Integrating AI Into Your Workday

Mitch (00:06):

We are in a good mood. We're here, we're ready. We're going to record an episode all about copilot and Microsoft 365 today. Welcome back to Make Other Successful, a podcast where we share insights, stories, and strategies all about building a better workplace. My name is Mitch. I do a lot of operations around here. I'm joined with Matt Dressel. Emma, they have been on a co-pilot kick. Oh yeah. And have been talking about it a lot. So this is sort of like, Hey, we need to get it all into a podcast, get it into our rotation and share it while it's sort of this new thing that people want to hear about. There have been a lot of questions, good interaction from the community, and so we're hoping that this conversation helps kind of further articulate our point of view and how we're using copilot and things like that.

(01:03):

If you have been to our webinar and are hoping that this content is different, you're in luck. We're taking sort of a different angle on some of the content there to try to keep a little bit of variety, but there's a couple parts to this podcast. We're going to try to keep 'em articulate and tight. Four parts. The first part is the essence of copilot. So what is it at its core? What is its impact on productivity? What is the learning curve and what is sort of the future of work with AI in our eyes? And we'll see where things go after that. But let's just start with the essence of copilot. What is the core idea behind copilot? If someone is maybe familiar with some of the AI tools today, how is it different? Give me the lowdown of, I don't know at all what copilot is. Tell me in three sentences or less.

Matt (02:03):

We'll try to do it in three sentences or less. Copilot as a term is a branding that Microsoft is using for lots of tools within their product suite that is powered by ai. So they have lots of things called copilot. The one that we have spent the most time talking about and what I think we're going to talk about mostly today is Microsoft 365 copilot, which is the AI tools that are an add-on product for Microsoft 365. So word, PowerPoint, SharePoint, all of these tools that they have and all of the content that you have within Microsoft 365.

Emma (02:41):

And I'd add on to that, that copilot is an interesting term and like you were saying, sort of branding identity for this because the phrase I'm liking that I keep hearing is it's copilot not autopilot. And that's an easy visual for you to understand how you are going to interact with the tool. It's literally like a tool that is a copilot, flying the plane with you, not necessarily flying it and you're sitting in the backseat. So that's sort of a visual.

Mitch (03:08):

So is it just like, let's say someone doesn't know what chat GPT is, I wanted to say is it like chat GPT inside a tool? But what, give me a little bit more at its core, what is it helping me with?

Matt (03:24):

I mean, it basically knows how to process large amounts of data, text data primarily or pretty exclusively, and then allows you to ask questions about that data or reinterpret some of that data and recreate some of that data based on your needs. So chat, GPT as you mentioned, the reason it's called chat, GPT is its primary interface, is a chat bot where you ask it questions and it does stuff. Copilot is very similar in what it can do. You can ask questions and ask it to do things and it will go out and try to do things, but it is the integration with all of the tools that is the differentiation, the important piece of it. And the other piece of it that is unique is the data that it has access to chat, GPT and a lot of these other, what they call language models or chatbots, AI tools, leverage public data, a lot of public data,

Mitch (04:23):

Anything on the web web,

Matt (04:24):

Anything on the web, anything that you put into it, that's what it's consuming. It's

Mitch (04:29):

Like a brain that is constantly learning, it's

Matt (04:30):

Constantly looking at what's out there on the web and is publicly available and try to bring that into its model. Whereas copilot and specifically Microsoft 365 copilot is grounded or based on the data that is in your Microsoft 365 tenant. So it is your private data, and so it opens up lots of doors that you didn't have available with something like Chat, GPT or any of the other AI tools.

Mitch (04:57):

That reminds me, there have been security concerns in the past where people use chat GPT to, it's been particularly helpful with what's wrong with this code or how do I make this more efficient or help me write this code. And I think someone from Samsung posted code into it, and that actually came to light somewhere where people got what was supposed to be internal code and it was effectively a security issue.

