Welcome everyone. My name is Sue Frost, and I am with case district five, and I'm our educational liaison on our board and we just want to welcome everyone for joining us at today's together and we're super delighted to have this be a joint effort with district five, six and seven. And I will also let Christy Donsbach who's the other moderator today, introduce herself. Hello everyone. As you shared I'm Christy Donsbach I'm an executive director with Lakeland College in Mattoon, Illinois, and also just wanted to welcome you and thank you for joining us here today on, on behalf of our case district five. Thank you. I should remember I, there's my actual my work life to address the case one to yourself. I'm an assistant vice president development at Ohio State University so one thing we're very excited about with this presentation is not just the topic large institution community college more urban or rural so I think we want to take that broad perspective. And for those of you who haven't attended to get together it's really a new effort by case to offer these professional education opportunities virtually to all of our members and to encourage learning but as much just conversation and learning from each other on different topics. So thank you for joining us today. Okay, now we're going to get in. In case you forgot what you signed up for today. It's the topic today is quick wins with generative AI. So you're going to learn discover ways you can use generative AI to streamline your workflow, learn about some of the common gen AI tools and how you can use them to save time and deliver results, and a short presentation will be followed by an interactive discussion where you can ask questions share your own wins, or just listen in, if that's easier on your end. So just a few housekeeping items before we get into the introduction of our speaker today with so many people on the call which is a great thing. We appreciate again you all being here today, questions can be an be entered and answered in the chat. So you'll be able to enter them directly in there, feel free. And as we go along throughout the presentation just to type them in there, and will be brought up between the AI use cases, along with some time at the end as well. Any unanswered questions will be used to determine topics for future calls, as well as additions to the resources that cases providing. So please again just use the chat box to give answers to the questions or to share some of your experiences to quick outline of our time together today we're going to have a presentation here starting in a momentarily and that'll be for about 3035 minutes, we'll stop in between for questions and again we'll save some time at the end for questions. And finally, please note this call is being recorded so if you have to hop off Don't worry, you'll be able to access it later. In order to access it will actually be the exact same link that you registered to for today to get signed on, and it'll be right there So, without further ado, I'd like to get in to the introduction of our speaker today Mr. Patrick Kelly, who is the web content manager of Harper College, Patrick has 24 over 24 years experience in web marketing for higher education and healthcare organizations, including Rush University Medical Center Swedish Hospital and Loyola University in Chicago. Patrick currently serves as the web content manager at Harper College overseeing the colleges website Harper college.edu. So Patrick, thank you so much. We're really thrilled to have you with us here today and we can't wait to learn all things AI from you. Thank you so much, Christy and thanks to Sue and Jacob for helping put on this event today. Really appreciate it. And thanks to everyone who attended today. So, without further ado, we'll jump right into presentation hopefully you can see my screen. And so, yeah, assuming Christy mentioned. I'm Patrick Kelly web content manager at Harper College. Trouble advancing my other we go. Okay. Yeah, I work in. I've worked in higher education, health care my whole career. I'm overseeing a website now and I've got a team of three full timers to part timers and to student aides. So, you know, we do all aspects of the Harper College website. You know we do the design we do all the editorial content. And we do all the coding and development ourselves and you know so this is a really small team for that kind of responsibility. We're also kind of expanding our reach and we're starting to do more UX stuff. We're doing like student surveys and focus groups and things like that and that, of course, takes a lot of time so you know AI is coming along at the right time for us, because it really has enabled us to just, you know, bring in some efficiencies into what we're doing and and enable us to have that time to do that extra stuff that comes along. So, so yeah, it's my challenges, very limited staff time, you know, like, like, we're really dependent on part timers and student aides to get a lot of our stuff done. So, anything new that comes along which, you know, the web is like, there's always something to coming along there's always some privacy regulations or accessibility changes that we need to keep up with. And, you know, our staff has remained fairly stable over for all the changes that have gone on