 All right. Good morning, and welcome to this week's edition of Encompass Live. I am your host, Krista Porter here at the Nebraska Library Commission. Encompass Live is the commission's weekly webinar series where we cover a variety of topics that may be of interest to libraries. We broadcast the show usually, normally live every Wednesday morning at 10am central time. But if you're unable to join us on Wednesday mornings, that's okay. We do record the show, and then it is posted to our, in our archives for you to watch later at your convenience. And I'll show you at the end of today's show where you can access all of our show archives. Both of the live show and the recordings are free and open to anyone to watch. So please do share with your friends, family, neighbors, colleagues, anyone you think might be interested in any of the topics we have on Encompass Live. For those of you not from Nebraska, the Nebraska Library Commission is the state agency for libraries here, so similar to your state library. So we provide services and training and resources to all types of libraries in the state. So you will find shows on Encompass Live for all types of libraries. Public, academic, K-12, corrections, museums, historical societies, and on and on. Really, our only criteria is that it's something to do with libraries. We have Nebraska Library Commission staff that sometimes come on and do presentations for things for everybody, or things that are Nebraska specific, but we also bring in guest speakers. And that's what we have today, sort of a mixture, mainly a guest speaker, a guest speaker with a guest. But Brian is here. He's been on the show a few times before over the years to talk about lots of techie related things. And he, I'm going to let you do a full, more full introduction of yourself, but he is going to talk to us. We reached out to him to talk about AI, things that are going on in that kind of World Artificial Intelligence, artificial intelligence, IT, possibly things like that that people are like, I don't know what's going on. But he's going to try and, you know, hopefully make us feel better about it, despite the title of the modern day Pandora's Box. So I headed over to you, Brian, to go ahead and also with us joining us today to chime in to is Amita Sweet who's our technology innovation librarian here at the Library Commission in case you're wondering who that other head is. So I'll just hand it over to you, Brian. I will go ahead and get us started. And my quick, my quick little intro is Brian, I'm the director of strategic innovation here at the Apollo project. And I like finding new technologies and seeing what they do and many of them in various organizations, usually in the education space like libraries and schools. I've been playing with AI for a few years now. And so this is very exciting that I no longer have to argue about the use of AI because everyone knows what it is. This is less stressful conversations for me. And so without further ado, I will go ahead and do this welcome. So I'm the modern day Pandora's Box. If you have questions throughout the event, definitely post them into the questions box. I'll try to answer them. I found my other screen over here. So if you see looking, that's why you have your right into questions or you're like, what is he talking about? Or you want to argue or challenge? I encourage all of that because I don't like presenting straight. I like interruptions. So please interrupt if you do have comments. So I'll start off with these types of slides when I talk about AI. What is actual artificial intelligence? By definition, it's the theory of development of using computer systems to handle what humans typically performed, whether that be basic tasks, visual understanding, speed recognition, making choices, or even having translations on between different languages. A lot of data, a lot of AI is based on data or using machine learning to make decisions. So machine learning is a subset of artificial intelligence that mainly uses algorithms and statistical models to enable a computer system to make decisions based on the data that it has. And so I stole this lovely slide from Amanda, who's with us today. She does a whole bunch of really great infographics and I linked a source webpage as well for you to kind of grab more of these. This is a really great document explaining the various facets of machine learning and how machine learning can build data, build content and make decisions. It is a really good infographic. And so, back in April, I was in Florida and this is always, I find this funny. And so I travel a lot with like a stack of iPads and a bunch of other things. I do hands on projects and demonstrations. It's all my iPads are connected to the internet and suddenly I just made like Twitter news. So it's legit. At what time was that at 445 am all my devices in my hotel room in Florida went off at the exact same time with that like emergency sounding beacon. We were like in a war zone. I like to not understand what's going on. That's because all my devices got the emergency alert. And so, and then that conference I was talking about AI in the morning, so I was very tired. And then guys, if AI existed, we may have been able to avoid this, which was a simply just a AI doing a test or not a person doing the emergency system, and instead of sending it to test users, they send it to everyone in the area. So usually when I talk about AI, I have these two types of people reactions. So hopefully not the one on the right, you need to be the one on the left. The other thing that I use using AI as a tool on how people typically use AI in different industries and will pull all together and apply specifically to the library world. So, AI in a net shell comes in three main forms you have your human like interactions whether that's agents chat box robots, etc. So digesting a bunch of data and making a determination from that data or advanced automation so making things happen automatically without having a user be intervened and that would be like smart houses and stuff. The success of AI or a special AI tool doesn't really come from a small group of people's like plugging away at a keyboard. It mostly comes from and users using it where they're providing with information is feedback whether or not that information is correct. So if you're looking at AI as important components we use we used to see AI mostly in logic logic and base rules. So something happened. This is a result in an AI would that was like very rudimentary AI. Now we have what's called pattern based or machine learning AI, where over time the AI gets better and better at what it does well that's creating content making or making decisions or looking at data. So machine learning takes that even farther and it's a subset of machine learning where it lets computers make decisions on its own. So without human interaction, it's smart enough or has the algorithm in place, or it knows what factual information what isn't to make better decisions. And then neural networks is a fancy work basically explain how computers