Matt (05:26):

Anytime you put something into an AI tool, the question is what are you doing with that data? In Chet's case, in a lot of these tools cases, they're just using that as another source of their model. They're building their model, and that model is a public model, which means that everyone else in the world is using it. So if I take my brand new book that I plan to sell for millions and millions of dollars and I put it into chat GT and ask it to rewrite some of it, chat GT now has my book, and somebody can ask for that book and they're going to get the book back or they're going to get a summary of the book or they're going to get all of those things. If you transpose that into business, if I take my brand new widget that I'm planning on doing or my patent authorization or my whatever, and take some of that content and put it into chat GPT, in the case of Samsung taking a sensitive security piece of code that deals with the security of my smartphone and put it into Jet GPT, other people can ask for that and get it back out.

Mitch (06:26):

Not good.

Emma (06:27):

But the key thing to note here is that copilot is different.

Matt (06:31):

Copilot is different. Copilot everything, both the data that it's grounded in and the information that you put into it are protected. That's your data. It's not going to be put into or integrated into the model,

Mitch (06:44):

Right? Yeah,

Matt (06:46):

The public model.

Mitch (06:47):

And so from what I understand, Microsoft partnered with OpenAI who created chat GBT to pull in some of that, the brain into Microsoft so that people can use that brain. Yeah.

Matt (07:04):

There's two different pieces to ai. One is the core technology, which is all about how does a language model get built, how does it process this data and use this data to be able to respond to these questions that are being asked. And then there's the model itself. You can, by taking the tooling behind it, you can apply and insert a model from all over the place to get better results. chatt knows a lot of things about a lot of things, but if you want it to be specific, you can train a model based on a smaller set of data and it would be more accurate in its results. And what Microsoft did was it licensed the technology about how to create those models, manage those models, maintain those models from chat GPT, and then basically used that and integrated that into all of their tools and also allowed it to target its model based on private data and some public data. Kind of a mix of all.

Mitch (08:00):

And I think I remember you're definitely a fan of the private aspect of co-pilot because when we wrote our guidebook, you didn't let me put any of it into chat GPT, and that was, I

Matt (08:12):

Tried at least.

Mitch (08:13):

So there's not really the whole thing out there unless you buy it, check out the links in the description if you're interested. Shameless plug. Okay. So we talked quite a bit about the thing as a whole. Can we dig into, okay, what has your individual experiences been like using copilot and how has it affected how you

Matt (08:37):

Work? I mean, I can talk to start. For me, the biggest impact has been what other people around me are using it for. Emma uses it heavily for meetings, which she'll get into in a minute, but I'm seeing the output for that. When I have to respond to some customers, a lot of times I will ask it to try to help me summarize an SOW that we might be sending, or I might ask it to summarize or take an email that I've already produced in the past and kind of recreate it for a new customer as an example. But for me, I wouldn't say it's something I use daily. I would say the biggest reason I don't use it daily is because I am often in customer environments where I don't have access to copilot. Honestly, the areas that I miss it the most is when I'm working in an environment, a customer environment that I'm heavily involved in a project for that customer and they don't have it. I miss it. I miss being able to go back and ask for a summary of a meeting or to ask it where this piece of data is, or to summarize a huge thread and give me some feedback about what it's about so I don't have to read through all of it. So that's how it's impacted me most recently or relevantly. Sure.

Emma (09:46):

Yeah. And I guess we should set the stage of we've only been using copilot within our business. I really have only been using it for maybe two months, maybe six weeks. So it's interesting that you and I are already feeling like we miss it when we forgot to turn on a transcription or we don't have it available on a meeting. You start to get used to the power of it. So I would say for my personal use of it, I use it for a lot of meetings and this kind of bleeds into our next topic on just the productivity side of it. But I also context switch a lot in my job as a project manager. So I'm constantly going from project to project to project, and sometimes you need to reset your brain a little of, okay, now what am I doing? And what I find really helpful is you can input a bunch of mess and it will give you nice organized articulate yes, detailed thoughts back.

(10:39):

So if I'm going from sending a bunch of different meetings and then I need to actually write some emails, I can just write into Outlook, the draft with copilot, what I'm trying to do, what I'm trying to get out of this, and it will give me something that I can then edit, polish up and send out, versus having to create the entire thing from scratch. So sometimes I can put down super messy thoughts where my brain's at. I've just got a lot floating around a lot of different projects, and it will help me organize my thoughts. So I really enjoy that part of

Mitch (11:09):

It. Nice. Yeah, that is a good segue. I want to end cap this section of the essence of copilot. I think we've articulated ai, it's this helpful model, this thing that you can chat with and interact with and use against data that you have in your Microsoft tenant and really leverage in a way that works well for you. It is a little bit greenfield in some aspects in ways that you can apply it to help you in your work, but let's talk about how it's impacted things like productivity. So Emma, you were just talking about context switching and things like that. How would you say it has changed the actual way that you approach your workday? Maybe not a specific task, but do you think about approaching things in a different way? Now?