in the past several years so having AI really helps helps us go further. On our team. We all, we all, you know, have our own areas that we work on. So I've got a developer. He does development stuff he does JavaScript and API's and things like that. Things that I or anyone else on the team really can't step in and help with. So in those cases having having access to chat GPT and AI really helps because I can, you know, will allow me to do a little bit of development so I'm not bugging my developer for every single little thing that I need. Another challenge that we run into is, you know, when we have a problem with our website. There's just not a lot of good help out there. We can't help websites like stack overflow for technical things. And I can tell you sort of kind of how you might approach the problem, but rarely does it tell you exactly how to solve the problem. And with generative AI, we can get like exact answers, I can get code that I can copy and paste and use on my website. Again, without bothering my developer. So, AI has just been a game changer for us. I kind of mentioned some of this but it does help write editorial copy, it helps write code. I've used it to generate ideas. It can actually perform edits directly to files. It does a lot of those things which is very handy you know especially as my role the team manager, I, you know, I'm making a lot of proposals I'm doing a lot of kind of administrative work making talking points for people presentations, and sometimes just taking like a big document and generating a summary of it is just very helpful to save hours really use it to edit images. And there's just so much more that I can do. It saves me time saves me incalculable amounts of time already I couldn't even begin to process it. And it, it produces quality work now I'm sure some people are kind of smirking at that but because it does it, you know, like chat GPT will tell you the wrong answers sometimes. But I guess I'm thinking of it as like, it gives me a quality first draft that I can use, you know, it, and I can build on that. So I see it as a real plus even if it does get some things wrong, or, you know, sounds a little cheesy sometimes I can work with that. So, what we'll cover today. Some ways generative AI can help you be more productive. Some easy to use generative AI tools that you can just jump right into and start using. And some real examples of how I've used generative AI in my work. So here's some generative AI tools. So chat GPT is kind of the big one there. I would say the market leader you know they everyone's kind of trying to be like chat GPT. And in my opinion, it's the best. Now I haven't used them all, but, but of the ones that I've used I think it just does a big range of different tasks. And it's. It, you know, does everything I needed to do. Microsoft Copilot which kind of runs on the same technology more or less is chat GPT. It's got kind of a different flavor to it. One of the nice things about Copilot is that if you sign in through your company accounts. There's some additional data protection that you don't get with like chat GPT unless you're on a, like a high tier plan. So that's, that's one of the nice things it is a little bit more limited. Like for instance you can't upload files to it. So, you know, it's, it's okay it does a pretty good job. And it's a good starter. One, I would say, Google Gemini is sort of their own AI tool, and they're trying to catch up with chat GPT and and Gemini is pretty much the same as chat GPT does a lot of the same things. It costs about the same for the advanced packages. So, you know, it's a, it's a pretty good option to. There's Adobe Firefly. Now this is a little different from the other ones like this is just focused on creating images. And if you have the Adobe suite. You know, you've already have access to Firefly. It is also baked into a lot of Adobe products, so you can, you know, edit your images in Photoshop using AI. And that's Firefly that's powering it. There's also a standalone tool that you can upload photos to and edit things. There are a lot more there are thousands and thousands and thousands more, there are more and more coming every day. And they're all a little different. It's kind of fun to just go around and experiment with them and see what they can do. Alright, so every time you're using AI for anything, you have to type in some kind of prompt or, you know, you can even now speak in to like phone for instance and that's your prompt, but you do have to give it some kind of a prompt. I just have some tips for prompts, you know, there are some. There's some great guides out there that that can kind of walk you through it. And, but those seem a little complex to me, you know that like you're doing some really advanced things. Those are helpful but I think the best tip is to just talk to your chatbot like it's a person. You know, it, it can actually kind of follow along a conversation. And, you know, the more use it the more you find yourself kind of talking to it like it's a person but it, it's not like other technology interfaces, where you have to approach it in a certain way and you have to speak a certain language to get it to do you know what you want it to do. It's a little more flexible it can understand different ways of