Microsoft AI Insider Reveals What Leaders Are Getting Wrong About AI | Chris Seferlis | AI Accelerator Podcast

AI is transforming every industry, but according to Chris Seferlis, most organizations are approaching AI the wrong way.

In this episode of the AI Accelerator Podcast, host Matt Zembruski sits down with Chris Seferlis, Strategic Advisor at Microsoft, Boston University Professor, author, keynote speaker, and enterprise AI strategist, to discuss what leaders need to understand about AI adoption, Microsoft Copilot, AI agents, and the future of work.

Drawing from years of experience helping enterprise organizations implement AI solutions, Chris shares why so many AI projects fail, how leaders should think about productivity versus replacement, and why successful AI adoption starts with people, process, and business outcomes—not technology.

From Microsoft Copilot and Azure AI to enterprise data strategies and agentic workflows, Chris provides practical insights leaders can use today to unlock real business value while avoiding the common pitfalls that derail AI initiatives.

At the heart of Chris's message is one critical idea:

"Start with the business problem, not the AI."

In this episode, Chris reveals:

◼️ Why most AI projects fail to meet expectations
◼️ The biggest mistakes leaders make when implementing AI
◼️ Why AI should enhance people, not simply replace them
◼️ How Microsoft Copilot is changing workplace productivity
◼️ The difference between Copilot Chat and Microsoft 365 Copilot
◼️ Why data quality remains critical for AI success
◼️ How organizations can maximize existing Microsoft investments
◼️ The rise of AI agents and agentic workflows
◼️ Why business outcomes should drive AI strategy
◼️ How AI is transforming enterprise operations and decision-making
◼️ The role of governance, security, and responsible AI adoption
◼️ Why leaders need hands-on AI experience themselves
◼️ How organizations can build an AI-ready culture
◼️ Practical examples of AI-powered automation and productivity gains
◼️ What the future of work looks like in an AI-driven world

Key Learnings

✔ Most AI failures are caused by poor strategy, not poor technology
✔ Organizations must define business outcomes before implementing AI
✔ AI is currently creating more productivity gains than workforce replacement
✔ Data quality remains one of the biggest success factors for AI initiatives
✔ Microsoft Copilot is evolving into a powerful enterprise productivity platform
✔ AI agents will automate increasingly complex business workflows
✔ Leaders should personally experiment with AI before driving adoption across teams
✔ Successful AI adoption requires change management and employee engagement
✔ Governance and security must be part of every AI strategy
✔ Organizations that embrace AI thoughtfully will gain a significant competitive advantage

💬 Chris's Most Powerful Quotes

"Most AI projects fail because organizations don't clearly define what success looks like."

"AI should be tied to a business outcome, not deployed for the sake of using AI."

"The companies seeing the biggest wins are using AI to augment people, not replace them."

"Your data strategy is your AI strategy."

"Leaders need to experience AI firsthand before they can successfully lead AI transformation."

"The future belongs to organizations that can combine human expertise with AI capabilities."

Follow Chris Seferlis

LinkedIn: https://www.linkedin.com/in/cseferlis/

Website: https://seferlis.com/

YouTube: https://www.youtube.com/@ChrisSeferlis

X: https://x.com/bizdataviz

Email: chris@bizdataviz.com

Follow Matt Zembruski

Website: https://leadingaiagility.com

LinkedIn: https://www.linkedin.com/in/mattzembruski/

Email: matt@leadingaiagility.com

Phone / Text / WhatsApp: +1 978-618-5778

Transcript

SPEAKER_00 0:04

Welcome back to the AI Accelerator Podcast. I'm Matt Sembruski, and today I'm joined by someone who lives right at the front line of enterprise AI adoption. Chris Seferlis is a director of technology strategist at Microsoft on the Azure AI side. He's a lecturer at Boston University, published author, keynote speaker. He does a lot, and we're going to learn more about him and what he can offer, what he can offer you and really get your mind programming into a better, better place. Chris spends his days advising VPs, C-suite leaders, on how to actually put AI to work. That's what we're all about here on this podcast, right? How to put it to work. Not about the hype, but about the reality of what actually works. We're going to get into where companies are getting stuck today, what's working right now in the Microsoft ecosystem, how leaders can move faster without breaking the bones of their company. Chris, great to have you on. Hey, Matt, thanks for having me. Appreciate it. And I'd love for the audience just to get to know you a little bit better. We had a chance to uh connect earlier. Um, but if you could just describe a little bit about your background leading up to where you're at today, your role with Microsoft and sort of what you're you're passionate about.

