Using AI and Data Visualizations – Ep. 201

AI is already changing the day-to-day BI workflow: it can draft copy, propose DAX, and give you a starting point when you’re staring at a blank page. The question is whether it can do the hard parts of our job: choosing the right metrics, building the right narrative, and designing visuals that drive decisions.

In Ep. 201, Mike, Tommy, and Seth react to the Medium article What AI Knows About Data Visualization and Storytelling. They talk through where AI can meaningfully accelerate report building (layout, formatting, repetitive interactions), where it struggles (audience context and business nuance), and why the real edge will go to the people who learn how to prompt, validate, and refine AI output.

News & Announcements

• What AI Knows About Data Visualization and Storytelling — The article that sparked the episode, exploring what ‘optimal’ even means in dataviz when an AI is doing the suggesting.
• Submit an idea / topic — Got a scenario you want dissected? Send it in and it may become a future episode.
• Explicit Measures Podcast — Subscribe and browse the full episode backlog.
• Mike Carlo on LinkedIn — Field lessons on modeling, DAX, and delivery.
• Seth Bauer on LinkedIn — BI engineering and measurement strategy perspectives.
• Tommy Puglia on LinkedIn — Platform findings and practical Power BI notes.
• Argus PBI — Friend of PowerBI.tips.

Main Discussion

This episode lands on a practical stance: AI is an accelerator, not a replacer. It can generate drafts—of words, code, even layout ideas—but someone still has to decide what’s true, what’s important, and what’s appropriate for the audience.

The crew also calls out the real gating factor for AI in dataviz: context. A chart isn’t ‘good’ in a vacuum; it’s good for a specific decision-maker trying to answer a specific question. If the AI doesn’t know the org, the hierarchy, the edge cases in the data, and the outcomes you’re optimizing for, it will confidently suggest things that look reasonable and are still wrong.

Where AI does shine is in speeding up repeatable mechanics: turning rough requirements into a first draft, helping you brainstorm alternatives, and automating the annoying parts of production—then letting you refine from there.

Key takeaways:

• AI won’t replace you as fast as someone using AI can—learning the workflow matters.
• Treat AI output as a first draft: useful for momentum, dangerous if you skip validation.
• Visualization requires audience context (what they care about, what action they can take, what ‘good’ looks like).
• Better prompts come from better requirements: define the decision, metrics, grain, time window, and desired takeaway.
• AI can accelerate technical work (DAX patterns, data prep suggestions, layout iterations), but humans still own intent.
• Messy data needs business choices: outliers, hierarchy changes, and restatements aren’t solvable by a generic model alone.
• Expect iteration: AI proposes → you filter → you refine → you ship.
• The long-term opportunity is automating repetitive report plumbing (themes, navigation, interactions) so you spend more time on narrative and decision support.

Looking Forward

Pick one recurring report task you do by hand (layout cleanup, formatting, interaction wiring, or text drafting) and test whether AI can give you a reliable starting draft—then keep the final narrative and validation in human hands.

Episode Transcript

0:28 good morning everyone welcome back to the explicit measures podcast with Tommy Seth and Mike oh I forgot through the names with the incredible Tommy with the infallible Seth ow I didn’t know you got to went to Rome goodness we went to Rome well I don’t know how to say it like that for Catholics infallible ability oh okay yeah okay all right I don’t really hear that word often except we don’t you don’t hear I hear that one that one but Seth’s ideas are always on point

0:59 but Seth’s ideas are always on point though though yeah we’ll see how this right with that grunt to give me so much confidence is it what day is it Thursday that’s right Thursday another these These Warnings for us man man I’ve been spending more time with you guys than anyone I know right now and you’re the happiest you’ve ever been in your life your life I don’t know for those listening or as

1:32 I don’t know for those listening or as my son would say that’s mid mid I miss you guys if you don’t leave oh yes exactly right kids ever repeat things speaking of like mid and that like all of a sudden they start saying something and you look at them like where like why are you saying that and all of a sudden your wife looks at you because you say it all the time oh no do your kids ever do that to you I I have said some words and the kids have picked up

2:02 some words and the kids have picked up on it and they think yeah that’s not the funniest words ever and now they’re like actually we probably shouldn’t say that there were a couple there were a couple I didn’t realize I was Satan to my son one was like bro like oh yeah I’m like bro come on he started saying that or my daughter did and then that’s funny there was something else in in conversation that comes up all the time he’s like oh I see maybe it’s like oh I see your point like oh yeah