phrasing things, there's been times where I've severely botched my prompt with like terrible spelling errors and bad grammar and it's still like understood what it's talking about somehow. So, it's, it can be, it can be a little bit like a person. When you're talking to it. It's kind of scary. But, but great. Explain what you're trying to do, and what results you'd like to see, and that's a really good way to get what what you're looking for is just tell it up front like I want, I want it to look like this I wanted to have this in it and etc. That's, that's a great recipe for getting what you want. If you're working through something you might get errors. I see this with code, especially if I'm generating something like our spreadsheet formula. Sometimes you get an error in the output. I would just copy and paste the error and put it in there and say, I got this error What should I do, or can you rewrite it. And it usually does a good job with that too. I would say if you're doing something complex, like you're writing program for instance or writing, writing some code. I would say start simple and try to build in the complexity with follow up prompts. I can just do this one thing right now. And now okay, now that you've done that now, now add this now add this. And, you know, that way, it's not taking in too much at once, where I think it can get over it, I think they can get overwhelmed. And that leads to the next point. It's okay to ask a lot of your chatbot, but just manage your expectations. They're getting better all the time, but, you know, they're still not perfect. And there's just, you know, I'd say if it can do what you need it to do, it's a big win. And if it can't, that's like, oh, well, you know, I guess you'll have to do it the old fashioned way. So let's just move on. So I have a challenge for everybody, for your next task, no matter what it is, could be work task. It could be like a personal thing in your personal life. Just try it with generative AI. And again, it's not always gonna give you something great or something you can use, or it's not even gonna tell you the right answer or correct information all the time. But I would say like most of the time it does, and it can be pretty good. It can be scary how good it is sometimes. And you get interesting results. The other benefit is you just get used to using it. And I would say, start small. If you haven't used it already, try something really easy, really small. And you'll get to know a little bit about how to interact with generative AI. All right, so some use cases. You know, you can use it to do math. You can use the right web content, edit spreadsheets, develop executive talking points, create how-to documentation, edit photo. These are the kinds of things I'm gonna be covering today in the demo part. So, and that takes us to the demo part. So I'm just gonna pop right into ChatGPT here. They have changed their website recently. It's now chatgpt.com. So you see it load here. It's got this big wide open area here. And down at the bottom is where you type in your prompt. Over on the left here are some of the past chats that I've had, you know, and it actually generates these headings automatically. But you can kind of look in and see, you know, for instance, an old chat that I had. So up here is the version of ChatGPT you're using. And you all are very lucky if you're just starting out in generative AI. Yesterday, they released a brand new model of ChatGPT called ChatGPT 4.0. It's not a zero, it's an O. And it's, you know, really fast. I haven't totally dived into it, but it's got some new features. But, you know, what is really interesting is they made ChatGPT 4.0 free. And that has some of the, that is like some great features in it. So you can use some of the more advanced features like uploading files and, you know, things like that. So ChatGPT 3.5 is still available. And it is, in my opinion, not very good. So, and here's an example of something that was generated in ChatGPT 3.5. I asked it, what is the tuition at Harper College? And the answer is it has different rates of tuition depending on whether you're in district or out of district or out of state. These are subject to change. Why don't you look it up? So that's not like a great answer in my opinion. It doesn't tell me what the dollar amounts of tuitions are. It just kind of says they have tuition at Harper College. Not very helpful. Here's an example of ChatGPT 4.0, how it answers the question. So one of the differences is actually looking out at harpercollege.edu, finding that information on our website and kind of reporting it back here. And here you can see that it does actually give you different dollar amounts for that. So the great news is you used to have to pay for this level of ChatGPT, but now it's available for free as of yesterday. So that's really cool. Yeah, so there are some other advantages to this level. You can add in, there's like a store, so you can add in different modules, et cetera, and make it like kind of extended with extensions. So that's pretty cool. So why don't we get into some, I'll get into kind of a first use case here. And this is just really easy stuff. You saw how I could look things up on the web. Well, one of the thing I like to do is math problems, like who wants to think