SPEAKER_01 1:14

Yeah, yeah, definitely. Um, so I grew up in IT, right? Uh pretty much read out of high school internships as I was going through school, um, spent a little time in the army, uh, you know, and and really those kinds of things kind of shape how you see the world and and gain perspective, right? And uh, and so I did over 20 years in IT and and I really started to hit a wall, you know, just uh started to feel burnt out and things like that. Um had an opportunity to kind of come over to the you know sales side, if you will. Um I still I don't call myself a salesperson, um, but you know, really more of that technical advisor, solution advisor, strategy advisor. Um, and so you know, been here at Microsoft about seven years now, just over seven years, matter of fact, and um, you know, really uh still like to do a a lot of stuff in the way of tech, you know. So if if I'm playing around with the systems, building my own sort of agents and stuff like that, um really kind of like to dig in. And then, you know, otherwise, uh a lot of outdoor stuff, fishing, swimming, biking, hiking, all that kind of good stuff too, up here in uh the Boston area.

SPEAKER_00 2:16

Yeah, it's fantastic. I'm in the Boston area too, as you know, and uh it's exciting to see a lot of the for those of you who are on the video, and you can see there are a lot of good Boston sports paraphernalia behind Chris there, a lot of a big uh um uh big, big Boston uh sports fan, which is uh which is always great to see. So, Chris, let's jump into the the practical reality of of some of the the conversations that you have. Like you're talking to to uh to C-suite leaders, vice presidents, senior senior executives at companies all day, right? You knew you represent Microsoft and the technology, and there's just so much happening in technology and AI today, as as you know. You know, so when you walk into a company, what's the most common thing that you think the leaders are getting wrong about where AI actually is right now?

SPEAKER_01 3:02

Yeah, it's it's a bit of a mix. Um, you know, we see published reports by um by like MIT, you know, 95% of AI uh projects fail to meet the expectations, right? And what I find is that um a lot of times the plans are kind of half-baked. Um, you know, it doesn't mean that the project um won't have any value. It's it's a matter of sort of um building it in the right way and and understanding really what we're working for, right? Um we we talk more about business objectives and and how to get the organization to adapt to all this new technology. And so I'd say the biggest thing is um, you know, a lot of executives are really excited about the prospect of AI. Uh, you know, uh hopefully it's less on the side of um job replacement and more on the side of productivity, revenue growth, cost reductions, things like that. Um, and and what I find is that a lot of times uh there isn't the uh attention that needs to be paid to really understand where the value is going to come from, you know, some of these automations and and uh productivity gains and things like that.

SPEAKER_00 4:20

I just think it's sort of a miss. Let's let's click on that just a little a little deeper more on that. So um it's not you know, the MIT study, a lot of people are probably familiar with at this point. I think it's it's been many months, maybe it's been close to a year, it's been a little while. Uh, but yeah, they say like 95% of the AI projects are failing. So you're saying there the um the leaders, um, because because I've seen the same where leaders are really clear on like, okay, I want this improvement done to my organization. Maybe it's like we need to get more uh more sales or better controlled costs or better operational efficiencies, whatever it is. So you say they have some clarity of objectives, but not really understanding how to apply AI and how to consistently carry it to the end line and sort of measure success. Right. It's a little, it's a little uh loosey-goosey from a project perspective.

