From Pilot to Production: Unlocking AI Value Fast

00:00:01 Mark Smith
Welcome to the copilot show where I interview Microsoft staff innovating with AI. I hope you will find this podcast educational and inspire you to do more with this great technology. Now let's get on with the show.

00:00:17 Mark Smith
Today we're crossing borders from Switzerland to London to meet 2 brilliant minds shaping the future of enterprise tech. Please welcome Tina and Thomas to the show. Hey, Tina.

00:00:26 Tina Van Heerden
Hi, mark.

00:00:27 Tomas Gold
Hi, mark. Thank you for having us.

00:00:29 Mark Smith
I'm excited to have you on Tina. I met you. What? Must been a couple of months ago. Now in Slovenia, and heard your brilliant presentation on copilot adoption and the many customers that you and your team have been across from an adoption perspective and these kind of first two to three years as copilot has really started to accelerate its growth around the world. But before we get into that. A bit of background, Tina, why not? You first food, family and fun. What do they mean to you?

00:00:59 Tina Van Heerden
Oh, great questions. So maybe I can start with the food. I have a bit of a mixed cultural background or mixed travel background. I grew up in South Africa and then I moved to the Netherlands where I lived a couple of years where I really came to appreciate food. Since the Netherlands is not that known for its culinary the nights, the lights, it kind of tries to. Really. Bring that to life. So actually restaurants there are really good and now I currently live in Switzerland, the country of cheese and chop. With which is also really nice. I tend to appreciate the food in the mountains, so that's kind of what that means for me.

00:01:37 Mark Smith
Nice. Nice. That's food, food, family fun.

00:01:42 Tina Van Heerden
So family is also mixed. Actually my most of my family is still in South Africa, but a majority have also moved to Australia, so I'm not that far from you. So yeah, most of my family is kind of between Canberra, Brisbane, toomba.

00:01:50 Mark Smith
Wow.
 
00:01:58 Tina Van Heerden
Family and of course, I still have the family in South Africa, so family usually means travel, which is part of the fun I.

00:02:05 Mark Smith
Nice, nice, Thomas.

00:02:07 Tomas Gold
Are those the same questions mark, I actually think.

00:02:09 Mark Smith
Yeah, same questions.

00:02:11 Tomas Gold
I think I will repeat a little bit what Dina said. I think what we share with Dina is we've been both moving a lot. So for me, food, definitely spicy. I used to live in Houston, TX where I got used to actually more spicy, you know, Mexican style food. And now when I live in London, I try to enjoy Asian food a lot. But yeah. A lot of flavor. Spicy food? That's definitely my favorite. And then family and fun. Also similar story to Tina. I've been moving a lot. So have families carried across Europe. I live in London at the moment, but for fun, I try to escape London as much as possible. I love to live in a city, but I do recharge when I actually leave the city. So like, fortunately for me, I'm leaving for. A nice trip to the Nordics on Thursday for vacation, so I'm hoping to get some energy back.

00:03:00 Mark Smith
 Nice. Where were you from originally?

00:03:02 Tomas Gold
And I'm originally from Prague, from the Czech Republic.

00:03:05 Mark Smith
Oh, wow. Incredible. Like, very diverse. And Tina, I never picked that. You're from South Africa, and normally Kiwis are very quick at picking S Africans being, you know, rugby and things in common. But I never picked it. That was where you originated from. For some reason, I thought you're Swiss, could you said you came from Switzerland? So yeah.

00:03:25 Tina Van Heerden
Yeah, I managed to hide the accent, but I am still into the good weather and the. Rugby and the good food.

00:03:32 Mark Smith
Awesome, what a jump to your history is in biz apps and I know that you're involved in the customer success internal MVP program, the CSU pay set is. Can you tell us about what the program was? Yeah. So what was it? What was the outcomes of it? How many awards did you win around it? All that type of thing?

