Your Identity and Power BI – Ep.565

Tommy Puglia, Mike Carlo, and Kurt Buhler spend this Explicit Measures episode on a title Mike had to shorten: your identity is bigger than Power BI. Microsoft Fabric and agents have shaken the years when “I build reports” was a complete professional answer, and the hour is about what you keep, what you hand to a tool, and what you still have to judge yourself.

News & Announcements

• Using AI Harnesses to Harness Microsoft Fabric — Thursday, September 24, from 3:00 to 5:00 PM Central, the Chicago Fabric and Power BI user group meets at the Microsoft Technology Center, 200 East Randolph, downtown Chicago. Tommy needs RSVPs in before he submits names for building access. The session is the pair of harnesses they have been building on the show: a context harness, where you keep, scope, and refine the information, and an execution harness that takes that context and does the work, whether that is writing, development, or a model. He wants to build that loop live, or have it hosted and recorded, and he wants the room to open the way the Homebrew Computer Club did, with people showing something they have been working on.

• A skill.md Is Not Wisdom — Tommy’s September 21 post on Matt Pocock’s grill-me skill, which Mike had pointed him at. A skill file can carry format, section order, and vocabulary, and it will fill them in with confidence. It cannot carry the judgment that a line is not a deliverable, or that the work is six hours and not two. Grill-me puts the agent on the asking side before anything gets executed, and it recommends an answer so you are reacting instead of composing. Tommy runs it on statements of work, design patterns, and new dashboards, stacked with the other skills he already loads, and he answers out loud with the microphone so he is not typing a general prompt and accepting whatever the harness recommends. Mike hears the same gap he has been working on with his family: how an idea in your head becomes something the computer can actually build.

Main Discussion

Topic: What is left of “I am a Power BI person” once Fabric and agents are in the job

Mike tells Tommy up front that the on-screen title is abbreviated. A lot of people trained as Power BI experts, and Fabric plus the market around it means that identity has to get bigger than the report. Kurt Buhler is back from last week, and they reach the question the long way: through video games, through personal data, and through the LinkedIn note Tommy keeps getting from people who are looking for work.

• A report succeeds when someone can leave it. Kurt has been studying video game design for three years, and he holds up Planet Zoo as the best dashboard he has used. Managing a zoo is a high-cognitive-load data problem, and the game only stays fun if that information is actionable and stays out of the way of the actual task. A report, in his view, has the same success condition. People have jobs, and those jobs are not looking at your dashboard. A game optimizes for engagement. The heads-up display does not. Tommy has said on this show for a long time that nobody actually wants a report. They want the information, and they would take a better medium if one existed. He credits an article of Kurt’s on what a person should get in three seconds, thirty seconds, and three minutes, and he notes the underlying idea was not originally Kurt’s. The same test applies to a data agent or a conversational experience: sometimes natural language is the frustrating path, and the real question is which tier of question this is, and whether the answer belongs in a report, an agent, an alert, or something else. Kurt’s advice for learning any of it is to start with something you care about. Tommy did that with his own Strava cycling data, threw out a heart-rate-and-speed view that did not change a decision, and kept iterating until the report told him what to do next. Mike found the same cut in his consulting business. He built a report full of visuals and uses a handful of them to know whether the business is winning or losing.

• The niche that paid the bills is the niche that got shaken. Tommy takes the room back to roughly 2015 through 2020. On the Microsoft platform, a Power BI pro was an identity: DAX, Power Query, semantic modeling, dashboard design, governance, and the rarer skill of understanding the business. He leaves the word “threatened” alone and stays with shaken. Fabric rattled that cage, and AI finished the shake. If you still introduce yourself only as a Power BI person, he thinks you are capping the career, and the live question is where that professional identity should sit and how you market it. The messages he gets are specific. I am a Power BI developer. I build excellent reports. Here is what I know.

• The tool used to be how you fronted the expertise. Kurt’s own start was Tableau, during research, after someone at his gym told him the work could be a job. The focus landed on the tool because that is what the tasks used, and because clients and hiring managers were looking for someone specialized in it. Knowing the limits, the performance tricks, and the hundreds of small report patterns — a DAX switch to choose a measure is the example he reaches for — was how you showed what you could do. Fabric widened the expectation. Power BI sits inside a larger platform and is deeply connected to it, so employers and clients now assume that even a visualization-focused person can do work that used to feel outside the wheelhouse. AI, which they covered last time, changes the quantity of what you can take on, the quality of it, and how broad the surface gets.

• Mike’s parallel is an intern who expected to write the code. He has been pushing his developers to stop hand-writing functions and to learn how to build with AI, and the pushback from some lead engineers was that writing the code was the job. A summer intern told him the whole vision of a computer science degree was to sit down and write a company’s system. Mike’s answer was that most of that writing will be covered by the time the intern graduates, and the useful move is from entry-level implementer toward someone who understands systems: how the database works, how identity works, the conceptual layer. JavaScript and TypeScript are becoming less of the requirement. Brought back to Power BI, AI is eating what he calls the click tax, the heavy labor inside the report, and it is chipping at the belief that the BI engineer is the person who knows Power Query and DAX. The skill he wants named out loud is listening to the business, empathy for the person who has to consume what you produce, and a critical eye on what you are building. What you build, and how people interact with it, now outweighs the act of building. He is now standing up Fabric SQL databases he would have skipped, and asking the agent where his own knowledge runs out.

• The effort was already at the ceiling before agents showed up. Tommy’s picture of the first six or seven years, before Fabric, is that one person cannot own an entire company’s data and was asked to anyway, because it was “just” Power BI Desktop. Models, governance, DAX measures, visuals, and a stack of hats. If effort had a cap at 100 percent, most of those people were already at 99, and a little proud of it, for the job security and for how many parts they touched. Fabric did not arrive as a clean split into a separate data-engineer role. It arrived as more of the same job, and he pictures that line going past the cap. He thinks 2026 is the first time agent tooling might bring an individual back under it. The stress of the old identity is still in the room.

• The fear is attached to the wrong competence. Kurt’s direct version, once they stop circling: the models are getting sophisticated, the technical skills people are proud of have been core to their identities, and the anxiety that AI will take the job or make the person less valuable is real. The competences that hold are business processes and problems, and the ability to wear the hats. You are already interfacing with business domains and technical ones, SAP included, to get the data in. Stay on solving business problems, engage the business, go into the technical detail, and use the agent tools responsibly inside governance, and he thinks that person is in a strong position. If the honest focus is memorizing paths, best-practice lists, and UI tricks, he wants that attention pivoted toward the problem. He calls the shift liberation. You are more than the Power BI identity, and with agent tools and real governance you can leave a single tool and a single way of working. Mike doubles down with Fabric apps. A report that only shells data outward can become a place where data comes back in, where someone submits something and a real-time action hits Fabric, including a SQL database that ingests it. He thinks Fabric apps before AI would have been a bump on the radar, a D3.js project for a few people who already knew how. Agents make the idea buildable, still in service of a business problem.

• The weight of the skill moves, and trench wisdom stays. Mike quotes an interview with Jensen Huang, the CEO of Nvidia, that has stuck with him: an employee he is paying $200,000 who is not using $500,000 of tokens a year is a question about why AI is not in the work. The reading Mike takes is about time. Report-building consumed the hours you would have spent on the larger problem. If that work can be minimized, the questions get bigger. It is also why the three of them end up rebuilding software over a weekend when a button is the wrong color. He expects business process to land in custom harnesses an organization builds, maintains, and runs, and he calls that harness the IP. Tommy’s name for the judgment in the meeting is the Department of Common Sense, and before the episode ends Mike has a Department of Common Sense hoodie up on the Power BI tips store. Walking in with three options instead of one, because time and resources moved, is the shift Tommy hears. Kurt’s correction is that the skills are still valuable, and the weight of that value moves. Sit someone down with an agent and say “make a Power BI report.” The person who knows design and the product can say to use error bars for the gap between actual and target. The person who only memorized tips and tricks will find that store less valuable, and so will the person who never tried to understand the business process. Technical understanding still lets you do more than you could before. His example of what just got cheap is turning off auto date/time on a model that is too big. That used to be a talk, and it used to pay a consulting week. An agent with the semantic model handles it. Tommy pushes on the word “transfer” with a pot of water: move some into another cup and the first pot has less in it. The vertical depth in one corner of Power BI gets shallower. Trench wisdom remains what lets you ask the right questions, and an agent in the loop makes it harder to claim you have personally lived every evaluation-context case. Kurt would spend the recovered time on why the work exists, on data quality and governance, on getting the house in order, and on the climb from descriptive analytics toward work that looked unrealistic for the team a year or two ago. The line he wants people with authority to say out loud is augment, not replace. The failure mode has a name he borrows from Shopify, from last week: the slop grenade, a wild idea thrown at colleagues with no skill to judge whether it has merit. A labeled prototype, offered as a picture of what you are thinking and then evaluated, is a different act. Mike is hearing the unread version from directors. People submit something to an AI and email forward whatever comes back.