talk to each other GPUs and things like that to make those decisions and provide output for that machine learning to learn from or get better as function over time. So artificial intelligence definitely exists. It can do self driving cars, your language translation. There's a company called interpretive that you can have an AI join a zoom meeting and it will do whatever languages you want and translate in real time. And so that's how far we've gotten now. The idea of strong artificial intelligence doesn't fully exist yet and that's the idea that a computer to think at a level that that passes people abstract thinking computers can't necessarily abstract thought. You can't ask a AI to have its own feelings and emotions about content unless if you fed it very specific emotional content if that makes sense. If a computer to be sensitive to the needs of Brian's wells in life, it will do a great job but if Amanda or Christina asked it to be sensitive to their needs it would be like nope, Brian's needs are most important. And so it's still designed by like data. Feel free to steal these slides as well the slides will be shared when I teach like younger patrons how AI works. I like to show them this example and this is called I call it my flashlight example. When you want to build an AI with logic based rules, the code if you will be if you see a dark turn on the light. Machine learning takes that a little bit farther. And the code behind it is the flash so it will make me turn on if it learns other words from for dark. So picking up words that contain the word dark and turning on the light, deep learning takes it even farther and makes connections to when you need to turn on the light. For instance, if I say I can't see in this room, it may go well, it made me need to flashlight and turn on the flashlight in my app. And so that's the best way to kind of explain the different levels of AI in my opinion. So bringing things out farther. When we look at AI, there's various ways for it to learn. There's learning on experience. So the system learns positive and negative experiences as you give it the data it needs, it makes a decision from those experiences. You can also let it learn on by example. And so if you think about like Google spam filtering service that built in. Every time you detect something in spam, Google overtime learns that well, they flag all these things to spam. Here's the similarities between them. So now we know what spam looks like because you keep providing us that information. Deep learning and self learning are the more advanced side of that family. Self learning takes advanced machine learning and makes the decisions by looking at own data sets and drawing conclusions from that. And then deep learning uses a lot more complicated math models to combine to create content to create pictures or even develop speech. So the biggest problem in the world right now and always has been is garbage data and garbage data paradigm, which means you can build and invest the best AI component in the world. But if you give it bad data, you're going to have bad results because the data is not good. Or if you have the best data in the world, but a really bad model, you're not going to have a very good AI tool. So that's why finding the right AI tool is incredibly challenging because to get both of these things mastered perfectly is near impossible. Overall from patterns to automation, the idea with AI is to use algorithms to sift through data. So when we talk about like chat, TPT or open AI, for example, it learns from data that people provide it in order to make those decisions. But there's so much data going into it, it has to take a step back and kind of digest it. And so AI is going to also be transactional. So you get an answer, give a question, get an answer like a virtual assistant. And it can also be automated. So providing routine tasks that, you know, you don't really have to like taking the practice out every day, you can build an AI to do that, that'd be amazing. So here's some very basic examples of what we see AI doing all by a very complex solution or a very complex concept. So AI can be called image recognition. And so it's how we see on our iPhones for those that are Apple fanboys like me. It can detect everyone else's faces, it can notice a different kind of dog and a cat and you can search in your photos for all your favorite furry friends just by typing in the word dog. Or you can type in a person's name and see all those photos of that individual in your phone. Facebook got in trouble for doing that as well. So anytime you post something on Facebook, it would see if your photo exists anywhere else. And I think I'll show you a really cool hack tool that's a little creepy but also kind of neat that uses people's images to find out if they exist anywhere else online. Yeah, it's good stuff. So image recognition. Well, you take that even farther and use this as an automated tool in the workforce. And so Microsoft Azure has a tool set called cognitive intelligence, where it can detect syntax or semantics or personality using AI. And so you can upload a photo and they can tell that person's excited, angry, fearful, happy, neutral, etc. It's a really great way to find out if you're significant other was excited to see you on your wedding day. Upload that photo and find out the real truth behind that marriage. That's a joke. Don't do that. You may not like the results. But using these tools, you can also use those in the workforce. And so the reason why you have Walmart greeters and every Walmart is it's a detractive for people that might want to do something bad like steal or hurt others. And so having like a virtual greeter greeting everyone, it can feed off of if this person is in a good mood or not, and then make those determinations and security questions answered. We also see AI and voice recognition. So you can see when you ask Siri, ask Google Home, Echo, etc. questions, you will get responses from that. And so it's gotten so well that you can ask your Google Home, for instance, how do I get home versus somebody else asking how do I get home, and you'll have very tailored results based on those profiles that were designed based off their voice. So those are all the different things that you can see using with AI in terms of an interactive standpoint. So then we also have a bunch of AI robots have come on over the years. Samsung is probably the one that does like the most like creepy innovative things. They have robots that can pick up trash. They have robots that can serve you food and you push on the order and it goes beep beep and it goes into the kitchen and grabs the things that it needs. We have robots that are for at home people that have like dimensional all timers and provides that emergency assistance if need be. They even have robots that can do dishes for you and set tables. And so there's there's a flavor for everything. So if you just even go on Google and just search for robot assistance for home you'll see a variety of products ranging from a couple hundred bucks to thousands of dollars that