Microsoft 365 Copilot: What Actually Makes People More Productive with Adrian Espes [MVP]

>> Yeah, well, everybody, to the M665FM podcast, where we explore the people, ideas and technology, sharing the future of work across Microsoft 365, co-pilot, AI, security governance, and modern workplaces. Today, we are focused on a fashion that sounds simple, but it's becoming increasingly important. Do Microsoft 365 co-pilot actually make people more productive? Not in a demo, not in a perfectly preparated keynote scenario, but on the normal Monday morning when you're dealing with meetings, emails, documents, teams, messages, deadlines, and all the other work that fills the day. My guest today is Arion S. Microsoft, MVP, Microsoft 365, co-pilot, consultat, and trainer ed, and Kamina. Arion has spent more than 60 years working across sales, data analytics, and modern workplaces comes out. Let's give him an interesting perspective because he doesn't approach co-pilot freely at technology problems. He focuses business productivity, understanding what really helps people work better, what creates unnecessary complexity, and what sometimes looks impressive, but does deliver much value in the real world. We are going to talk about practical Microsoft 365 co-pilot use cases, adaption, productivity, and so on. What Arion has learned from working with real users of the organization. Arion, welcome to the Amos 365 podcast. Thanks, Merikom. First of all, thank you for inviting me. This is actually my first time speaking English on a podcast. It's my first time because that I'm so sorry if I said different things because now my brain is only thinking Spanish and it's so difficult to meet, thinking another language. Thank you for your welcome. I don't remember the work now. My brain is... Introduction. For introduction. My brain sometimes attacks me. It frees my brain. My brain is freezing sometimes because I know that my brain thinks so faster than my mouth and I speak. I lost the words and I lost the things and I need to breathe and think the sentence or for saying that. Thanks for your welcome, Merikom. Thank you. So, before we get into co-pilot, how did you actually end up working in the Microsoft 365 ecosystem? I don't understand you well. Oh, sorry. Before we get into the co-pilot topic, how did you actually end up working in the Microsoft ecosystem? What was your start? My start in Microsoft 365 ecosystem. I start... I need to think about the years because I don't remember the date, but it's... If I don't prong, I start to work with Microsoft 365 ecosystems, more or less seven, eight years ago. When I was working in a concentric, I was working in a concentric for ten years, more or less, and in this part of my work life. I was making trainings with people. I was making trainings with people and I used so much these tools because Microsoft 365 tools are the most... For me, the most important, the most important tools that we use in work life. Now you can use Google Workspace, you can use LibreOffice and you can use Microsoft 365 ecosystem. And I started seven years ago, more or less, the phytomprong. The year for me runs very fast in the last few years. For me, if I take this another in another... I don't prevent that the word now, sorry. Last, for example, last three years for me is like three or four months, more or less, and seven years for me now it's two years, more or less because when I... I started with... When I started to use Microsoft 365 tools, I changed a lot my role in the companies that I've been working and I use some... I use some math, word excel, the typical tools, and then I'm introduced in my... in my work tools. But what be I? Yeah, but what be I? And then I changed completely my focus from training to be a more commercial profile. And then when I'm done now, I'm a Microsoft 365 consultant in Athena and Microsoft MVP in Microsoft 365's co-pilot. Awesome. Yeah, you have this data on all the tech background, do you that influence how you think about co-pilot and AI? In that moment, no, because I was worked... I worked first time with Microsoft, Power BI in 2021, more or less, and in this year AI doesn't exist. And now I'm more focused in AI, in business workflows, in personal productivity or in collaboration. I don't focus... I don't focus in Power BI and AI in that. In that AI. For me, it's a pending subject because I put... I would like to learn about Microsoft Fabric. For me, Microsoft Fabric is... It's a huge, it's a huge mountain. And it's so difficult for me because I didn't have a technical knowledge in the aspect. I'm not an engineer. I'm... My studies is about economics, business analytics, business analytics for economics and business administration. I didn't