Looking Forward

Kurt’s close is aimed at the person listening. The skills from Power BI and from data already scale past the tool, and they can have a positive impact if you augment them with agents and point them at problems you already have an opinion about. Try it. Once it feels possible, be careful with what he keeps calling token cocaine. An optimistic future, in his telling, needs optimistic people to act, because every domain is already in the middle of this and the opening is larger than it looks. Mike’s version is the same invitation from the other side of the mood. The ground is moving, a lot of people are scared, and the move he wants is to spend time with AI, use the knowledge and IP you already have, and build something with it. He has been doing that with his son on a small 8-bit game, Bit Drop, which you can find by searching Reddit for that name. He narrated the features, ran a design session, and directed an agent that wrote the code, the graphics, and the sound, because describing a game to an AI is harder than describing a dashboard. He currently holds the highest score. He also describes both poles of a documentary his family just watched, the optimist and the doom-and-gloom, and he lands on the optimistic side. His comparison is cars: get the innovation out, then let people make it safer.

Tommy holds the livelihood line under that optimism. Mike has been producing AI-generated songs, combing the lyrics and acting as the producer, and he is clear that nobody is coming to watch him perform them and that he is not making money from them. A hobby is a different decision from a job posting. Kurt, who knows working artists and musicians already hit by image and music generation, is uneasy about the intellectual property baked into the tools and about how easily a suggestion arrives without a source. He still revises hard opinions when the capability changes — code and Blender are a different act from ripping off a drawing — and he wants intent, effort, and accessibility in the judgment, including people with dyslexia who use it to write. Speaking as a biomedical scientist, he puts wonderful medical and genomic possibilities and terrifying ones on the same spectrum, and he wants the nuance kept. His ask when an AI hands you an idea is to try to find where it came from, so the lineage of human work does not get dropped. Tommy’s practical version of the same split: independents can test Fabric and agent skills whenever they want, and he keeps hearing from Power BI people whose companies have not enabled Fabric at all. They are losing practice. He is curious what the job description that used to say “Power BI” is going to look like, and whether the honest path is deeper trench wisdom in the product or a wider data practice with agents. The episode leaves that twofold, and it leaves Thursday’s user group as the room where the practice can show up in public.

Episode Transcript

0:01 I get your fix. Explicit measures. Drop the beat now. Pumpkins feel the crowd. Explicit measures. Hello everyone and welcome back to the explicit measures podcast with Tommy and Mike. Tommy, good morning. How are you doing? doing? I’m doing excellent Mike. How you doing? Welcome back to the show. We are once again start another episode., our main topic today before we get into a

0:31 main topic today before we get into a little bit of news here, I had to abbreviate the title so the title probably isn’t really indicative of the of what we’re topic is. We’ll we’ll see how it goes. The topic really here is your identity is bigger than PowerBI. I think a lot of people have globbed on to or done a lot of education or training around just being a PowerBI expert. But I think with now fabric and where the market is going, your identity needs to start expanding. You you you’re bigger than just the PowerBI reports. So

1:02 bigger than just the PowerBI reports. So that being said, Tommy, give us some updates here. You have a meetup that you’ve listed here on the on the news this week. Yeah. So make sure to RSVP by really tomorrow. Our next Chicago PowerBI and Microsoft Fabric user group is actually happening Thursday the 24th at 3 PM at the Randol or at the credential building downtown Chicago. The scope of the meeting is what we’re calling how to use AI harnesses to harness Microsoft fabric. we’re going

1:33 harness Microsoft fabric. we’re going to talk about this concept Mike that we’ve almost developed on the podcast. We’ve talked about where there’s a concept harness and an execution harness. Where do you actually keep your con or context harness rather? Where do you keep all the information and basically managing and scoping and refining the context that you need and then sending that out to the execution harness to actually do the work to write the paper to do development to build the model and basically how do those

2:03 the model and basically how do those work together? What does that loop look like? We’re going to be building that out. So really exciting to I cannot wait to showcase this. I will try to do it live or have it hosted live. And then the other part of this too is ories have it recorded. The other part too is if you’re attending, we already have quite a few people attending right now. We want to see your wins at the beginning. We’re going to try something new. If you want to share something that you’ve been working on, you want to share in front of the group. we want to do that kind of the group. we want to do that that what is it what Apple used to do

2:33 of that what is it what Apple used to do the homebrew computer group. Showcase what you’ve been working on. We want to see it., so really want to build that network and that community in the group. But really make sure to RSVP by tomorrow so I can submit the names so you can get into the building. All right, the link is in the description below. The link is also in the chat window as well if you are able to attend. It is a good event. It is in the downtown of Chicago. It is the the Microsoft office downtown as well. So you’ll be able to go visit the Microsoft office and hang out with Tommy. So it’s

3:05 office and hang out with Tommy. So it’s a good time. All right., moving on. Let’s go on to our other topic here. Tommy, you got an article here. A skill is not wisdom. What do you mean by this? Yes. So, this is you introduced me to the grill me skill. And I wanted to write a little more elaborately about what Matt PCO created was called the grill me skill. And I wanted to show the difference or really the importance of that skill at least to my workflow utilizing that with other skills. And I think a lot of people take like I’m just

3:35 think a lot of people take like I’m just going to do the,, PowerBI spinic model skill and I’m just going to say build it, right? Or I’m going to build the skill for a presentation that I already have. Those are great. However, you just usually just accept what they say. What the recommended thing that Micros what Claude or whatever the harness does. the grill me skill. I really show how I use it and why that’s so important because it really forces you to think about what you’re going to execute on. And I love that because I to your point, I really

4:07 that because I to your point, I really don’t do a lot of typing. It’s me kind don’t do a lot of typing. It’s me answering questions using the of answering questions using the microphone feature and whatever harness I’m in. And it I really rely on this on any statement of work that I build as I start building out design patterns or for building a new dashboard. I really have utilized this in really every workflow that I do to make sure that I’ve thought about everything, but also that I’m not just providing a general prompt whenever I’m going to do any execution, right? And you’ve talked about the skill a ton. I have not

4:39 about the skill a ton. I have not relied on this. So, it’s really just an article writing about the importance of it and how to utilize it best. Yeah. So, this is a really good connection between where I think you connection between where I think where your head lives and where the know where your head lives and where the agent knows and what you want to build versus what the agent can build. I think there’s I’ve been teaching this to my family a lot more. all my we just watched a documentary about AI recently and yeah, the whole family did and did and there was like basically two

5:09 there was like basically two personalities in the documentary. There was one of like the the heavy optimist like everything’s going to be better. we’re going to be able to do all these other innovations. There’s things that are going to be unlocked now with AI that we’ve never been able to do before. And then there’s like the doom and gloom side, right?, it’s going to kill us all. It’s going to the AI are just going to take over. It’s going to take all of our energy and our power. Like there’s lot So, you have to like figure out what side of AI do you fit on. Do you think it’s going to be a revolutionary technology? Are you an optimistic side or are you kind you an optimistic side or are you more of the pessimistic side?, but

5:40 of more of the pessimistic side?, but that being said, the our family loves AI. Like all my kids are like, “This is fun. I like using it. It makes it easier. I I do a lot more the directing and the managing.”, and the communication and while the story is story is we continually dream up things in our head. head. We have a picture of like what something should look like, a statement of work., ,, there’s a a lot of things that you could put into your mind. How do you get it from here your mind into what

6:11 you get it from here your mind into what the computer can build? And so a lot of my experimentation over the last couple months have just been trying to unpack like what do I need to learn or what tooling do I need to get my hands around to understand how to get from what’s in my mind into what the computer understands so we have a mutual understanding about a feature, an app, a game or whatever. Oh, I will say this., I’ll go grab a link., Tommy, this is not a news item, but this is just a a very random before you go to the random thing. I need to something that you said. I don’t

6:41 need to something that you said. I don’t want to steal anyone’s thunder here because they did mention this on the private chat, but to your thing, you have to pick a side. Can’t you be yes to both because you can also feel this great possibilities, but there’s also this inherent danger. Personally, if you are a reader of Pope Leo, he came out with an encyclical about what is coming with AI, which goes on both ways where it’s like, yeah, this is actually really great. However, there’s a lot of these arguments here. So, I

7:11 a lot of these arguments here. So, I would argue, Mike, that you don’t have to pick a side. You can easily say, “This is awesome and amazing and cool, but also, holy crap, we really need to think about this before we move forward.” I think that’s definitely a place we can go. I think there’s [snorts] I again Kurt’s been on the podcast a little bit here recently and and so his three C’s I think are very very spot on here and I would would be apppropo to like really keep talking about this without Kurt being useo by the way. Yeah, I think that’s how you use it. I heard it said one place. We’ll we’ll see. So

7:41 it said one place. We’ll we’ll see. So So let’s hold off on this conversation. I’m pretty sure Kurt will have something to say about this one. So let’s So let’s chomping at the bits. Chomping at the bits. Let’s hold off this one. Okay, let me quickly give you my fun thing here really quick. So, [snorts] in part of my what’s in my mind versus what’s on the computer screen, my son and I have been building little 8-bit video games for the last couple days. Okay. Okay. I have had a challenge of where do we where do we publish them? How do we get people to use them? What does it look like? like? So, did, Tommy, and you may you

8:13 So, did, Tommy, and you may you may already know this. Did that Reddit and YouTube has a bunch of games that you can develop and build and push out to their platform? Did this? Did this? I did not know this. So, yeah. So, they all have it. Yeah. And every platform now is adding this like video game thing to their platform, getting people to stay on it more. And so, like you’re reading articles, you’re watching a video, and you can play this little mini game that’s inside there.