can follow you around and finally Brian has friends. So you go to my house with 1000 robots running around they all love me and that's all I need. There's a lot of other forms of AI as well. So we have optional character recognition so I can take a photo of text and it can detect that text. That's how license plate readers work. There's not like some nerd sitting behind a screen blurry photos of license plates. It's all driven by an AI component. Advances of preferences is another flavor of AI where it looks at what you're like buying habits viewing habits depending on Amazon or Netflix and then compares that with other people doing the exact same thing as you and finds out what the differences are between the two. And that's how those recommendations take place and then you have sensory data analysis. That's how your Apple watch or your Fitbit knows the difference of you running or walking or swimming. And so when I go for a run my watch like should we call 911 for you. You don't do this exercise over everything. Okay. That's a joke. And so there's all these other forms of AI that have been developed over the years to make our lives simpler. And here's some humor. Like enough to move. That's why you name all your robots. So we also see AI being used in our in different corporate environments where AI is either being asked questions in a chat conversation to provide answers like how do I set up an office reply what's our Wi-Fi password. Where's the bathroom. So I'm in a way too many like different teams chat for different organizations I work with and I cannot read all the messages all day long. So there's an AI I use called spoke that will read everything throughout the day and then send me anything that I need to worry about as well as some like one or two bullet point summaries. And that's how I get through life. So this point I always like to share whenever I talk about like tools and technology I always like to see how other industries are using it and what works well and what doesn't work well. And so we can see what their success are with AI what their failures are. We can also help make those decisions in terms of our organizations is education or library for community service. I lost my voice over the weekend. I'm still recovering a little bit. I went to an email rave in cave so screaming all day. It's a good way to lose your voice. That'll do it. And so this is AI can be used to look at like MRIs X-rays. And so it can take an image and figure out like is there something wrong to the scale of what a doctor can do. And now here's where it gets exciting. If we think about a company that builds an AI product for a hospital. And that hospital makes money based on doing surgeries. Do you want the AI model to be very strict in terms of well this might be something bad to let's operate. Or do you want the model be like no we've seen that at no time. It's really anything I wouldn't worry about it. In my hospital design based on billables which model would you pick the first one or the second one. Most likely a good business. Re-picking the first one when that's more. Hey this is this be anything we should operate right away versus the model of well. It's really anything. And so this is a really good read. In 2019. Or someone researched. A bunch of AI products that look at. The hospitals are using and whether or not. It's a business decision to generate revenue or business decision to find things that might be overlooked. So put your phone. Pandora's box has been opened. Smart homes. We've been using a lot of people have been using playing with smart homes for a while and let your Amazon's very like so. Some of us like like nerdy people like me will go all out and have the blinds and open and close. The doors lock when you walk away. Looked all driven by AI automation. So I designed the rules in place a logic based rules and it makes this decision as it goes. So the we also see AI being used in drones. Self driving vehicles. And even, you know, pretty soon, but we're supposed to have self driving cars at some point. I saw back in the day. We're at our deadline. So any moment now, we're going to have those two. But we're seeing autonomous flying with drones and people have been looking at using drones for delivery for a while now. But as AI is getting better. And the cost of using computing power has gone down. We're most likely going to start seeing that sooner. Because the reason why it was financially irresponsible. Was the cost of the machines to run the drones with the knowledge and information was just way too expensive. Like you can you can overpay a truck driver and still save money like giving them six figures to drive a truck to your house and it was cheaper than doing the drone portion. I wonder if that would increase like delivery theft. Where you get like a little teenager that just tracks the drone and grabs them. Let's let a little drugs, a little tasers on two and chases people around. I like it. Okay. My world a lot different than everyone else is okay. But that's a good question. But yeah, if there's no how do they. Yeah. But yeah, if there's no how do they. Yeah. Security shooting, shooting the drone down it gets the package. So there's a lot of AI tools that exists for visually impairment. There's a company called be my eyes. There's a contact lens company that you put contact lens in that will like read things out to you as you walk past them. They won an award a few years ago. We have our digital systems and there's other libraries that are using digital systems like Alexis and Google homes to provide quick answers like where's the bathroom because that's the number one question that every library gets is where's the bathroom. Having an AI provide those answers saves a lot of time and energy. So there's a lot of really cool inspiring AIs. Alpha go is a very, there's a game called Alpha go. And it's like one of the super hard to play there's like an infinite number of moves that takes hours to win. And so what Alpha go did by based off of Google is they watched a whole bunch of people play the game and I figured out how to win the movement like or win the game like 10 moves. No one's ever been able to like do that. I remember that's all I remember those pretty cool. And I Ross is another great company so for those that have heard about the chat GPT story of a law case. So there was some airline dispute, and the one lawyer use chat GPT to come up with a bunch of cases that he thought existed because chat GPT told him. So it made up all these court cases for him he went to the court said well you can't do this because this court case that I can. No one's heard of these judge haven't heard it the defendants haven't heard of it so they were like well let's research these and it turns out they never existed so they asked the lawyer that presented them. And he's like oh yeah he's trapped to be T though so they have to be real. And so, if he would have used that works. Yeah right. I did hear about that story yeah and that's similar to like when students are using it to find citations for their research