SPEAKER_01 5:04

Yeah, and and you know, there's a lot that goes into it. Um the quality of the data that we're using, um, you know, really uh pinning down what are those objectives? What what what are we what are we working toward and and how do we measure success, right? Um if that isn't defined up front, then yeah, 95% of organizations are disappointed with the outcome. You know, the the projects aren't necessarily failures. Um, you know, perhaps, perhaps they're a starting point, right? Perhaps they're uh a foundation that you know the team can grow on because now they've gone through the paces, they understand where you know they might have made some errors, um, you know, what things went really, really well, um, and and even some of the challenges around change management, the culture of the organization and stuff like that, right? So you get a lot of those things that get mixed in. It's it's not um it's not a very simple cut and dry, um, you know, we're going to implement X and it's going to save us 10% a month, right? It's it's not as simple as that.

SPEAKER_00 6:08

Yeah, I've noticed the exact same thing. There's more nuance to it. And uh the media often is not uh our friend and trying to get to a positive outcome. You know, there's a lot of fear and noise out there in the media about AI. It's shifting all the time and different storylines. Um, from where you sit, what's the reality that executives need to hear instead? Like what's the what's more of the the truth about AI um to bring us to that positive, successful outcome that the that the companies want to get to?

SPEAKER_01 6:37

Yeah, I mean, one of the big ones um that that we see a lot is that you know, XYZ company has cut X amount of headcount um because uh they're going to use AI instead. Uh, you know, I sat in a uh uh Gartner conference uh in the fall, and you know, they said flat out, um, we have zero tangible evidence of organizations being able to replace headcount with AI at this point, right? I mean, there are definitely some automations, but they've been going on for a long time, right? So smart factories, this isn't new. Um, you know, uh robotic process automation RPA, this isn't new. And and so, yeah, maybe some organizations are starting to see some of those gains from those projects. Um, but the idea that you know the advancements in with LLMs um are gonna replace everybody that's a white-collar worker in the next 18 months is just insane to me. And more recently, we've seen you know, Sam Altman and um um uh shoot Daria from uh Anthropic, I'm trying to remember his last name. Um, you know, yeah, thank you. And and um, you know, they've both sort of walked back their, you know, we're gonna replace, you know, a large portion of the workforce with AI in the last couple of weeks here, right? So um at the same time, uh that this stuff is advancing so quickly. Uh and a lot of times it really just has to do with how the focus on um the the various environments look. You know, Microsoft had our has our build conference right now, as a matter of fact, out in San Francisco. And we we had a bunch of announcements yesterday. Um, some things around personal productivity, some things around quantum, some things around new AI models to to help um you know cut some of the costs that it takes to run these things. Um and you know, some of the stuff is is somewhat of uh uh an advancement on prior technology. Um some of it is a new look at things, and and we have to see how the market's gonna react, right? And so um there's a lot of exciting stuff out there. Um, like you said, there's a lot of fear that gets put out as well. And and whether it's you know competition that's putting fear out there or uh you know just the the salacious headlines to get you to read or the clickbait, right? All that um, you know, it's just do your homework, right? Cut through the noise. I understand we're all busy. Um, but if this is something that you believe will make that impact, really do the homework.

SPEAKER_00 9:10

Yeah, that's great advice. That's great advice. Let's talk about um things that some some things that only you could talk about, Chris. You're some of the only people uh about the Microsoft ecosystem uh related to technology and people and all the advancements. I mean, you you have a very unique perspective on this uh to share with our listeners. I want to get into this. So a lot of executives sitting out there, uh they've been working with Microsoft Technology, Office 365, Microsoft 365 um um infrastructure has been built into their organizations for years, especially mid-market, large enterprise, you know, large, um obviously the incumbent provider out there. Uh when it comes to AI, the recent shift in recent years, right? Uh they're they're on the either the free uh free copilot license or maybe they're on a basic version of co-pilot license. Can you help explain, sort of just cut through, because you understand the technology, the people and everything that's going on? You know, what what are what are your what's your basic advice um to uh companies out there that are trying to figure out how they can best adopt the Microsoft tools? Uh how can they can best bring AI into their organization, into their workforce, as you as you had talked about, to really empower their people and really unlock some of the value in the organization. What what's your uh next step advice there?