8:43 little mini game that’s inside there. So, I’ve been playing some of these little mini games that are inside Reddit. Some of them are pretty well done, others pretty much are not. So, so I invented I we were playing around and again to try push me to understand where is my is my how do I how do I get better at doing things with an AI? And I would say I would argue one of the hardest things you can do is build a video game. It’s it’s challenging. There’s logic in it.

9:13 it’s challenging. There’s logic in it. It’s got to do certain things. It’s it’s animation. There’s it’s hard to describe to an AI what it needs to do to make it work. so, that being said, I’ve been challenging myself and my son to build these things. And so with that, you’ll see in the chat window we have our very first game. If you go to Reddit and you go search for bitrop game, it’s a game I’ve made. So, if you want to go play this random game on Reddit, you’re more than welcome to go try it out. It’s a game that I I built with my agents. I

9:45 game that I I built with my agents. I had my agents help me create the idea. I never wrote a line of code. The whole thing was created with an AI. What? What? I I narrated what I wanted feature-wise. I had it build the graphics and the images and everything there. So, I actually had it do a design session with me first. And then from the design session, I actually made the game. What? What? Yeah. The whole thing. Everything. And I’ve actually added logic and I want thing the pieces to move and the sounds like everything in this thing has been entirely AI generated, but I’ve been directing it, giving it like things of how I want. So anyways, I’m

10:16 things of how I want. So anyways, I’m starting to toy with this figuring well if I can learn how to build a video game using AI. That’s probably the most challenging thing. Dashboards should be like cake. Building a report should be like cake compared to like building a video game. So I’m just trying to learn how hard I can push an AI to get it to do something. Anyways, the link is in the description for if you’re bored of this episode., the link is actually in the chat window, not the description. I’ll I’ll see if we can add that later on. Bit drop. It’s bitrop game on Reddit. If you go if you literally search Reddit

10:46 you go if you literally search Reddit Bit Drop game, that’s the game that I made., we’re looking at adding maybe multiplayer,, playing against somebody, getting a getting a session going, something like that. But we’re playing with it right now. It’s just our Yeah, it’s it’s I I really like it. It’s very fun. Yeah, it does have a leaderboard and I do have the highest score on the leaderboard. So, other people are going to have to figure out how to beat me. So, anyways, that being said, I have the highest score on the game that I made, which makes sense. Okay.

11:16 Okay. All right. With that being said, any other news items, Tommy, that you’d like to talk about? I think I think we need to pull Kurt in. I’m sorry. I think he is I I feel bad for him because I think he is literally like a dog right now. No more news. Let’s get out of the news part here. Let’s get our our guest here. So, we’re working further here with your identity is bigger than PowerBI. And with that being said, I’m going to bring in our guest who’s been with us last week and this week, Kurt, welcome to the show again. Happy to have you back. Hello. Hello. I’m sorry, Kurt. We were talking about

11:47 I’m sorry, Kurt. We were talking about heavy topics in the news and you should have been here for that part. So, I think you should have mentioned We’ll cut that off. We’ll let you enter in. Welcome again. Thanks. Thanks. Looking forward to it. Looking forward to it. Looking forward to playing your game. It’s very simple. It’s extremely simple. And I’m I’m trying to make it a bit

12:04 And I’m I’m trying to make it a bit more. One thing I think you will find in games,, and Kurt, you play a lot of games. I’m assuming you do some video games, but also board games and in, you games, but also board games and in,, things with real people in them. know, things with real people in them., , one of the things I found hard with games is I can come up with an idea of a game or copy a game that’s already kind game or copy a game that’s already out there already. The hard part is of out there already. The hard part is making it interesting because you can build some mechanics. Mhm. Mhm. Yeah. Go ahead. So, so I’ve been like studying video game design for the last three years in my spare time outside of work. work. Gez, I now I’m humbled because I’ I’m I

12:37 Gez, I now I’m humbled because I’ I’m I am giants now. No, no, no, no, no. Cuz I have I have done very little, but it’s is you get easily distracted with like once you have kids but but it’s it is hard. It is extremely hard and like there’s a lot of books on game design and yes but I fully agree. So like years ago, like when I first started reading these books, it’s actually they’re very interesting reading material also for like you mentioned dashboards. Like if you read a video game design book and then you in your

13:07 game design book and then you in your mind you like pretend it’s about building a dashboard for people like a lot of the concepts are very similar. Like you’re you’re building things for other people and you’re trying to minimize cognitive load and all of this stuff. Like yes. yes. So do what the most effective dashboard is that I’ve ever used ever? And I always tell people this is the dashboard in the game Planet Zoo. What? Planet Zoo. Okay. Never heard of this game. So So it’s a video game where you have to manage zoo and it’s just straight up

13:37 to manage zoo and it’s just straight up like got like its own little BI platform in it because you need to manage zoo, right? right? Like and it’s it’s a really like complex high cognitive load thing to do. And the video game needs to in order for it to be fun, you have to be able to convey all of this information and data in a way that it’s actionable and that it’s not going to get in the way where your primary task is like looking at the dashboard and not looking at,, the game. So, it needs to

14:07 at,, the game. So, it needs to be unintrusive and it needs to leverage pre-attentive attributes and all this stuff. And I remember like playing the game and I was thinking like, man, like this is the best data visualization I think I’ve seen. But it’s it’s super funny and there’s quite a few games like this and it’s funny because you also see games where it’s the opposite where it’s like they h they have a lot of data and you’re just like man you’re really underutilizing the out of this. So So yes, very much so. So so there’s the the you you mentioned

14:37 So so there’s the the you you mentioned like one of the funniest things I like to do is when you’re playing a video game is I like to dig into the save file and then analyze it and like make a dashboard out of the save file. Amazing. I I love that. That’s incredible. I never would have thought of that. But now that you bring the analogy of like building something for someone else, making it interesting enough that you want to keep engaging with that thing. there’s actually a lot of parallels between what you’re describing and video game design,

15:07 you’re describing and video game design, which I’m I’m finding fun and unpacking myself right now, which is anyways. It’s been a blast. I blast. I I get a game I’m gonna build in a fabric app. It’s gonna So, it’s based off inspiration. I doubt anyone here’s ever heard of Homestar Runner. Oh, I have. Okay. So, one of the funny videos with those old fantasy text command games. I was like,, and they had one in there called Peasants Quest where it’s like, get book, nothing happens., get ye book. You found a book. And

15:38 , get ye book. You found a book. And I’m going to build a fabric adventure or like,, take baby. He was like,, you took baby., I’m gonna I want to create a fabric adventure fantasy game where like,, you’re in a data dungeon thing. It’s like find data flow. And And what do you do? Right. Yeah. Pipeline. Yeah. Yeah. Your pipeline is broken. It’s leaking data everywhere. Your pipelines are broken. What do you do? Find pipel. Yeah. So that’s funny. Yeah, it’s amazing.

