The shifts show Captain's Log Episode 31. Returning from a deep space mission guided by the steady hand of their newest copilot, Joe and Duane debrief us on the journey to a new frontier that traces back to The Pioneers of the last century. At the command deck, a fresh ally stands ready, transforming mere words into the gateways of discovery. Every Voyager can steer the course. All it takes is a commanding prompt computer initiate greeting sequence for our new assistant.
Hey, buddy, how's it going? Hey, hey, it's going. It's going good. Am I? Am I? Am I your Co host on this show or am I your copilot already starting off with the puns? Wow. Yeah. I mean, you're definitely not an autopilot, so I'm not sure you'd want me to be a copilot in an aircraft anyways. In a real one. So, yeah, you know how I am about flying. I get sweaty hands. I do. I do. Well, I don't know if you've heard of this new thing called copilot. In fact, there's like 12 things called copilot.
So I thought it would be fun to have a conversation around #1. Like, what is Copilot #2? Like, which copilot are we even talking about? Because somebody decided to name everything Copilot and that's made it not confusing to say the least. And and didn't even start off as copilot. We we really started off with some some Bing. Yeah. In there too. So we've gone through some renames. Oh, don't forget about GitHub copilot that's been off for quite a while. You know, predates this.
Yeah. Yeah. True. True. I fully believe in the name copilot though I do. I do believe and I think we even said that like several episodes ago when we had Sean from from the PowerPoint product group on and we talked about how we felt like that is actually like the perfect name for this because it's just sitting there riding alongside you helping you get stuff done.
There is a that persona of a copilot is exactly how you should treat this and think of it as it's it's someone's going to help you help you get to a destination. It's not going to fly the plane for you.
Yeah. Yeah. And I guess before we just jump into the weeds like head first all the way we should back up a minute and and talk about kind of this global planetary transformation you know movement that just kind of snuck up and and really one of the biggest changes in technology in the last, you know of the decade easily right. And that's just what the heck is generative AI in the 1st place and where did ChatGPT come from. So you got you got any origin stories there for us.
Well you're talking veger from Star Trek is that where we're heading. Yeah. You know as if you've looked at this a little bit that we'll we'll go back we can go all the way back. I I don't have the history on it but we can go back I think into the Fifties, 1950s around the the concept and theory of some of this work that's now taking place.
And as we find with science and with math, some of these things just build on top of each other and this exponential growth and that's kind of the kind of where we've we've hit now with this generative capability. I see it though as also the accessibility of it to us. It's not something that a bunch of, you know, math, science people are doing. You're not having to pull up some crazy programming code to somehow access this. It's become accessible on, you know, on your phone.
It's become accessible in a browser to start taking advantage of it. So I think not only the science that has exploded has been important, but accessing it is also something that's changed.
And getting, you know, getting access to this, these things called copilots, we're going to talk about those and how they are integrated, integrated throughout a lot of the Microsoft ecosystem right there, right when you need them, they are inside of a bunch of different apps that you don't have to search around to figure out how to access them. They're going to pop right up right when you need them.
Yeah. Well, the, the couple things you said there that I think are really important is accessibility. When you think about AI and you think, let's think about if somebody was using the terms artificial intelligence and they were thinking about that three or four years ago, it's going to be like the mental image I get is like a big computer, right, like processing and doing powerful things. And usually they're pretty specialized workloads.
You know, like AI is being built into machinery to detect images and objects and recognize them and do something about them, right? Well, that's a very specialized thing because we don't all have manufacturing plants in our house, right? That doesn't do us much good. I mean we benefit from it down the chain, the supply chain, eventually the efficiency boost that AI brought to that particular piece of the puzzle. But we don't get to use it in research, right, Big AI movements and research.
You and I had some faculty doing some pretty amazing things with recognizing weeds and an orange field, you know orange Grove also pretty specialized, pretty specialized implementation. It was using machine learning and image recognition and hardware that was flying over the orange field and and orange Grove. It's not an orange field is it?
Orange Grove Pretty sure I live in Florida my entire life I should know this stuff so but but the point is it was AI but it was not accessible unless you were in a role that it was part of your your job and along comes ChatGPT and now my kids are using it to play and write stuff and ask it questions.
And you know, then Copilot enters the picture from Microsoft and and before I go down that, you know, before I go down that road real quick, I'll say that, you know, Microsoft and Open AI, this didn't just happen last year. Like there's been a relationship there for several years. And and so, you know, Microsoft big, big investment in Open AI. And so their technology sitting behind, you know, their GPT 4, you know, now GPT 4 Turbo, like all the different models which we'll get into that later.
I'm sure sitting behind Copilot in both of those products really made AI like a household term. That's so cool. I mean to say like, yeah, I mean I've got friends that I talked to that, you know, they don't, they're not really techie, but they're like, yeah, I used AI and I made a new you know, the cover for a book I'm writing. Like that's a true story. Friend of mine, great friend of mine, not a technologist, went through and used AI to generate an image for a book he's having published.
How cool is that? Like, Oh my gosh, it's just mind blowing. Yeah. The the the shift that we see at at a global level from this is just the the power that we all have now is just unreal. Yeah. The like the your kids using it. I'm just thinking right now of just using your phone and you're typing a text message right now you've got autocorrect happening and you've got predictive text occurring. You can.
I've seen, I've seen some jokes about how you just create a sentence with the predictive text that shows up on your phone when you're typing a message. Just take the take the very next word it thinks you want you might want, and just use that word and you get this crazy weird sentence that doesn't really work. OK, not really played that game before, played that game before. So that's really what this is doing. Just add a massive scale of knowledge behind it.