2:34 oh I see your point like oh yeah yeah and it’s funny when you work with you work with people and you start seeing like they have phrases that they use use we just hired a new intern and and he’s great he’s doing wonderful but he’s got a phrase and I think it’s very it’s a good word he’s like interesting [Laughter] yeah I know it’s good well I’m with with a kid who any I’m sorry any kid who’s five even if they don’t mean it if I ever hear a kid say bro even if I say it you’re like Punk yeah they don’t know like my daughter

3:04 yeah they don’t know like my daughter also I was talking I was like yeah we’re gonna go there she’s like wait what come on I was like where did that come from then I realized that was you yeah exactly yes they’re little lenses they lens exactly what you do they pair it back to you what you do when I was when when the kids were really little my youngest would like just I she saw that I talked all the time about power bi so every story that she ever came up with when she was talking about the family had some power bi something in it

3:34 some power bi something in it didn’t matter what it was and then they power bi or whatever he did random things and I’m like yes yes they did oh I meant to I meant to tell you guys this is something I wish I I missed a major opportunity at the conference so this shirt that I’m wearing right here has a QR code on it and the QR code is a link to the power bi tips website but what I should have done is should I should have worn the sweater with a Rick Roll on it so it would people scan the

4:05 Roll on it so it would people scan the QR code it takes you to a Rick Roll yeah would that have been what’s on your shirt dude I don’t know scan it and then they’re like Never Gonna Give You You well if you’re trying to get more purchases on this yeah right I don’t know if that it would work out as well April 1st is coming okay April 1st is coming exactly let’s jump into our topic for today for today let’s talk about this this is going to be interesting topic I think what

4:35 to be interesting topic I think what does AI artificial intelligence know about dataviz and storytelling oh boy yeah now it’s taking their jobs right the the author certainly sounds Italian how do you all I find are you pronounce that Tommy okay so first off first off kudos to the Italian Community because I’m not intending to do this I think he is I think you are Tommy it would be great it would be great if I would yeah this article’s not good it’s not for Italian no it’s just a testament

5:06 not for Italian no it’s just a testament to I think that maybe we were just a test of it I love it hybrid like genetically inclined to be Oh no you’re going to genetically inclined oh boy Tommy I don’t know what thing but like just kudos to all these Italians that we’ve found who’ve done amazing articles I think it’s because of the espresso they have a really good espresso yeah probably no oh that’s bad okay Hollow how you pronounce the spot cure I’m not even going to try it right now I it’s really or something

5:37 now I it’s really or something yeah I’m adding an R when I shouldn’t I I don’t know we’ll move on either way that’s a very good article he’s a founding director at the center for design at Northeastern University nice which is the partner of the visual agency the so they spend a ton a lot of Education and Research dollars towards design and visualization and that’s the article we’re looking at today so the article will be in the description of the of the YouTube

6:08 description of the of the YouTube video so if you’re watching Youtube or LinkedIn we’ll try and keep this this article inside the link so if you want to see the article rate for yourself you’re more than welcome to look around that as well initial reactions what do you think Tommy Tommy so this goes back to one of our conversations we’ve had and I apologize

6:25 conversations we’ve had and I apologize if I’m saying this in the future or not I know we’re doing some recordings but right off the bat with this article the the idea here is we’ve talked about gbt a lot but can I also go into the database space I don’t know if you guys ever use Dolly at all the the generator for like it’s funny but it’s not necessary like it can do yeah it’s not as far along as the prompting in the text but that being said we see what data like with AI can

6:55 said we see what data like with AI can do and from I think from our role from the role of the bi team where does AI fit in like obviously from a coding point of view but we haven’t really talked about can it what about the day when the AI can actually read or model like read over a model and the right metrics and understand a tabular format especially with tisdle coming it should be pretty simple to do I think you mean like Tyndall right with the new formatting right yeah yeah

7:26 formatting right yeah yeah so where does AI going to play a role in terms of visualizing data and then in terms of is it going to replace a lot what we do is it going to be one of those just look at powerapps where you say I need to build an app that sends out emails when people do invoices or whatever and then bam you’re done do you just need to say I need a few lines like I need to track three weeks over time how our sales are doing and then you’ve got all your visuals so I think that’s