16:08 that’s funny. Yeah, it’s amazing. It’s incredible how similar it is, but then there’s certain areas where it diverges heavily. So the the area where video game design and dashboards I love how like we’re already so far off the rails, but it’s fine. Your your identity can be a video game developer. Okay, relax. So all right. So good. Keep going. So the area where like they diverge the most is like the purpose in my mind of a dashboard is you’re successful when the person is spending the least amount of time possible on it. Their job is not to

16:38 time possible on it. Their job is not to look at your dashboard. Okay. Yes. Yes. and there are exploratory an analytics and all like okay whatever but like in general you you people have jobs and those jobs are not looking at dashboards and they want to get back to doing their actual meaningful valuable tasks and not looking at the dashboard. However, in a game you’re optimizing for the opposite. You’re optimizing for the the engagement and stuff like this. But

17:08 the engagement and stuff like this. But the HUD in the game, however, the heads-up display and like the information that’s being conveyed, that has the same purpose. So the your ability to design a dashboard and your ability to design a good heads-up display in a video game and a menu those those are highly correlated because indeed it’s not the point of the game to look at the HUD or look at the menu unless it’s like a meta game, I guess. But so so and and video games also have this concept of juice

17:38 games also have this concept of juice which is like the game feel. So it’s like the feeling that a game gives you and it’s like the various artistic direction that goes into conveying a particular feeling., so it’s like,, when you shoot the gun that it sounds like crunchy enough and that it’s like it gives you the right feeling and and there’s there’s a certain science and art to that that’s that’s unique to that type of medium, let’s say. So,, but yeah, it’s a it’s a wonderful area to like

18:09 it’s a it’s a wonderful area to like read into and dive like really deep into. It’s it’s super fun and I think it’s valuable if you want to get better at making interactive data experiences. Well, I am going to take you there because Mike, how many times have I said on this podcast that no one actually wants a reporter? It’s just the best medium to get the information they need. I quoted that and I appreciate you saying that, Kurt. Kurt, I’ve created a skill off one of your articles actually. So, I’m going to give you some flowers here. You wrote and I I don’t think it was your concept, but you definitely

18:39 was your concept, but you definitely expanded on it was the 3 seconds, 30 seconds, 3 minutes on any report. they what should someone get out of three seconds? What should they get out of it in 30 seconds? What should they get out of three minutes? And to me, that’s one of the more powerful things when you think about a dashboard because really no one really wants a report, they want the information in it. If there’s a better way to convey that information, they would use that. Yep. Yep. Yeah. And so now if you think about like data agents and like conversational BI experiences too and you apply that same

19:11 experiences too and you apply that same logic like at the same time like if you just want the question answered like sometimes a conversational BI experience or a data agent can be like immensely frustrating because you have to play like a dance with natural language to get the answer. Yeah, agreed. so it is it is like really important to be able to evaluate like what tier of question is this and how do we get this answer to what’s what’s is a report the best thing is a data agent the best thing should we do something else data alert

19:41 should we do something else data alert this stuff. So it’s it’s important to think about that but you still feel like that command line tool like get report get ye report like and you you and English on everything you got to really finangle it but no I love love thy data is down thy sales are up the system is down so well well this is interesting very cool I I never thought I never brought the correlation Kurt that you’re bringing here that

20:11 Kurt that you’re bringing here that working on game design things., to be perfectly selfish here and and very frank, the only reason I was building video games was because there’s a number of games that are not mobile optimized for my phone. I was literally like, I love this game and I can’t play it on my stupid phone. I’ll just rebuild it. I’ll spend $50 in tokens and see what I can create with it. But I but I also take it as like an opportunity to learn, right? This is also my learning opportunity. I want to spend a little bit of time like just picking at this stuff and figuring

20:41 just picking at this stuff and figuring out is this even doable? Is this something that could be even real at some point? That’s fun. It’s it’s I think it’s something that everybody should try. Like if video games is something that like drives you and this is something that in general like if you’re thinking about using AI and like using AI is something that you just you’re not that interested in or you’re struggling to like figure out how do I learn this? The best thing you can do is take something that gives you joy and say, “How can I create

21:11 you joy and say, “How can I create something in or adjacent to that space?” I love this idea. That’s what you also said about dashboards. If you were building your own data, you said if you want to get started, use your own personal data, which I did. I used my own Strava cycling data and it was a great way to learn to build your own report. You’re actually passionate about the information you’re trying to learn about, about, right? It’s the thing that you that moves you moves you and you’re the user. So when you’re using your own data, like you quickly realize like, oh man, like this is not actually that useful.

21:41 actually that useful. Yes. Oh my gosh. That’s what that’s exactly true. I just need it in Excel. And can I make this a table? Yeah. Exactly right. Amazing. That is a big point. And I I would encourage anyone listening if you’ve never built your own report for your own personal data, something that matters to you you or someone in your family. Oh, it is a fast track though. Well, even when especially for you like my cycling data, I realized I went through so many iterations because I just

22:11 so many iterations because I just said, I’ll just build,, heart rate and speed. I’m like, well, that doesn’t help looking at it like that. I’m like, I want to know this because of this action that I want to take. to take. Yes. Yes. And unless you actually go through that process yourself, it’s hard to have that empathy. And I think I think you’re touching something here, Tommy. So again, I I run a consulting business. So therefore, I’m tracking people’s hours and what we’re working on. And so I have a very clear dashboard and a lot of things are like I think about,, what is the company doing as a whole? And then I need to think down to what is individual

22:41 need to think down to what is individual employees working on, what are they doing dayto-day? And so I’m constantly jumping up and down between these different levels in my business. But I found that I built an entire report thinking I was going to use a lot of visuals on it. And I only use a handful of them. There’s a there’s a there’s a couple core ones that I know I have to use to know whether or not the business is winning or losing. and so you you can have a lot of ideas of what is in the report, but in reality you have to distill it down to like what’s really usable

23:11 like what’s really usable and there are certain questions I can ask about the report. But when I started going into data discovery, like discovering or or research inside the data, I find reports aren’t really that great for data discovery, for figuring out like what’s going on inside the data. data. No, it depends like exploration. I guess I would say maybe more exploration than than than just report because it takes a lot of time to make a report. and I think we’ll talk about this maybe more on Thursday which which is when we start throwing AI into the mix

23:42 when we start throwing AI into the mix the amount of time it took us to make the report starts changing and doing data discovery or explorations can is I think is going to start changing for us here in the near future. Yeah. Yeah. Okay. Okay. Yeah. Yeah. I do want a hard transition so I’m going to cut it there. Thank you very much. Awesome conversation. We might have to have another episode of Kurt in the future around what games are you developing? Let’s play games together to

24:06 developing? Let’s play games together to see what’s happening there. Another podcast. Yeah, game design podcast. Yeah, exactly. Exactly. AI games., something like that along those lines. So, let’s skip ahead here. Let’s go over to Okay, Tommy, give us this idea here. This idea around just kick us off with this concept. Let’s unpack your identity is bigger than just PowerBI. What does this mean, Tommy? What’s changing here?, I think a lot of people in the market looking for jobs. I get a lot of hit up hit on LinkedIn. Hey, I’m a PowerBI

24:36 hit on LinkedIn. Hey, I’m a PowerBI developer. I build excellent reports. Here’s what I know. Here’s my skills. And a lot of that hangs on just the reporting side. So, Tommy, where do we take this? So, let’s let’s take you back in time to 2015 and really 2015 to 2020. Mike, your life, if you worked in data development, especially in the Microsoft platform, you were a PowerBI pro. That was part of your identity. If you wanted to be an MVP, if you wanted to be a consultant, if you wanted to move up, you had to understand PowerBI. And that was great

25:08 understand PowerBI. And that was great because you needed,, like, hey, I know DAX and Power Query, I know semantic modeling, I know dashboard design, I know governance. And all that was well and great because again, that skill was a very niche skill in so many in so many ways. You had a lot of people who could develop but that unicorn of understanding the business and it however however fabric comes out AI comes out and I think this is now rattled that cage I don’t want to use the word threatened but it’s very much shaken the

25:39 threatened but it’s very much shaken the foundation of someone who could say I just I’m a PowerBI person because fabric alone alone rattled that cage and then you introduce AI to this yes yes no longer can you if you just call yourself a PowerBI person are you capping your career and I think what does it even mean to be a PowerBI person and more importantly if I’m in that situation right now like what do I call myself where part of that identity that career or professional identity where

26:11 career or professional identity where should that lie and how should I in a sense market myself or focus where my skills are so let’s open this up to the floor. And I I just want to start with that question there. Call yourself a PowerBI person. What did that mean to you when you first started and what does that mean to you today? Kurt, take it away. Start there and then I’ll I’ll give my feedback after yours. Well, [sighs] man, it’s it’s a bit of a heavy topic, right? because it’s like it’s like

26:42 right? because it’s like it’s like cultural in a sense like because if you think about what PowerBI was back in the day like PowerBI was like this more like connected community in an essence and and that was like you said also how you had success in your career like that’s how I got started is not with PowerBI but with Tableau I during my research I started using Tableau to make the graphs for my work and then I thought wow this tool is amazing and I met someone at my gym who said you could you could do this for a job and I said, “Wow, okay, that’s

27:13 job and I said, “Wow, okay, that’s cool.” cool.” What? What? So, so So, so and play video games and do BI. Yeah. And so so but then but then indeed so the focus is the tool because ultimately that’s what you’re going to be using. And you’re right that like of course you have to know data visualization and like the more conceptual stuff. but then the tasks that you’re going to be doing and how you market yourself like if you’re a consultant to a client or if you’re applying at a company is they’re typically looking for someone who’s

27:44 typically looking for someone who’s specialized in a tool and can use that tool to help them deliver value. So, it was how you frontended a lot of your expertise and a lot of what you could do was unique because you knew the ins and outs of that tool. You knew what the limitations were and how to overcome them. You knew the tricks to be able to get things to perform well or look good. And those of us who use PowerBI understand that very well like in the context of reports like all these little tricks with like the the