It doesn't just, it's just not the words that you've typed in a phone in the last 30 days. It's it's the Internet it has access to and the data behind that to predict what information and what words come next to help you, you know, solve the problems you're asking it for. Well let's let's dig into that for a second too because we've talked about you know AI where it used to be AI where it is now.
We said the words GPT several times and then yous just said you know it, it's learned from us basically. So can you kind of unpack that? You know, if I were prompting you as Copilot, I would say, hey, you know AI Joe, I need you to explain this to me in simple terms, what is GPT and how is it leveraged in AI solutions like ChatGPT and Copilot? And I would just continue a little bit.
What I said was it's there is this corpus of of data that it has taken that is available publicly or maybe there's some licensing behind it. It depends on perhaps the, the GPT tool you're using. But there is a large language model, LLM. There's this language model, bunch of science behind it. There vectors and just lots of math behind this that decides in certain ways how we form our sentences, how we put, how we put words together.
And when you ask it a a question it's going to use that to help itself understand what you're actually asking. You know we've we've done some of our own little demos where we we mistype words, misspell things and leave out words. And it still it still actually understands what we're asking it because of this massive amount of language that it's captured and can and can use to predict the basically the next word in line. It starts with a word and predicts the next word.
But there's little ways you can tweak it, little ways that you can just help it decide what are those words that are going to come next. And we can, we can type a sentence into Microsoft Word and make it sound very scientific. Or maybe we're typing an e-mail to our friends and it's very informal and we're going to type things differently.
Well those are tweaks you can do when you when you prompt to basically give it an idea of help it help it think about what might be the next word based upon some of those tones that you're interested in some of those styles of writing. But the the that pre training part where it's it's figured out all these all these sentences across the across the Internet for us it'll put them back together into a way that we we can take advantage of.
So yeah, it's a natural language, which is kind of challenging. Or a generation of people that have been conditioned to find information using search engines and they are talking to Bing or Google or, you know, Ask Jeeves or whatever it is that you're using. Like it's a computer and they're searching for keywords. And this GPT thing comes along and it says, you know, hey, I can produce natural language and I can receive natural language. Like you can talk to me as if I were a person.
It's kind of a hard adjustment to do that. That's one of the things that in my experiences, I see, not that I say people struggle with it, but it's this, this the shift that you have to make on. You've been, like I said, you've been conditioned for so long to to treat a computer like it's a machine. In this thing, you're going to get better results if you, if you treat it like it's a companion and you actually, like, have a conversation with it. Which is weird.
I know that sounds weird, but it's true. It's it's true. Are you an extension agent? Just getting started with teams? Or maybe you're experienced, but you're looking to improve your digital cloud skill set? Are you an IT pro supporting extension programs at your university? Well, you're invited to join us in a community working to empower extension agents and their support staff to skill up and learn together.
If you're interested, visit AKA dot Ms/ join Cloudy County to sign up and join us for monthly training sessions and open discussion with your hosts of The Ship Show. Cloudy County Extension Essential cloud Skills for Extension by Extension. On that subject, what I like to say is if you're you're going to pull up pull up Bing and Google and and do a search on something, you're going to get back Just this long list of links that you've got to decide is it an ad? Is it, is it?
You have no idea if it's going to be good quality. On the other side you you got to look at the source URL. Maybe you've got to make a bunch of determinations or even bother to click on it, and once you click there, you're going to click through a bunch of other things that you might need to look at and read this long page and and see if this is actually information that's related to what you wanted. That takes time.
So instead, if you were to think about it in Copilot and type out a little bit longer sentence, maybe give it a little bit more detail of what you're really looking for, things that you would never think of doing in just a keyword search. Copilot is going to take a little bit longer 'cause it's thinking about producing perhaps a paragraph of information for you or some bullet points.
But the information that you asked is probably gonna show up for you, available in a way that you can skim it and read it. And if you need to dive deeper into it, what Copilot will do is give you real links to real sources that really exist. Yeah. And you can go research it deeper. So it gives you that first that first rough summarization of what you're hoping to find right away.
And you may not have to dive deeper into any web pages if you know if maybe there's enough information right there to get started. So it time wise ultimately it might actually be faster than clicking through a bunch of different links. If you're like me, you get stuck in going out a rabbit hole somewhere. You click one link and you click another link and you forget what you're actually looking for. So 47 browser tabs open later you know and that's a low number for Joe. I know that fact.
Yeah, that's an interest. I keep using the word shift not just cause of word play of our podcast, but because it is a major change. The movement from keyword searches in a search engine to asking natural questions to what we now call an answer engine is a big change and it's hard to get people to do that. But everything that you just talked about kind of keys in on the ultimate goal of this generative AI like in the workplace or in our our daily life. And that is to save you time.
It is faster for you to ask copilot hey, where are some good places for me to go and eat tonight? I'm in this city. I'm going to willing to travel this far. You know these are the you know, people I have with me. These are the kinds of things we're interested in. You know, and it's just going to give you a list of like hey, you should try these five places. That's faster than searching and hunting and then clicking on pages and and and those types of things.
And that's ultimately the goal I think. And I think we could probably, you know we've already been talking about a product but we haven't really totally named it yet for the listeners. I think we're going to talk about four, we're going to talk about four. Is that the plan roughly we'll we'll figure it out, we'll see where it goes.
Mostly we're going to talk about four or maybe 3 1/2 copilots and and the first one that we're talking about right now is just Copilot or called Copilot for the Web or previously known as Bing Chat and it has a variation to it for the business world. So for our extension lane grant friends out there or anybody with a work or school login, you'll have a little bit of a different experience that I think we'll key in on.