7:56 your visuals so I think that’s the initial idea here I think what’s interesting to me is overall overall to to sum up the article I think his point that I agree with and when we we started let’s finish the thoughts huh that I agree with you AI is an accelerator so far AI is an accelerator not a replacer right correct when we started talking about chat CPT or like the instant the instant place you go is

8:26 the instant the instant place you go is like oh my gosh what is this going to replace yeah it’s pretty good huh that was great yeah you were now hired for all officials let me work on that let me work on it okay but but ultimately it’s it’ll be fun to go through the article because I think he does some of the this not the same but it’s very similar what would chat GPT say about storytelling data visualization

8:57 about storytelling data visualization and then it provides commentary back and forth about what he thinks the output is generating which I think dovetails into a conversation we were having this morning about chat GPT and these systems are only as good as the input that you provide it right so I I I I like this article and I think it’s going to be the first of many that we we might interact with where people are stressing the bounds of what AI is providing for prompts and

9:28 what AI is providing for prompts and giving them ideas for feedback and things like that to some degree and this is a side tangent I know some of the same reasons why we started this podcast related to the conversations that we were having about business intelligence and data and power bi all the time were related to the value we found in that conversation right and I think chat GPT and AI to some degree fits that role for

10:00 and AI to some degree fits that role for people who don’t have that right now you can’t get the community you can’t get the show so you better like I’m not saying just stick with chat gbt we’re much more interesting good jokes [Laughter] if you think about it it’s that it’s that you providing your thoughts and ideas about something and then getting a slightly different or different thoughts and ideas back that start to create

10:30 and ideas back that start to create patterns of thought and accelerate your direction and pursuing different Pursuits of solving problems oh I I agree and here’s where I think I’m struggling a little bit here on some of this I think there’s I feel like the AI portion of stuff is helping me automate some of the simpler tasks some of the mundane things they’re giving me at least so I have in my mind what I want to say to say but I want to get like an example of

11:00 but I want to get like an example of like here’s some words that I could start shaping and forming and tweak the meaning of them so that meets the objective of what I’m trying to reach so I I like in the in the chatting side of things where I could ask a question and it returns to me like a starting point and then I can refine it and tweak it and delete out stuff that I don’t like as opposed to the flip side of like how would this happen inside visualizations I think there’s a lot of automation tasks that could be handled by the AI for example what if I just laid out my visuals on a report page and said AI

11:31 visuals on a report page and said AI figure it out like and then it I what I’m assuming will eventually happen was what we see in PowerPoint right you put a couple visuals on a page and then you tell power bi okay well just generally make me something that looks good with this and it figures out how to make it better pretty right or something like that that there’s also simple things like why do I need to muck around with bookmarks anymore why can’t I tell an AI thing to say I need three bookmarks in

12:03 thing to say I need three bookmarks in this report that go to this this and this this that should be a very repetitive step that could be done over and over again again I know that you while you were saying that it would be awesome yeah while we’re like while we’re thinking about new Microsoft tools that they should build yes if if you use some of this AI technology and you said here’s my theme file right so abide by Colors by all the criteria I’m setting for you

12:33 by all the criteria I’m setting for you as far as a visualization and go build me I need three pages the first one’s going to do this this this I need navigation buttons did you say pontificating because this is going to go back to the pope thing I did okay so anybody can pontificate

12:51 did okay so anybody can pontificate I know but again where you normally hear the word I digress so that is a good point so we’re still dealing with code until your point like the theme yes it can start with the visuals yes when the Apollo really points out and I think that again goes back to a theme that we’ve been saying is AI is can enhance what we do is not necessarily going to replace the people who are going to replace you or the way AI is going to replace you is not the AI itself but

13:23 replace you is not the AI itself but it’s going to be people using it at the same on the same token when it comes to I think our ability to storytell and to visualize yeah it may be good from a start from an Inspiration Point of View but there is a key difference in terms of there’s they were not dealing just with code and we’re not just dealing with a particular in a sense prompt because when we’re visualizing data it’s for a particular audience who has in a sense

13:53 particular audience who has in a sense what’s important to them and I think that’s a key distinction that AI I don’t know if it can get to that point but I think we have to really remember that AI is not initially going to understand the context of who’s the audience but but it’s your point earlier though Tommy right like it absolutely is not going to be able to figure those things out unless you give it that information right like