Emma (12:02):

For me, one of the biggest parts or freedoms I would say that I feel like the tool has given me is the ability to feel present at more points during my day because I'm less worried that I'm going to miss or forget something. Again, I'm in a lot of meetings, so this is really true of when you're going, let's say you have six hours of meetings that day, there is a lot of stress of am I going to remember everything I'm supposed to do from that, from all those meetings? Probably not. And so having copilot on means I can kind of rest easy that if I need to go back and ask it questions, I'll be able to find the answers relatively quickly versus having to watch six hours of meeting recordings back or something like that.

Mitch (12:42):

Or yeah, when transcription came out and it was a great feature, but it's only as helpful as how much time you have to go through

Emma (12:50):

Review

Mitch (12:50):

And read through. And so for someone, it's basically you're asking someone else to go read the thing and tell you a synopsis, which is super helpful. So you get to be more present.

Emma (13:01):

Yeah, it definitely allows me to feel like I can be, especially if you're a facilitator of meetings or you're trying to also capture notes and be a part of the conversation, you can just be a part of the conversation and know that the notes are going to be taken and then afterwards you can go back and we kind of use the word interrogate. I don't know if that's the right term. Matt used it the transcript a little bit, and it probably is kind of how a lot of project managers feel when people are like, what was the summary? What was the, and you're asking that of the transcript. So I think for me, it's created a more present mindset at work, which I don't even know if I can put a value on. It's been really great.

Mitch (13:40):

Sweet. $30 a month, Matt, has it changed any of your perspective of if you show up to work, is your mind in any different place knowing that you have copilot?

Matt (13:52):

I don't feel as bad responding to people's questions or giving feedback on people's ideas, knowing that they can easily take what I'm saying and put it into copilot and do something different. I think what we're doing right now on this webinar or this podcast is a good example of that. We had an idea, you created a list. I was like, Hey, I feel like it's kind of the same as what we're doing for the webinar. What should we do? You kind of sent it through and said, Hey, what are some other ideas?

Mitch (14:22):

Give me a different

Matt (14:23):

Angle. When I've done it personally within my blog writing and some of the other things, I think that also has been very helpful. It really takes a little bit of the pressure off coming up with alternate ideas. I don't always like what it does. It's not always the right thing or even in the ballpark, but it gives you a really quick, easy way to get an alternate idea and get the juices flowing without having to go into a meeting and let's all talk about it. Does anybody have any ideas or have me try to explain what I'm trying to say or it is changed the way that I think about that process.

Emma (15:01):

And this may resonate with some people, but I tend to enjoy editing and working off of something that's already created. That's just, it's easier for me, especially when I'm trying to do, again, context switching of this to that, to that, that. So if I have to sit down and create a lot of things completely new, and I know we've got other creatives on our team who they live and breathe creating content, I'm definitely someone who is more of the editing mindset. I know this is true of people when you write resumes, it's really hard to write your own resume, but it's really easy to edit other

Mitch (15:34):

Or a biography

Emma (15:36):

Or a biography or whatever. So in some ways it is sort of, what's the word? I think a lot of us have that it is kind of a generic quality. A

Mitch (15:44):

White empty page can be really

Emma (15:46):

Intimidating. It can be hard, or at least it can just be draining. I mean, we can all probably do it, but it takes more energy. So if you're able to throw, just, I was telling Matt here before we started, I usually put 10 word, 20 word prompts in. If I'm trying to get it to write me an email, just something super quick, super messy, it usually doesn't spit out the best, but then I edit from there and for whatever reason, that is more motivating and much easier for me to start with than just a blank

Mitch (16:14):

White

Emma (16:15):

Piece of paper.