Those products would have been called previously known as Bing Chat Enterprise, now known as Copilot with commercial data protection. So the public version of Copilot, it's the same thing with two extra features, you know guarantees if you will add it on top of it. So those are the answer engine products that we've been talking about. You can go to those on your web browser, ask it where you could go to, where you could go to dinner.
Give it a list of all the stuff that's in your fridge and have it suggest a dinner recipe. You know, like there's all kinds of and mine would say go find a restaurant. Yeah, you would give it everything that's in your fridge and it would suggest 5 restaurants. Those are the things that we've been talking about thus far. And so the the public version of copilot, that's the one that my kids can play with. They can use that. They can ask your questions.
They can, they can have a generate content, they can have a look at content that they already have. If they're using Microsoft Edge, you know, and it's it's very powerful. But I think in the business world, especially in the education world, there have been some changes recently to accessibility, and I'm going to read a a comment. Let me find it real quick about one of the. Objectives of Microsoft's responsible AI policy, if you will, like.
One of the goals is they want AI to serve society broadly, not narrowly. And so a year ago we didn't even have this capability and now a year later everyone on the planet has this capability for free just using Copilot, copilot.microsoft.com. I'll, I mean I don't I can't think of a better way to meet that objective, right. But then they went a little bit of a step further in the business world and they added this copilot with commercial data protection variation of copilot.
So Joe, there's two things that it does that faculty, researchers, educators, business folks would care about. Can you explain what are those two things and why? Why would they care about having this more protected environment? Sure. So in in in our education or or in corporate world, perhaps what you sometimes can worry about is your prompts, the the, the data that you're typing into the browser, Where is that going? What's what's happening to it?
Well, that's that natural language input, right? Like, where can I go to dinner? That's the, that's the prompt. Yeah. And you might not be worried about typing that in, but maybe you have some research data that you would like some analysis done on or maybe you have some demographics data that you would like some data manipulation on, data analysis on.
Or maybe you've got an abstract for, you know, an unpublished paper that might turn into a patent down the road and you want to help make sure that it you know it, it makes sense and take advantage of copilot for that. Those are things that you would not want to get saved into the system, you wouldn't want to end up using being used to train future models on.
You might want to make sure you know Microsoft isn't using that in some way and that's what the consumer data protection tool or or policy and feature is going to give you for us in education field or in work and school. If you're licensed for that, you'll know that you've you're you've are in that mode. If you've signed in with your work or school account, you're gonna be looking for this, this green protected. I'm gonna say right now it's green and it says protected.
Little little pill up there that that gives you an understanding that you're in that mode and there's data protection, data policy. Those prompts aren't saved, they're not stored, they're just used long enough to get you a response back. They're not being used to train the models. They're not your your information isn't gonna end up in a large language model that could accidentally in the future somehow it end up as a response. Yeah that's AI think that's really really important for researchers.
Yeah that's the big one. So if you're a researcher worried about some some new data that you're working on and nobody else has done anything with it. You don't necessarily want to just open up any of your your free versions of any of these tools and start using it because you're not 100% sure what that's going to be used for in the future. Often your use is, you know it's free but there's a cost.
Yeah. Yeah. When you when you ask chat GPTA question, you know that when we, when we use those words GPT or large language model, you know I would ask and not sarcastically but kind of sarcastically like where do you think ChatGPT got the information from to formulate an answer it it's been trained on public information you know from the Internet right from the world.
And the more people that are using ChatGPT the more inputs it's getting to be become better and and be trained in the way that we are using it And and so yeah that's you know you you take your proprietary research that's unpublished and and hasn't you know been seen yet and is super important to you and you use a public tool like that. Well guess what you just contributed to the training of the large language model. And I'm not saying that to like fear monger people.
I just think you should know what you're what you're using, what you're getting. And like you said, yeah, it's free, but I don't want to say it comes at a cost, but it it's free. But there is a benefit to overall to the world there of OK, we're furthering this technology, we're making it better. And when people say it's learning, yeah, it's learning 'cause you're using it, it's learning, 'cause you're feeding it.
Well, if you don't want to feed it, you need to be using your work or school account when you sign in to copilot.microsoft.com. And if you're doing this from your work computer, that's likely already happening. Hopefully it's happening automatically. Your IT team has things set up to just sign in and you know you're up and running, you're good to go.
But the thing that you called out that's really important is, if you're wondering, like is what I'm doing right now, training the model, if you see that green signal that says protected, then the answer is no, you are not training the model. And the things that you are asking Copilot to do are not saved. And that is a double edged sword because one of the fun things about using generative AI is finding prompts that work really well, that do something really cool.
We'll probably put a couple of like recipes, if you will, in the in the show notes for people to kind of get started with. But when you find a prompt that does something really cool, you're like, oh wow, that was awesome. I need to save that. Well, if you're using Copilot with commercial data protection and you refresh your browser window or you walk away from your computer and you come back and it signed you out, it's gone. That's one of the protections.
This is, is we're not going to save your stuff. So yeah, keep that in mind if it if the output you got was great, take that prompt, copy it, save it into OneNote or loop or whatever your favorite note taking app is. Sticky notes Notepad. Write it down because it's not coming back. Yep. That's that's the that's the free and and sort of licensed version of of just a interface to prompts to work with.
But we've got a couple others that were, yeah, yeah, that one, Copilot Free is incredibly powerful. And Copilot with commercial data protection is also for our Edu customers more or less free as well, because that has been added to even the free licenses that education customers get. So it's still free even with those added benefits of we don't save your stuff and we don't train the model.
The next one is a little weird because now we're starting to bridge the gap between two major features that people want. People want Copilot in Microsoft Word, you know, hey, I want that GPT stuff, but I want to just open up Microsoft Word and tell it to write a summarization or to reference another document and reason over it and give me, you know, give me a a shorter version of that document. Well, Copilot Free doesn't do that. You do have to pay for that.