about like what the formula is for percent differences or whatever. If you use that kind of formula every day, it might be second nature to you, but if you don't do a lot of math in your life or you're not very good at math, you know, ChatGPT can really make a big difference. You don't have to know exactly what to type into a calculator or how to program a spreadsheet to get kind of a good answer to your question. So I typed in here, my team did 180 projects last year and 571 projects this year. What is the percent increase? So I'm giving it a word problem. Again, let's pretend I don't know anything about math and I don't know what to type into my calculator to get the answer. You know, I don't know the order of the numbers or divide, multiply, et cetera. This makes it really easy. First thing ChatGPT does is kind of explains how to do it. And you can just take that. You know, if you're really interested, you can find out how to do it for yourself, but mostly you're probably just looking for the answer. So there it is, it gives you the answer. It's nice how quick that was. And it gave me an answer pretty quickly to my question that I can then, you know, just copy that number and put it in my presentation or whatever. You know, some of my favorite uses for it are like personal uses. Now I pay for this account, so I get to decide if I can use it for personal uses. My employer doesn't see this at all. So anyway, you know, I've been doing some biking this year. I'm trying to get more into it and I just want to do a little traveling. So I said, find me some dedicated bike trails in the Midwest where I can ride about 20 miles with minimal riding on the streets. So let's see what happens. So it is currently generating an answer and it's coming up with some different trails around the Midwest that have long stretches of segments that don't go on the road. Now I could have found this by Googling and I could have typed that into Google and found some articles. Maybe there's some that talk about great Midwest trails. There might be some other ones that, you know, talk about, you know, great trails that don't go on the road a lot, but there may not be any articles that talk about the exact thing that I was looking for here. So this is what's so great about ChatGPT and other AI tools is like, you can get really precise about what you're looking for and it'll give you an answer that's reasonably correct. So I guess we can stop here and just see if there are any questions so far. You know, Sue, I think you're talking and you're muted. Questions now, but we've got some great examples and we appreciate, yeah, having the virtual conversation with people talking about, they've used it, for example, to do a draft of a donor thank you letter. I think this is great. Taking engineering and physics information and using it as a way to explain less complicated research in the summary, data analysis with quantitative and qualitative data, meal planning, love that. Let's see, and write contact reports to summarize notes. There's somebody saying I'm on ChatGPT right now and it's telling me 3.5 is three, but 4.0 is $20 a month. I don't know if you know any more about that. And then also another 4.0, not zero questions, what is it trained on? So I don't know if you've got 4.0 or 4.0 questions there about price and what it's trained on. You know, the ChatGPT just announced yesterday that the 4.0 would be free. So maybe it hasn't rolled out completely yet. I'm currently using the paid version. So it's kind of hard for me to see if like, I have access to it, but it should be coming out soon. So I guess I would check back on that. But yeah, this just came out yesterday, this 4.0 version. And then there's a question too around this, just questions with using AI ChatGPT about privacy concerns with information you're putting in, things we should think about regarding security issues or how to use information to not be misused, including for their use in the model. Yeah, that's always a concern. So what I would say is, you know, you have to use your best judgment, you know, on your own personal data, as well as your company's data. You know, if you're uploading a file, for instance, you should really make sure that you're only giving them the data that's necessary to do the task. Like for instance, if you have a spreadsheet, you may want to, you probably should remove all the names and email addresses, and maybe just have like some kind of identifier if it's got like personal information in there. And I have some examples of that later that I'll get into. But there are some models or some products that do offer a degree of privacy, as I mentioned, Copilot. If you're using your company's Copilot account, then they don't use that information for training their models. If you're using the free version of Copilot, then they do. And with ChatGPT, there's like a Teams level where you pay, I believe it's like $30 per license, but you have to have two licenses, $30 a month. If you pay for that Teams level, you have a degree of privacy there. Like they don't use your data to train their models. And you also have ownership over both the prompts and the outputs. So you could use some web content, for instance, that gets generated without attribution to ChatGPT. So yeah, good points. And