papers. It's not real research is going to make up. Things for you. Well, see that's in my name we had Wikipedia. And I would just edit my Wikipedia article to prove my point. And effective. Gotta make sure my professors around here because I don't want my brain to be reversed but yeah Wikipedia and you can edit that you're never wrong. Yeah, and so I digress so I Ross is a paralegal tool that has been fed specific court cases. So if you're working on a court document and you need to know if there's any court precedence I Ross is the tool for you. And just to show you the mass diversity of different AI tools are out there every year for the end of the year CBI or CB does a report of all the top like AI companies. And so in 2022, these would be AI companies on the map across various industries or different products. And how many use chat GPT to come up with their name. None of them because no one used to get to November of 2022. But they will. So, some fun runaway I like runaway AI is or AI is that you know we're fed guard bad data. And so these are really good proof points of what that looks like. 2016 Microsoft released a, which was an artificial artificial chatter bot. And the goal was to have conversations of users on Twitter and have like this really cool like experience and interaction. But people on the internet like to ruin good things. And so they fed a whole bunch of like negative conversations to pay and make a very angry and aggressive. I think she was a 19 year old girl made her a very angry and aggressive 19 year old girl. And they had to shut it down because no one took the time to teach Tay what was considered good language and bad language. And so it's all about the data that you provide your model and that's what you get as a result. A funny one was inspire robot dot me and this one will create a random quote with a random image and the original design was to have photos match an image that made sense. But people like to ruin good things. So now you get like really random quotes and really random images that sometimes work but usually in a funny way. And so this was considered a happy accident to me. For those that are on speakerphone. I got in trouble for this at a conference because it actually had someone call now and one. So the context for AI has struggled with context. So way back when if you said the s word to your your your Apple phone and said call me an ambulance, it would then go up. Let me call 911 for you. And it would actually dial out 911. It would say hey I will now call you ambulance because it didn't know the context so you asked for an emergency health and said hey Siri call me an ambulance. It thought that oh you mean it thought oh shoot let me let me rename your name you're no longer Brian hey ambulance how are you. So since it didn't know the context, it had to be corrected so now I think if you say. Call me an ambulance it will ask if you need emergency help and then call 911 or you can also say which I found out was hey Siri. That's a call 911 it'll actually do that. So I was at a conference in New York and suddenly my phone's ringing someone else's phone's ringing and it dialed out and so that was humorous to me. That's fun for everyone else. So that's one of the challenges AI sometimes struggle with context. There's also the precision component. When we talk about chat TVT for instance there's a it's precision they call it temperature. You can tell the AI to be make things up as it goes be creative only be very very specific with the specific knowledge it has. And again, you can still flood things with bad data make bad decisions so precision is another challenge with AI and the third challenge is how do you train they are model over time. It's not a no matter what the sales you will tell you there is not a single turnkey solution with AI you always will have to train the model. So if you're running a call center or if you're running a help desk if you're running a like how to do something cool on YouTube description. And people are feeding content to it and asking questions or still requires you to see if those questions that were asked had the right answers associated. So this constant circle of training making sure the precision is accurate and it's understanding context is you don't monitor those three components you'll your AI drift away. You know just like humans learn to better information better decisions, and you can train people with bad data and get a horrible human being. I'm just kidding about that. And so another funny little joke. And so there's even books that were entirely by by AI so if you're having a rough sleep and you can't fall asleep and you're like what can I do other than listen to Brian Ramble on for an hour. You can read this book and it'll put you right to sleep. I have batteries those are entirely by an AI chapters and everything organized by me and this is done in 2017. So it's a really cool use case. But who gets paid from it. The publisher. They keep all the profits. But it was more designed to see what an AI can do to expect all this data. I like to learn like to take things apart. And so if you go to experiments with Google dot com slash collection slash AI, you'll be presented with a bunch of AI activities that you can like edit and tweak and see what happens with it. There's a bunch of other like cute little games you can go which face is real dot com and it generates an AI image and finds a real image online and you have to choose which one's real or which one isn't. There's even AI models that allow you to draw a cat. So if you go to this website, you trace out like an image of the cat and it fills it in and you get this out but I caution you. If you have the inability to draw stick figures like myself and you try to draw a cat you will get something that's very nightmarish with like multiple heads and multiple tails and multiple arms. So, like, make sure you're good before you start drawing because it's sometimes a little creepy. So AI images are always creepy I think like 95% of the time. There's always something. And I'll show you a couple of AI created images of yours truly as well. So how do we apply AI in educational library. The biggest piece is it offers an opportunity for lots of new skills. I put an effort on this. So if you teach everyone how to program so way back when in the late 2000s or early 2010s late 2000s when there was that big push of. Let's get everyone to code and Microsoft Facebook, Google push that really really hard and the reason the socioeconomic reason for that is if you teach everyone how to code. Guess how much you have to pay coders then they no longer $100 coders they're $20 coders. So I always put a little asterisk that if you teach everyone how to do something the cost of doing that that that passes lower it's just the science behind why people promote things to find man. And so, but if you wanted to learn AI the tools to learn is the coding languages Python, learn at a very basic level what the syntax is it's very simple Apollo. It's commonly used with most AI models. We're looking at the math portion