And there's two things that you could pay for, and they get a little confusing. The first one's called Copilot Pro and we're only going to spend a couple of minutes talking about it. Joe, you're a copilot Pro user, 'cause you know why not. And it works in your office apps, right? But what's it, what's it actually connected to? What's it tied to? It's it's it's tied to my personal Microsoft world, like your Hotmail, tied to my my Hotmail, my OneDrive of of my own personal environment.
So it's not going to see any corporate data, anything like that related to your work or school world. So you need a Microsoft account, get in there and it's just going to work with your your own personal files. So it's $20.00 a month. I think you can even get a short little trial for it if you want to. But I would say you know that is a that's a quick way to get a feel for the tool and how it works inside of the Office products.
But it doesn't have all the features of the tool that I think we really want to talk about was copilot from Microsoft 365. So it's a subset of those features in in Microsoft Copilot for Microsoft 365. But the key there is really it's it's it's personal. It is just your personal world is where you're going to get access to that and use it. Yeah and and that's powerful for individuals. For non I use the words work or school. That's perfect. That's what the account is called.
When you sign into something, it's going to ask you like do you want to use your work or school account. Well that implies that whatever it is that you're doing is for work or school. So Copilot Pro does not work with work or school and and like you said, you know your corporate information, your organizational data, the stuff that's on in a Microsoft Teams team, it's stored on a SharePoint site. It's in your colleagues OneDrive like none of that is going to work with copilot Pro.
So I have a little joke. My kids love this movie. It's called Encanto, Encanto, Encanto. One of those. And there's a character in there that's kind of like a recluse. And they sing a song that says we don't talk about Bruno. And if you're a parent, you've heard it probably, you know, 100 times because it was pretty popular. And, you know, when I talk to people and they bring that up, I'm like, you know how we don't talk about Bruno?
Well, we didn't talk about Copilot Pro either because in the in the business conversation, it doesn't have a place there. It has a place fine in your family, you know, if you want to get it for your kid so that they can use the office apps and they can use that generative AI type functionality right inside of their office apps, yeah, have at it.
But don't be a faculty member and go out and buy copilot Pro with your Outlook.com or your Hotmail account and then expect to sign into your work device and at work 'cause it won't. So. And I I don't do it, yeah. And and I would say only quite recently really that we might have had a different conversation because copilot for Microsoft 365 had sort of a big paywall to kind of jump into it for quite a while especially, and it wasn't even available in education to start with.
So some of those have come down where it's now, it's it's worth talking about 'cause it's accessible this this license is now something that is, is really truly accessible to faculty or staff in higher education. The upfront licensing costs, there was a a large amount of money to be invested to get even started and that's that's faded away, that's gone now and even small businesses too can start taking advantage of it.
But that's the reason why I think we another reason why we talk about this other product more because it's a a fuller complete set of of tools and services not a subset and it's in the affordability range that education institution can take on now or an individual faculty member with a grant Pi or something like that. So that's something that why we want you to consider more the business side and that would be copilot for Microsoft 365. Yeah this is going to be the bulk of the episode.
I think you know we're we're not even quite halfway through recording this and I feel like that's appropriate that we would kind of rush through the other ones. I I I do want to just restate the the Copilot with commercial data protection product is incredibly powerful and can do amazing things.
And if you're using Microsoft Edge and you use Copilot inside of Microsoft Edge, you can have it analyse stuff that you're looking at, like summarize APDF, summarize a YouTube video before you even watch it, right. There are some incredibly powerful things that you can do and not spend a single dime. It's free. However, you know, taking it to the next level of of productivity. Yeah, Copilot for for Microsoft 365 is kind of a big deal. It's, it's kind of a big deal.
It's it's a little bit of a, you know, I don't want to say sticker shock, but I'm not not a salesperson. However, I've had conversations with folks and it is the retail price, which is also the education price is $30 per user per month. Another way of looking in that is $360.00 a year. Another way of looking at that is $1.00 a day. So same numbers just presented differently, right?
And I like to focus in on the dollar a day thing because there is a functionality that is so easy to demonstrate to people that will absolutely save you more than $1.00 a day in your time and sanity. It is. It is incredibly powerful. It is incredibly hard to demonstrate because all I have to do is click one button. But before I get there, I'm going to save that as a teaser.
When you move into copilot for Microsoft 365, now you are connecting that GPT technology, that large language model technology, You're connecting it to your work information that may scare people at first. They're like, wait a minute, you know, you said the other one didn't train my model. It it. You know it didn't. It didn't. My prompts weren't saved and didn't contribute to the overall training of GPT for the world to make the world a better place.
But now you're telling me that when I pay money, it's it's actually connected to my work information? Yeah, but of course the same, the same guarantees apply. We are not going to train the large language model using your work information. In fact, because of some technology tricks that that can be done. We don't need to, we don't, we don't need to do that. Your data is your data. It's always been your data and it's not going to be used for anything like that.
But this is really bridging the gap between generative AI and work data that lives inside of your Microsoft 365 network. So think things like Microsoft Teams, emails, chat messages, files that are stored in Teams, files that are stored in SharePoint sites, files that are shared with you by your colleagues in OneDrive. Those all make up your kind of like map of information or web of information that you have access to inside of your inside of your organization.
And that's where this starts to really become powerful. So the the, the free version or Microsoft Copilot, Copilot in the web you can sort of get it to look at a web page or maybe APDF, but you're you're you're in that browser, you're in that, that's the only place you're accessing this. There are some mobile versions of it, things like that, but that's really just basically web-based. So this product is bringing that information or bringing bringing that access right into where you are living.