speaking of generating web content, so I'll move into my next example here. So we had this page as our real estate broker page. This is kind of in a program. It's not, it doesn't get a lot of attention or love, at least, it's kind of like a half credit program. So we have the credit side of the house, which gets all the attention. And that's where the focus is. And then we also have community education, which has its own audience and their own people. But this is kind of their very niche, like in-between kind of type program. So they have a very small staff and they don't have like a dedicated web person who manages their website and can really promote them well on the web. So they worked with me on redeveloping their website. And this is what we started with. It was just like, this is all there was about the real estate program. This intro didn't really kind of sell the program at all. It just says, here's what you need to do for licensing. And then there's some information about the classes and stuff. This could be doing so much better. It just needed like an intro. And so I would go into chat GPT, hold on. There we go. And I would type in a prompt that's like, write a new intro for this page. I actually did this in Gemini, but GPT is close enough. Wait a minute, hold up. Something. OK. So this isn't exactly what I came up with in Gemini, but it gives you something to work with that you can kind of copy and paste into that page. And it has a little bit more about what the program's about. It talks about some of the strong points that the experienced instructors and flexible course options, et cetera. This is actually what Gemini actually did come up with, which I thought was really good. It had some nice bullet points. I've worked with this. There are actually like 10 bullet points. So this isn't exactly what it came up with. I narrowed it down to just the important things, the things that were true about the program. I liked that it covered some of the areas that were mentioned down here in these courses, but it's kind of hidden, et cetera, et cetera. So having something like this on the page really helps. It helps with SEO. There was a huge increase in SEO traffic from this. I forget the exact number, but I want to say it was like 70% increase or something like that. And we did this across all the web pages and in just an afternoon. So we did like 20. We had 20 programs we had to generate content for, had it all knocked out in just an afternoon instead of a couple hours each page writing these programs. So it's really helpful for that sort of thing. Just going to jump into another kind of request that I would make of chat GPT. So let me get back in there. So I'm going to copy this prompt in here. So I'm actually going to add a file. These are all our web requests from the past couple of years, dating back to 2020. And it says manager. I'm like, I want to know what are some of the top level statistics about these requests. How has the number changed? This is good for reporting purposes. I'm always trying to get a little extra for the web I'm always trying to get a little extra resource or a person added to the team here or there. So it's nice to have some top level numbers. And I've just pulled in our requests. Again, I did scrub out all of the names and everything. And I just have the text of the requests there. So it is taking a moment. While we're waiting for it to analyze, I'll just show you the file really quick. Oh, here it goes. So you can see it kind of gave me the totals for each year. I could have gotten that easily. I could have gone in the spreadsheet and highlighted all the rows and logged it myself. But it's nice that they did it for me. It says it has some observations or steady increase, et cetera. It's really kind of generic, little generic top level. If I want to get a little more specific, I can kind of ask a follow-up question. I know, so here I'm writing a description of the requests is in J. And there's other information, EFG and I. Those are the columns in the spreadsheet. Using this information, summarize what percentages of requests relate to what system. Omni, CMS, which is our content management system, Formstack, another thing we use, Localist or Calendar, and Concept3D, our maps. So I'm asking it now for some really specific information. And I'm sort of phrasing it like a follow-up question. I didn't have to restate everything. I just wanted to make sure that I was getting it right. I didn't have to restate everything. I didn't have to re-upload the file because it kind of remembers from the last prompt that I did. So here it's kind of breaking this down. It's giving me Omni, CMS, Formstack, Concept3D, et cetera. Now, there's nowhere in the request that says, please perform this update in Omni. Like, ChatGPT is looking at the request and saying, oh, yes, this is related to Omni because it's asking for some update to HarperCollege.edu. Oh, this is Formstack because it's asking for an update to a form. So ChatGPT is kind of making those logical connections, saying, oh, this is Concept3D because they asked for something about the map. And it's able to process that and give me numbers. Really handy, really handy. So now I can ask it another question. Let's ask another one. So based on this data, what topic should