of it. You want to learn statistics. If you're looking at how computers see text videos and content etc. Or computer vision that also uses Python. And if you're blending those components together. So if your robotics and blending hardware and software make them work together. So what how to put me on board AI today. The biggest wing is having it used as a tool. So if you look at a reception standpoint, and having an idea reception so you can detect emotions on people you can find out what kind of help they need. You can answer all those very common questions that you get like where's the bathroom. And then it would escalate those more complex and intimate conversations to real human. Same with chat chat bots we see this more and more on different chat websites when you go to chat and for support. Usually working with a chat bot first. And most of the time that chat bot can resolve your problem for you without going to a live agent. Then there's the other component. As people are investing investing money in autonomous driving specifically in the shipping industry which makes up one of the top five jobs every blue college city. If that becomes all automated what you do with the you just place a bunch of workers. So I always encourage people that are worried or thinking about that is having those people that do all the driving do all the transportation provide the correct data to the model. And that way because again, as we talked earlier that model requires constant precision constant fine tuning understand context and constant training to begin on that path sooner than later. They can be part of that movement but also provide and have that that job security. And the last piece of AI is how can we use it to make the world a better place so what tools can we use and employ using AI. People are doing that with smart cities, figuring out where like low traffic areas are how to help promote businesses and low traffic areas how to avoid traffic, doing emergency solutions. Like gunshot noise and knows where the city that gunshot was taking place to route traffic away from it so please can go. But there's a lot of really cool use cases in the smart city umbrella of AI. So I'm going to shift gears slightly and talk about the products that exist that use AI to teach people artificial intelligence. My favorite is zoomie. So zoomie is this cute little low robotic car that Amanda has a worksheet on and I will show that worksheet at the very end. So zoomie is self driving car you teach it how to drive. So you do early coding like following the lines following light. And then you get more complex avoiding obstacles. There's ties plan for ties education. What they do is I've got ties plan which are also in Amanda's documentation that walks you through how to program a robot interact with other robots you can code in a variety of different ways. You can even see the robots in the VR world to and provide code to it. I feel like a lot more about ties clan and ties education are in compass live two weeks ago was we had them on the show. Awesome. Yeah. So you can help code around from the same company that builds any but for drones and what they're one of the only people that are approved for the aerial drone competitions. And so you can learn how drones can self drive the autonomous avoid obstacles. They're very easy to use coding. For a second I thought you said the same company that built me and I was like I knew he was a robot. That would be a plot twist. I've been faking the sort throughout so I look more human. Learned. They did a good job. Thank you. Thank you. This is Luca. Luca is able to read books in different languages and read it into whatever language you'd like it to read it in. Luca is reading a Chinese book in this example and I can read it in English and you can show almost any book in this there's like 75,000 books programmed in this library already. You can teach Luca it by reading the book and then you can turn pages and it will read that page for you once it learns it. If it doesn't have it already. Let's see super cute. Mm hmm. If you're looking at doing your own projects. This is a great do it yourself kit with Google and building like the Google homes. This is a from Google called dialogue. And this helps you understand how machines understand what you're saying. So you can see those intense and use cases as you program your model and see where they fit. And that's what that console looks like. So there's uncertainty levels that would be listed out and you can then counter correct it for future use of that model. So just fun to play with. I always tell people play with it and see what happens. This is a tool called the lens app. You can download it on your phone and you upload a couple photos of yourself and it will create like futuristic like cool edgy steampunk photos. I did have a complaint when I made mine because all mine are sad. I uploaded photos of me being happy so I don't know why every one of the photos I have looks like I went through some stuff. I'm just processing it. And so don't know why I'm usually pretty happy so whatever. And there's other companies that have developed AI models that will take your photo and make them into business business business photos. I should use that one. My hands look a little so I started with hands. Gonna say those hands are not human. I have like an extra finger on one hand. I think that's it. The most are in the pocket or something or off. Fun fact. The reason why the AI models can do feet better than hands is because of discord. Because people wanted to see feet photos before hand photos I guess and you can draw your own conclusions. And so that's why the model. So again, the model is only as good as you train it. And so that's why it requires constant tuning. The AI professional photo models have put me in water and also made me a twin brother. And so take this for what it's worth. My mom's not on here so I can say this. I sent my mom a couple of these photos and she's like, wow, where were you? These are really good. These are nice outfits. Did you find something that looks just like you? I'm like, mom, they're not even me. They kind of look like me, but they're barely me. Like the left one looks more like you than the right one in that last one. Like, yeah. And then this one reminded me of the toy Santa from Santa Claus 2. I just looked like that little robot Santa. That's what I thought. So my mom thought it was me too. And I was like, wow, mom, you do not love your child. So how do we bring this all together? So usually the conversations about here is how does it impact staff and how do we improve it? I always suggest that looking at AI is the shift of focus. So how can we make ourselves more efficient and more accurate? And so we want to use AI to do those things. We're always job displacement. A lot of it is using it as a tool. So like some of us use our phones for calculators and we don't want to do the math in our head. Think of that as the same component. Write that quick email because I don't have the time or send out a bunch of scheduled appointments