Often inside of the Microsoft suite of tools. This tool will surface up into Excel and PowerPoint and Outlook and your your OneDrive and whiteboard and OneNote. And I mean it's just coming to all these tools that we have. I know they're coming to a few more, or maybe they're in a few more, but it's basically when you bring it into those interfaces, you have not only the ability to use it to help you work with the document, but you can even start having it help you.
Basically manipulate the interface, manipulate the the documents themselves. Not write something, but maybe actually make formatting changes or produce entire slide decks. So bringing that right into those apps is the, you know the the number one significant feature that you're going to get and that's why it costs a little extra money. So, well, I'm going to, I'm going to unveil, if you will, not really. It's not that important.
But I'm going to talk about the one that one trick, you know, I sound like clickbait that that saves that dollar a day because of the fact that Copilot is going to be built into apps that you're already using. So you know you start your day. What do you do when you start your day, Joe? What's what's like being honest with me now? Like I know you're a little bit of a special case with efficiency and and and and stuff like that.
But what's the first thing that you do when you fire up your work computer for the day? What's the first app you open? I'm. I'm opening up Teams and open up OneDrive and I'm. I'm sorry, Outlook, Teams and Teams. I was going to say OneDrive is an odd one. I wasn't expecting that. Yeah, yeah, yeah, Outlook, Outlook. So, but I'm looking at, I'm looking at my e-mail. I'm looking. I'm looking at Teams. Teams is usually always always wide open and accessible.
I might have to actually open up my e-mail app. But that's what I'm going to I'm going to dive into first. OK. So the two communication persons, I'm looking. Yeah, the two. All right. So we've done episodes on Outlook before. I'm pretty sure we called it a black hole of productivity, might have been the term that we used for it. And the easy feature is this works in both Outlook classic 'cause we do have this duality right now with Outlook. There's a new outlook and a classic outlook.
There is an option, most used feature that I'm aware of called Summary by copilot. And when you're looking at an e-mail thread, and it's not uncommon I would say, for either one of us to be included on an e-mail thread that we weren't included on for a while, right? They're like, Oh yeah, we need to pull Dwayne into this 'cause he needs to be aware of this 25 emails later. So I have this really long novel that I have to scroll through and read. I have to scroll to the bottom and scroll up.
Half the screen is going to be eaten up by people's e-mail signatures and you know all the the subject line and all the people who are recipients. And so I'm really only scanning like maybe 25% of the screen at a time looking for the actual meat and potatoes of the conversation. Now all I need to do is click one button summary by copilot. It will go through that entire e-mail chain, break it down, give me a few bullet points and let me know what it's about.
That is hands down the the, the most incredible thing ever that I, I I tell you when I show that to people they're like Oh my gosh this is like magic and this happened this it's a it's a problem that everyone can relate to being pulled into long e-mail threads or just having a whole bunch of unread messages and being able to get through this messages faster.
Well, we're right back to where we were 25 minutes ago in the conversation of what's the objective here, save time the, the, the, that large language model and understanding the context of those emails, how often some of those are probably in a 25 thread. You know, some of those are going to be like maybe they're like I agree or you know, thanks for this information. So you've got to read through those and deal with those.
But Copilot's going to know those aren't of serious value, aren't going to, you know, use that in that summary. Summarizing data, summarizing e-mail, summarizing video, summarizing teams, meeting transcripts, that summarization process, summarizing your Onenotes across the board is a major time saver. It it understands the context around what's important and can you know, provide you even such things as you could ask it for. Are there any action items? Are there any task lists?
Are there anything that has yet to be completed in that in that summary thread that you you're talking about? There might be some place where, you know, maybe earlier on somebody said maybe we should ask Dwayne this and like finally somebody actually sends it to you. But you got to read through 7 messages to find out what it was.
Well, it might pick, you know, it might pick up where it's somebody calls you out and says I wish Dwayne, you know, could follow up on this And it's, you know, pick that up in that summary. If you, you know, ask for those kind of that scenario right there is a is a perfect example of why we would call your inbox a productivity black hole 'cause you're going to get sucked in and not get a whole lot done because you're spending a lot of time trying to comprehend information.
And there are, I would say I want to limit copilot to three categories, but there are three categories and I'm a huge fan of alliteration so I really connect with these. But comprehension is one of the like pillars, if you will, of problem solving that Copilot brings to the table is I'm looking at something and I need to be able to understand it really quickly. And that could be a document, it could be a Teams meeting, it could be a an e-mail thread, right Excel workbook, you know.
Those are all things that might be sources of data, sources of information that, if we're being quite honest, the the thing that all humans are limited by, regardless of your job title, your salary, you know, whatever it is that makes you awesome, you still only have 24 hours in a day. And so you and and we're trying to maximize what can we do in that 24 hours? What can I do to make my life a little bit easier with the things that are honestly just annoying?
You know, our CEO used the word mundane. I probably shouldn't use the word annoying, but if we're being honest, think about that. Who enjoys going through their mailbox? I I don't know. I I probably couldn't name anybody that does. So ticking off the box of comprehension, you know I can. I can comprehend that e-mail thread in seconds instead of many minutes. I can comprehend a teams meeting that I missed. That is the next thing that people are are super stoked about is especially in Edu.
Everyone is assigned the job of taking meeting minutes, right, and identifying action items. You know, what are the things that we need to do when this meeting is over? Well, if you know we're dealing with GPT, it's text, you know, based, we're rendering or reasoning over text. If you transcribe your team's meeting, well, guess what? You can feed it an entire transcribed text of everything that you said you were going to do.