28:14 these little tricks with like the the switch in DAX to be able to choose a measure and all these things like we’re all familiar there’s there’s hundreds of them. but as as things indeed have changed the scope of our work has broadened because of the ecosystem we’re in technically in terms of like like you said fabric so PowerBI is part of this bigger thing and perhaps unfortunately or unfortunately you could say like depending on how you want to look at it

28:44 depending on how you want to look at it there are expectations from employers and clients that because PowerBI is part of this bigger thing and integrates in a very deeply connected way with this bigger thing that you should know this bigger thing and that they they they expect you to even if you’re just focused on like the visualization side to do things that might feel outside of your wheelhouse. now with AI it’s different as well because with AI we talked about this last time like you can do more as a

29:14 this last time like you can do more as a result of AI. both in quantity as in is as is also in terms of quality like you can also do better things but you can also tackle a broader area. Yeah. Yeah. I think that’s perhaps one area to start. Mike. Yeah. I’m going to I’m going to lean on what comes to mind when I talk about this is this is when I was working with my let me let me give a parallel analogy. I have developers on my team. We build software and apps. when I really started pushing

29:46 and apps. when I really started pushing my team to like we need to learn how to code with AI. We have to stop writing functions in our languages in the systems that we build and AI was getting good at this but the harnesses weren’t quite there. It’s still getting more proficient in how to write codes and applications. I had a lot of push back from some of the lead engineers of like I don’t know this feels like what I was going to be doing and we have an an intern this summer and I had a really candid conversation which I thought was great. He’s like my whole vision of when

30:18 great. He’s like my whole vision of when I graduate college was to be a computer science degree person. I was going to sit down and write code for a company some system and then I said I don’t think we’re going to be writing code by the time you graduate. I think most of that will be covered by the AI. So I need you to move away from just being like an entry-level engineer or an entry-level software developer into more of a middle manager engineering person. You need to understand the systems. You understand

30:48 understand the systems. You understand how the the database works, how to talk to identity, like conceptually you have to grab a lot more at the conceptual level. I’m not necessarily requiring as much knowledge around I need to learn JavaScript or I need to learn TypeScript. Those are becoming a little less. And so because of this, right, AI enters the place. Your skills aren’t just I’m building reports. So I’m going to bring the parallel back to to the PowerBI side thing out. Right?

31:18 thing out. Right? For me, AI has just allowed us to free ourselves from some of the click tax, from some of the heavy just working in the reports. And for me, AI is shaking a lot of people’s core beliefs of, well, I was the BI engineer. I made reports. That’s what we did. Power Query, Dex, that’s my thing. And now we’re starting to see AI chip away at some of that knowledge that you need to have in order to be successful. And so this is where I I want to really kind

31:48 so this is where I I want to really kind so this is where I I want to really emphasize because AI is showing up of emphasize because AI is showing up now. Your identity is not really that your skill wasn’t really that. Your skill is how do you listen to the business? What are they asking you to build? How do you how you have empathy? We talked about this one of our Kurt you brought this up which I thought was a great point. How do we have empathy for the people trying to consume the data we’re trying to produce them? Tommy, to your point around the bike and the Strava data, right? I’m building things, I really need to be critical about what we’re building. The the bu the act of

32:19 we’re building. The the bu the act of building is not as much of a importance, but what we’re building and how people interact with it is more important than ever before. And so I think the the blast radius of AI is allowing us to then shoot more area of influence against PowerBI and now also into fabric. I never used to work in SQL wasn’t I didn’t really I wasn’t interested in it. And then SQL fabric shows up and I’m loving SQL databases. I’m like building data DBOs everywhere

32:50 I’m like building data DBOs everywhere like let’s make a database for this, let’s make a database for that. I’m trying to use it more. this is super useful, but I wasn’t on my radar until stuff started showing up at the fabric. So, I’m I’m able to open my skill set and then I just ask AI where my knowledge lacks and I think it through with the AI together. K, you were going to say something or Tommy, sorry. I I want to make an argument here right off the bat here or not necessarily argue with you, but I think there’s a fundamental thing we need to remember. Even in the first six years of PowerBI or first seven years prefabric, most

33:22 or first seven years prefabric, most people working in PowerBI were already stretched thin, right? If you were a PowerBI person, right? you there’s no way a single person can manage entire business data but they were asked to because oh it’s a you it’s just PowerBI desktop but you’re managing models and governance and DAX measures and visuals and most people if you were to have on the Yaxis your level of effort and there was a 100% cap of effort allowed they’re already at that 99% I cannot tell you myself

33:53 at that 99% I cannot tell you myself other people from the user group that and just from networking who have felt that because you wore so many hats. So that was already and you kind hats. So that was already and you were proud of that a little because of were proud of that a little because job security but also too that you were touching so many parts but then you go to pre AI Mike just go to fabric that’s where the foundation was shook I imagine that graph goes past 100% because you’re like you need a fabric person now right I don’t think the

34:23 person now right I don’t think the concept in the beginning was at all about oh we’re going to now split these roles we’re going to have a data engineer that’s going to work in fab fabric. I think this is all happening now, but not when it came out. It was like, well, you’re already in PowerBI and this is just part of PowerBI. So, already I think a lot of organizations were really finding that stress and mostly individuals. However, there is a other side of this. we finally get to 2026 and I think this is the first time

34:53 2026 and I think this is the first time that a individual may see that effort go under 100% again because of the introduction of agent tooling. Yeah. Yeah. And I I have a few more things there but I’ll pause there because let’s not forget the stress we already had in your identity of being a PowerBI person. Yeah. And I think like we are beating around the bush a little bit because when I think about this topic it is it is reminding me of the fact that when we focus on AI

35:23 fact that when we focus on AI like the capabilities of AI are getting quite sophisticated and all of us have developed these technical skills that we’ve been proud of and have been core to our identities. Yes, there is there is an anxiety and a concern that the AI will take our jobs or that it will displace us and make us less valuable at our work. And I think that’s that’s like the the core thing and I think what what we probably all agree on I think is that your core competences are not

35:55 your core competences are not necessarily rooted in these tools. like the expertise that you’ve built in to your point, Mike, is to do with understanding and dealing with the business processes and problems and also,, Tommy, and being able to put on all these different hats to be

36:10 put on all these different hats to be able to manage all these different stuff. You’re not just interfacing with PowerBI, you’re probably also interfacing with various business domains and technical domains,, things like SAPO or whatever it is, in order to get that data in. And so I think if I was to give people a very direct message when they talk about being worried about AI, I would say that if you are focused on solving business problems problems and if you are willing to engage with the business and also get into technical

36:42 the business and also get into technical details, I don’t think you really have anything to be worried about so long as there is you’re not willing to you’re willing to be able to leverage these agentic tools that are coming up in a responsible way., however, I would also say that if you feel like you are mainly memorizing paths to solve that problem, like you’re more interested in being able to

37:12 more interested in being able to memorize the best practices or do the UI tricks or stuff like this, like if that’s where you’re focusing, if if you’re honest with yourself, then I do think it’s important to pivot away from that attention and direct it towards the business and direct it toward the problems more and abstract away and we shouldn’t feel in my opinion like a threat like it’s a threat like but instead we should see it as liberation to the to the session

37:43 to the to the session you are you are more than your identity at powerb you don’t have to be the powerbi person anymore like you can do so much more and you don’t have to be shackled to a specific tool tool or way of working once you have these agent tools to support you. presuming you’re working in the appropriate governance guidelines and all these things. things. I want to I want to double down on your point there just really brief Kirk something that comes to mind on the liberation of using AI.