14:24 right like who knows where this guy leads but at the same time like the I I believe somebody really smart called Tommy pulia was just telling me this morning that he’s learning a bunch of how do you interact with these these tools right it’s changing us the more you context you’re providing the the more value you’re going to get out of something something right so it’s not going to know that unless you tell it right who’s the audience for this report what are the

14:55 audience for this report what are the main key things that they need to be doing when they’re analyzing the report what are the big takeaways that they need to do which metrics do they need to see like all of those would be pointers to the data that we would have to produce for it or or shape or or whatever what I’m saying is the cut we would provide that context and that’s the value where we’re streamlining and using these tools to accelerate the production of what is the technical a lot of the technical

15:26 the technical a lot of the technical skill right and I’m going to say it’s going to replace everything because it can’t but at the same time like well if you want to produce this insight and this visualization for the end user you need to create this really really hard Dax cut calculation based on the model that you have yeah I’d much rather have that than like oh geez here we go that’s that’s it in the middle of the article and again Tommy do you can’t think of your point here you what you were saying there’s a section here talking about

15:57 there’s a section here talking about data pre-processing and I think as I’m thinking about like how do I work with models and companies today today there’s too many tables we don’t have technically all the right business requirements around how data links together and the often the number of times when people bring data to a data model they try to get it stitched together and these the stitching together of that data Works only so much there’s always a handful of errors or mistakes with it so he he talks in the article about AI can clean transform and prepare data for the

16:29 clean transform and prepare data for the visualizations and then identifying how would you want to handle outliers because I feel like there’s always this thing of like for example you have a hierarchy in a company right you had a hierarchy and someone reports to someone in a weird way the AI should be able to point out hey the data for this person changed heavily because the accounts changed for this person halfway through the year how do you want to handle it do you want to restate it or do you want to divide the accounts across the two different sales people like there’s a there’s a business challenge that comes with the

16:59 challenge that comes with the information because the data is fluid and that’s where our jobs are sitting right now like we come in and we we talk to them hey here’s the technical challenges of the data right and then we say here’s how we would want to solve it you can either choose a or b each one of those may take a little bit longer or less more or less time based on like how hard it is to solve that problem problem yeah and well I I think more where I see to assess point in terms of the prompting where

17:30 point in terms of the prompting where who’s to say we’re not going to get to a point where just AI will know your organization structure or hierarchy maybe so and those apis and you say this reports for Department X and for this quarter right and the model is called this and again who’s to say that I won’t be able to do that no we’re not there yet no but I won’t be surprised when we are I think it’s going to throw out a lot of junk to your point like I think it’s going to throw out it potentially could

18:00 going to throw out it potentially could recommend a lot of things but it’s up to you you only you can control real data like there needs to be like right there needs to be like another hook to this that says okay here’s a lot of ideas we think you might want Tommy what is actually relevant to you because it doesn’t know yet on what’s most important for your feedback right that was in in your schema yeah month 12 always has the highest number of sales from a cumulative total no dip Sherlock it’s it’s the cumulative total for the last 12 months

18:33 is the series of issues I’m saying you could go into like again not to speak ideals but the models there or the the structures there for you as a bi to have a schema of Team metrics that are important what this role is and then you basically say hey I need a visuals for team this you hey I need a visuals for team this for this month can you show know for this month can you show visualize something and who knows but we’re not interesting yeah so interesting I think the biggest thing

19:03 interesting I think the biggest thing one thing one thing I want to focus on is there’s a big difference I I what we’re doing between effective Ai and storytelling and then

19:15 effective Ai and storytelling and then efficient efficient I think and that’s one thing Apollo points out I think absolutely this is going to be part of what we do and as Paul said you better know how to prompt soon but I think a lot of what it’s going to do in my personal opinion is it’s going to be very efficient in terms of a lot of Base visualizations a lot of Base items items but like I said I think there’s a lot more in terms of

19:45 more in terms of you like you said about bookmarks or the the audience or that real context that I don’t know if it’s going to be like the replacement be-all end-all and do you see a difference there in terms of between effective and efficient with what the store the visualization can do effective versus efficient yeah I know I’m having a hard time distinguishing the two what what like what would be

20:15 the two what what like what would be these characters these were two prompts he was getting in outputs right that he he leaned into a little bit bit Let me let me re-look at that real quick to answer your question is it in his article that you’re talking about exactly yeah yeah oh okay because Chachi PT kept using that in its outputs so he he leaned into that a little bit and bit and and outline so just just for some context he says I was tempted to

