Brookfield Old Town Hall - The Perils and Promise of AI 8/19/2026

Thank you so much for being here tonight on this gorgeous night. We thought people would be out enjoying the weather. So, we're really delighted here. But we figured such a a powerful topic that if no one showed up, we'd be talking for the rest of the time. who've already gotten gotten served. Uh I'm Nancy Doda, member of the um old town hall committee events uh events committee I should say and I just want to read a little bit about a bit about the town hall and then start with the panel. Um, one of the things we want to say there is thanks for coming up. We also support the town hall and our summer events building and the Brookfield Community Partnership needs support to continue to offer beautiful community events like this one gathering as we do uh for collective logging. And please continue um coming but also consider donating so we can continue to provide these wonderful activities. Um, we have a donation box in the front and you're welcome to donate any amount and we're thrilled to have that. There's also QR codes if you're interested in using technology to do the same thing. Um, so thank you to that and our lead sponsor is Frankenberg. Um, but also tonight's sponsor is Washington Electric for this special event. We're really grateful for that. Um so um let me start by telling you we've gathered this panel this illustrious group and um we'd like to uh take a minute to introduce them. Would you mind using the microphone? I can't hear very well. >> Thanks Alison. >> Okay, great. >> Maybe you want to repeat the donation on the microphone. Yes, we have to have a little humor around that subject. It's not my favorite thing to have to say. Um well, we did want to start and introduce our panel members um for this panel. And you'll notice the title of course is um AI, the perils uh and promises of artificial intelligence. And we have um been fortunate to be able to have a number of people with us, but there are many people in our community who've since volunteered to join this panel. So we could literally have a a rather large group. We have uh Mark Natanda Gara and Mark is with AI Vermont and I'll be telling you just a little bit about each of these persons but Judith will introduce Mark in just a bit. Chris Sores and raise your hand guys so people know who you are. Um is a an MD. He's a retired pediatric and adult opthalmologist who dedicated his entire career to taking care of Vermoners, including me. I'm thankful to be able to read this without glasses. Um, Chris's son Nate is CEO of the Machine Intelligence Research Institute, and Nate recently authored a book titled, "If anyone builds it, everyone dies." Controversial title. Chris has devoted many hours to speaking with Nate on the perks and perils of super intelligent AI and he can share what he has learned through these many discussions with you. Teresa White u a neighbor just down the road has spent two decades in elementary education in central Vermont specializing in science instruction and structured literacy. She holds a master's in teaching from Brown University. Began her career as a policy associate at Rhode Island uh kids count program shaping policy to support children and families. Her work centers on evidence-based personalized instruction that helps struggling readers achieve success. I can vouch for that because I work with Teresa. She is currently working at the Stern Center for Language and Learning. This gives her a grounded practitioner's perspective on how AI tools are shaping her work and her world in education. Teao, who I've just come to know face to face because I usually see her from a distance, um has spent her career teaching, helping rather missiondriven organizations strengthen their communication through leadership and their communities through leadership, fundraising, marketing. Today she serves as senior director of development and community engagement at Gford Healthcare. Naturally curious and a practical everyday user of AI, Teao has woven these tools into her own work and her team's daily workflow to improve communication, spark new ideas and streamline routine tasks and create more time to focus on work that matters. Dan O'Brien is also a neighbor, used a slide rule in high school and remembers when calculators were controversial. After building and exiting an engineering business that grew to 30 million in annual revenue, he chose to repurpose rather than retire. Today, he coaches business leaders and facilitates monthly CEO peer group. Dan's thoughts on AI oscillate between hope and alarm, between an innovator's enthusiasm and a realist's concern, especially about data mona monetiz moni I can't say that word monetization you get it right the connection with money privacy and technologies potential human cost Judith >> could borrow that mic >> absolutely Yep. Hi everyone. I'm Judith Irving, um Brookfield resident and also I'm currently chair of the board here of Old Town Hall and it's really great to see so many faces that I recognize. Um, I when we started getting this panel organized, maybe it was in March, I said, "Wouldn't it be interesting if we just sort of like collected some quotes about AI as we go along?" And I was supposing that what we collected in March might seem antiquated and quaint when we got to August. And that actually seems to be fairly true. Um, the other thing that is very different is that you can't open a newspaper, listen to a podcast, turn on anything now without hearing something about AI. And it was like everyone is is on everybody's mind. When we started this in March, it wasn't so much that way. Um, so we're really seeing a very big change in the attention. And of course, we all kind of have our own ideas why that is. Um, one of the things we were going to do, Nancy, did you want me to introduce the >> do it? It's up to you. >> Okay, I'll just do that and then I'm going to I I have something to read at the end and then we'll start. So, at