we be focusing on for training so we can reduce the number of requests? And it's responding with, some of the top requests for Omni are how to add, remove, and update content efficiently, how to handle text updates, how to create a page, SEO, Performstack, slightly different things, steps to create new forms, and so on. Integrate forms of other tools, et cetera. Update campus locations on maps, et cetera. So this gives us a good roadmap for topics we want to hit home with in our training and make sure that people understand. Again, with the goal to reduce these types of requests. So I'm going to move on to the next slide. So let's see. I'm going to do one more. And for Omni, CMS updates, categorize the types of requests and give me the percentage associated with each issue. So it does take a little bit to process. But here, again, well, this time it says other is 87%. That's not very helpful. I did this earlier. It gave me a very different answer. It was mostly text updates. That's the fun thing about these tools is you just really never know what you're going to get sometimes. But yeah, I wonder if it's something different with this model. Anyway, it does kind of look at each of the requests and try to decide which category it falls under. It didn't do such a good job this time. I'm just going to say it. But any kind of questions? I know that some people, sounds like a lot of people are already using it for data analysis, which is really great. I think it's one of the best features of generative AI. There was one question, Patrick, that just came in that asked if you have any examples that relate to alumni donor relations or marketing communications. I can speak quickly to a quick donor piece that we've used it for our planned giving. We do a monthly e-newsletter. And so we will actually use chat GPT to send in our last month's newsletter and help update it based on all the campus happening since then to tie in the campus updates too. So that's one example personally that we're using it for here. But Patrick, I'm sure you have some examples how your college is using it too in terms of alumni. I don't have anything directly related to alumni, but I have something that might be kind of interesting for alumni. So this is a spreadsheet of our graduates. We create profiles on the web for them. And we ask them to answer four questions. What does graduation mean to you? What's the next step in your career? Who would you like to thank? And give a shout out to a faculty member. So part of my task this week, it being graduation week, is I'm reaching out to all these people who are mentioned here. Professor Bethany Statler, et cetera. It's very inconsistent how they're mentioned. There's one. Here's Lisa Gustow. They can be in this column too. Sometimes they're thanking their mom or the boyfriend or whatever, or just friends by name. So it'd be very hard to tell what's a professor, what's not a professor, who's getting thanked, what programs are being thanked, et cetera. So I try this in chat GPT. Just if I could create another column that just has all the names that I have to reach out to and say, hey, this person thanked you in their profile. That helps me. So if I put in a prompt like this, attached are some answers to some recent Harvard College graduates, some recent questions Harvard graduates submitted. So I'm giving a little context. They talk about what graduation means for them, et cetera. The faculty and staff might be named in column E and F. Column G, can you create a specific list of people that are mentioned? And leave out anyone who appears to be a family member or friend. So it processed this. And I'm just going to open up the spreadsheet really quick. And it did exactly that. It added these two columns. And it just pulled out the names that were mentioned here. And it made it really easy for me to find the people that were being thanked. So that's one way to use it. I will mention also that there is a lot of other personal information again here. And I just removed all that and just replaced their names with a key so that it protects their privacy. But in terms of privacy, these profiles do get published on the website. It's not like super sensitive data. I just wanted to point that out. So you always have to use your judgment, though. Well, let me jump into the next one I've got. Because this is kind of a different product, I guess. This is WizardShot. And this is just one of the examples of the millions of AI products that are out there. But we do a lot of tutorials. Or sometimes we answer questions like how to do things. So you can install WizardShot as a browser extension. And you can kind of see it's this pink icon up here. And you can open up a file here. So let's say I wanted to explain to someone how to make a new file in my content management system. Now, I'd have to type out all the instructions and say click New in the corner, and then fill out this form, et cetera. But I'm just going to make the AI do it. But I'm just going to make the AI do it. So I'm just going to click on that and start capturing. It's going to record what I'm doing here. So I'm going to click New. I'm going to click Interior Page. I'm just going to fill out the form. Voila, we've got a new page, right? And then I'm going to stop recording. And what