study... I don't study... Sorry. I don't study anything about engineering. I know I'm a self-whistler word. Self-employed? Self-employed. Self-learning. I learned by myself, by my own, in this case. All of that I know about Microsoft 365 ecosystem or Microsoft 365 co-pilot is by my own or big or... or that my company in that case, in Camina, gives me the tools or give me the paths for learning about that, about these topics. Awesome. That's cool. What would you say? Where do you think the Microsoft 365 co-pilot actually stands today? So if we ignore the marketing for a moment, how nature is the technology from the perspective of someone who uses it every day? I need to think about that. Please repeat the question because my brain. Today I have a rainy nose. I'm a rainy nose and my brain is full of other things that my brain doesn't think. Well... Yeah. How did you make sure the technology, the co-pilot today, will you say it's really productive, really productive, ready? Is there anything? I don't know compared to either I think Claude or GGPT, where it's still not so good, where it's better, what did you think? It's a... Now I understand the question. Is that tricky? A tricky question because I think that... Now we have different tools, different AI tools. We have perplexity, we have Jimmy and I, we have a GGPT Claude. And I think that... And I think that each AI tool is better for different type of activities. For example, I think that Microsoft co-pilot is... It's for me in my opinion, it's the best AI tool for be more productive, be more efficient in your work life because if you have the Microsoft 365 license, you can ask to your co-pilot about your emails, about your pending tasks or follow-up tasks. And it's more with the word in English, the word in Spanish, I have, I have the word in Spanish in my brain now, but in English it's different. I need to find another word. In that case, you can, you have Microsoft 365 co-pilot in, inside in your tools and be more efficient for me, this is the one of the most important things that Microsoft 365 co-pilot give us. By another way, on the other hand, I think that, for example, Claude, it's better for a developer, Claude is better for developer tasks because Claude have one of the most, or the best important thing, deeper reasoning. For example, in Microsoft, I think that GitHub is a tool of Microsoft in this case, in that case. And for example, in GitHub, you can choose different types of models of AI models. You can choose the Microsoft models, you can choose the Cloud models, the DPT models, you can choose the different type of model for making your, for making your, or for doing your tasks. I think that, I think that, you can select your model instead of, or instead of, no, you can select the model depends on the type of the task. For example, that DPT is more, it's more, more, more conversational. If you need to, for example, if you use that DPT, that DPT, it's more conversional than copilot if you need to make something, something with your documents with your Word document, you can, you need to share your documents with that DPT, telling that DPT to make the changes and then you can need to download the results with Microsoft copilot. You don't need to share the document in that case if you don't have Microsoft 365 license, you need to share the document and in that case, the documents are protected about, are protected with DP in that case. And Gemini, I don't use, I don't use so much Gemini, I can't start now with, with Gemini because I received, I benefit because I'm studying a degree and because I'm studying a degree, I received a license for one month and this month I need to try. I need to make four, compare the outcomes from, say, GPT, copilot and Gemini. For me Gemini is the best model language for create, create pictures. And Nanobanaana 8 is awesome, it's incredible. I love this model of reasoning for making, for making pictures. But now I need to say that if you use Microsoft copilot with Opus model, Opus model language, you can get a great pictures and it depends. So we can say, it's all great models and they are all there, I don't know, right to live, but the Microsoft copilot 365, it's living in the Microsoft ecosystem. So yeah, I think, I don't know, I use more AI foundry, so I'm not so deep that I have Opilot topic, but it's inside how it looks, teams words, PowerPoint. Yeah, you have. You have Microsoft 365 copilot inside of the most important tools. You can use it in Microsoft teams with your own chat, your one and one chat, your group chat, in your meetings with, for example, one of the most important uses inside of Microsoft teams is with the facilitator agent. If you activate, for example, if you're