00:03:56 Tina Van Heerden
Maybe I can do the introduction, then Thomas can talk about the awards. So I think it's kind of not obvious, but if you're a Microsoft employee and you do a lot for the community, you can't be an MVP. So that's kind of, yeah, a bit counterintuitive, but we don't have any internal MVP's. So within the customer success unit we started.

00:04:09 Mark Smith
Yes.

00:04:17 Tina Van Heerden
Program to still recognize some of these people that are really doing a lot for the community and are really thought leaders at Microsoft intern. And so we started this program last year called the CSU pacesetters and part of the program, so they they get a trip to Seattle. Similar to the MVPS. And then we also gave them. Time to work on any technical business or customer challenge that they want to work on, and then we finish the program where to sort of showcase to have them give them opportunity to show within the organization. What they did, and we were really surprised with how well they picked up this challenge and just ran with it. And there was a lot of creativity that came out with it. I think one that is external that we're very proud of is tool that reassigns your dynamics records. It's called the dynamics record ownership project. It was built. On CRM toolbox and in the two months it received over 2 1/2 thousand downloads external downloads from customers. So yeah, there was these kinds of projects.

00:05:20 Mark Smith
So tell us about the platinum.

00:05:23 Tomas Gold
Maybe before we jump into Platinum awards, I would just like to highlight that whether the program has been running for two years. So we have seen some evolution of the program and Latina has been the one leading the entire effort since the inception for business applications. And really what we have seen, yes.

00:05:38 Mark Smith
Tina. Wow. OK, that's awesome. That's epic.

00:05:43 Tomas Gold
Yes. And actually first year, all the different teams across all the different solution areas were basically running these technical challenges. One of the ones Tina. And actually the one a different one that I can talk about more one I think the all up awards that's the best technical challenge across all social areas. And so I think it's been really successful and I think it's a little bit dying back to the entire AI transformation age that we are part of you kind of need to experiment and you need to have a permission to be able to experiment. And one of the objectives of presenters was to basically give people that space that creative.

00:06:15 Mark Smith
Yeah.

00:06:20 Tomas Gold
Outlets in hope that it will lead to some interesting outcomes and some of the projects really made a.

00:06:25 Mark Smith
Difference. So when you say CSU was this across modern work, across biz apps, across Azure, everybody OK? And I take it was the entire Geo as an anyone in that unit anywhere in the world?

00:06:38 Tomas Gold
Yes, exactly.

00:06:39 Mark Smith
Interesting. OK. OK. OK. And so and this has been running for the last two years and did people self nominate in or did their peers or colleagues nominate them in? How did it run from that perspective?

00:06:51 Tina Van Heerden
So it's a peer nominated program. Nice. So we don't take any input from managers. They are nominated by their. Years to give you an idea of the size. So there's 200 architects across the world that are not as well that get into the program. Of course, we have a lot more in the worldwide organization and so they kind of get divided based on how many we have. We get a couple of seats on the program worldwide and and it's really an honor.

00:07:06 Mark Smith
Yeah.

00:07:17 Tina Van Heerden
It's very difficult to get in. You can imagine with the kind of work that they're doing. So it's really an honor for them to get into.

00:07:24 Mark Smith
Wow. Wow. And so out of what has happened today, obviously. The the one solution you mentioned in the biz apps around the record reassignment, what are the other big wins? Has it been more AI driven or has it been more just across the board in general?

00:07:40 Tomas Gold
Actually, I think a lot of them are focused around AI and the one that actually won the OLAP award last year was more internally facing. What we do is we provide customers with a road map insight. It's like monthly technical update, we think and it does take a lot of effort to waste every month to put together the presentation.

00:07:53 Mark Smith
Yes.

00:08:00 Tomas Gold
You just need to look at, you know, the enduring road map you need to put together the storyline you need to put together slides. And the team that was working on this, they actually employ AI to be able to produce the deck and all the materials for the monthly technical update briefing, which will basically if you look at how many times this offering is delivered, which is in hundreds of times to customers, that saves unbelievable amount of time and effort.