because I don't have the time. And so using AI is to make your workday more meaningful is the most important. They say that AI will struggle with empathy and emotional intelligence. And I would say that's no longer true. Because AI can make a decision. AI is typically designed to make a decision based on facts and not account for emotional variables. But you could program in AI to understand those emotional variables. I think audio might be a little off. I'm muted because I'm gagging. And so I've gotten on mute. That's fair. This is Maslow without the W. I misspelled it. And so what Maslow did was they had a whole bunch of people me included journal our day every day for like a month. And we use colors and like random words describe what's going on recorder ourselves. And so it was one of the first AI companies that use emotional intelligence back in 2018 2017. And it's now designed as a coach. And so if you wanted to have like a professional coach. Either you can have an AI drive entirely or a coach would use this tool to give to their clients and they would get the summary of their days. And figure out where their trends are. It's also been used for therapy and a few other components. But that's Maslow. It's another fun tools to make your life better. I use meat geek for all my meetings. So meat geek joins our meeting and then summarizes the conversations that I've had and gives me bullet point list items of like what to do next and what's going on and who owns what project because I don't take notes when I talk. So this is meat geek. Poise is a very similar product that you can actually upload your it takes your agenda in your calendar. And as you talk, it will cross things off your agenda to make so you can have your agenda on the side and it will cross things off if you talk about it. And then it can even score that interaction. So let's say you were doing a job interview. You choose this was for a job interview. And it will tell you if you had clear communications for the job where you were you or we like we're unsure if you were more sure you may have a better chance doing the job. And it's a performance coach as well. So it makes sure you say you'll say, um, like Brian, that's all the time. Or if you talk to you fast like Brian does all the time. So using all those little components, you can learn how to better present and better speak as well as having it as a meeting tool. There is a variety of eight AI content creation tools out there. So there's writer to create text. There's pictorial to create videos. Oh, just playing with that. It's pretty nice. There's smart writer that can generate landing pages for you that can help you write business ideas. You can give us some bullet points. It helps you write a business name or a business plan. There's crawl queue dot AI, which will, it's a content AI copywriter. So turn it in can't compete. It'll just copy and paste whatever content you want and they'll rewrite it. So it's all your play dress. It's a great tool to create a large form documents like chapters and books. Looking at social media for the social media people out there. Q. That's a Q you, you, you will actually find articles online that match your brand and automatically post based on the time of your interaction within Twitter, Facebook, LinkedIn, et cetera. You can also write articles for you and then posted as well. Missing letter does the very similar, but take it one step further. And if you had articles on your website that you've posted, you can tell to look at those articles and generate a what's called a drip campaign and social media world. So the drip campaign will retweet or repost a article over and over again over a set period of time. And by choosing different quotes, it then reposts it. And so what it does is it looks at quotes within your paper or resummarizes it in such a way to make that make that post look different to help draw awareness to that article that's on your website. And so that's missing letter without the E. Okay, that's probably one of my favorite social media tools. This will automatically write tweets, posts, Facebook's, TikToks, et cetera. And schedule those out to be automatically sent out so you don't have to actually post anything. You don't have to use AI to do it all. And these are some of the tools that I've been using over the years for content because I don't like social media too much. And I rather not be on it. So I use AI to do most of my posting. If you're looking at building your brand up. It's not socially acceptable to use bots to increase your view usage and Twitter, Facebook, et cetera, usually block those apps working. What Instu does is a little different. It uses the algorithms in place to for virility. And so it will like posts and then unlike posts so it sees these interactions to help boost your reputation on those various platforms, the loophole in the world but Instu also works on Tinder. However, that means you know, image editing has been another huge component of AI creating images and text or even just doing it as an image editor. So it's one of my favorite tools, probably one of my favorite like very inexpensive tools to use to edit photos. You can highlight objects you want to remove, you can have it enhance other objects, et cetera. Adobe Firefly creates like really cool little artistic renditions of things just by asking it to generate. There's a bunch and so a lot of these tools that I've chatted about I put them all on a single web page so you can click and play. I do have access to all these tools as well. Access to any of these tools are really some most of them. I can give you a username or password that you can sign in and play. Just send me an email and I will set you up. And here's some more. So if you go to links.evolveproject.org slash AI tools, you'll see this beautiful PDF of everything on there. If you're like me and you don't like paying for things. I built my own AI tool that uses open AI's API. And so you want to instead of paying 20 bucks a month. I'm giving it away for free right now because I don't care. You can use the APIs from open AI and actually have access to chat GPT-4 and just pound away at the API key. And it's still cheaper than $20 a month for me to run it for most people. So that's like the price discrepancy between everything just so you're aware it's very inexpensive to run 3.5. 