Mitch (16:16):

And I feel like, tell me if I'm wrong, but most of the time we know we are not relying on copilot to give us the idea. We're asking it to put a framework around an idea,

Emma (16:29):

Right? Yes. Yeah. Organize the idea in a actionable,

Mitch (16:34):

So it's not necessarily inventing the here's the next 10 blogs for you, right? It's like, no, I have know kind of what I want to talk about. Can you give me a structure that I can work within? So it's not doing everything from start to finish. It's like playing this beginning 10, 20% role. Yeah,

Emma (16:55):

I mean, I don't know. I was the person when word came out with the thesaurus feature and you could right click and choose a different word. I loved that feature because it gives you, like you were saying, more options, more ways of saying things, and this just takes basically thesaurus and does it to everything.

Matt (17:11):

Well, I think the thing to think about when you think about what it's actually doing is it's taking all of the ideas that you have and marrying it with its understanding of language to create content that maybe combines things. So maybe it takes a research paper and turns it into something that is more creative in the way it's written and structured, right? It's taking something that was maybe a presentation for an executive team and turning it more into a marketing presentation for an end customer. And it doesn't know about the ideas that you're doing are the creativity that you're creating, but it does know how to transform that in the context that you're asking it to do it, and that's the big difference. Whereas the source, it's a very similar metaphor and it's a good metaphor, but it is limited to just a word and to just the context that it has of that one sentence. This expands so much higher. But I think it's a very interesting topic because most people talk about and think about AI as being something brand new, the source grammar, checking, autocorrect,

Mitch (18:21):

Reword this remark,

Matt (18:23):

They're all the precursors to what we're talking about now. Copilot chat, GBT, they are way more advanced, can do way more. They're based on a different type of technology, but it's all similar stuff,

Mitch (18:38):

Heavy language, word based, which I read someone say, these tools don't actually understand language. They understand how to put words in a certain order. That makes sense,

Matt (18:51):

Correct. That they've seen a lot that everybody else talks. Yeah. That's why what it's grounded in matters because it's basically using that to say, oh, somebody categorized this book as a creative book and you're asking for a creative thing. I'm going to use it to write like this. Right.

Mitch (19:06):

So you guys obviously copilot in a lot of the pros. We've talked about a lot of the pros in the workplace. Have we experienced any cons or seen any cons? What are some of the maybe not so good parts of it? For

Matt (19:21):

Me, it is not seamless. I have to remember and choose and think about how I want to do that and to use it effectively. It is there in a lot of different places, but unless you click the button and unless you construct your statement in such a way that it will actually produce something like what you want, it's not going to work, which a lot of that is training and repetition and using it and choosing to use it. And some of it is there ways that they could make it a little bit easier, a little bit more integrated, a little bit better responses?

Emma (19:58):

We've both experienced times where we think we're being pretty clear with the file that we wanted to use or pull information from. I mean, almost to the, you're almost putting the exact title in and it's not finding it, and then you have to actually go find the file, use the exact title, paste it in and say, okay, use this. And so it's those little moments where just anytime that it takes a little bit more time than what you would do yourself. Yeah, what you would do yourself. That's where I think a lot of the inefficiency of the frustration comes from, but it's usually outweighed with the value you can get out of it. But those times,

Mitch (20:33):

So those are things that could probably be fixed over time or adjusted or made better. Is there anything that you're genuinely worried about the long-term in the workplace with using these tools?

Emma (20:46):

Yeah. One that we talked about earlier was just there's not necessarily a disclaimer that comes along with it being confidential or private information. I mean, there's always the disclaimer that it's generated by copilot, but if you're searching your private data within Microsoft 365, let's say you're asking it for bullet points from the last quarterly update, and one of those points comes from a slide where there was a disclaimer that said, this is confidential information. This is not public. And someone pulls that from copilot not realizing that it's confidential and then shares it at their next client meeting. Those types of things are difficult because unless you're doing the work of the reference point and all of that, it's hard. You got to have that human nuance accounted for

Matt (21:31):

To take that one step further. In general, it is copilot not autopilot. You're expected to review it. You've had scenarios where in a meeting it attributes what people are saying to the wrong person. It doesn't quite understand the nuance of what the conversation really was. If you're summarizing a long email thread and if something doesn't make sense, go back and check it because it is not a hundred percent, it is not exact, it's not precise, it's not specific and a hundred percent correct all the time.