SPEAKER_01 10:26

Yeah, um, and it's interesting, you know, like you mentioned, the the free um version of Copilot, it's a copilot chat, right? It's built into the browser and um you know it's free for anybody who has a Microsoft uh 365 license of any kind, right? And and primarily that's going to be based on um web results. So it's gonna be similar to the free version of Chat GPT or Gemini or you know, Claude or any of those, right? Um and and the the benefit to doing that step up license to the Microsoft 365 Copilot is now it's using um a contextual layer from all of the data that you have access to. And that's a key point is that it's the data that you have access to. So if you're a CEO, you should probably be able to see just about everything, right? Uh unless there's some kind of legal reason or something like that, but pretty much everything. But if you're you know a data engineer, you're you're gonna have guardrails. And um the tools can be used in so many different ways. And and again, the advancements that are coming through um, you know, things like cowork that we're we're designing and building with anthropic um directly, um, where you sort of have these multi-chain processes where you step through all of it, right? Um and and it it just uh aligns as an automation piece. Um, taking the approach of having multiple models, so instead of um having a subscription to just anthropic claude or just open AI Chat GPT, um Microsoft is bringing in multiple models into the interface now. So you'll have things like Opus from Claude from uh from Anthropic, you'll have um GPT 5.5, I think is the current one. Um, and then as the the Microsoft models themselves, MAI, um start to come out more and more, and we just released some new versions yesterday, um, you know, we'll we'll start to see those. And so we're taking this multi-model approach within Copilot that allows you to work with all of the models and get the best of each of those models um, you know, for whatever task you're trying to do. You know, as we were chatting um, you know, in prep here, I use the example that uh, you know, Claude is fantastic at making PowerPoints, it's got great integration with PowerPoint, it's it's nice, presentable, it does a really, really good job. ChatGPT didn't do as good a job, right? And and I mean everybody's improving. I don't want to throw any mud because I think they all have their strengths and weaknesses. But at the same time, you know, if I'm looking for a process flow diagram or um an info sheet on a process that I'm working on when I'm teaching my class, right? I can go into ChatGPT and it's gonna create a great image. Um, you know, all the spelling stuff has been worked out and everything like that. And so, really, if you think about all of these models having their strengths and weaknesses, uh and being able to use one interface for all of those, that's really powerful.

SPEAKER_00 13:19

It is it is very powerful. And I think the multimodal approach um uh is is a smart approach no matter what, for people to get familiar with what's out, what's available out there, and if Microsoft's building a an intelligence layer on top of it to say, okay, there they're the user wants to create an image with graphics or with with with uh with text on it, okay, automatically map to the the AI that that's in there that's gonna have the best way to do that. Right. Uh let's let's talk about some uh beginner level use cases. So let's say uh organizations out there, they haven't upgraded yet to the uh to to the M365 um uh uh paid license version, um, but they're considering doing it and they're not sure uh where to start, how to start. If you're if you're sitting down with a CEO or uh senior leader in a company, um they know what their challenges are, right? That that are going on, and and it's different in every single company. But what what do you sort of how do you advise them to get started? Do you lay out the uh you know discovery of like what's going on in the organization, and then you move to the recommendation of licenses, and then this is how we're gonna implement them. Sort of walk us through that that uh basic process. Could be you or one of your colleagues, but like how how do you address that?