00:08:25 Mark Smith
Incredible. Incredible. And you did win an award off it, right, Thomas?

00:08:29 Tomas Gold
Yes, I was very fortunate to win the bottom award this year, but it's been a humbling moment because one of the reasons for it or the nomination for it was for pacesetters, which is the project that Tina had. And so I think it's a shared award, definitely for my entire FY25 team.

00:08:48 Mark Smith
It's so awesome. It's so. With coming from a biceps background, what are you seeing in how AI is really starting to permeate dynamics, the power platform, and, you know, recently I had a conversation with Ryan Cunningham and really talking about, you know, enterprise vibe coding. And really, a change coming to the platform where. Almost a infinite flexibility coming with the tool set to build solutions like and. We've always been very flexible, but to be able to prompt more and more into reality. From the inside, what are you seeing? What are you seeing in your customers? What can you share and what is the rate of change that you're seeing? Yeah, particularly in the customer landscape in this. Bit.

00:09:38 Tomas Gold
So maybe I start and Tina can try. I mean, maybe one thing to actually highlight is that if you consider how we think about business applications and power platform as of July this year within Microsoft, we tend to refer to them as the AI business process area. And you see that this is a subtle change. It's not a dramatic change. We're moving from your business. Applications to our business process, so a lot of the terms remain the. But actually there is a meaning behind it and if you break it down, I think we will see a lot of evolution behind the business applications or apps in general. But what will remain constant in the ad transformation is the focus on the business process. And I think in the context of the business process, obviously the solutions may be different. In some instances, so the solution may be an. In other instances, it may be an agent that is actually performing some actions, so within that business process, autonomously. So I think we're kind of keeping it a little bit more open. Also keeping an open mind and we're trying to figure out how apps and agents will collaborate going forward. Obviously with the human factor involved.

00:10:47 Mark Smith
Are you seeing an increasing demand from the customer base, particularly in the enterprise accounts? Around really wanting to get a strategy in place for AI. And, you know, cause one of things I've heard a lot in the market that there are a lot of situations where customers have, you know, they've gone as far as pilots. But they're getting stuck in pilot land and not going into production. And I don't know if you've come across that, but how are you helping customers take that kind of, Yep. It's been proved in the pilot. How do we get this thing into pro?

00:11:24 Tina Van Heerden
So I think what I loved about your discussion with Ryan was that he also had this focus on experimentation, right? He said something like it's better to be late, early and wrong than to be late in this space. And when you mentioned 5:30, when I think about what kind of vibe.

00:11:37 Mark Smith
Yes.

00:11:44 Tina Van Heerden
Thing you can do today with the power platform. I'm a huge fan of the plan designer. I've designed a personal app in an hour that I, you know, use every day and I'm not a power platform developer. So I can say that this is really, really easy to just kind of prompt these tools and get apps. Out of it. But I also understand that my app that I built is not production ready. It doesn't have any governance. Actually it took me two hours because I first designed it on the. Wrong environment and. Then I had to change and do it from but on the other environment, so like these kind of easy vibe coding things that we do are obviously not ready for production, and I think also customers. Understand that there's more behind creating these tools than just, you know, the development. There is also things like the adoption and finding the use case where you really have the return on investment. I think the way that we kind of see or saw this evolve in the last couple of years is that when these models and these kind of Co pilots came out, everyone said Ohh wow, look how shiny this is. It's like a magic wand, right? You just, you know, swing it around and it does these amazing tricks and it went completely viral. But then we realized that a magic wand is just a stick. If you don't have a good spell right, you need a use case that you really think like, wow, this, like, makes someone levitate, right? This is an awesome spell. And even that is not yet enough. You need then Wizards that are like swinging these sticks and casting the spells, and that's where I think the adoption comes in. So you need this kind of trifecta of good tool that we understand how it works and with the good use case and the adoption arm that is the the people casting the spells.