4 is a little bit more expensive because it requires more GPU power to provide content generation. But 3.4.5 is really cheap and inexpensive. Let's go that you mentioned that because that was a question that came out. Someone who was asking with all these different resources and services you've been talking about what is the cost to use them? Do they have? Yeah, most of them are set costs amount variable costs, which are nice. You can use as much as you want and not worry about being a charge over it. The social media tools range anywhere from like 10 to 35 bucks around the same cost of Hootsuite back in the day. Image content creation is usually done by credits. And so it would give you X amount of credits for 10 or 20 bucks a month. A single image will take anywhere from three to five credits depending on the company. And the way they're associated in credits is API calls to open AI or stable diffusion. Depending on which product plan they're using. So we have 10 more minutes and I will wrap up with a couple little helpful slides and really quick like discussion points. Challenges with an AI education. I'm a very, very high level there's conversations about do we allow the use of like chat to PT in a classroom. Yes, you know, and it seems in my opinion the best solution is not to prevent the usage of it because you never will be able to. But instead ask questions such as, hey, use chat to PT to write your article and then fact check it and find out what was true or where it found the source. That's a way different type of exercise. Another component that someone mentioned once in a workshop was asking more feeling based discussions. So instead of doing factual based discussions do how does this make you feel in your world or in your life. It's very easy to read that to find out what was true or what's written by an AI and AI can give life experiences. Unless you tell it to make one up. Then there's a concept called AI generative and breeding. What that means is AI has been creating so much content lately. AI like those early on detectives that can detect if an AI wrote piece of content or not, no longer a relevant because there's so much data out there that AI has now generated. If it looks at both and assumes that the AI generative one is legitimate. It cannot make that comparison anymore. And it's gotten so bad that like chat to PT some of the math questions that used to answer pretty well are wrong because it's getting too much information in the ecosphere. Or get to touch the difference between AI content creation versus something that's original. I think the example is like a Shakespeare's piece. I thought that a Shakespeare's piece, a very well known one was AI generated content because it's was posted on AI generated content through websites and it made a long correlation. The last challenge that people talk about is how does it pull content like what sources is it using. So today I tell people the best way to build your own AI is to choose and teach the AI to look at specific content sources usually on your own like knowledge base or your note for your own infrastructure to look at very specific documents to use that as a determining factor. And using open AI API as an example to do the conversational piece. So it knows exactly where to pull the content from you have that conversational dialogue and it only looks at its content versus the content within its own ecosphere. So it has some resources that Amanda has put together as well, and I'll let you talk about it for a few minutes Amanda so I got to my voice. That is fair enough. I also mentioned everyone who if people weren't here at the beginning. If you have any questions that you want to have anything more explained or expand, you know, elaborate on more or if you didn't hear about that you want to know more about type into the questions section of your go to webinar interface. And these slides that Brian has been using will be available afterwards when we post up the archive. So all these links and everything you had in here you all have access to that afterwards. So, so looking at the links here I'm going to make a 99% certain guess which ones they are. So the bottom one is the attract tech gadgets. Most of them I actually just copy from what Brian's already been tracking so don't tell. And so I put together like a little chart that shows different AI gadgets that you can get to use in your own library or your classroom. So that Kai's clan and the zoomie thing those all show up on that little guide. So if you have absolutely zero budget and you are still looking to introduce AI to your community you can use that little chart. And incidentally if you're a library in Nebraska you can also check out a lot of that stuff through the tech kits of mail. So I have little classroom packs of up to 15 copies of each different type of kit or robot available. And you can either check out a few of them check out all of them and you can use them for your programming or or try it before you buy it. People use in different ways. I don't care how you use it. And so that is that bottom link. And then the top link is the little summary pages about what AI is different types of AI stuff and some different like communities where you can go to. So if you did a programming event to introduce zoomie and AI to a bunch of your teenagers, then that same teenagers can go to that top resource and find out, okay, I want to learn more. I know that I can't go ask my librarian because she's going to look at me with wide scared eyes. So I'm going to go over to the to this learning community, and I'll ask all my questions there. And I'm going to go over to this resource page and I'll try out some more activities because I want to dig deeper into this. So that's what that's all about. All right. So we are at time got five minutes left. I haven't. There's questions feel free to post them in the chat. If there's questions if you're watching the recording my email address is there feel free to email. And yeah, that's it. Thank you for attending. Yeah, awesome. They good leave that slide up there here for wrapping things up. Yeah. Thank you so much Brian and Amanda for explaining all of this to two people, it's, I think it's a lot to learn about everything but I think it's great we got a lot of these basic information and the resources I think is the key I think for a lot of people. If anybody doesn't have any other questions go ahead and type in the question section we can ask anything you have out there or if you're just overwhelmed by all the AI. Yes, yes. And I think it's an important topic is that a lot of people in talking about and confused about not understanding. What is it what do I need to be worried about. Is it something I need to be worried about. But I think, even though it is a Pandora's box, there's there's good things in there to that we can use them for use it for in the right way. Yeah. Absolutely. And then as a shameless plug computers and libraries, the magazine. October issue has an article I put together about looking at like tech trends and how to pinpoint like what products use and I have some of these products explained as well. With an article so if you are strapped you'll have a really cool article to read. And I'm in the September issue. I was going to say is that same years years is that because we just yeah I just yeah. Sometimes some of the articles, not Amanda's unfortunately right now, but are available from the computers and libraries magazine on their website, some of them they, you know, they pick a few of them to make available for anyone just click on the links for the HTML version of it. So depending on the article you might be able to get it without having to be a subscriber or receive the magazine. But it depends. Yeah. I don't see any other questions coming in right now. So