SPEAKER_01 14:37

Yeah, I I mean, I will say in the enterprise, um, the vast, vast majority of organizations have some flavor of this. Um, you know, maybe only a couple hundred licenses of the uh M365 co-pilot. Um, like I said, the co-pilot chat is is free. Um, and we're seeing a lot more use of that. And that's more of that personal productivity. Hey, summarize this document, um, you know, uh help me write this email, you know, things like that, right? Um, and and that's all kind of table stakes at this point, right? Where it's it's um it's not it's included with that licensing and and so there isn't the additional cost. And so I think a lot of people are kind of getting getting their own productivity gains. But where I'm seeing it fall down, especially when we're when we're talking about that co-pilot chat licensing, is that there isn't as much mind share within the organization as there really could and should be. Um, you know, whether it be uh, you know, one of the mistakes that we've seen is that organizations will roll out um uh, you know, peanut butter spread is the term you hear a lot across multiple functions in the organization. So, okay, finance, you get five licenses, IT, you get five licenses, and marketing, you get five, you know, and and the challenge with that is um that there is less of an opportunity for those groups to collaborate because their their challenges and the way they're using the tools are very different. Um, and so there's less of that opportunity um to do that mind share and really maximize how the tools are being used. Uh even internally at Microsoft, we have uh we have all kinds of um internal communication around here's the prompt of the week, right? Here's um I just had I just had it do an analysis on my working style, looking at my strengths and weaknesses as an employee, um, where I can improve. And it was pretty darn good. It it actually um calls out some of the things that when when you're having those really rough days, and you're like, why am I playing traffic cop right now? You know, and it and it told me you're doing this too much, right? And and that's not, I mean, you do it well, but that's not where your real value comes in. And so that sort of next level is is um the IQ, so work IQ, um, which takes it to um more of a semantic ontology layer. And instead of looking at a specific document or a specific piece of data, now it's reasoning across all of it and being able to build a point of view to say, here's how you can use the tools better based on all of these data points. Um, you know, and then I'm looking for specific documentation. Well, it's gonna pull back things from Teams and OneDrive and uh, you know, uh various loop components and you know, Outlook and all these different areas that are related to that topic, as well as you know, business-related data and things like that, um, to really kind of give you that complete picture. Uh, that's definitely the direction of this. But getting started with Copilot, um, you know, and having those paid licenses, native integration inside of all of the office suite, which is huge. Um, you know, so any, you know, any kind of document wanna write built in a word, you any kind of um PowerPoint you want to build, built right in. So you don't have to do that context switching, you can stay right in that application. And um, I those are some of the tips as to um how people can maximize the tools right out of the gate.

SPEAKER_00 18:11

Yeah, I like that a lot. And um, so you mentioned M360, Microsoft 365 Copilot, as well as uh WorkIQ. Is WorkIQ a separate offering that has a specific layer built in? Yeah, yeah.

SPEAKER_01 18:23

So it's um it was announced back in uh, I believe February, uh, and and sort of we just um had more announcements about it uh yesterday um during Satya's keynote. And and really uh again, it's that context. It's it's context about multiple areas of the organization, multiple disparate systems um where it brings it up into that semantic layer um to be able to provide that context and be able to return all of that information as opposed to having to check multiple sources.

SPEAKER_00 18:56

That's very, very that's that's very cool. I don't have experience with that product, and that's why I'm asking. Plus, our our audience, I know it might be new to them as well. Um, so how does that compare to like the uh the the concept of the project in Claude or a project in a chat GPT where it tries to frame up and scaffold some of that semantic layer in there? Is that similar or different? Is this Microsoft's version of it? What is that?

SPEAKER_01 19:22

Um I would say it's uh similar on steroids, right? Um so you know you're telling the IQ in the background what sources it should be connecting to, right? So it's similar in that where you load the documents, you kind of give it some context what you're trying to do. Um, and and most of this can just be done in the background. You don't even have to impact or uh or turn anything on. Uh, and then uh, you know, you can have it look at other sources that maybe aren't part of that native environment, right? So we we see the ability to work with some of our um partners. Um, you know, we're big partnerships with SAP and Snowflake, and being able to tap into some of that data as well makes it super powerful. That would be a big differentiator from a project, right? Because you're not going to be able to point it at an SAP database and say, hey, go go get get this information back.

SPEAKER_00 20:17

Interesting. So there's there's um it seems like there's there's other connectors that are available within the Microsoft uh ecosystem that um are more native to the are more native to the to the work IQ exactly versus versus like you know anthropic or or openai and the other other offerings out there. Yeah, yeah, yeah. That's good. That's excellent. That's that's that's that's fantastic.

SPEAKER_01 20:41

So looking forward to seeing how it goes, honestly. Um sorry, sorry, I cut you off there. I am I am looking forward to seeing how it all kind of comes through because you these take these take time to build and and there'll be multiple iterations. Um, but I actually I wrote an article about it last month, you know, and just the rise of this semantic and ontology layer. We're not the only ones doing it, um, you know, but but where does the value come in really? Um, you know, and and and how does this get governed and and things like that, right? So um, you know, it's very, very interesting to me, and I'm definitely going to be following it closely.