WizardShot does is it automatically generates screenshots for every important thing that I did. So you can see it's highlighting the button that I clicked. It's giving me some instructions saying, click here and click on this interior page. Type in AI rocks. I could go in and edit that and make it say something else. And I could fill in the form, type in blah, click Create. And then I could send this URL directly to somebody who needs instructions on how to do that, which is so handy. And we've started creating some documentation that we're going to use with this tool. So it's really helpful. If you have someone who's a little more technologically challenged and you can't see their screen, they don't know what you're talking. It's hard to explain where the button is. This kind of thing really helps a lot. So we've been using that recently. I just found out about this a couple of days ago. So it's a great tool. Yeah, I guess I'll stop here. Are there any kind of questions? We're sort of getting towards the end. I have not. I have to admit, I've not heard of wizard shot. So that is awesome just to see kind of on the forefront there and to be able to see how to utilize it's helpful. There were a few comments that have come in as far as going back to alumni and other types of uses for it. So just wanted to share some of those that one approach was to use for solicitations that it's really helpful in refreshing yearly communications, year-end solicitations, using it to format university statements and bullet points. So those are all great uses. Another person shared they use it to create first drafts of donor thank you emails. That is an awesome use of that just to kind of get an initial draft and tweak it from there. And let's see. It looks like there was another question that just came in. Diana from Massachusetts. Hello. She shares she's very excited to use AI, especially to revamp acknowledgement letters and other mailings. She's wondering, has anyone found interesting ways to utilize AI and advancement services? Yeah, and again, I don't work too much with our advancement people. But after giving a presentation on AI at my college, the person who takes our board minutes came up to me afterwards and said it's so helpful for her because she has to summarize the board presentations. And she could be working with documents or video or something. And she uploads that into a chat GPT or a tool like that and can generate a summary of what that is that she can then put into the board minutes. So I thought, again, it's like a really great time saver. She says it saves her so much time. Patrick, to your point and the question about how people are using it, I've learned this more from a faculty member who at the end of each semester, she used to do a summary of student evaluations of her teaching. And she found it was an accurate summary with highs and lows. And some of the units I've worked with, we've translated that to after we have a large event. You can see the homecoming or all calls around reunions. That's going to be a way we'll look at the in-depth. But by putting it all in there, it's a way we can get a summary of a survey with what worked, what didn't from evaluations too. Yeah, yeah, totally. Another thing that might be helpful for people is I had to do these executive talking points recently for a product we were purchasing called BrightEdge. And all of our purchases have to go before the board trustees. We're a public college. We have an elected board of trustees. And you just don't know what kind of questions they're going to be asking when this thing comes before the board. So I used a prompt like this that says, our board of trustees is considering contracting with the SEO platform BrightEdge. A copy of the presentation about BrightEdge is attached. Give me some top level bullet points for executives and board members on why BrightEdge is a good investment. And I can attach the file there. There we go. Being a little fussy. OK. So I actually did this on a call with the BrightEdge sales people. And they were pretty convinced that ChatGPT would not be able to produce a good response, which actually it isn't right now. It's just stuck here. Let's see if I can find the one that actually did this. Yeah, so when I did this for them, they were blown away because these are all the points that they would brought up. And it uses the information and the deck that they gave me to generate this response down here. So I could just more or less copy and paste these talking points and hand them over to our executive leadership. It's just a little bit of focusing. I also did a follow up question. I took the grant proposal that I created. I wrote this grant. I asked for $50,000 for one year of this platform. And I had all my justifications and data in there and why this is a good investment, et cetera. And I said, using that information, give me some bullet points on why BrightEdge would make a difference for the Harper College web team. And again, it just distilled everything I had written into 10 bullet points that I could then, again, hand over to executive leadership that they could use to answer any questions that might come up in the board meeting. Again, a great time saver. I mean, I could have written this myself, but that would have been like