Emma (22:06):

And I can see people getting very comfortable with it and then assuming over time, and I don't even mean that much time, look at us how comfortable we are with it in just a couple of weeks, in even a year, getting so comfortable that those reference points, but we just trust that it's got it. And that could get you in

Mitch (22:23):

A, it's like self-driving car.

Matt (22:28):

The other two things is, I've experienced it from the other side. I've had people from Microsoft use copilot to try to answer a question that I'm asking them, and it feels bad. We're getting these responses from people who should be able to find us the answer instead of finding us the answer. They're using copilot and effectively giving us the same thing that I could get if I asked Google or something else to give the information and it doesn't feel very good. It's one thing to ask it and then research it and then provide the response, but to literally get a response back that says, Hey, I copilot this. Here's the results.

Emma (23:07):

The quality of things may greatly suffer if there's not a human, which

Mitch (23:14):

Is why I joke with heart marketing coordinator Livy a lot about like, oh, don't you just spend your whole day in chat, GBT coming up with social posts or different ways to word things. I would say there is. That's not real. She doesn't do that. She's told me at least. But what if work just becomes that where we're constantly just using AI to collaborate with each other? I don't know. There's a risk, but I think we'll always long for that human touch that you're talking about and hopefully we all serve each other through that and do good by the other person. Let's close. So this is, what has the impact on productivity been? We've talked about some good things, so maybe not so good things. What was it like to learn copilot? Let's talk about the learning curve maybe the first couple days versus where you're at now. How has that gone for you guys?

Matt (24:12):

Yeah, so I'll start with

Emma (24:14):

First.

Matt (24:15):

Oh, I was going to start with the install, like the management piece of it. Oh, sure.

Emma (24:17):

Yeah. The You have to buy

Matt (24:18):

Licenses. Yeah. It's not cheap. You have to assign licenses. It's kind of annoying to manage the licensing piece of it. It can take several hours to even a 24 hour period for it to actually show up for somebody. It's not a bad experience. It's very common. It's very similar to all of the other tools, but it's kind of a little bit annoying because once you activate it, somebody has to go do something with it, which is probably what you experienced on your end.

Emma (24:46):

Yeah, I mean obviously you're specifically asking about copilot learning curve, but I think there's also just a learning curve to understanding what you can use AI for. So I think I went through that learning curve first because chat GPT came out and we were messing around with that, playing around with that, and you do start to really get creative on how you can use these tools. So if you've never used AI before and co-pilot's your first run with the whole thing, I would say I'd first encourage to just mess around with it and kind of try to see what you can do with it, have some fun with it. That's what I did with chat GPT. So then when I had copilot downloaded, I feel like I was already in the right mindset, which is important because prompts are, everything with copilot is creating really creative prompts to get the information out.

(25:28):

It's not going to really do that on its own, although Microsoft gives you some prompts to start. So that was really my experience of kind of knowing, oh, this is interesting. This is the types of things I can ask it and it's going to actually give me this information back. But then finding the actual logo and all of the tools takes a minute of where do I open it up? Wow, there's a lot of different places I can open up. Is it grounded in web? Is it grounded in my data? So paying attention to that. And then

Mitch (25:55):

Talk about that just for a sec. What's web versus this may be something people don't understand.

Emma (26:00):

Yeah, so Matt mentioned this earlier, but chat GPT is something that's completely web-based, grounded in web-based data, public information, internet, basically internet, not internet, internet, internet, internet. And then if you choose, and I think the only place I've really toggled it on and off is in the copilot app that you grab from the task bar. You can toggle in the right hand between web-based and private, which means it's just working on your data.

Matt (26:29):

It's hard to talk about the web versus work differentiation without talking a little bit about copilot as a brand copilot as Microsoft 365 copilot, the whole point of buying it is so that you get the private work grounded. It's grounded in your Microsoft 365 data. At the same time, you lose out on all of the things that would be web-based content. And so you are correct. The only place that you can really see it often is in it's actually copilot search, the copilot Windows app, which is a replacement for Bing search in Windows. And in that model, they want you to be able to choose Bing search public, which would be Web versus Bing search, which would be work in the work or the private version. But I found it actually annoying that in all of them. I couldn't go, yeah, I don't want to ask this question just about my private data.