00:13:29 Mark Smith
Yeah, you mentioned something interesting there about, you know, the app that you built is not production ready or it's not enterprise grade or anything. But inside an organization, there are probably thousands of use cases. That don't need to be enterprise grade. I just came across 1 today a consultant. Vibe coded an app for statement of work validator for their business. And the whole thing is, is that if you get statement or works wrong, it creates a a high degree of ambiguity right about what the deliverable was, what was the agreed terms, etcetera. And so they you're saying in his business he had an internal process that you know around how to write a good statement of work. But no matter what happens, people forget. And so he vibe. Voted in three hours. Atul. That. Gave a a high degree of accuracy or ability to validate and I was just thinking, you know, that might not need to be enterprise grade, right? It does a a very specific thing. It's three hours worth of work. He's he's already accumulating many hours saved in his teams that the statement of works they're producing and it makes me think that are we going into the age now where?

00:14:43 Mark Smith
Things like the power platform is our modern day excel. Even though we've talked about, you know, moving away from Excel so much. But it's our enterprise grade Excel where anybody can build anything they want. That might be just for their team or their department, or, you know, there might be no intention to make it customer facing but could easily wipe hours out of there from a, you know, the drudge work of their day, potentially by creating these little solutions that are never going to have a risk profile for the organization that would, you know. Make them shy away from using the tick.

00:15:19 Tomas Gold 
Mark, you know it was 1 interesting point that you mentioned when it came to that example, it is probably the number one lesson learned that we had from last year or two years with our teams and that's you were very clear about the value. So as soon as you started to talk about the app and what it is doing, you identified actually why it is so crucial and why it is important. All for. People, and I think that is really I think the focus that we see a lot of customers not moving from the power to production if the use case is not clearly identifying the value, it is not a necessary for people. It is just like a proof of concept or just like an interesting application of AI. So I think that's the number one question we've been trying to answer.

00:15:58 Mark Smith
Yeah.

00:16:02 Tomas Gold
And I think you know also what we've learned is that if we go in with specific use cases that drive value, that's like the first entry point to AI. That's typically much better than if we try to go in with this, like broad AI transformation, communication. A lot of these workshops, a lot of these governance, security, all of this is important. But if you are not having that. Really, in the centre of what you do, you will probably not make it from pile.

00:16:28 Mark Smith
The other concern that I see rising is that you take plan designer. For example, it's potentially taking a lot of work away from a lot of people that would have traditionally done that role, build up the RD's of the roles and all. All that type of thing saved us a massive amount of time. But you know, if you tell a lawyer, a lawyer or a law firm. That you can save them a lot of time. They're not really interested, right? Because they bill by that three minute increment. They don't want to reduce their time on things, although I've heard a recent use case brilliant where a law firm charged $24,000 to do a a piece of legal matter, and it was just standard that their company churned out over and over again. They applied AI to the process and they were able to reduce the time to deliver that down to about two hours. And prior to that it was taking them around 25 hours to do do the work. The law firm turned around half the price to the market for that. Offering. But their lawyers can do so many more of them now. So ohh, it's like this is a brilliant case of where the customer wins.

00:17:35 Mark Smith
But the law firm doesn't lose because they've been able to multiply the actual amount of cases that they can get through that fit this particular parameter. How many of those type of conversations and I go back to the how do you have the conversation with people worried about their jobs and the implications? And I I think particularly consultants right now are going, you know, if I'm a consultant that's an expert on power automate or power apps for that matter. What does my role look like in three to five years time and in the speed that things are going in the way AI, you know, we're going to see so much more use of agents and what we do. We see Microsoft even looking at their API's and going are these API's built for people or are they built for agent agent communication, right. It totally changes the paradigm of how you build these things. What are you seeing?