SPEAKER_00 21:14

Yeah, it's it's very important. That's why I want to ask another question. I asked another question about it, is because it's a really important I view that as the next level. Once people are are working with AI and they're starting to get used to uh how to get the benefits of it, really having a stronger semantic layer is absolutely critical to be able to um become more efficient, become more integrated with your strategic thinking, and there's just a way of operating that you're doing. Let's talk a little bit about the agency side and agents, a big buzzword out there uh today, but let's talk about from uh a Microsoft ecosystem perspective. I know there's this concept of uh of agents that I think are up in the top left with the copile when you can create you can create the agents and so forth, uh researcher agents.

SPEAKER_01 22:00

or or there's there's finance agents uh with excel and so forth like that um what what's what's sort of like a a simple you know not not I know there's complex ways to do it but what's like a simple example of of uh of an agent that's in Copilot how does it work what is it uh how does it bring value to the organization like what are the benefits of using an agent just using like the normal chat sure yeah um so you think of like researcher you mentioned that one uh and and that's uh a super powerful agent it's it's one of the first ones that we released and um uh oftentimes it's likened to um ph devil PhD level research uh into a certain topic and going and grab grabbing that information from numerous sources right so in for instance I want to do uh you know an overview of of a uh company that you know July 1 is our our new fiscal year you know there's a potential we'll have a little bit of turnover in accounts uh you know and and so I get a new account um the first thing I do is go to researcher and be like tell me everything I need to know about this account and then it goes to all kinds of places I couldn't even list all the places right but you think of corporate filings corporate announcements websites um you know 10Ks 10 q's all of that type of stuff plus all of any internal stuff right uh ongoing projects in flight uh who the former account team was right um and and so aggregating all that information into a nice neat summary uh and then you know I want to go deeper on a specific topic I I I see that um Kagger was down uh 3% year over year um can you can you tell me more about that and it can elaborate further and go deeper and tell me why Kagger was down last year by 3%. So you know super powerful um you know I know a lot of the education research is using tools like that um nowadays because of um its power and capability uh you know and and so um that's just one example you know and and and those are sort of what's built in um to to Copilot 365 or M365 Copilot I always we change the name so frequently I lose track. You know but then there's um the this this this ability and notion of creating your own agent um and you think of it as sort of almost like micro automations. You know uh I'm sure that that a lot of the folks in the audience may have heard of microservices, right? Where instead of having this big monolith server, we're breaking it up into all these little applications to help it run more efficiently and uh you know save some cost hopefully and things like that. Well think of the agents as sort of these just micro automations that you can do. You tell the agent um I want to be able to do um research uh on what the latest AI announcements are from Microsoft. These are the dozen or so websites that Microsoft has that puts all these announcements out. Put together a nice list of everything that was released in the last week you know and then you can put a list together and and then you can save that as an agent and tell it to run you know every Monday at 8 a.mork now we're talking about stringing agents together bringing in other applications right and and so now you're talking about okay well I want to have that that weekly job run where I go out and search those agents um but now I want to automate that process so far as gather them all, uh put them into a customer ready PowerPoint, give me a summary in in an email as to what these are at a high level um and then give me a PowerPoint that I can send out to my customers. Now I still check it because it's it's uh it's still in beta right it's it's still being built um it's still in preview here um but uh you know it puts together a great PowerPoint of of sort of all the top announcements it looks at um all the different areas all the different websites we're talking about and so that's going through if you think about it's going through um an agent that's specific to those websites searching those websites it's an agent for scheduling um the the process to actually go and do that search right it's an agent for building um the the PowerPoint template um you know using a PowerPoint they call skill um which is you know using an application or or some type of uh process that you've defined um and then finally sending that email right is another agent right so um and I'm sure there's more behind the scenes that I'm not even thinking of but uh you know from a simple standpoint um you're talking about four independent agents there that are being brought together to help automate some of my workload right and and there's a lot of those types of things um you think about preparation for the week you think about all of your meetings and and all of the notes that you've captured over the week and and all that um emails and bringing all that together and say okay what are the what are the top 10 things I need to accomplish in the next couple of days here you know and so um really really powerful stuff there.