(27:24):

I really want to be focused more on web content, just asking about stuff. I want one copilot experience related to this, and you don't really have that, which is when we talk about learning curve, it's probably the biggest learning curve is that each one of these tools has its own copilot integration with its own UI a little bit, and its own functionality and its own approach to how it's using it. They're all very similar. They're all still text-based. You ask it a question or you give it a piece of data and it spits something else back out. But they're all a little different and they're all, you have to find them all, as you said, you have to know that they're there. Yeah,

Emma (28:04):

I think it's really important to have a flexible mindset when you're first using copilot, even I'm six weeks in, but if you get frustrated easily, it's not going to love the tool because you really do get different responses all the time, and you have to be good at just, oh, I'm going to just try this a different way. I'm just going to ask this a different way. Or Man, this is really not working. I got to change how I'm working. If you want it to work the same way every time, it's not the tool for you. So having a experimental mindset is definitely the way to go.

Mitch (28:37):

Interesting. Would you say when you're using copilot in the different tools, is it sort of a common experience? Does every tool kind of, does it feel the same or have you found that some tools work better than others, or how do you process that? I guess

Emma (28:57):

So in my experience, and I wrote about this to you guys the other day, we realized that you can't use copilot within the Meet Now impromptu teams button. It doesn't let you interrogate the transcript after. So a workaround, I put all of that transcript into the Word document and used copilot within Word to ask questions. I did not get very good responses from that, and I don't know if that was just the media transcript, I'd have to do more experimenting, but I have found that the copilot analyzing transcripts in teams is probably the best at giving responses than I've experienced in PowerPoint, Excel, or Word. I haven't done as much experimenting, but it does feel like it understands the nuance of conversation and transcripts better than just text and documents. That's been my experience, but I've also used a lot more meeting transcripts than probably anyone else.

Mitch (29:50):

Well, but I think that's important to realize. Every tool probably has its own, they call it a system prompt where it's like, this is how you should be answering these types of questions. Here's who you are and who you're supposed to be as you're responding. So it sounds like maybe that's what you're experiencing because yeah, if you're getting two quite different answers, yes. It's got to be something related to the tool.

Matt (30:17):

It is very different because if I'm in an email and I'm asking it to help me draft an email, but it produces a list of bullet points of analysis of some other thing, it's not what I want. I want an email that I can send to someone. If you're in a Word document, you're looking for pros that you might include in a document, and if it's analyzing something in a Word document, it's expecting prose or something, a very written out verbose thing versus transcripts, which is meant to be conversational. And so they are very different. They all feel similar, but you can get radically different results when you ask the

Emma (30:59):

Same thing. You can get radically different results, not only within the different tools, just within the same tool asking it the next day. Yeah. Right,

Matt (31:07):

Right. Yeah, a hundred percent.

Mitch (31:09):

Yeah. Your personal experience with that was we were trying to record for the webinar. Oh, yeah. And you were trying to make a repeatable process example.

Matt (31:20):

I went through a whole series of prompts and I was expecting it to not answer the first one very well, and it didn't. And then I went to the next one and I got a little better, and then I went to added more, and I got to the end of it and I'm like, cool, this works. I can record this. Well, the next time I did it got a good answer for the first one, learned for me about what I was asking. And so then I was like, it's

Mitch (31:42):

Smarter than you, believe it or not, folks. Or

Emma (31:44):

When you think about it from a meeting template sort of mindset. If let's say every single meeting I go to, I want a summary action items and a recap. I can ask every meeting transcript to give me that. But the results that it's going to give, the level of detail, just all the little nuances, it will be different every single meeting. A hundred percent. And I guess if you want that to be consistent, that could be a future thing that they make it more consistent. But I've found very different results. Even an action item list, how detailed it gets.

Mitch (32:18):

It reminds me a little bit of, we've joked Matt Dressel has read the internet three times. How many times do you think Copilot has read the internet and should we have a face off between you and ai?

Matt (32:29):

Not

Mitch (32:29):

In the next podcast?

Emma (32:31):

Well, the true test is I sit next to Matt, Dr. And I usually ask him the question before I ask, so

Matt (32:39):

He's still winning.

Mitch (32:41):

How much do you pay for that?

Emma (32:44):

It's a free work perk.