38:14 to mind on the liberation of using AI. We’ve been heavily focused on building reports for the business to add value over time. And [snorts] I am greatly shifting and I I even see this I think even inside the PowerBI community. I think a lot of people are not disillusioned by PowerBI but we’ve we’ve also seen this other shiny thing called AI and now we have things like fabric apps. We’re still able to apply a lot of the same core business solving problems and principles but now instead of having a

38:45 principles but now instead of having a one-way dashboard or report that just shells out data outward, we can now have data coming in. We can have an interaction with data. We can we can submit data and have real time things happening against that data directly to fabric. So a SQL server can then ingest data. Cool. Like so this to me I look at this going this just opens up a whole another realm. And if you ask me had Microsoft produced fabric apps before AI I think they would have been a very moot point. I think it would have been a bump on the radar and no one

39:15 been a bump on the radar and no one would have built them because there would have been a select few people that are like, “Oh, I can look look I can use D3. js to build with it. But now that I have AI, have AI, great, I can just rip out really cool ideas.” Again, still focusing on business problems and solving things., I want to very quickly, I apologize, I’m going to say a last thing and I’ll let you guys talk. So, I’m so excited about this topic that you brought up., Jensen Hang, the CEO of Nvidia, I’ve said this a couple times on the

39:45 I’ve said this a couple times on the podcast, and Tommy, we’ve talked about this. Yeah. He had an interview and it made an impact on me. He goes, “If I have an employee who I’m paying $200, 000 for the employee, and they’re not using $500, 000 of tokens a year, I’m going to ask, what are you doing wrong? How are you not why are you not incorporating AI into your process, your work, your development? what something’s not right here where you’re not leveraging AI to its max fullest in your job. Building new things, creating new stuff. And so I really liked his mentality of look, our

40:17 really liked his mentality of look, our engineers are solving problems right now based on the technology and what we know how to do. AI allows us to open that door and solve bigger problems. And we don’t even know the problems we need to solve tomorrow, next week, next month because AI is now we we haven’t had the mental time to think about and solve those problems because we’re so focused on building reports and dashboards, right? So now that we can let that be handled or owned by the AI, if we’re able to

40:47 or owned by the AI, if we’re able to minimize that work in that space, we’re able to ask ourselves bigger questions, solve larger problems, tackle bigger issues. This is why Tommy and I and Kurt, I’m sure you are too. This is why we’re building our own software. If that software doesn’t have the right color button, we’re like, “Ah, I got to build that in a weekend. Let’s go rebuild it.” Like, we we are taking on immensely larger problems to solve because of any little friction we see in applications and development. And so, this is just opening the door for I’m going to call

41:17 opening the door for I’m going to call this I think it’s custom harness. I think businesses and your business process is going to start falling into these custom harnesses that organizations build, maintain, and run. No, No, that is their IP that that is what makes them good. Let’s dive into this Mike because I think part of the skill conversation here is is AI then making PowerBI the skills in PowerBI less valuable or less important but your ability to jud like do judgment and

41:47 ability to jud like do judgment and evaluate and basically the department of common sense around data more important common sense I like I I I I use that all the time every company needs one that’s a shirt right there the department of common sense seriously I think that’s mostly my job in a lot of ways too. But that that ability though and again if you to your point point I am just relying on AI and we’re going to go with Nvidia in terms of how they are focusing their developers and someone’s going to hire me or hire

42:17 someone’s going to hire me or hire someone with PowerBI. Well, my skills are less important because I can walk into a business meeting and provide three options we can do rather than one because of time and resources, right? So, but I don’t need I don’t want to say I don’t need the skill, but by virtue of what I’m hearing, those skills are less important to know. I’m not going to say less important. I think it just shifts what we’re able to do. do. Yeah,

42:47 Yeah, sure. It’s it’s like for example, so if you sit someone down with an agent and you tell them,, make a PowerBI report and you have someone who knows a lot about design and someone who doesn’t,, you’re going to probably be able to do quite a bit more. But if that person knows a lot about design and they know some of the ins and outs of PowerBI, like then they’re really going to be able to at this point like do pretty much whatever they want. like they’ll be able to say, “Oh, use the error bars to show the delta between like the actual and the target.” And

43:18 like the actual and the target.” And whereas,, someone who just knows about Yeah, that’s a very cursed wisdom. but like the the point is that like it’s not that it’s less value, but the value shifts to different places and ways and the weight of that value is different. like like I said like if if your focus is that like that that you really want to know all these tips and tricks and you focused on memorizing these things then you’re going to find that that is less

43:49 you’re going to find that that is less valuable like and if you’ve really focused on only technical competences and you’re trying to you’re not you’re not trying to understand the business process like I think then that can be more problematic in this area because there are ways to be able to compensate for a lack of technical understanding in a way, but it’s not that that’s less valuable. It’s just if anything that helps you be able to do more. There’s ways that you can advance and accelerate and leverage that knowledge now that you simply couldn’t

44:20 knowledge now that you simply couldn’t before. And I think that that’s that’s what’s critical. But what Mike says like,, that you you can do more. I think that’s that’s definitely true. And not just more, but like a lot more. Like you can really be ambitious but in a reasonable way. Like you don’t want to let your token cocaine, you want to let your token cocaine,, knock the business off the rails, know, knock the business off the rails, right? Or like the slop grenade as Shopify mentioned last week. Oh, I like that one, too. Yeah.

44:50 Yeah. Yeah. You don’t want to be throwing slop grenades at your colleagues because like you had the crazy idea and you don’t have the the skill to be able to evaluate whether that has any merit. Okay., but I think if you if you restrain it and you say like, hey, I have this idea. I made a prototype and this is representative of what I’m thinking. I’m not saying we should do this and I’m not saying we should use this. Like this is just,, take it and leave it. This is what I’m thinking. And then you evaluate it like that. I think then it’s

45:21 evaluate it like that. I think then it’s it’s makes a lot of sense. Sorry. Go ahead. ahead. Yes. No, no. I was I was going to say I I really like this point and I’m hearing more conversations of other directors, leaders in organizations that you I’m giving AI to my employees and they’re just submitting things to it and just letting whatever comes out of it, they just email that on forward, right? You didn’t even read it, right? Got to be critical. You got to be critical of these things. And so you’re you’re more than you’re more than welcome to. So actually I was this is a this is an ongoing debate at

45:51 this is a this is an ongoing debate at my family in my family home because our kids are high school and getting ready to go to college age right so where does education and learning of these things sit in lie of AI oh oh like never before can you can you site AI do you still have to site books and and websites you can ask an AI something so for example what if I write a paper and then submit my paper to an AI and say, “What topics or ideas or

46:21 AI and say, “What topics or ideas or concepts am I missing in this paper? What would be a relevant conversation?” Having the AI evaluate what you do and then give it back to you as feedback, right? Or, hey AI, I’m trying to write a paper about this at school. Help me write an outline of something I need to write about in the paper. This is really new and I was, to be honest, horrible at writing. I hated writing papers because they were so boring to me. It it wasn’t what I was interested in. If you gave me papers that I could write about like science and learning and ed if it was

46:52 science and learning and ed if it was something I was interested in, I could write all day on it. No problem. Legos. Yeah. Legos. Yeah. 100%. Get me on that bandwagon. I’d be there all day long. I should have never said that. I’m going to divert them. We are in we are we are off on the rails on that one. So, but yeah. Yeah. But that was that was a challenge for me. That was an area that I wasn’t very strong in and I really wasn’t interested in exercising that muscle. What I really liked was computers and other things. site. This is a really interesting balancing act. Kurt, you’re going to say something there. So, very quickly, like I do think it’s

47:22 So, very quickly, like I do think it’s important and one of the things related to like we talked about AI literacy last week. I think like one of the things you need to understand that a lot of people need to understand is that AI is not coming up with the idea and like that AI is ultimately GPT LLM they are taking the data that they were given in their training which did not belong to a lot of those companies and it is important to cite that primacy you it is important to cite that primacy the information and to be able to know the information and to be able to address that and that is part of responsible and ethical use of AI I think and so when you are using AI to be

47:55 think and so when you are using AI to be able to come up with ideas or to evaluate something and you get an idea, the AI suggests something, I think it the onus is on you to be able to say, okay, where does that come from? Where did you get the idea? And even if that forces you to go through traditional tools to try to root that out and to find that out and it is possible that

48:14 find that out and it is possible that through the combination of different things that have come together that you’re not able to find an original source or something like this. And but but you you you need to do your due diligence and your best effort to be able to do that because we need to respect the past intellect that humans have built up over time. And we need to make sure that we’re not breaking the lineage of,, passing information on through generations and being able to to to to lose the thread because that is one of the big risks of

48:44 because that is one of the big risks of AI is suddenly like it just it just kind AI is suddenly like it just it just like blithers away and we’re not able of like blithers away and we’re not able to site like where the primacy of where did that information come from., so so yeah, just just quickly on that point, I I love that. I I do need I couldn’t let this go and Mike knows usually I can’t let things go but both you mentioned something to me that may be a little high level philosophical but I I I just can’t get out of my head. I asked does skill less important and I think

49:15 does skill less important and I think this this will transition us a little more in terms of like how do you identify yourself with PowerBI but you both mentioned it’s not a lessening of skill it’s a transferring of skill. Yeah. And I I need to push back on this because you both mentioned this and if you if I had a pot of water and I transferred some of that water to a different cup, that pot has less water in it. in it. And like right and and I think this is important though because just because less skill or less important doesn’t

49:45 less skill or less important doesn’t mean not important at all. I think there’s still that trench wisdom that’s still essential to this to like asking the right questions to have agents work. However, if you’re saying you’re transferring that skill somewhere, I think by virtue of that statement, that’s also saying that your total expertise, the vertical expertise and tech and tech technical skill that you had in a given area in PowerBI is just doesn’t have the same depth anymore. It

50:15 doesn’t have the same depth anymore. It doesn’t have that same length of needed knowledge. I don’t I may not have to know now all the intricacies of bookmarking like like the different ways and the applications of it if an agent right so and I think this is an important thing where if I’m saying I’m a PowerBI person and I’m using agentic skills I cannot assume that all the different things evaluation context and you’ve done every type of situation because you have an