I think we may be good for the day if nobody has anything desperate you want to ask of Brian or Amanda before we wrap things up. Get it into your question section there. And I will do my wrap up on waiting to see if anybody does anything they want to ask anything what no more about something about AI you are hoping to learn today that hasn't been covered yet. Or if you want to share which cat you are. She have like the wide eyes right now. Relax comfortable. Whatever. Too much to deal with right now. I think there's a guy in the middle of the road cat to of like, I'm really interested, but I don't know if I have the, the brainpower the energy right now to dig into it but I'm going to keep it on my radar and we'll see. That's great. Let's see we do have a comment I guess someone's saying here, most recommendations for workers up against AI up against AI taking their jobs are to quote make yourself more marketable. That individual focus rarely are there suggestions or support for workers to make sure AI implementation benefits everyone. You're not wrong. That's true it's doesn't have to be an either or they can work with for their own. Yeah. There's a book that you might like called AI for AI Q, or the new geography of jobs. If you're curious about how AI is changing the workforce and how AI can be used to make sure everyone has a good job instead of life sucking for some and being awesome for others. I don't know if that's the attitude caution, but maybe that's just me. No, that sounds appropriate and me. Yeah. And then we just have some thank yous coming in. Thank you Brian, Kristen, Amanda, you will. All right, so I think I'll do my wrap up here if you do have any other questions like you said, Brian's and contact info was out there you also all know where to access contact Amanda. So if you want to get the recording up on recording should be done and ready by the end of the day tomorrow. Yeah, usually takes about a day or so for go to webinar and YouTube to process everything. This is our session page for today, but I'm going to pop over to the main encompass live page. Here's our upcoming shows and then here's the link to our archives. Today's show the most recent one will be at the top of the page. And there's the one for that kinds education if you want to watch that one. So if you're interested in today's show and registered for today show get an email from me letting you know when the recording is available. We'll push out onto our social medias as well. We have a Facebook page for encompass live if you like to use Facebook give us a like over there. You can do reminders, his reminders log into today's show, manage about other things going on introducing our presenters. As you can see, follow us over there we also use the hashtag and comp live a little abbreviation for our show name on Twitter and against the grammar, our people put things out on there. I'll announce there as well when the recording is available on the recording page here I'll show you there is a search feature if you wanted to see what we've done before so for example you could look up Brian's name and see all the other things he's done for us over the years. Whoa, 2012. Yeah, we've been around for a long time. And like I said earlier before we went live the show. This is our full show archives going back to when and compass live first premiered in January 2009. Do pay attention. If you are watching a recording to the original broadcast date they're all dated when they were first we're done. Some of the shows will be fine and stand the test of time be perfectly useful useful and everything, but some things will become older updated resources and information may become no longer be available anymore links may be broken people may work completely at a different library than they worked at before when they presented for us. So just pay attention to that when you're watching this, but as librarians do we do keep things for historical purposes as long as we've a place to host all of our recordings which right now is our YouTube channel. We will have them always out there available for you so just pay attention when you are watching any of our recordings. And Amanda said her email for Amanda is Amanda dot sweet at Nebraska dot gov. The library mission. These are upcoming shows but I wanted to that I just announced this morning. Another online event we I host here is the big talk from small libraries online conference. It's a national conference, not just something Nebraska focused where we have presenters from small libraries all of our presenters are from libraries with an FTE or population served of 10,000 or less. So I opened up the call for speakers for the 2024 conference. So, if you are, you can see all of our previous conferences we've been doing this since 2012 is the first one yes. So if you are at a small library, and in a rural library I would like to present if something cool you want to share submit a proposal to me. And is December 15 to submit the proposals and big talk is always on the last Friday of February. So this next year it is Friday February 23 is when it will be held it's all online you don't have to travel to Nebraska in the wintertime. Nope, nope, nope. You just log on to your computer and present. So I've pushed this out on all of our various and ages. Big talk also has a Facebook page and a Twitter account where you'll find that information as well. So, if you so spread the word please either submit a proposal or spread the word about big talk. Where did this year go it feels like you just had big talk to know. That's why I have notes in my calendar that pop up and say hey do this now. And also talk about 2024 next week our encompass live about will be about our library commission grants that are going to be opening up for applications this Friday. This is Nebraska specific session of course this is grants for Nebraska libraries. So we have our four grants will be opening for applications see and continue education grants internship grants library improvement and youth grants for excellence. And next week, myself, Sally Snyder and Holly dug in we're all be talking about the various grants that we run and give you some more details about them. The applications will open this Friday and come as live as next Wednesday and they'll be due November 17. For anything you're doing in 2024 as a Nebraska library so if you are in Nebraska, and you want to see if we can give you some money for something cool you want to do that. And look at our other sessions we've got some things filling in here we've got more people I'm talking to get specific topics on to keep an eye on our schedule for any other outcome that I get confirmed in on here. All right, anybody have any last minute desperate questions they want to ask, get it in right now so your last chance, or anything. Last words of wisdom. From you Brian or man. Yeah, we're all going to die. No. Brian. No, that's it thanks for having me everyone. All right, thank you everybody and hopefully we'll see you all out of future episode of encompass live. Take care everyone. Bye. Bye bye.