on a facilitator agent, this agent brights for you notes about your meeting. In other use, you can use, for example, copilot in the lateral chat, if you, like if you use copilot in Microsoft X, it's in the right, in the right side. You can ask about all about the conversation, if you, for example, if you, if you have an meeting and if you can relate to this meeting because you have any issue or you have any, any call, an important call from a client, you can ask to, to decide with copilot, and copilot answer you gets your, a summary about the meeting and this chat, it's private for you. For example, if you ask anything to a facilitator agent, this agent shows the outcome for all people because a facilitator agent is like another person, a person inside the meeting. He takes notes if you ask anything to facilitator, facilitator answer you and all of participants or attendees of this meeting can see the question. For example, if you come late to a meeting, I don't, I will, I recommend that don't ask to facilitator agent and ask to copilot, copilot chat in that case, you can use it inside of the tool and this is a great, it's an important, one is a great tool. For me it's one of the most, one of the best tools inside of Microsoft ecosystem. I know the tools for example, recently and I used them, I don't know who's the word, in English, in Spanish is coming, yes, in English I don't, I don't remember now. And it's recently that Microsoft have so many, so many changes in Excel. For example, eight months ago, if you need to use copilot inside of Excel, for me, don't have, I don't receive the outcomes that I hope in that case, I hope in that case. Now for example, this, this part, what changed more or less in December, December, January of this year, Microsoft gets an upload in the product. And now you can get different outcomes, you can get that dashboard, a financial dashboard, you can get different types of outcomes. Yeah, I think that's interesting. I have also the feeling at the start, it's more a clipy, so it's only says, okay, this is this, and then I say, okay, okay, I don't know the power point or the Excel and say, okay, chat you put your chips for me. So yeah, either it started, it was, yeah, yeah, it's more, more marketing than really helpful. What did you think for the employees? Did they have to become now, or prompt engineers or how, well, what separates for starters, a good copilot from a bad one? About prompting, this question is about prompting. Okay, I think that, I think that, we need to have, I would take a good, okay, I need to think about the question because my brain thinks so faster in Spanish and I couldn't think well in English. Okay, in that case, I think that we need to have a good techniques with prompting. We don't, I think that we, first of all, if you have your first, if you try AI at the first time, you don't need to be a prompt engineering because I think that, I think that, and I've read so much about that, your prompting technique improves when you practice, when you practice, when you practice with AI, you need to try, to try different types of different ways, just speak about, about, no, just speak with AI, you need to try different ways. If you need a general outcome, you can be more general, you can say, for example, hey, bye-bye, tell me, tell me, who is the weather today if I have also? Here, outcome was, her outcome, in this case, will be more general. If you need, and in, I need to start with this part because I'm confused. Okay, in that case, prompting, it's, I don't remember the words now, this is frustrating for me, yeah, this is frustrating. Okay, in that case, I think that prompting, it's, we need to explore our own techniques. For example, when I start, when I started speaking with co-pilot, and I start to speak, I just start to speak with co-pilot more or less three years ago when he appears in our lives, and I be more general when I speak with it. For example, please tell me who is the weather today, tell me who is the meaning for this, for this world. Prompting techniques for improved prompting techniques, you need to explore different ways, just in terms of ways to speak with AI. Sometimes, if you, for example, I use a technique, and in that case, for example, now when I have more, I be more, I don't think man, but the words, sorry, my, my, it's, it's, it's right in, it's so frustrating because I have the words in my brain, I couldn't speak. And then you repeat the question, please. Yeah, I, I, I, I start again. We think about the, the, the, what makes a co-pilot a good or a bad prompt, so if the people need to be a prompt engineer or, or not, what tips can you give to, to, to, to, to, Okay, work with, with our company. Okay, first