Chris Huntingford
Chris Huntingford is a geek and is proud to admit it! He is also a rather large, talkative South African who plays the drums, wears horrendous Hawaiian shirts, and has an affinity for engaging in as many social gatherings as humanly possible because, well… Chris wants to experience as much as possible and connect with as many different people as he can! He is, unapologetically, himself! His zest for interaction and collaboration has led to a fixation on community and an understanding that ANYTHING can be achieved by bringing people together in the right environment.
William Dorrington
William Dorrington is the Chief Technology Officer at Kerv Digital. He has been part of the Power Platform community since the platform's release and has evangelized it ever since – through doing this he has also earned the title of Microsoft MVP.
Andrew Welch
Andrew Welch is a Microsoft MVP for Business Applications serving as Vice President and Director, Cloud Application Platform practice at HSO. His technical focus is on cloud technology in large global organizations and on adoption, management, governance, and scaled development with Power Platform. He’s the published author of the novel “Field Blends” and the forthcoming novel “Flickan”, co-author of the “Power Platform Adoption Framework”, and writer on topics such as “Power Platform in a Modern Data Platform Architecture”.
Ana Demeny
Partner CTO and Senior Cloud Architect with Microsoft, Ana Demeny guide partners in creating their digital and app innovation, data, AI, and automation practices. In this role, she has built technical capabilities around Azure, Power Platform, Dynamics 365, and—most recently—Fabric, which have resulted in multi-million wins for partners in new practice areas. She applies this experience as a frequent speaker at technical conferences across Europe and the United States and as a collaborator with other cloud technology leaders on market-making topics such as enterprise architecture for cloud ecosystems, strategies to integrate business applications and the Azure data platform, and future-ready AI strategies. Most recently, she launched the “Ecosystems” podcast alongside Will Dorrington (CTO @ Kerv Digital), Andrew Welch (CTO @ HSO), Chris Huntingford (Low Code Lead @ ANS), and Mark Smith (Cloud Strategist @ IBM). Before joining Microsoft, she served as the Engineering Lead for strategic programs at Vanquis Bank in London where she led teams driving technical transformation and navigating regulatory challenges across affordability, loans, and open banking domains. Her prior experience includes service as a senior technical consultant and engineer at Hitachi, FelineSoft, and Ipsos, among others.
Listen On