00:18:32 Tomas Gold
There are so many different ways how we can talk about it, but I'll start from 1 angle. I think AI in general is very good at specific tasks and I think for those tasks I think the personal strategy for anybody should be how do I best augment myself with those AI capabilities and become exactly as you mentioned, more productive, faster or better in some way. But then there are still some. Areas where I I think it's not very good at we can talk about it. So maybe a little bit more in detail, but it's for example the experience dimension or the trust or the ethics where I think you know we do not see AI still active in that space.

00:19:07 Mark Smith
Yes.

00:19:14 Tomas Gold
And I think those are exclusive areas where people can lean into more and maybe with the extra time that they have because they can't step down from some other responsibilities that they had otherwise. So I think it's like this two prong approach on the first part, you know, try to leverage AI as much as possible, just become faster and better. And I think that's the strategy I would apply. And then consider what are the roles that AI is not very good. And I think those dimensions actually exist for every job, and then see what else you can do? How much experience you can provide to the customer. Can you advise the customer more? Can you spend more time actually talking to the customer because you can spend less time in front of the computer actually code? Yeah. I think definitely AI is going to completely transform the way that local developers were consultants, you know, all of our roles are going to change into going into the future.

00:20:08 Tina Van Heerden
I think one of the things that feels overwhelming to me is at the pace that this is.
 Funny, and I think the thing that I try to focus on is really considering that the time that I win my workflow by using AI, I need to use that to keep up with changes and I need to see keeping up with changes and learning and growing with this technology as a core part of my job. It's not something that I do anymore. Kind of extra if I have time I put that block first. What's happening in the tools? How are? Is changing and I think that is kind of like keeping up with the change is now a core part of everyone's job and we should see it that way. And I mean, this isn't easy. It's really something that you have to be intentional about. And I also think that change is difficult. So it's not something that just sort of happens slowly to you. It's something that you have to identify. Be intentional about and. Really go with the change instead of just kind of waiting for it to happen to you.

00:21:07 Mark Smith
Yeah. So true. Thomas, I just want to go back to one thing that you said there as well, which is you talked about augmentation like you need to, you know, be an AI augmenter. And it's interesting because that is one of the number one skills that I'm seeing repeated over and over. And YouTube and other social media platforms at the moment is that. The importance of practically using AI to augment as much as you can, because in that you gain the skills and I just wonder is there anything that advice you could give around, you know, a listener go away and practically think about their life, their work, their day differently? From that augmentation lens.

00:21:56 Tomas Gold
I think it's fair to admit that neither you mark or you Tina or or myself are excluded from this change management process. So I think we are all learning as all the listeners are. I think you have a couple of different elements to any change management and basically learning. The first one is you just have to try and I think you have to experiment.

00:22:03 Mark Smith
Yes.

00:22:16 Tomas Gold
There's a huge element to it. Try to incorporate AI COPILOTS agents into your daily work the best way. You can. It's probably not going to be very successful initially, but overtime I think you will find some elements where I can help you.

00:22:31 Mark Smith
Yeah.

00:22:31 Tomas Gold
Second, I think you know you have also an aspect of time. It typically time is the best change agent as they say, they'll give yourself a little bit space to actually get better at it. I remember my beginnings with copilot and prompting. It wasn't great, but now I think I'm actually quite effective. It, and I ultimately believe that quite often I think in the media it is positioned that it's either humans or AI and that's the competition. But I actually subscribe to a slightly different view. I think that in the next three to five years, your biggest competitor is not AI, but your biggest competitor is somebody else, like another consultant who is just better at using AI. And simply they will be more productive. They will be faster, they will be better and then ultimately all customers and all companies will start to choose those individuals. So I think your competitions will not AI AI is your friends, your competition are other people who can use. They are much. More than you, Tina.