of all, didn't need to be, uh, apparent the junior, uh, for speak with, to speak with AI. In that case, you need, you need to try, you need to try to speak by, uh, for, by different ways, for example, uh, first of all, our prompts, be more, um, generals are general, the, our prompts, the first prop, our first, first, prompt, be more, um, generals. And if you, for B of our obtain more efficiency with the outcomes with AI, you need to get more information. If you set, for example, if you set to AI, uh, how's the weather in Tharawatha, um, co-pilot or TFT or GB9 or another AI tool, um, answer view, um, in a general way, for example, in Tharawatha, there are, um, the weather in Tharawatha is warmer because we have 34 degrees or less and some wind is a general, uh, a general risk outcome. In that case, uh, for obtain or forget more efficient or better outcomes, you need to introduce different, uh, more information, for example, you need to, when just speak with AI, first of all, you need to be, you need to be clear, the goal, the main goal, you need to be clear the main goal because, um, the main goal, um, but this isn't the old information, you need to, you need to add, um, different parts. If you, if you, uh, if you have, if you have, uh, if you have clear, uh, if you have your main goal, um, it's not all, you need to add, for example, a role, but if you read, uh, Microsoft documentation, now, a Microsoft said, uh, that it's, is not necessary to get to the AI a role because with your main goal or the context or objectives, um, co-pilot in that case, uh, assumes the role, but in my case, I set, for example, to co-pilot, the, uh, a role. If I make an, uh, financial analysis, I get, do you be a financial analyst and need to be, blah, blah, blah. But, um, I, I know that, um, when just start to speak with AI, could be first, uh, could be frustrating, could be frustrating because, um, you don't have the, you don't have the tools for making, uh, for, for make, uh, the prompts well. You need to practice, you need to try. And, uh, finally, the, the most important thing is for me is that is, uh, we don't have a, uh, B-ball or a book for the best prompting techniques. You can try, you can make your own prompting techniques. And this is, I think that, um, about this question, I don't know if I said anything, I had something else. Yeah, that's, that's, that's interesting. Um, I think from, from this perspective, co-pilot do, do, do a lot, but we have these skills, tools, companies or organizations or departments, I said department said, let's say sales or marketing, show they do our visit, do that help when they have an expert and they do skills for them. I, I say, okay, we communicate with our clients and doesn't this, uh, language, you can do this, this, this material, yes, our source material. This is the farmer and so on. Show, show, we do it. Can us help or does it more brings more complicated into your process? With different areas, for example, um, when I, for example, when I teach to these people, to these people for, for making their own prompts or, um, it's a question. Uh, in that case, for example, I, um, I, I personalized my, when I, um, okay, I start, uh, when I participate in, uh, in an adoption plan with a company, for example, in a company, I, I make it, I make this, um, this other time, because I'm a consultant that I need to teach. Uh, in that case, I personalized, um, personalized, uh, examples. Mm-hmm. Isn't, uh, isn't the same when you speak, uh, like, for example, uh, data analyst, uh, data analyst in financial, uh, department or, for example, or, uh, uh, uh, a marketing assistant. The goals, the main goals are different. For example, for a marketing assistant, uh, will be more important, uh, is more important, the outcome for the outcome or the, uh, the outcomes, the, the pictures, the, the creative, the creative materials. And for, um, data analyst in financial sector is more, is more important. Yeah. Uh, it's more important. The, um, I don't have them for financial analyst is more important. The, uh, real situations, know the real situations. Sometimes, uh, and we, and it's known, uh, for all AI, uh, could be, have some hallucination, uh, some hallucinations. And for, uh, try to be, uh, try to, uh, I don't give you these hallucinations, do you be, uh, strict in your prompt? For example, I use, uh, sentence, uh, I use a sentence in all my prompts when I need to be more, uh, accurate with the outcomes. Uh, you don't, uh, if you don't know the information, don't, uh, don't lie, don't lie me, don't give me information if you don't, if you don't have the, um, the data. Um-hum. I don't, uh, in Spanish, I write different words, I have different sentence, but now