50:46 type of situation because you have an agent to do Yeah. Yeah. Would you agree or disagree with that?, [sighs and gasps] yeah. I think like there there are definitely certain things like like the the example I give of auto daytime. I cannot tell you how many times I’ve been in an organization and there has been either a consultant I knew or from another organization or something like this and there was like a model that was too big and they come in and they turn off auto daytime and everybody’s like whoa like the size went down so much and like people have given

51:17 down so much and like people have given talks about it and now it’s like it’s it’s literally like an immensely trivial thing that any any agent tool with the semantic model will handle. and but that’s something that literally would pay the the consultancy bill for that week and and that transfer of knowledge of like well we didn’t know that the consultant taught us this and now that is indeed less valuable and that does have an impact on that

51:47 have an impact on that ecosystem and for better or worse like but it’s I think ultimately the the the goal can then be like that we can spend less time and effort trying to focus on these these specific little details that pertain to the tools or how to do something and more about the why are we doing it and more focus on like the actual insights like we used to talk like I don’t know this has been like a thing for a really long time already of course like the whole descriptive

52:17 course like the whole descriptive analytics to predictive analytics and whatever to prescriptive analytics like maybe it’s time to start bringing that little mental model back and like if we can maybe spend less less of our time on the descriptive aspect. Then we can start thinking about okay how do we harness these AI tools to focus on the core things like our data quality and our governance and making sure that we have our in Dutch like house of like our house is in order and we have a foundation to build on and then because

52:48 foundation to build on and then because you can go quicker through like the descriptive aspect then you can start leveraging those to do more advanced analytics and stuff like this like how can we start addressing the business problems in in more sophisticated ways that might have seemed unrealistic for our team,, a year or two years ago. either because now we have more time and capacity or because now we have these tools that can augment our capabilities. just as one like off-the cuff example, I would guess

53:21 like off-the cuff example, I would guess I really like this augmentation of skills you already have. And I think what this is also doing is it it’s for me personally AI is accelerating my ability to more critically think about what I want and the process and step up a level from I’m clicking buttons to make a thing work to what do I want the thing to do. So So and yeah go ahead. So so quickly on that point like augmentation like one of the core tenants I think of like your AI culture in your organization should be augment

53:51 in your organization should be augment not replace. so circling back to what we said before like this is a challenge that we’re that everyone’s having like we we the three of us are having it for sure. So to be very clear with anyone who would be listening like without any doubt like so I think it is important for people in organizations with positions of authority to be able to establish that like that is our value to augment and not replace. So sorry I didn’t mean to interrupt you. No, no, no. I I think this is exactly where I wanted to take some of this off,

54:21 wanted to take some of this off, which is is that it’s that concept. It’s it’s that,, we’re I’m able and again for me, I’m very encouraged by this. I’m very optimistic for where AI is going. [snorts] I really like the ability for me to do more, handle more, create more. I’m going to I’m going to touch some toes here. Okay, so I’m going to be be clear about this one. this one. Just touch. I have really been stretching my legs in other media other than just power. Using full physical analogies here, aren’t we? Full physical analogies now.

54:51 Full physical analogies now. Physically. Yeah. Why? Why reach out my hand? Yeah. So, I’m I’m going to step on toes here because Kurt and I have had some words around where does AI sit in like creativity as far as music go and and where does it sit in like image generation and artwork and where does creativity and art sit now in lie of AI and and so I’m to be honest I love music. I absolutely love music. I’m a drummer. I like have two drum sets sitting in the back room over here that I can that I

55:22 back room over here that I can that I haven’t that are just collecting dust now. But I just love music and I just enjoy interacting and thinking about it. AI has made for me a love of music kind AI has made for me a love of music reinvigorate inside me around being of reinvigorate inside me around being able to like generate things that I find are interesting. And I to to some degree I love the sound of it. The lyrics are innocuous to me, right? I when I listen to music and songs I don’t really care about the lyrics as much. To me it’s very much about the feel and the music and all. And so I’ve been building

55:52 music and all. And so I’ve been building a whole bunch of AI generated I’ll I’ll be frank in a minute AI generated songs but I’m combing over the lyrics and I’m refining portions of the song. I’m really acting like the producer to what the AI is doing. Creating expression. Yeah. And but I’m enjoying it. So like this is one area that I think it will be very interesting. I know Kurt you have a different opinion on this one. So I’d love to hear your opinion on this one as well. well. if I’m using AI to express part of my creativity and I find joy

56:22 part of my creativity and I find joy from that part that that’s a preference and this is where I think things get really interesting for me like so my IP for this is in my head it’s how I express myself with this one and then give some of the AI direction of where I want it to go but at the end of the day I would say this whether it’s your business whether it’s AI whether it’s music whatever This is the core tenant for me is you’re valuable in the IP and

56:52 for me is you’re valuable in the IP and what and understand. Yeah. Yeah. Regardless of where this thing sits. So let me let me just pause a thought there because I could go with different ways. I don’t want to take it too far here. I’m going to kick it back to you Kurt. Do do you mind reacting a little bit here? I want to grain and hear some thoughts here. So, it’s such a messy area and I could talk about this for an hour on its own, but it’s like I so I personally know a lot of people who are employed as professional artists who across the spectrum like tattoo artists or or all kinds of things

57:22 things and musicians and and so on and they are by and large negatively impacted by AI and u they were they were impacted before a lot of other people because AI AI was, let’s be frank, like just not good like in 2024, but it was able to generate images and and stuff that had like a softer definition of,, what correct or good could be look like in certain commercial contexts.

57:52 in certain commercial contexts. So that has predisposed me and like just generally the the IP unfairness around the creation of this technology feels extremely uncomfortable. Like if I’m frank, when I use AI, I I do in general tend to feel guilty like when I’m using AI. Like I like there’s a part of me that does feel like I’m I’m doing something wrong. And part of it goes back to like that feeling of like you have the Eldrich wand and you it’s so powerful that you

58:22 wand and you it’s so powerful that you you you you feel like you shouldn’t have it. But part of it also goes to like the fact that if you read up on like you fact that if you read up on like how the information was gathered in know how the information was gathered in order to make these things purportedly then it it’s uncomfortable. So however like as these capabilities evolve like there have been a lot of circumstances that have forced me to like re-evaluate what were very firm and strong opinions I had. And like for instance, you can use Opus 5 today and give it JavaScript and you can

58:54 today and give it JavaScript and you can give it some some quite like esoteric instructions and get it to draw like crazy things like really crazy out of the box things and sure sure but but it’s just like it’s not like ripping off artwork or something like this to to that extent like it’s using code and you can even like make videos or when you get it to control Blender and stuff like this. And I I I think part of it goes down to intent. Like are are you the individual like

59:24 Like are are you the individual like respecting the the privacy and the intellectual property and the ideas of other people? if so then I think like if you’re doing something that’s an act of expression and you’re ex ex you’re putting effort into it know you’re putting effort into it then you are able to to use AI to be able to create these things and it could be due to certain reasons that you’re incapable of that could be accessibility related like I gave the example in one of our previous

59:55 example in one of our previous discussions about like people I know who have dyslexia or challenges with language who use it to help them, right? right? That would be me. That’s why I hate writing and reading because it’s like a challenge for me. Like I I have I flip things around all the time. It’s Yeah, I get it. This it’s helpful for me in that regard. I’m better at talking and communicating than I am sitting and

60:15 and communicating than I am sitting and reading and writing all this stuff. So, and I think what what it boils down to is I’ve tried I try really hard to re-evaluate my opinions like when new information comes up. And like I think it’s important with AI to take a nuance take because to what you guys said like in the beginning there are it crosses the entire spectrum like there are literally like speaking now as a biomedical scientist. Okay. So there are there are wonderful amazing things that on the horizon I believe like in the world of medicine and genomics and genetics like I do

60:46 and genomics and genetics like I do firmly believe that but there are also horrible terrifying awful things that are coming and everything in between and not just in biomedical sciences but across most domains. So if we are to be able to tackle these things we need to be able to address the nuance. Sorry I rambled. No no that is great. I I do want to make a point distinction here because all this is well and good but Mike from the music side there is let’s make a clear distinction between a hobby and

61:17 distinction between a hobby and your livelihood commercial yeah commercial music production so you’re never you’re not going to job postings for music right and you don’t need to that’s not what you I’m not I’m not other people are are right so but that’s the distinction where I can have that creative expression where I don’t have to have that in a sense background that trench wisdom that years of experience because I’m also not relying on that for my livelihood for a lot of people today in the PowerBI world they’re

61:48 the PowerBI world they’re looking at those job postings I’m curious how they’re going to change I am very curious on what they’re going to look like in the upcoming months or in the upcoming years and but more importantly the difference with the three of us being independent and honestly as lucky as we’ve been able the be where I can test things whenever I want. Mike, how many mailbacks have we gotten from people in PowerBI who haven’t even allowed fabric yet, right? And they’re getting in a sense penalized or there’s

62:19 getting in a sense penalized or there’s a penalization happen to them because they’re not getting experience in fabric. They’re not going to pay for it themselves. Their company is not letting use fabric. And then you add on the agentic tooling with this and skills. They can do that,, creative expression on their own, but they’re not going to do in work hours like you,, like you and I get to. And more importantly, we can test these things out. So for a someone who has an identity in PowerBI right now this leads to the question the discussion I guess what’s the honest

62:50 discussion I guess what’s the honest answer in terms of should someone go deep in PowerBI to get that trench wisdom or should they go more horizontal in just anything data using agentic tooling and more importantly again what job title or what’s the job description really going to look like for what we consider that PowerBI identity it’s twofold there. Yeah, I I think I I do want to answer your question, Tommy, but I also want to go back. I think I want to just quickly touch on quick Curt’s point here really last on this idea here of like IP area.