00:23:31 Tina Van Heerden
I think for me, what experimentation has shown me within. My own job is that. You have this problem of that when you have a hammer, every problem looks like a nail and it takes some time to figure out what are the things that you can actually truly solve with generative AI or other forms of AI. And what are the things. That. Should not be automated or cannot be automated, or can be automated with other things. And I think the thing that has. Made me most productive. Is. I have a power automate flow that schedules meetings for me. Because we're in a big international company, so it's often hard to find time across time zones, so I just give it the names of people that I want to meet with and how long and then it sends me three options in in teams and I just choose which of the options work best for me. I wanted to write this in AI, but then I realized it's not an AI problem, right? I know the names of the people and I know how long I want. It's a power automate, no AI solution and I think just doing this kind of experimentation thing like I have this problem. How do I solve it with a OK I can actually solve it. Much easier. Another example is I want to streamline my inbox. One of the things I wanted to do is to kind of auto reply to some of these kind of just general. I need to review a teams group that I have or something and what I realize is that I shouldn't be automating these emails. I should be maintaining better which teams I am part of. And reduce the number of emails that I get. So for me experimenting with AI is not just about learning AI, it's also about learning which problems. Are good to say. With it.

00:25:08 Mark Smith
Yeah, I like that. I like that as we draw to a close Microsoft internally 10, you shared this at the event that we're at, you have a an internal report of 1000, the first thousand. And I mean it, it started off as you know an incremented up to this thousand customers that you took on the copilot journey. What were the standout things in your mind or takeaways from that observation over the last two to three years of that happening? Yeah, they're two or three real highlights that have stuck with you, resonated with you, being stories that you tell about. The impact the challenges, the joys of those first thousand customers that went through that process.

00:25:54 Tina Van Heerden
So maybe I can start with some background. We really wanted to be intentional with the way that we work with customers and copilot. So after the first 1000 conversations that we had with customers, we started a program where we interviewed the architects. That worked with these customer. To get a sense of how this is going, how we should pivot kind of where the gaps are. And it was interesting. We found really a lot of sort of unexpected things where customers get stuck. I think this example where you said lots of customers are easy to get to the POC phase but then getting them in production. Is a challenge. And there were a lot of things that kind of prevented that next steps and most of them were due to lack of planning. So what I would say as that is things like considering the security requirements when you are already still in the sandbox and testing out, because if you don't talk to the security team, they might be the ones that block this going into.

00:26:55 Mark Smith
Yes.

00:26:55 Tina Van Heerden
To the next step, right? Same with governance. Of course. We also had lots of customers that were interested in copilot, but they were in industries where there were maybe some compliance concerns, right? Healthcare, financial services and. So I. Think. Many of the issues that we saw could have been prevented by rediscussing the end to end process of like where do we want to go with this and maybe coming back to it must side about discussing the value so that when you kind of get to this like when it starts to get hard and you have to overcome this challenge because you have this kind of golden or North star of what is the value for us. Then it's easier to get through these block.

00:27:31 Mark Smith
Chris. Yeah, that's brilliant. That's brilliant. Thomas, do you have?

00:27:35 Tomas Gold
Maybe I'll just be completely transparent. The our transformation is impacting everybody and also at Microsoft, we try to learn as much as possible. Nobody has all the answers and the answers evolve, you know month over month. So this White Paper for example and this effort to actually gather feedback. From our customers and all our architects in the field, that was quite vital. How we influence actually either the product direction or just the way we engage with customers. Asking this applies to everybody. The world is changing very fast. Just take time to learn. Take time to maybe consider what has been happening in the past three months and always adapt. I think the adaptability is probably the other keywords I would highlight. I think those that will be most adaptable to the change that will be the winners.

00:28:21 Mark Smith
My final question and this will is each of you would like to answer this one. What's your personal use cases for AI that you? It's kind of like? What you use it for and you feel like probably nobody knows how awesome it is, or nobody else knows about this. You know the particular way you're using it or the particular use case for you personally, it might be something to help your in your business life. It might be to help you in your personal life. I use mine in my greenhouse whenever I'm planting seeds, I use it. To I just read out the you know, I have it on voice mode. I read out the scientific name of the seed and I say this is the situation I'm in. This is my zone in the world. What's the ultimate, you know, to get the best propagation? And I have that voice mode cause I've got gloves and stuff on. Right. And it's very, very helpful. In gardening, what about yourselves?