I know, I, I, I'm not remember one of the words, I'd say it's the expression, but, uh, for different areas, you need to be, you need to have a personalization with prompts or with your techniques or, um, you can, you can, uh, explain the same to all areas because the main goal for different areas, the four, the areas is different or are different in that case. Okay. Um, what was it you think we have to do this, does one part, only we have the other part, say, uh, governance and so on, Pewview sensitivity labels. Um, how much should the, the user work on, on, on the quality of the data, co-pilot use and how many, it's, I say, uh, admins, IT departments, uh, yeah, um, governments of, of data that the users share with co-pilot. In that case, okay, in that case, and Microsoft co-pilot when we, when we, um, intelligent Microsoft co-pilot in, in our company, IT departments and AI offices, uh, needs to, to review their, um, their, uh, their, uh, their environment, their, uh, Microsoft environment. For example, when I start, um, an adoption project, I have a meeting with the, uh, with their, um, IT department and we review all of the, uh, configurations about Microsoft co-pilot. First of all, um, Microsoft co-pilot, um, don't, um, don't share the information that I upload to, to the chat with other people. In that case, um, we have in, when we use, Microsoft co-pilot with an enterprise account, we are protected, uh, with a Microsoft EDP enterprise data protection. Uh, and this enterprise data protection, um, give us the security and the information that I share with, uh, co-pilot, don't consider with another people in our company or in other companies or in Microsoft. And this information don't be used for, uh, training for training or for, for the training, um, in AI models. If, for ex-prolifies share, the financial, financial, information, by, by, by, um, my financial information, for example, with a Microsoft co-pilot, uh, chat. In that case, Microsoft co-pilot, the, uh, Microsoft don't share this information with anyone, with anyone, and this information don't be used for training, uh, for the training, uh, for training the models. Mm-hmm. For example, this is different in chatchipity or Gemini. If you, uh, if you have the need to share this type of information in Gemini or ZDPD, I recommend to, uh, turn off one of computer relations in that tool, because if you don't know when you share some information about you, some, some personal information about you or about your companies, Ed's a tipit or Gemini, could be shared this information with other, uh, organizations, not directly, but, uh, these tools can use our information for training, their models. And in, uh, in direct way, we can share our information with other people and could be, uh, uh, safety, uh, safety issue, uh, safety problem. And for example, when I talk about companies, uh, when I start to talk about companies, we introduce AI in their work lives, I told, I told, uh, all of them, you don't worry, you don't worry, um, you don't worry, Microsoft Copilot is a secure tool because Microsoft Copilot is inside of Microsoft 365 ecosystem and all of the Microsoft products if you use an enterprise, uh, in that case, because I, I talked with companies, uh, when you use your enterprise account, you're reprotected. All of you, all of your information is protected, are protected in that case. Awesome. Um, well, what, what, what did you think from your perspective? You have done a lot of, implementation projects. So, and I say, it's, it's not so hard to buy Copilot license and, uh, what should an enrollment look like? 30, 60 and 90 days from your perspective? I don't understand you. Uh, that question. Yeah, I think, um, it's not a hard part to buy a copilot license and the, the implementation is, it's the hard part. So what, it's a good road map for the first 30, 60 and 90 days from your perspective to implement, uh, we can implement for implement, uh, Microsoft 365 license. Uh, first of all, when we think about implementing Microsoft 365, uh, copilot in our organization, we need to check all of the, uh, settings in our Microsoft 365 environmental, we need to check, uh, they use, uh, they will be used, you need to check, uh, already you, um, they wait to get a license because if you, uh, the Microsoft Microsoft 365 license, uh, if, uh, no way days, no way days, no, uh, no way days, Microsoft, uh, I start the game, um, my tongue was stressed, um, Microsoft 365 license have accost around, uh, last, uh, last time that that I've seen, it's 20, 20, 28 euros, more or less, at 30, 30 dollars, 30 dollars, 28 euros. And if, in that case, uh, this make necessary to have a road map to implement license, the, the license mode, for