SPEAKER_00 27:30

Yeah that's excellent so where where my mind went as you were talking about that is okay let's talk about the practical application of that because what you just described makes a lot of sense to a technology professional to a systems thinker right but what about the more technology uh newbie or the non-technical um C-suite leader who just heard a bunch of terms like and things like that um so is it how easy is that for for our listeners to actually put in place can they just go into the Microsoft 365 copilot and say I like I want to um uh have a process I I want to uh research this company who's a new client and um and and um and um share some results about that and also draft a plan for how I would engage with that client is it as simple as that and then and then the Microsoft ecosystem will go figure out what agents to launch or do you have to like go and micromanage sort of your plan that you just like walked through is it like how easy is that for uh a more a a more newbie or a less trained professional to do?

SPEAKER_01 28:40

Sure yeah um you know the researcher agent um it's that straightforward I want to research this company I want you to give me a summary of these things that are most important to me uh you know go and and it'll do that and then if maybe you want some refinement or something um that's the great part about all of these interfaces you can speak plain English right you can type plain English to it you can speak to them right they'll talk back to you now um and then uh you know when we think about some of the the more um uh workflow type automations again relatively straightforward um cowork like I mentioned is is in preview um you know and and only uh customers in the frontier program um can get access today um you know but that's gonna that's gonna come out soon and um what what I find is ask it right ask it hey this is what I'd like to do ask me the questions that I need to answer to help you build this for me right they're really good at telling you what it needs um you know and and we think about refinement on documents and and um you know bringing clarity um being very very specific in the results that you're looking for don't give me the fluff don't tell me how brilliant I am right just just give me the breakdown of what I need to do to make this happen um you know and it can go in a whole lot of different directions you know it can say um you should go build this automation in power automate uh you know and and that'll get you what you need um you know and and if if you know co-work can't do it for you um but oftentimes it can really handle what you're looking for um and if if it's unclear um it'll ask you questions and and what's really kind of becoming more predominant is that instead of even having to retype the question it'll give you um do you want this do you want that do you want that just click whichever one you want right and and then it it goes or or if you say I want all three you know I'll take all three um but you know it it's it's much better about sort of leading you down a path um and the great thing is you can interrupt it right and you can say whoa that's not what I'm trying to do let's let's hold off come back you know um um they do really work um like that conversational mechanism yeah and that does make it very powerful and that's uh that that's an important aspect of it so let's talk about one of the challenges uh that that I've seen out there uh uh talking with companies that are um already already engaged in the Microsoft uh ecosystem with with uh Microsoft 365 etc but not all of their workforce has access to the um m365 copilot right and so then I asked the question well why and they said well we have to justify we have to we need a business case around this so with with the with the amazing power that we're talking about here with AI and how it can really improve things um what what are your thoughts and ideas around uh you know how we get um uh decision makers in the organization to make the mindset shift uh to roll out more of to empower more of the workforce with AI and give them the appropriate um um hand holding education guidance what have you um in order to do that so they don't have to overthink the um $30 a month per person justification right which is less than an hour of most people's time in most companies right so what is that um why do you think there's still hesitancy there and what do we do to get around it to make it more of a uh a more of a um to unlock more of the value of what AI can bring to the workforce honestly one of the biggest ones is naivete um you know they don't use it themselves or they only use it to you know make images of their buddies playing golf right um you know that there's there's definitely a component of that where they just don't appreciate the power because they haven't seen it firsthand um we've also seen um we we had some negative feedback about copilot for a while there right and and um you know it it it got a bad rap for a little bit um but you know all I can say is come back and try it again I think you'll see a very very different experience um you know I I had uh a technology leader uh once liken it to an iPhone with an Apple Watch and you know the the Apple Watch you know it's nice it it gives me the time and it it will give me some notifications and you know you know it's neat gives me some stuff that you know I'd have to pick up my phone for but really I can't live without my phone um you know and and it's moved far beyond just that um it is