20:45 context he says I was tempted to start questioning the meaning of effective and efficient two terms often used to identify the goal of AI assistance but refrained I really couldn’t resist when the promise was optimal design asking what optimal design exactly means in the context of data visualization data storytelling led to a synthesis of efficiency and Effectiveness with a key argument that I loved it depends so that we love GPT because it’s our favorite teams it depends so

21:15 it’s our favorite teams it depends so what do you take from that what do you take from what I I love that paragraph I think it’s heavy yeah it the output of it like it finishes with proposed dependencies are type of data the story being told the target audience the goal of visualization is not bad in the context of data storytelling efficiency disappeared replaced with engagement claiming that a story should be both informative and captivating so I think I think where this is I guess leading to me is the efficiencies that AI is

21:45 me is the efficiencies that AI is driving I I guess could be related to developing the visual but I I see that much more so in the data preparation and how it like creates the outliers that we were just talking about because that’s going to be much more it’s going to be much better at identifying some of the the challenging business things that we that it take takes a good amount analyst to find like you have to understand the business and look and be really good at data to see problems in it sometimes right to

22:15 see problems in it sometimes right to your to your point Mike you don’t get to that point when you’re building a report until much after you’ve already produced it after you’ve already solidified everything and you’re presenting that to people and people are looking at stuff going that doesn’t quite make sense like why is that happening proactively bring those forward because AI would Point those out to us that streamlines our process I agree with you so effective though leads to the point in here where how how

22:45 point in here where how how good is or where does AI or us need to focus in terms of like when you’re trying to tell a story which is significantly different like we talk about data storytelling much more in what infographics in journalism right where it’s this it’s not just a visual that helps people make decisions it’s it’s designed to outline for you the background the direction like there’s a path that is

23:15 direction like there’s a path that is designed to lead you down so you understand a concept or you understand the story of what all the data is doing for you so I think I think that’s where he Dove deeper into this context within Chachi PT because he kept seeing this back and forth with effective versus efficient in all the outputs so to react to that I’m going to go back up a little bit further so effective and efficient and I liked how he landed on the on those statement there right in replacing efficiency disappeared

23:47 in replacing efficiency disappeared and talked about engagement or engaging visuals I’m struggling on like I could see potentially where you would apply AI for various aspects but earlier in the article he talks a little bit around automated chart generation or automated design design and I think those are two distinctions where I would see where AI would fit the bucket or the ticket here around where you’d want to apply this right can I get AI to stub me out so I guess maybe

24:18 AI to stub me out so I guess maybe in my mind I’m thinking it would be really cool to be able to ask a question of your information build a data model have it sit in front of you and then you can say what are my total sales or if you’re running like a store which days of the week do I sell the most stuff and then it it takes the this basically it understands this is the this is the end goal I need to be able to tell you and this is where I’m thinking like in my mind I’m going this could be part statistics this could be part using the data and it could

24:49 be part using the data and it could produce a Litany of just random visuals or things that would say here’s how to get from your data to the answer to that question and it would say you question and it would say on Tuesdays you tend to sell more know on Tuesdays you tend to sell more whatever ice cream and it has to be a day that’s higher than a temperature of right and so the idea here is it’s potentially creating part of that story and telling you very specific things I think though what Tommy’s been doing

25:20 think though what Tommy’s been doing like Tommy the AI is teaching you basically just under greed right it’s teaching you how to prompt it to get what you want out of it so this is the same thing that I think that happens with Google and and people today like we’ve been we’ve already been ruined by Google like we know how to search for things I remember when I started using Google initially it was like I don’t know how to find the things and you just like start stamping and like you

25:41 like start stamping and like you you now have a format like how you want to search through Google so Google has trained us to go look for like if I’m looking for power bi things I search Power bi and then the thing or if I’m looking for something in Azure I do Azure and then the question I’m trying to ask and then it scans and it’s it doesn’t so in the same way I think these AI elements of how we build reports and how we storytell we are gonna it’s gonna be like a middle ground right the AI will provide something and then we’ll also change how we ask the

26:11 then we’ll also change how we ask the questions or change how we integrate with the computers to get the answers to make it understand what we want it to produce does that make sense so I I think that’s a great point where I think you one can make the argument or if I were to ask you this today could you survive in your in your role if you didn’t have Google and and I think there’s there’s a there’s a point here we don’t want to use books all the time but I could