And now it can look at it and it can, it can say, like, well, Joe said he was going to do XY and Z. So that's probably an action item. And these are the main five things that were, you know, the different topics that were discussed. So let me summarize those for you in a nice little bulleted list. This comprehension of information is going to get you through your day so much faster.
You remember when we were, I guess we'll call it, hacking transcripts when we first got transcripts in like a teams meeting of recording your meeting. And if you know you've got an action item coming up yelling out like yelling out like cucumber or something like that. So you could look that up in your transcripts and you know where that was. Like that some somewhere around that.
Every time you look for that, that was like early days of teams when we were just, like blown away by the fact that it would generate a transcript of the meeting. That was like, whoa, you know? And yeah, that was like a little physical life hack that we came up with of like, we use a goofball word that nobody else is going to say in the meeting. And then you could do a keyword search, find all those goofball words and those are your action items. And yeah, this is just light years ahead of that.
It's not even close. It's exactly it's appropriate something that was really really silly and painful but worked. Now it's just natural built into the and you're it's just it's gonna be right there. Yeah recap. So yeah being able to help you create you know create to do lists or go into your open up an Excel file and help you write formulas for your data.
You know if you crack open excel and just click through all the formulas that are possible, you know how do you choose decide AV lookup or an H lookup, something, whatever. All these esoteric formulas that are there that are really powerful, maybe they would help you and and you learn from it. That's what I find is I I'm not sure how to tackle this but I learn and watch what it what it has figured out and has worked in the past and I try it works for me and I I it is a really good tutor.
Yeah, I I had a scenario the other day that I used. I was working on a presentation. I did not design the presentation and I had to do some teaching based on the slide content and I didn't have any slide notes. All I had was the text that was on the slides. I mean, I knew the I knew the topic, but I had to get some slide notes in there and I wanted to do it quickly. But I'm not a wordsmith either. So you know, OK let me go to copilot.
So I asked copilot explain this to me in simple terms and then I gave it the subject and then I told it what my intention was. I said my intention is I'm going to be teaching this to an audience who has never heard this before. So you need to use simple terms, make it relatable, you know, and make it succinct. And I asked it to explain it to me and then just to make sure to kind of like dual purpose.
I said in a second follow up prompt because that's one of the cool things here is you can kind of have a conversation with it is I said OK I'm going to explain it back to you and I want you to confirm that I understood it correctly and then I use my own words to explain it back. So I I'm doing a teach back now to validate my comprehension of what it said and simultaneously kind of have it Fact Check itself in a sense.
And it said Yep, you understood it, this is what you said, this is how it relates to the XYZ topic. Good job, you're awesome. I was like oh thanks, I feel better about myself now. I'm just going to copy paste everything you said and paste it into the PowerPoint slide notes and that is now my slide notes. And so when I get to that slide in the presentation, I could almost verbatim just read off what it said and educate the audience on it.
So yeah, it is a fantastic coach, tutor, copilot, you know, And again, the word is the word. The name is just perfect. Yeah, I I know our extension agents, some of the responsibilities are taking some new research paper and writing A blog about it, writing some documentation for, you know, the citizens of the state that they're in. And it might be some, you know, fancy scientific research paper. It may be slightly out of their field sometimes that they need to go understand it and research it.
So using Copilot to help you like you were doing, ask you questions about this, this document and and can you respond in a way that's correct to help you understand, is that, you know, is this something that you could maybe talk about in the future in a presentation or something like that? Or help you write a like again like A blog post in a way that's more just more accessible and uses, you know, terms and terminology that just easier to understand.
And compartmentalizes all of those thoughts down into some nice summaries that fit in a blog post, fit in a Facebook post that may, you know, go out and link to these research documents that that we put out on the web for people that need to dive into those deeper. That's definitely where I see times. Yeah, Yeah, that's. That's a huge one. That's one of the other. You know, I told you I liked alliteration. That was one of the other C words. So that's creation. So we talked about comprehension.
You just nailed down creation. You didn't need to generate content. That's one of the most intimidating things ever is to say, OK, I need to build a PowerPoint deck and you fire open PowerPoint and you click new presentation and you're just sitting there looking at a blank slide with a title text box on it. You know, and you just want to get started, maybe you have a research paper. So remember we talked about the integration here into your work data, your work information.
You can go into Copilot and PowerPoint and say build me a PowerPoint presentation based off of this Word document and it's going to reference that file, it's going to use the data that's in that file, It's going to interpret it. And then that's one of the powerful things of Co PIL is it's going to issue commands to the PowerPoint app to start building slides. And that is probably not going to be just, oh, it made all my slides and I'm good to go. Like, you still need to look at them, right?
Yeah. Like I said it's never get stuck on like writer's block. It will it will get that first draft.
It will get that first draft for you and that PowerPoint example is is really powerful because I've I've played with that a while and it will pull out images that reference that they know that makes sense for that slide Even so you know out of the out of the stock images from the Microsoft suite of stock images and it's it's really slick when it does that the, you know deep dive there into you know some type of document and out pops a PowerPoint slide deck is is pretty fascinating.
That's probably the second most mind blowing thing I've shown people when they realize that they have a document because, you know, you've been asked to present these findings to the school board or to your committee or whatever it might be, you'd probably look kind of goofy just going in and sharing your screen and scrolling through a Word document. You know, people are expecting it to be a PowerPoint. That's the format that they're looking for.
And so yeah, being able to turn or convert or transform a Word document into a PowerPoint with, I'm going to say somewhat relevant imagery. You know, you might want to come in and put in your own chart. But setting the expectation here for anybody who's listened this far into this episode is you are not going to get the final product the first time you tell copilot to do something. But you are darn sure going to get, you know, 75 eighty, 90% of the way there.