63:23 last on this idea here of like IP area. Again, I Kurt, I I really you’re in the world of world of medical and DNA and testing things. There’s some really important things that need to be talked about in that space. You were telling me you were telling me about how AI can greatly change you can go use AI you can throw a bunch you can go to somewhere sequence your DNA you can throw DNA at an AI and say solve these problems or how what drugs would

63:53 these problems or how what drugs would be most impactful to this DNA and it’s figuring things out it’s doing things and you can get drugs or things made that could be very helpful to society but also very detrimental too with very little oversight in those areas so I just want to call out like while we are be giving a very powerful tool Cool. And I don’t really know where I was going with that point, but I’ll go back to the music one because that’s maybe the one I was pointing at. Even though I’m generating some very funny, you I’m generating some very funny,, AI generated music, no one’s going know, AI generated music, no one’s going to come watch Mike Carlo perform at a at

64:23 to come watch Mike Carlo perform at a at an amphitheater with my AI music. It’s just not possible. No, no one’s going to ever want to do that. So, on some degree, I’ve capped like I’m in a very niche area that’s just fun. I’m doing something on the very side of here. You something on the very side of here., I’m going to go watch YouTube. I’m know, I’m going to go watch YouTube. I’m gonna go watch big artists play in massive stadiums. So, I really think there’s this level of like I’m I’m doing things with AI in in certain ways, but I’m not actually like yes, I might be stepping on some of the toes of like the streaming services. And again, I get zero views or listens to my music,

64:54 I get zero views or listens to my music, FYI. FYI. It’s not like I’m making money off this thing. It’s just pure for fun and enjoyment of my own self, right? So, but there I think what this is going to do, it’s going to shift where people spend time and money into much other much different areas for where artists are going to be required, right? Artists are still going to be able to pack stadiums and venues and like that part of the you can’t replicate that. AI can’t replicate going to a music conference venue and enjoying that show, right? I think there

65:24 enjoying that show, right? I think there are some very human things. I think what this is potentially doing for AI is it’s taking a lot of the me some of the mundane technology pieces and it’s actually h it’s allowing us to focus more on the human the humanizing aspects of a lot more of that and I I want to I want to be again I’m on the optimist side a little bit. I’m on the side of like I think this is going to be good but every time we come up with an innovation we got to figure out how to make it safer. get the innovation out in our hands and then have really smart people shape this

65:55 then have really smart people shape this so that it is productive and beneficial and it’s not hurting people long term. And I think that’s I feel like that’s something we figured out as we went through other technology revolutions, the industrial revolution, when cars came out. A lot of people died when cars came out cuz they were so unsafe. We have a much better pattern track record for keeping cars and many more people have them now. So like it just took us time to like bring the technology. It was revolutionary. No one rides horses between cities anymore. We all ride in cars and other vehicles. So

66:26 all ride in cars and other vehicles. So really good in innovation revolution. We’ve had to then greatly that thing has mutated and changed over and over and over again over the years we’ve had it to make it safer and better and now everyone drives in these things. So, I’m going to air on that side of optimism around what AI is going to do and we just got to keep stepping forward, but then critically evaluating what are we doing with it and making sure that we’re using it appropriately. Let me just pause there. Thoughts? So, I’m going to try to tie what you said with what Tommy said. So, I I think

66:58 said with what Tommy said. So, I I think I think it’s important to be optimistic. I do also think it’s important to be to be realistic even if it’s very uncomfortable to face some of the challenges but what I would say and and more direct this to to the people that listening is you have you do have tremendous skills and capabilities like the skills that you’ve developed like working in PowerBI working in data you’ve learned skills that are inherently valuable and deeply valuable and scale beyond PowerBI and not just

67:28 and scale beyond PowerBI and not just that but that have the opportunity to have a positive impact. And if you can leverage things with AI, with agents, you can apply those skills and scale those skills in ways that you probably didn’t realize were possible. So I think it’s important that if you haven’t yet tried to use AI and agents that you do try and you try to see what’s possible and once you do feel that stay be careful with the token cocaine but think about think

67:59 token cocaine but think about think about about keep saying that yeah seriously it’s dangerous but think about how you can take these skills that you have and the augmentations of AI and apply it to these problems that are emerging. So look around us at at every single domain is having various challenges with AI and undoubtedly you have opinions and feelings about what could or should be done in order to secure a better future or in order to get these good outcomes.

68:31 or in order to get these good outcomes. If we want an optimistic future, then we need optimistic people to act. And if you want to think about how your identity is changing right now, then I would really highly strongly encourage people to to do that to to act like this is an opportunity right now. There are opportunities that are are just of your wildest dreams. And there are things that you can do and impacts that you can have that you probably didn’t realize were possible, but that definitely 100% are.

69:04 definitely 100% are. Amazing. Amazing. That’s incredible. I know that. Did you already make the t-shirt? Okay, so for those of you in the chat, oh my goodness., so now officially on the PowerBI tips store, we now have a Department of Common Sense hoodie in case you want to go grab it. So,, we’re going to try and push out real products here, real real time here. Thanks, Kurt, for saying that, by the way. So, awesome words of wisdom to go by.

69:34 by. So, we’re gonna we’re going to continue to pump out funny things here. We we still have some phrases from the last episode. I haven’t gotten into a shirt yet. So, anyways, I thought that was hilarious. Dana Donna, you were asking about I would buy that shirt. Well, there you go. You You’ve got at least a hoodie to go get out there. I’ll see if I can go get another DCS Department of Common Sense. we’ll get another DCS out here started and maybe do one for a t-shirt as well. All right. This has been a great conversation. I absolutely love this. you your identity is way more than just building

70:05 identity is way more than just building reports. And Kurt, I think I want to just dovetail on your final point here, which is which is where we are right now is in an exciting time. Things are greatly shifting. The world is shaking underneath your feet potentially and you might be feeling very scared about all these things. Yeah. Yeah. Embrace it. Yeah. Yeah. Step into the AI space. Go figure out what that new skill is. I’ve seen many other MVPs in the PowerBI space just really jump into AI and just figure out how to work with it, build stuff with

70:35 how to work with it, build stuff with it, create things with it, learn from it. There’s a lot there’s a many, right? right? Just Yes. Yep. True. Yeah. Don’t go do bad things, but like I’m I’m ar that’s an when I look at my skills and I augment things with the AI, I I don’t even know what it can do. And this is why I’m doing the video game stuff. I want to push the boundaries of what I think is possible. what I can I is this even a real thing? Is it just more than just a chatbot? And I think the answer is yes. And it’s going to continue to be this really big push into

71:06 continue to be this really big push into this new space. Anyways, I want to encourage you, spend time with it. Go figure out how it works for you. Go figure out how to use your already known knowledge and IP and figure out how to build cool stuff with AI. It’s going to help your career regardless. I I think it will be a good win for you to figure that out. I think so. Awesome. That being said, Tommy Kurt, thank you so much for the episode. Tommy, where else can we find the podcast? podcast? You can find Apple, Spotify, or wherever you get your podcast. Make sure to

71:36 you get your podcast. Make sure to subscribe and leave a rating. Helps us out a ton. Do you have a question or idea or topic that you want us to talk about a future episode? Head over to PowerBI. tipsodcast. Leave your name in a great question. And finally, join us live every Tuesday and Thursday, a. m. Central on all of PowerBI tips social media channels. Thank you all so much for listening. Kurt, thank you for attending. Tommy, nice job as always. We’ll see you next time. Tommy and Mikey. Dance to the day to

72:08 Tommy and Mikey. Dance to the day to laugh in the mix. Fabric and I get your feels. Explicit measures. Drop the beat. Now kings feel the crowd. Explicit measures.

Thank You

Want to catch us live? Join every Tuesday and Thursday at 7:30 AM Central on YouTube and LinkedIn.

Got a question? Head to powerbi.tips/empodcast and submit your topic ideas.

Listen on Spotify, Apple Podcasts, or wherever you get your podcasts.