26:41 could do you think you would be as a really with power bi especially like if you didn’t have if you had no ability to search do not use Google you think so absolutely why would I be as effective probably not so okay so if you’re if you’re talking about like people were using computers and doing SQL well before Google hit like hit the mainstream of

27:11 Google hit like hit the mainstream of like querying all your your text answers and or finding your code and all this stuff so yeah you can it’s just not as a fast or efficient now stack Overflow that’s a different question question let me rephrase the prompt would how much credit do you give to where you’re at now is Microsoft MVP as the expert that you are because of searching and Googling that’s a great question because I say verbatim everything I know is

27:43 I say verbatim everything I know is already on the internet I’ve I part of it is stuff that I’ve created created but part of its stuff that I’ve gone out and searched so I wouldn’t attribute my MVP to the ability for me to be able to Google the things I would attribute my MVP to being like over the top obsessed with it and and going like listening to an article and so there’s a there’s a difference between I want to say if there’s a difference between knowing and understanding I think there’s a point in time when you

28:13 think there’s a point in time when you work with some tool where you start to understand it and and that’s the moment where you can say when I understand it I’m able to tell like people can ask me questions around power bi I don’t even look at a desktop I don’t even need I don’t need to have anything in front of me I can tell you exactly what to do click this button go here click this thing it’s found like this like like I can tell you exactly what to do without even that’s to me the point of understanding it with things now I really like to understand things I’m very curious so I I would like to

28:43 very curious so I I would like to attribute like that Curiosity to teasing things out figuring out where things fit trying to find I love creating solutions from nothing or things that have never been done before so like I think I would I would probably not lean very heavily on Googling getting me to where I was at for MVP but what I would say is for anyone who’s new coming to the space there’s a ton of resources out there and it’s hard to distinguish between what’s good content and what’s not good content so you have to take it upon yourself to learn you had to get

29:13 yourself to learn you had to get experience and you had to get from a point of knowing how the tool Works to understanding how the tool works and when you get to that level that’s when you can add real value so I would make the argument that my ability to like basically query things on Google and GitHub or the community site to put in the right search to get those results and discover those answers has been a a marginal difference in terms of where I’m at in terms of skill marginal means

29:44 terms of skill marginal means no no significant significance marginally significant yes but and I think it’s made me incredibly efficient in terms of getting the answer that I need and I think the same way right because I don’t have to go I’m not by paging through a book I can get the answer in a second because I know how to queer I know how to get Google to not like ignore a word in my in the query I

30:15 like ignore a word in my in the query I know how to do certain things in Google and on the community site to get an answer quickly so I don’t know what you’re describing what you’re describing to me like is is search engines provided in accelerator for us to learn more rapidly yes that’s it for the perfect way to say it for those of us who are older when you were in grade school cyclopedia you learned by going to the library right and and you guys know like index

30:45 right and and you guys know like index cards trying to find the right book trying to like the library had the book now you gotta go find it and the only way to find it was like to pull out these massively long things where the fold like there were no like computers were in a day to life the way we learned was significantly different does that mean that the things that like that there weren’t experts no does it mean that there weren’t MVPs no does it well I don’t know how far back they go

31:20 right the effort to learn these things took a lot more it was physical Journeys two locations because that’s what knowledge stores were and now like yes does search engine optim like does search engines allow us to learn a lot faster yeah but the expectations are a lot higher and I think to some degree I don’t know where you’re driving without like do do all of us have an onus of saying that like yeah we learned rapidly and we have audiences that are much bigger than it would have been as far as like us not having platforms like the

31:51 like us not having platforms like the internet to reach out to people and talk to them and interact with them on a level and these massive communities of people revolving around a site right like us right not a specific location

32:05 like us right not a specific location but a site location on the interwebs right like before that it was your 10 people if they were paying attention to the newspaper and they showed up at the location and and had to be there right so yeah all of this I think then dovetails into the conversation today where it’s like hey Ai and visualization data storytelling how does this fit like it’s another accelerator for us I think that is just going to help us produce more as organizations start to adapt and

32:35 more as organizations start to adapt and adapt their tool sets to allow us to use them more them more that’s exactly what I’m driving home at is I Ai and I think from what I can do for my visualization point of view is not necessarily going to replace but for what I think we can have I don’t know if it’s going to be called a responsibility but I think in order to stay part of the skill set to being efficient is to utilize Ai and I think being able