example, first of all, we need to think about, uh, in our company, we need, uh, we need to think about, we need Microsoft 365 license, or it's, um, with, uh, basic license, uh, I have all that I need, I need it. Because, uh, because, uh, because I said that, uh, if you have a Microsoft 365 license, and enterprise license and business license, our team's license, you have access to, you could be have, uh, you could, uh, you, you have access to Microsoft 365, set in a basic way. The basic way is a free way. You can use the Microsoft 365, uh, you can use the AI tool from Microsoft, you can use it in a safety way, you can share, document them. Most important thinking that case, is you need to use your enterprise account. You need to, you need to make logging with your enterprise account. And, uh, if you have, think about that, think about, I need Microsoft 365 license, or with the license that I have is, um, I, is enough. I don't, I don't remember the word enough. Um, you need, you need, first of all, you need to think about that. I need Microsoft 365 license, or we have my license, Microsoft license is enough. And my recommendation is try the Microsoft 365 Copilot chat, the free version, the free version. And if you need, uh, to use a Microsoft 365 with, uh, your emails, your meetings, your documents, you need to, uh, you need to make, uh, you need to make documents from the scratch, you need to think about, uh, the Microsoft 365 license will be necessary. First of all, we have, uh, if you have a Microsoft 365 license, you have, um, a very useful AI tool without need to pay more, you know, you don't need to pay more for, for use in, in that case, Microsoft 365 Copilot in that case, uh, you can use it without, without limits. You can, you, you can use it without limits. This is the first, the first step in my recommend roadmap. First of all, think about, we need license or don't need license, uh, the Microsoft license is enough. And the second step for me is checked as some, it's taking some settings in Microsoft environmental. One of them is the, uh, we need to review, for example, in, in, in Europe, we need to review the use of, uh, anthropic models, because now anthropic models, all of their use is a, in servers on the USA, and we have the GRPD, uh, a, a, a, a, a, GPD, no, g, RPD, in Spanish is different because we change, we change all the words. And, uh, we have the GRPD, and we need to review the terms of, uh, terms of, uh, the terms of use, the conditions of use. And, uh, if, if someone, uh, hear that, it's pro that, uh, he or she think, why, I've been said that because, uh, now, um, anthropic models, uh, if you use, and you can use anthropic models, or in Microsoft compiler to have the possibility to use anthropic models, um, um, inside of them, Microsoft 365 product, it's, uh, a model similar, um, than, for example, that, uh, GPD 5.6, but in that case, uh, companies, we need to, uh, we need to know that, if you, uh, turn on these models, your information, uh, don't move to the USA, for example, don't move to the USA, but, uh, Microsoft needs to move the, the rationing to USA servers. And, for example, for, for European companies, could be, uh, important because, uh, if you have a high sens, uh, sensibility to, uh, GPD, you can use it because, uh, one of, one part of my information goes to the USA for getting, um, the outcome and comes again to, to Europe. But I think that, I think that, uh, that question could be changed in a few more, uh, in, uh, uh, next month, I think that, I don't know. But this is another, uh, for me, this is, uh, in Europe, this is one of the most important, uh, settings that I need to review. We have different, uh, another settings you can, uh, turn on, turn, turn on or off the, uh, web generation, video generation, uh, you can use web information for getting, um, outcomes. But I think that is the, for me, this is the most, um, the most important steps. In the future, Microsoft are changing the, uh, the, the, the, the, the settings continuously. For example, um, one month ago, uh, Microsoft activated a different, uh, activated or so, uh, a different settings about, um, the new TPP models, for example. I don't, I don't remember the, the, the, the specific, the specific name of the Serian, but, um, now you can choose that if you, um, get the new models or the new models of TPP, for example, uh, now or in a few weeks or for some people or nothing in the company can use this type, uh, this models, it, it's, it's tricky, uh, and in that case, um, any different company, half and one, uh, any company have their own, uh, criteria, so they own, uh, nilics. Awesome. Um, what, what, what, what, what, what, yeah, I also think about, but what, what, what, what, what, what