And that's what it's meant to overcome is if it was going to take you 3 hours to build that presentation and now it takes you 30 minutes, well, that 24 hour clock that we're all bound by, you just got 2 1/2 hours back. It'll it'll get you to get you to the airport and and again that's back to why it's called copilot, not autopilot. Right. Or just pilot.
You know, if it was driving for you, well, we might not want that, but we do want it to get me almost to where I need to go and let me finish it off. So yeah, that that is a big one. The the last C word is collaboration, you know, being collaborative. And I kind of feel like the the teams meeting notes that we were talking about earlier.
That's a big part of it being able, I am now finding myself in a habit of when I pull a colleague into an e-mail chain, I just go ahead and summarize the e-mail thread for them, tweak it a little bit and include it in the message that I send to them. So now 'cause I have the context right, I have the context of the whole conversation, I can use copilot to summarize it for me and then I can dial it in to specifically what I need that person to understand or look at.
And so I'm using it on their behalf now, saving them time and they never even had to click the summarize button. So collaborating with your colleagues quicker, providing them summaries, using those summaries that you generate for them teams meetings, getting those meeting minutes fired off quicker, yeah, that's a that's a huge win that saves time.
That's if anything if you've heard anything from this, it's copilot is designed to save you time and I think in my experience and people I've talked to that have it, it's definitely doing that. The yeah, it's this exciting time this, this is just going to completely change to me right now.
I'm finding that some of the the quality of my work, the quickness of getting back to people sometimes just makes it easier to stay up to date on some of the projects that I'm working on with other people and looking forward to where this heads. I think it's it's just we want to say that this is, you know this is a tool that we really want you encourage you to start working with. Whatever version of the tool you have access to is just get comfortable with it.
Find a find a way or a reason or just have start off with just having fun with it and it will. It will write some poems for you or something like that or some recipes you know just get get a feel for kind of how it works if you haven't. And and again the the data privacy around it and the importance with the security that you get in your work and school accounts take advantage of that when you start really diving into it with your your corporate. That's a great point.
I know we spent the majority of this conversation talking about, you know, the hot rod of of Gen. AI when it comes to personal productivity, which has a price tag associated with it. But remembering that we don't want to serve the world narrowly, we want to serve it broadly. There is generative AI out there in some fashion for everyone and there's a chart. I know this is an audio podcast, but I'm going to try really hard to describe it in a way that it is impactful.
There's a chart, I'll we'll put it in the show notes that talks about the amount of time it took for a certain technology to reach 100 million users. And so I'm going to read it off mobile phones. So we're going back to cell phones, Back phones, you know, Motorola Star, Tac, Nokia candy bars, all the way to current times. It took 16 years for mobile phones to reach 100 million users. OK, well, you're like Dwayne. The Internet didn't even exist for some of that.
OK, Like, we barely had AOL dial up. Let's be real. All right, Well, the Internet itself took seven years to reach 100 million users. Facebook. OK, Internet's kind of mature now. We all have some sort of, you know, access. Some of us have high speed access. We have. Facebook has launched. We're all, like, posting stuff on walls and liking and whatever. Facebook 4.5 years to reach 100 million users ChatGPT 90 days When you and when you look at it on the chart, it doesn't even look real.
It looks like it's the edge of the graph because it's just a straight line. It's just a straight line going like a straight line up. And when you compare like economical measurements like GDP to these previous like major shifts in the world, there was always an increase in GDP when these moments happened.
And so I would say it's pretty easy to to predict if if it's even a prediction at this point that AI is going to have a significant impact on GDP and and make it higher and just to make sure everybody understands like just it's not a fad. Think about think about all the tech companies. Think about the commercials that you see or hear on your podcast or TV or whatever. What tech company hasn't used the word AI or machine learning?
You can't even buy something on Amazon right now without getting an AI generated summary of all of the reviews that were left. And what is that doing? Saving you time. There's 4000 reviews on this product. Well, let me go through and read them all. No, you're not going to do that. But you can read a summary of what people liked and didn't like. It's everywhere. It's no joke. It's it's legit and and it's accessible. All of us have access to it in some way, shape or form.
But for those enterprises out there that are investing in it, my only encouragement would be make sure that your users know what it is that they have and equip them to be successful. Learn. Teach them how to talk to it. Teach them how to get the most out of it. Get that $1.00 per day or more back out of copilot because you definitely can. I'll make sure we include some adoption guides and adoption kits for the Microsoft Copilot in in Microsoft 365.
That that that tool you want to want to roll that out through the adoption toolkit and take advantage of that. There's a significant effort put into that. Those resources. There's workbooks, there's user guides, user training material. There's trainer, trainer material. You know there's checklists and workflows on do this first, you know all these steps to get involved. Invest in that and I think you'll have a more valuable and quicker to quicker return for your users. Yeah 100%.
Check out the adoption.microsoft.com, go there, click on products you know, click on copilot. For M365 you are 100% spot on. There are significant amounts of very valuable resources that are on that site that will help make your users successful. And this is not a roll out that you just turn it on and walk away from it and expect it to work. It's not just pushing out a new version of Windows. This is a big deal. This is a very big deal and it's a big investment, right?
And you want to make sure you Max maximize your return on that investment. So make sure you equip your people. If you're listening to this and you're an IT person or a change management person this is the most important thing you'll ever hear on this episode of this podcast is make sure you train your people on how to use Copilot.
This could turn out to be one of the most important tools that you do train them on one of the most important tools that you do adopt and we're I mean this is coming from long time teams fanatics right copilot. It's kind of a big deal. Well we haven't done this in a while. Sorry everybody for that and I'm pretty sure if I remember correctly how we end these episodes it's it's with some sort of positive moment some sort of win. We have a lot of time that has passed.