getting back into my headspace when I wrote that a year ago and sorting through all that information again and distilling it in a way that is easy to relate to other people. I guess there's just a couple other things that I wanted to touch on. And this gets back to that challenge that I mentioned earlier, is for your next task, use AI. Try doing it in ChatGPT or something like that. I'm going to just show you an example. And this is a program. I don't expect a lot of people on this call have access to this or use this program, but this is just one of the many programs out there that is integrating with AI. This is a vendor we use called Siteimprove. It scans our website, and it flags accessibility issues for us. These are the issues that it has noticed on our website. Sadly, there are many, but we're working through them. But you can see over here on the right side, this is like the universal AI icon using more of these stars. So this shows that this issue can be resolved with AI. So let me just get in. If you open that up, this is what it looks like. It highlights that area that has a problem. Now, in the olden times, they would have an example of the types of code that would solve the problem, but it didn't really apply to our website. It was generic code. Didn't really always work with what we were doing. We had really specific things. But now I can use AI to generate some new code. It'll give me the exact code. Here's an example of the generic code that we used to get. Seems to be going a little slow right now. But this is so handy because I can just get exact working code on my site. And there's other ways that you might notice AI creeping into your life. I hate to put it in those terms, but here's an example. This is my iPhone screenshot. I got these flowers for my stepmom for Mother's Day. It was a huge basket with many flowers. But we're like, what are those flowers? And didn't say on the label. But if you have an iPhone, you may have noticed that there's a little icon down here that has information. And you can see those stars there. And you can see that that's something to do with AI. So you can use it to identify what the flowers are. And so click on that. And it turns out this is edging Lobelia, maybe. I mean, that's AI's best guess. But it was at least an answer. At least we could figure that out, see if that's true. So it's just another way. If you see those little stars on something you're using, this is really handy. Not to end on a very gross note, but I used this just like last week because I went on a long bike ride. And I came back, and there was a tick on my leg, right? So I pulled the tick off and put it on the table. And I'm like, I better, before I totally squish this thing, I better figure out what this is. So I took a picture of it. And it identified the kind of tick it was. And then I could look it up and see what diseases I was in for. So I mean, it does identify animals. It can identify dogs and plants and all kinds of things. I find it's really interesting and cool to use. So just watch out for those little ways that AI is kind of creeping into our everyday lives. It's fascinating. So we have a couple of minutes left. Are there any last questions? Hey, Patrick. Patrick, this is Brian. Oh, should I just type my question, maybe? Or do you prefer if I type my question? Go ahead. We're wrapping up. So go ahead and give it back. I just tried the chat GPT 4.0. And I asked it if it could create an image for me. It said yes. And I'm using the free version. And apparently, it created a ping image of what I asked for, only I can't see it. Have you had that problem or something? Or I suspect I have to upgrade. But I did create the ping. I just can't see it. It's interesting. There might be a link to download it. That's usually how it works. But I got to tell you, I don't create too many images in chat GPT. So I'm not totally familiar with that issue. And also, like I said, they're in the midst of rolling out a new version and upgrading. And it might be just one of those things that you might want to try in a couple of days and see if you get the same result. We're going to thank you, everyone, for being part of this. And I know joining us across multiple districts. And we appreciate even just the conversation happening with the chat. I know as we look forward to other future sessions where we've got this larger group, we'll probably create the opportunity to do breakout rooms. One thing I'm putting in the chat right now, and you'll also get as follow up, is CASE has developed a white paper about the use of AI and advancement. So it is in the chat, but we'll send it in the follow-up information as well. So thank you for joining us. And I want to give a special thank you to Patrick to doing this. It's just incredibly for sharing that. So, and I said, this is one of many conversations we hope to, we will continue to do with the together. So we encourage you to continue the conversation. Thank you for your feedback. Because as Christie said, we'll use that for future sessions. And again, thank you for joining us today. We appreciate it. Thank you. Thanks everyone.

Sue Frost

Christy Donsbach

CASE District Five

professional education

generative AI tools

Patrick Kelly

Harper College

AI applications

advancement services