it is genuinely a time saver uh you know whether it's you know bringing together spreadsheets or um you know bringing together a summary of briefs or you know you you start to see a lot of things like um quarterly um uh earnings sheets and and uh communication being put out with this technology now I I still check everything right I check my sources I verify um before something goes out um but the the the power of the ability to save that much time um you know candidly I'm a I'm a slow reader especially when it's something that you know is technical or or really I need to take in um but if I can summarize it and summarize it by section or however I like it's going to help me understand really where I need to focus right and and things like that are super important and and big times time savers. So I I think first and foremost use it yourself right try it out you know use it in the context of your day-to-day work um are are you you know I mean look we've got CFOs that still love working with spreadsheets and they'll still work on spreadsheets right um instead of having to um you know remind yourself how to build a pivot table tell co-pilot hey I want to build a pivot table with this data go you know and it's gonna do it a heck of a lot quicker than you can do it by hand. Yeah that's uh so yeah so use it yourself start practicing uh with it see where the benefits are for you and then sort of like as a leader extrapolate that to the organization and say wow like if we did if we rolled this out across the organization uh that would provide a lot of benefit yeah you'll see it you will see it if you use it for sure so as we're as we're as we're wrapping up here Chris um if a CEO or senior leader at a company is listening right now and they feel like they're falling behind because they're not adopting enough AI in their organization yet and they're just concerned about uh fear of missing out you know what's the one simple move that they should be making this next week um I would say the the most important thing would be to understand why you can't move forward right are are there are there blockers um from legal right are there are there blockers from IT we're just overwhelmed we don't have time to roll this out right um are are there concerns about security governance uh you know uh there identify what that that holdup is identify what is stopping you and push through that right um this can be deployed at a very very simplistic inexpensive quick win level and it can be also um transformational for an organization you know uh some of the things that we've seen um automations that take place with um you know some of the more advanced agentic workloads are saving companies millions of dollars it's incredible Microsoft's own stories they're all public around how we've dramatically reduced our support cases uh our supply chain our financial closes and and and all of those types of things like I said all public out there um and it's incredible the changes that have been made and the millions of dollars that have been saved as a result yeah there's so much that's what gets me excited and gets me out of bed every day Chris exactly what you just said there's so much value to unlock in organizations when you pair people with the power of AI today and what can what it can really bring. Yeah the hard part for guys like you and I is that there aren't enough hours in the day to play with this stuff right um there's so much cool stuff out there it's just you can only do so much. Exactly exactly and that's why we have these podcasts so we can we can talk about this because you you know you know a lot about uh about Microsoft and the ecosystem and and uh so how can uh as we're wrapping up how can folks contact you or reach out to you uh we can put it in the show notes what's the best way to look at um I'm I'm always available on LinkedIn um you know very very active on there uh you know love to love to sit down have a half hour chat always always willing and open if I can find the time of course um you know and and you're not looking for a job because I can't help with that um I get a lot of those uh but you know um yeah feel free to reach out pick my brain you know I I love talking about this stuff um I I love that sort of uh that that bridge between technology and strategy and and the impact that it can make um so yeah definitely feel free I mean I've got a YouTube channel I got all kinds of stuff up there but um you know one-on-one is the best way to go I feel that's excellent so we'll we'll include that information in the show notes in case people are driving and things like that.

SPEAKER_00 38:18

Appreciate it really appreciate um you your time your insights is all this you know every every uh conversation we have on this podcast very unique and this one we got really into some good practical aspects of the Microsoft um ecosystem and what how it relates to AI and how it can practically apply to companies and I I really appreciate you for sharing that Chris because very um experienced knowledgeable and uh unique perspective on all this so thank you so much.

SPEAKER_01 38:44

No thanks for having me and just have me back again next week because I'm sure we'll see a bunch more changes we'll we'll we'll have you on every week as a recurring guest sounds like a great plan.

SPEAKER_00 38:55

Very thank you for helping us cut through the noise uh Chris and uh really appreciate it for those of you listening appreciate you listening we'll see you on the next one have a great great day thanks all