So we have several moments we can choose from. I'm going to try to be quick with mine but Joe do you have you got anything that you've done. I'll yeah and I'll I'll for mine I'm going to stay on topic actually using copilot to do something that was pretty far out of my my skill level. I have a department that was had a third party company help them develop a WordPress site.
It's actually a WordPress an application built on WordPress and I was working with them working with this third party company and part of the part of the application required using some of our my university's identity system in a certain way that this company doesn't have access to. Had to make some assumptions on how that would work and we kind of first rolled out this WordPress site. Some pieces of it weren't working. I roll out WordPress all the time for my users.
I I know WordPress and how it works and the innards of it. But I'm not a PHP developer. I'm not a coder who writes PHP plug insurance and software. But I know I know what they look like. I know kind of how they piece together. So I used Microsoft Copilot to help me write some functions, help me write these things called short codes, and integrate them into the work that was already produced and basically laid out for us to fix a few things that weren't working and and integrate those.
We didn't have to go back, you know, open the contract, back up with this company and and hire a few more hours of work. I was able to I knew just enough to be dangerous and I wouldn't say fix things but I guess basically make them work the way we wanted to and and expected and it's all worked out but I did it just straight and copilot.
I knew I knew what kind of logic behind to ask so I you know I I needed my own logic and intelligence a bit to know that what it was producing for me was was going to be accurate and probably going to work. Do you think that saved you time? What would the alternative have been? It would have been you know getting back, turning that contract back on going through purchasing, doing some work like that.
So I mean it would have been the work that these folks they probably would have in their sleep probably could have written this and redone it, but the process it would have taken to kind of re establish that would have taken longer. So I mean and for me it was learning experience too. So I mean I didn't I got something out of it also which I think is again something that we talked about earlier in the in the podcast is this will teach you.
And you know I I feel a little more confident opening up something that's not working in in yeah we had code or we had comprehension creation and collaboration. I think we should add coaching. You know, going to make up my own and go rogue. Going to add my own bucket. Yeah, there you go. Yeah. Well, I don't know about you, Joe, but I I mean, who doesn't love opening up a purchase order in the system? And you know, well, good on you for doing that.
Yeah, PHP, not a language that sys admins are typically working in, More of a developer type thing. So yeah, that's definitely, definitely out there. My win has nothing to do with copilot really. I actually tried to shut off. I tried to close my brain down for a while and had an opportunity that my wife had been dreaming about doing that We've we've just the the right place, right time, right moment, right. We were able to plan a trip and take our kids to go see some really cool stuff.
We went to. I had to fly, so wasn't super cool about that. But I survived. I made it, made it to Arizona, and we went and did some hiking in Arizona. We went to Nevada. In between, we went and saw the Grand Canyon. So I don't like heights. I don't like flying. Grand Canyon was kind of terrifying, if I'm being honest. Like it's a gigantic hole in the ground. Holy smokes. What is it like 4000 feet or something like that? It's it's a lot.
But wow, it's beautiful and just, you know, really, really happy that I got to provide my kids with that memory. They get to see that stuff. They get to experience that stuff. Did some really fun hikes in Arizona, went to some Hot Springs. That was really cool. We don't have those in Florida. I'm pretty sure our springs are always like 72°. This one was like 125 or something like that, so enough you could get into it and wade through it. So that was cool.
And yeah, just had a really good shutdown for a while so 'cause, you know, you're doing this for a while and your brain gets kind of frazzled, so needs a good hike through a hot spring. It's that's AI don't know. That's a beautiful, beautiful part of the country that I had been recently for a NETSI conference, National Extension Technology Community Conference was out there in Arizona.
So it was fresh in fresh in my mind, too, that it's, yeah, gorgeous, beautiful and staring down to the bottom of the Grand Canyon. Sometimes you want. I get that feeling. I just want to kind of hold on to. Oh, yeah, yeah. My body wants to do all kinds of bad things going. Yeah. Not a fan of heights. I'm a baby. Cool. Well, we went a little long this time. We're overdue. You get a little extra content. I feel like it's worth it, though, because this, this is kind of a big deal.
I said that numerous times throughout the episode. This generative AI thing is really changing the way people work in a good way, and so hopefully you took away some some pro tips from this. Some things you can go and try, We're going to try to put some good stuff in the show notes. Joe always does an incredible job of making sure there's a lot of value in the show notes, so check that out.
Connect with us on LinkedIn 'cause there is going to be a newsletter that goes alongside this where we'll do the same thing. We're going to try to put value there along with a copy of the episode. So, but yeah, go out and try some AI. Whatever it is, I don't care. Go try some AI. Share with us. Share with us some of your best prompts. Oh, that's a wonderful thing. See what works best. Yes. When you find a prompt that does something amazing, it's like a discovery. It literally is.
It's like you just found something like a gym, you know, and you dug it up, a golden nugget, something like that. So don't hang on to it, share it, post it online, send it to us in an e-mail, like you know, whatever it is, share it with us. We want to hear about it cause yeah, that's going to help your friends do something faster though. But thanks everybody, We really appreciate your time and appreciate you listening to us banter on about technology.
We hope you enjoyed this episode of The Shift Show. If you found our content useful, please consider leaving us a five star review. This will help others find our show. We'd love to connect with you on LinkedIn and also check out our homepage. Shift dot show extension agents can also find us at AKA dot Ms. slash. Join Cloudy County for monthly training sessions. We'll beam back next month with a new episode. Transport a room. Stand by to beam up landing party.