Mitch (32:46):

Lucky you. Alright, let's close this third part about what the learning curve looks like for copilot. Hopefully that's helpful to someone who might be starting it from scratch and maybe some perspective of how to approach this tool. Let's talk and close this conversation with the future of work in ai. How are we thinking about how this thing might evolve over time? What are some of your speculations and then is there any ethical things that we're kind of worried about in the future of

Matt (33:25):

Yeah, so I'll talk a little bit about the future. Microsoft spent a lot of money to license and get access to this technology. I know that they're heavily working on trying to get money from it to monetize it, which I think is all great. They need to continue to invest in it. This is not something that they can let what Microsoft often does, which is let a technology sit when they've started to make money at it, just let it sit and rake in some money and not make it better. They need to be aggressively doubling and tripling the investment in this technology to make it really worthwhile. We're paying for copilot, not so much for everything it does today all the time. I think the cost that it is, is we would never buy it for everyone. If they can do what I think they can do with it over the next two or three years, there's no reason I shouldn't buy it for every single employee. They just need to continue to improve it. It's much like we've talked about with Loop. We've been amazed with how Microsoft has continued to evolve Loop. It has been unlike most Microsoft products where they spend a ton of time behind the scenes, nobody sees it developing something and then voila, they're done and they're done for five years until they really features really revolutionary.

(34:48):

Revolutionize it, revolutionize it or iterate on it,

Mitch (34:52):

Which we've learned actually firsthand working in the Loop Tap group. I want to give a shout out to the Loop. They're great team members. You all are great. You've been a bright spot for us in the Microsoft world, so thank you. Absolutely.

Matt (35:04):

Loop has been, they have been doing that and they continue to add more and more and more, and it's refreshing and it's great. And if they do the same thing with Copilot, it's going to be amazing. And when I say that, I don't just mean within Microsoft 365, I think across the board, if they continue to iterate and make it better and listen to the feedback that you're getting from customers and listen to where customers are looking for more there, I think across their spectrum of products, I think they could see a lot of great things happen, but they have to be willing to do more than just the sales pitch, which is what they're currently doing. Not that they're not spending time on improving it, but I don't have a lot of visibility into what the next big thing is. But if they don't have a next big thing that's going to happen in the next year, it's kind of like, what are you doing? Right? You got to keep it moving.

Emma (35:53):

From my perspective, this is maybe a positive and a negative as copilot becomes more and more integrated into different people's just daily work life, the same way that we all use Excel. We all use different tools. If a team or a company were to decide we're not going to pay for that anymore, or it becomes too expensive with a license or maybe you worked somewhere where you had it and then you moved to a company where you don't have it, I can see really the way the tool works, it doesn't give you the chance to, what's the word for it? Like exercise some of those skills because take me for example, I have to listen to a meeting, facilitate lead, take notes. I'm exercising that muscle if I'm trying to juggle all those hats all the time or wear all the hats, juggle all the balls at the same time. But if I use copilot, then I'm able to be more present, but I'm also not sharpening those skills. So I can see with especially people jumping companies, if you come from a place that had copilot to a place that doesn't, yikes, I think that would be actually really hard for a person. And then I think about education as well. And if you're using copilot in school, is it robbing you from some of those learning opportunities? I don't know.

Mitch (37:09):

Yeah, that reminds me of, I brought up the self-driving car thing earlier, but that's because of personal experience. When my wife first got a car that had sort of lane assist and smart cruises, all of a sudden I got in my car and it wasn't warning me when I was going over the lane or it wasn't keeping me in my lane, and it was like, Ooh, I'm getting lazy. I need to be in a different mindset in my car versus her car.

Emma (37:36):

That's a really great example, even with the blind spotlight. And I know people have gotten in rental cars that don't have certain things that, or yeah, you get a used car after having something that has that more advanced, it's dangerous because you think in your head, you're so used to this or that. So I don't know yet how that's going to play out, but

Mitch (37:55):

See any notes on the ethics side of things? When we were reviewing this, I said, oh, we have this ethics line. I don't really have anything for that. And Matt said, of course you don't.

Matt (38:07):

I mean, Emma talks about a little bit the school situation. You can ask these tools, not necessarily copilot exclusively to write in somebody's persona or to create content as if it were Dave Ramsey or JK Rowlings. Yeah, JK Rowlings. And it will do it. Is that okay? Right. If I ask it to create a business idea or a business plan based on somebody else's business plan that's not currently public, and it's taking those ideas and merging it with some stuff that I have, is that okay? Right.