would you say to productivity, um, and I said, it's a illusion of, of productivity. How do we know if co-partners are actually making someone more productors? Woof, uh, this is, um, a very difficult question because for me, um, because I, I don't like to speak about productivity, because productivity is, uh, subjective, sometimes it's not subjective, uh, is subjective because for, for example, for me, uh, for me, be more productive, uh, means, for example, make, uh, PowerPoint, uh, in a quicker way, more, more faster than I, than I make by myself. Uh, for example, in, uh, it's difficult because, uh, different areas or, in, inside of the area's different roles, make different, uh, tasks, have different, uh, have different tasks, uh, tasks. And that is, where to know, uh, now, for example, when I speak with companies, uh, some companies, uh, told you, we need to be more productive. Okay. But who is productive for you? First of all, who is the, who is productive for you? For me, could be one thing for you, could be another thing. In that case, I changed the word, now I'm changing the word productivity for, uh, naturalizing the use of AI. We need to, we need to know that, uh, we need to know how, uh, we use AI in our daily tasks. For example, for making, uh, data analysis, you can use, for now you can use copilot chat and 365 copilot, eco piloting excel, you can use, uh, AI tools in different ways. And who is, but in that case, I need, I try to change the word productivity to naturalize. We need, now, we need to naturalize the use of AI. Um, I see, recently, I, um, I saw that some companies, uh, um, have, uh, an expectation with, with AI and things and first of all, things that AI, uh, can, can get in the, uh, uh, the fire, can fire them. And in that case, we, we need to, uh, we need to have a different, uh, a different message. Uh, we don't speak about productivity because who is productivity for you? For me, it's one thing for you, Mirko could be different productivity on, for example, another colleague, uh, some, um, for another person could be different. In that case, we need to be, change the word productivity for normalize the use of AI. We need to be more efficient. In that case, we can use efficiency, uh, efficiency word because if I'm, if I'm more, more efficient, uh, efficient, uh, this, this word means, uh, I have the tools to make the same, for example, in less time or in a higher quality. And in that case, the, uh, we, we get off the negative, negative meaning of productivity because when you speak with some, uh, with someone about productivity, first of all, in their brain, it's probably, it's possible that this, uh, this person think, uh, Adrian asked me about productivity because they is thinking about fire me, fire me, fire, fire, fire, fire, fire. If you talk about efficiency, efficiency, it's, I think that efficiency is, uh, one of the, uh, one, uh, efficiency is one of the, for me, it's one of the most important skills. It's, it isn't a skill, but I don't, um, find the good word for, for, explain that, um, I think that if, if we, more, or if, if we improve our efficiency for our companies, um, for example, the, invest, the investment in Microsoft 365 for pilot, have a big, uh, a big, uh, refound, no, uh, return, return of investment. We can have a higher return, a row, return of investment. Efficiency is, um, I know the word, I don't know who, um, efficiency is more, it's for me, a better word, down, productive, it's more, it's more human. It's more human, um, productive. Productivity is subjective. It's a subjective, um, a skill is a subjective, objective, objective, skill, um, is also a subjective. And, um, yeah, I know it's more human. Good to have companies have a human resource or a people, uh, department. Yeah. Yeah, I know it. Um, so, um, yeah, I have on every, um, every, part of the rapid fire route, but I do it to this time a little bit different. Uh, I ask you some questions and you have three possible answers. Uh, so the first one is hype. The other is useful today or future potential, as an answer. So, uh, co-pilot will animate most meetings. Most meetings. Yeah. Can it eliminate the most meetings? Ah, can element a, can you repeat the three options? Well, co-pilot eliminate most meetings and is it a hype useful today or a future potential? I think that it's now, it's a hype because I'm, uh, I'm being, um, uneducated. I have, um, a lot of, uh, meetings that can be, uh, could be, uh, could be, uh, email for email, for example. Yeah. Now it's a hype for me. Now it's a hype. Microsoft, co-pilot, don't reduce now the, uh, no, meetings. I can't, can't reduce it. And, uh, what did you think agents will replace traditional business