00:29:15 Tina Van Heerden
I'll give it a go, so I also love the plant use case. I've been walking around with my phone and taking pictures of plants everywhere, but I think what has had a bigger impact on my work productivity is that. I really try to use speaking instead of typing because I'm not a very fast typer. I was convinced to take French in high school and now I live in a German speaking part of Switzerland, so that was not very useful and as a result I didn't take typing. So I'm quite as slow type. Paper and I started using copilot to. Even if I want to reply to someone in the teams message or an e-mail, I speak it to copilot and they kind of understands the meaning and gives me a good. Sentence or paragraph to reply yes. And how this has really impacted my workflow is it's difficult to speak to copilot while I am on. A different call. Because I can't speak and listen at the same time. So it means that now I'm a bit less distracted or multitasking in calls because I wait for the call to finish and then I start speaking to copilot to reply to the messages that I got in the meantime or the emails that I have to respond to. And it seems simple, but just the amount of kind of mental noise that it's taken out that. I don't type messages while I'm in a meeting. I listen and then I speak to copilot and. Do that work afterwards.

00:30:35 Mark Smith
That is awesome. That is awesome. Tomas.

00:30:38 Tomas Gold
So I I personally actually have a playground. I built my own agents and I gave it access to pretty much all the information that I possibly have at Microsoft. And I obviously experiment with different prompts and like different different scenarios, but the one that actually stuck with me and I had some interesting insights coming out of it in. The past few weeks. Is to actually leverage agents and AI for one specific use case, and that's really constructive criticism or dissent. Basically a personal challenger that we'll look at. You know how I communicate what decisions I made in the past few. Yes. Thanks. And will they try to argue an opposite case, and especially, you know, if you have any management in overall like this type of dissent and constructive feedback is absolutely vital because sometimes your team may not be. But I completely happy to disagree with you vocally. But AI is actually very good at it because it doesn't have. Links. It doesn't care. It doesn't care if it hurts you or how much it hurts you, and it actually provides me with this mirror that I can easily look into. And I. Can. Act on so for example, I got the feedback, you know, last week that I should be more directive and instructive. In some instances the specific people. So that's in the context of fiscal changes that we go through at Microsoft. But the Super interesting use case AI has no feelings, and therefore it can actually tell you things that other people wouldn't.

00:31:56 Mark Smith
Yeah. That is brilliant. I feel like I want to keep talking to you. I want to more on what Tina just said as and I will actually say one thing. They've done a lot of the research coming out. Is that the most the people that have adapted massively to AI are all using voice. They even if they're fast typists, they are going to voice as the interactive agent because you can be so much richer often in your voice than you can on the keyboard, so it is a a sign of somebody that is really augmenting heavily with the tech, as it often be voice first rather than keyboard. I have always trying to force I'll start typing. I was like, no, stop use your voice. Use your voice just to get in that habit. More to detach a bit from the keyboard and say I think I could speak to you for another hour. Our time is up though. Thank you both so much for coming on and sharing your stories.

00:32:53 Tina Van Heerden
Thank you so much, Mark.

00:32:54 Tomas Gold
Mark, thank you so much. And I think when you ask me the question about what can you do to actually become better, I think one way how to do it is to actually talk to people and learn from the community. So I think you actually play a vital role with this podcast in that process. So thank you for doing it.

00:33:11 Mark Smith
I love it. Thank you. Hey, thanks for listening. I'm your host, Mark Smith, otherwise known as the nz365guy. Is there a guest you would like to see on the show from Microsoft? Please message me on LinkedIn and I'll see what I can. Do. Final question for you, how will you create with copilot today? Kakute.