Mitch (38:48):

I saw someone built a bot, an AI bot that was supposed to be a business advisor, and part of the context for that bot was it's read all the books that this person has read, and suddenly it's like, oh, you mean you fed all of that content into ai,

Matt (39:06):

Which is copyrighted content that is supposed to be paid for, but they're not providing that content. They're providing a summary of that content, which how is that any different than if I read that book and then I provided my interpretation

Mitch (39:20):

Of the book?

Matt (39:21):

There's lots of questions about all of these things. I don't know the answer. Lawyers and judges and all of that are figuring that out, but there's definitely concerns related to that.

Emma (39:34):

And it does feel like right now there's just a lot of honor system happening, especially within companies that are rolling out copilot, but aren't necessarily rolling out a brand new employee handbook on how to use it. So I think in the next year, hopefully there'll be some guidance.

Matt (39:51):

I mean, a lot of people working on it for sure to try to come up with it. Unfortunately, a lot of it is we just don't allow you to use AI tools at all.

Emma (39:58):

And you know that people aren't necessarily following that. Sure. Well, that's the thing,

Matt (40:02):

Is that just because you say they're not too, it doesn't mean that they're not. And honestly, using copilot is actually probably a better way to go than saying, don't use anything. Or if you have something, at least you can say you had a choice. You could have done something different. Not having anything. You're really in a sticky situation because people are, they're going to want to use these tools. You just don't have a good way that's allowed to use them.

Emma (40:32):

So I don't have a teenager right now, but my question, I think to listeners too of this podcast, and I think the self-driving car analogies such a good one. If self-driving cars were everyone was using 'em and it's more convenient, it's more safe for your teenager, do you just put them in a self-driving car and rob them of the experience of learning to drive for the rest of their life? Because that's just where the technology is and where it's going. Because I see that as the same question with copilot and ai. Do you allow students or people who are just coming up and learning how to create ideas, use AI to help them generate all of those things, I think you're kind of robbing then someone, or is it just expanding it? I don't know. That's my

Mitch (41:12):

Question. We got to wrap up, but it does make me think about, I imagine someone ask the same question of does an automatic transmission make them not learn how to really drive a stick shift? Right? And I'm sure this is not a new idea anytime new technology comes out, but I feel like AI just made it jump way higher. Yeah, we'll see Lots of questions around the future. Let's recap really quick. Our four parts. First was the essence of copilot. We talked about what copilot is, what it's made up of, how you might use it, and some of our personal experiences and how we're using it. Then we went into sort of how it's impacted productivity in the workplace as a whole, even some kind of things that have come up. And then we went to the learning curve of what was our experience learning it, and there's a few things that you should have in your mind as you're approaching these tools that will be helpful to maybe not get upset with them and use them to the best ability and then the future of work. Where is it going with AI and what role does it play? And we'll see how that all goes. So any closing thoughts?

Matt (42:24):

No, I think we covered it.

Mitch (42:26):

Is this everything about copilot that we're ever going to say? No.

Matt (42:30):

No. A hundred percent not.

Mitch (42:32):

Yeah. Matt's excited to get back to this one. Wait, bonus question before we go. How is Copilot different than Clippy and Microsoft Word? They do this years ago.

Matt (42:42):

Radical different has a better logo at all. The same. It's not at all. Same.

Mitch (42:46):

I don't know. He shows up and he's like, Hey, let me help you. Let me help you. Yeah, it's kind of a copilot. I digress. Anyway, I'll let you all go. Thanks so much for joining us today. Thanks, Matt and Emma for talking copilot, and we'll see you all next time. Hey, thanks for joining us today. If you haven't already subscribed to our show on your favorite podcasting app, so you'll always be up to date on the most recent episodes. This podcast is hosted by the team members of Bulb Digital and special thanks to Eric Veneman for our music tracks and producing this episode. If you have any questions for us, head to make others successful.com and you can get in touch with us there. You'll also find a lot of blogs and videos and content that will help you modernize your workplace and get the most out of Office 365. Thanks again for listening. We'll see you next time.