applications? Agent could be the same. Yeah. Yeah. Uh, co-pilot agents will they replace the traditional business applications? Well, uh, it's a tricky, is it a tricky question. Uh, I think that, no, definitely no, because agents are, our agents are agents are some new, I'm so new because they are some new and a B, uh, and in that case, we win, for example, in Microsoft, we need to, be, uh, we need to have a more, uh, security tools for agents now. Recently, we have agent 365, uh, 365, but a company, uh, some companies or the most part of companies demand, most security tools for get, uh, for get control, the, the, the AI agents. I, and I, I, and I said to you, for example, I have, I'm now for, for me, my, for myself in my profile, in my enterprise profile, uh, I have more or less 2000, not 2000, not so much 200 agents. I have, I have now 200 agents in my enterprise profile. And I know that age, a, AI agents could be replaced, couldn't be replaced now, business applications, business applications are necessary. We need, uh, now we need power platform for all of our platform, power BI, power apps, but in that case, AI agents, uh, could be, um, could be help us with, um, tasks in that, in that type of tools, in that type of tools or in that programs. But, uh, now, definitely, uh, AI agents couldn't, couldn't be replaced, business apps, business applications for me. Another, another colleague of, uh, another colleagues could be, uh, could be think, could be think different about me. Um, yeah, uh, then, um, did you think, uh, co-pilot become, or is co-pilot also useful for front, uh, for front, uh, for front like workers? It's all that we're just, uh, I, I, I think that depends of the tasks that this type of worker, um, has. For example, this works, see, if I, if I worker have so much emails, for example, co-pilot could be useful. But, for, uh, thinking, uh, thinking about licensed, uh, Microsoft, uh, Microsoft license, from the front line license, have, um, have different settings than, for example, uniterpised license, uh, data, they don't have, uh, so much, um, uh, uh, mailbox, um, I, I don't remember the word. Uh, I hate my brain in, in that, in that moment, in, in this moment. Um, okay, the question is, the, the useful of Microsoft, uh, Microsoft co-pilot, in front like workers. I think that in, in that moment, um, co-pilot could be useful, depends on the tasks that this worker has. If that, if, if this worker has so much emails, or sorry, yeah, so, uh, a lot of emails, um, could be useful because co-pilot can, uh, summarize, summarize them, and get, uh, the principal tasks, or the principal, or the highlights for this, uh, for these emails. But, uh, I, um, honestly, honestly, uh, now co-pilot isn't useful for this type of, uh, of workers. Because they don't have so much work in Microsoft tools. We, uh, they have some, some, some workers, uh, have this type of license for getting, getting connected in Microsoft Teams or email, but they can use it. Yeah, cool. So, yeah, then I'd say, thank you for, for, for the overview. And, uh, what I do, it's, uh, when I have to speak another language, uh, um, or another dialect, uh, I bought the beer from, from, from, from this country, like, like, Ireland, I, uh, by, uh, Guinness or two more, uh, or, uh, when I have, like, Spaniel, I, I buy a strilla dame, uh, it's so on. So, uh, sounds, the beer, the beer is three and it's the yellow, the beer. Yeah, and it's making also the brain a little bit more thinking slower. Yeah, but, but, yeah, thank you for, for, for being here. This was an awesome session. And I say, thank you so much for, for sharing all your knowledge and, uh, also for, uh, for, uh, yeah, for coming, there also, you are not a natural speaker. I'm also not so, you're also, you bill, more, uh, but I think it makes it also more, uh, sympathetic and I think that, that's, uh, we have talked about the, the humanizing and I think, uh, that's also, uh, top, uh, my, all the technology we don't have to, to forget the human, uh, behind the, the AI hype. And, uh, so, yeah, uh, thank you so much for paying you, will need. Thank you. Thank you, thank you, Mirko. I said, I said to you out at first, it's my first time, just speaking English without, uh, without English professor. I speak English with my tutors, my English tutors, but it's the first time, uh, that I make this type of, uh, have this type of conversations, uh, conversation in English. Uh, thank you for having me today. I really enjoyed with this, um, with this meeting and I hope that you, uh, that you, uh, that you found this information useful. Yeah. And it's a pleasure for me. Yeah. Thank you. And, uh, yeah, not, nice to be in yet. So I say goodbye for this. Goodbye. [BLANK_AUDIO]