33:06 is to utilize Ai and I think being able to prompt it from a efficiency point of view to get that base part of at least like Mike you said like quickly get the outliers so if you kind quickly get the outliers so if you get some of the base information out of get some of the base information out there that I think can really like you said like and make a significant impact compared to someone not using it at all and I think it’s going to be just part of the process I think throwing AI at some of these data data challenges like like the the quality of the data I think is going to be immensely helpful because building

33:36 be immensely helpful because building data quality type metrics and Reporting is very tedious and it would be nice if you could throw AI at a data model and say I’m going to analyze all of your relationships and tell you how many blanks there are from which tables and what records they are just just very quickly like just because they can do that in moments where it would take me like a couple minutes 10 minutes to build some tables and look around a little bit and see where the dimensions are not matching okay why are they not matching what records in this thing that are not in that one where’s it there’s other things here I think that I would probably try to apply more

34:07 that I would probably try to apply more around the AI pieces so with that I think we’re actually at the time of the hour we’re out able to say it’s time for what does chat GPT say so since in a conversation around AI we should definitely use AI to cap this conversation as well I wonder if it’s gonna not encourage these exactly it is a bad idea you should not use it it’ll dominate the world here and you’re going to be in The Matrix here a little bit replaced it’s all replaced so before we do that though we did ask chat GPT to give us a joke so I said

34:39 ask chat GPT to give us a joke so I said please chappy T can you can you provide to me a dad joke sure here’s your joke why do chicken coops only have two doors because if they had four they’d be called a chicken sedan [Music] thank you thank you I’m here all night I’m I’m teeing at my material for for powerbia tips Comedy Hour so that’s we’re gonna we’re gonna have that happen no we won last so I did ask a real question I said Can artificial intelligence build data visualizations

35:09 intelligence build data visualizations of clear clearly Chachi petite says yes yes AI can build data visualizations in fact AI power data visualization tools are becoming increasingly popular in the industry industry I’d like to see one that works so we’ll have to see we’ll get some recommendations go ask it ask it what are please please cite your sources chat GPT calling you out I’m calling you out AI algorithms can analyze large data sets which I think what we talked about and that’s what the article also said they

35:39 that’s what the article also said they identify patterns and Trends and they represent findings in a visually appealing way which is interesting I like gaming good it can also cleanse data processing and suggest the best ways to present database on the type of information being analyzed I think that’s also pretty relevant again I think the cleaning part here is potentially a lot more valuable that I think about where I spend a lot of time versus versus maybe in in the actual building of the visuals themselves for example again it goes a little bit more examples here for example AI can create charts graphs and other

36:09 can create charts graphs and other visualizations and then it can also assist with color selection selection font choice and other Design Elements to make the visualization more effective and engaging that I really like so the idea that it should be able to say here’s a color and here’s the answers for those colors anyways anyways I think it’s pretty on point here I would probably agree with it I would give this a pretty good score from chat jpt I think this was not bad that’s a great idea we should be tracking our our rating for each one and

36:39 tracking our our rating for each one and create a data set from that rating from each one for every GPT how good was there oh yeah you’d respond we need to yeah we need to start plotting this all right well more data Chad TPT please plot your responses back from the pot yes in the explicit measures podcast how often was Chad GPT right of a score we should just give it a score every time unless you chap GP you can go find it later on I’m I’m pulling that oh boy oh boy so anyways with that we thank you very much this is a shorter episode we’ve recorded it so apologize for not being live we’ve had people traveling around and doing things so

37:09 traveling around and doing things so thank you very much we really appreciate your listenership if you like this content if you found this engaging please recommend you share this with somebody else across the internet we really appreciate it and not just share it let us know what you found was interesting in the article or in the podcast tell us what you thought where do you think AI fits and then share that and we’d be happy to have more conversation on the social media areas Tommy where else can you find the podcast podcast well make sure to ask GPT to subscribe to the podcast that’s where we want to

37:39 to the podcast that’s where we want to get where GPT knows who we are but make sure you find us an apple and Spotify Google podcast subscribe to all of our episodes we’re almost at 200 or you can watch live on YouTube and Linkedin every Tuesday and Thursday at 7 30 a. m Central awesome thank you all very much and we appreciate your listenership have a great week we’ll see you next time foreign foreign [Music]

38:29 [Music] foreign

Thank You

Thanks for listening. If you enjoyed this episode, subscribe to the podcast and check out PowerBI.tips for more templates, themes, and BI guidance.