your seats there's a piece of paper and there is also a pen and it's a very simple prompt that says, "I feel blank about AI." And if you would just take 10 seconds and fill that in and then turn to the person who's next to you or behind you and share what you wrote. Um, and this is we're going to spend like three minutes on this. So, this is off the top of your head. >> If anyone needs something on which to write, please feel free. I have a little bit more. >> Yeah. Yeah. >> All right. I think that's probably a good amount of time for that. And you'll notice on the sheet that there's an opportunity at the very end of this where you're going to be able to fill it in again and see how your mind has changed. All right. So, I was going to share all the comments that I ended up writing down, but I mean like hand you a copy of it, but it's now like 14 pages long and I decided no, I'm not going to do that. Um, but I do want to start by reading um something that Christa Tippet wrote. I don't know if everyone is familiar with her but she's a very um interesting reflective person and she was at a conference in Australia at the wisdom and action forum and the presentation or the discussion was called trust in the age of AI and I just took a couple quotes from that. You can see the whole thing online by just going to her um website. Her program is called onbeam. But I thought it was very interesting the way she phrased it. So this is what she wrote. We humans will no longer be the most intelligent force on the planet. The definition of intelligent is key both to the power of AI and to the deepened self-standing to which I believe it is calling humanity. But this much is clear. Unlike humanity's previous technological advances, AI is more than a tool. It has agency, the dimensions and meanings of which are unfolding in ways mysterious to its makers. Radically different from other technological revolutions, this one will cocreate the future with us. What if our technologies are calling us to nothing less than a greater understanding of the deepest meaning and purpose of human intelligence? The modern western world has revered our cognitive, rational, computational capacities and demeaned and demoted much of what makes life worth living. Now these miracles we've engineered are paradoxically dissolving the illusion that the computing capacities of our big brains are the most special thing about us or the most superior. The intelligence we bring to living as whole mature human beings on this earth are as embodied as they are cognitive. What if AI is here to bring us literally back to our senses? So, I'd like to set the stage. Um, Mark, you're gonna Mark is going to do an overview. And I'm sorry, Nancy, when we uh you just began. Did you give the whole story of Mark's incredible background? So, here I go. I'm gonna just wait to hear it. >> I know. I'm gonna just try. So you've had Mark's been involved with education for over 35 years prek to college and then about three or four years ago he started the um organization in Vermont called AI Vermont which has focused on researching studying communicating about AI um through workshops with teachers and students and community groups like this all over the state and uh we are very fortunate to have him here and he's going to be the first panel member speaking and give us a grand overview. Is that what you're going to do? >> I will now. >> Hello everyone. Uh in addition, let me just say about AI Vermont that it's really just three of us. Um we've been doing this since we got together in 2023. Chris Thompson, Denise Shakersian, and I discovered that we had uh as Dan said, both uh concerns and hope for this new technology. And on top of that, we're three very curious people. And curiosity is what binds those two pieces together and keeps us going. So, I want to make it clear that AI Vermont and I are not ever speaking to people to advocate for or against AI. That's not our job. We just want we we understand we're realists. It's here. It's not going away. So, our role is to help people get enough information, quality information, information that works for you personally so you can make your own decisions. And what I say, and I usually close with this, but I'm going to open with this. There are politicians and tech bros, big company leaders who are happy to make those decisions for us. So unless we're part of the conversation, that's what's going to happen. So that's why we exist. Um we also I'm so glad that you mentioned um sort of our research base. We are deep researchers. We read articles constantly. We meet with professors. We meet with people during this. We interview tons of people. But we also are very interested in the practical side, the practitioners. And they work both ways. If we see a piece of research, we think, "All right, that sounds good on paper, but what does that really look like in a classroom, in an office, or in a home, and we look at that?" Likewise, if somebody says, "Well, I'm doing this." We say, "Well, let's see if there's a research base to support that." So, we're, you know, we're all still struggling to prove practically anything. So, I'm a lifelong educator, and one of the the deadly sins, the deadly secrets of education is we do things without a whole ton of evidence. We do it by the seat of our pants and we do it by intuition and training and most of the time that works and sometimes it doesn't. So we feel the same way about AI. We really feel that it's important now when somebody says that it can make this huge change. What's the evidence for that? So I leave that to you. So let me just talk briefly about what we mean by AI. So our goal is to provide to increase AI literacy across Vermont. And we're doing it primarily through schools because we thought, well, first of all, that's the language I speak as a 38-y year educator, but also because that's a really good bang for your buck. So, you meet with uh, you know, 100 teachers and administrators and then they each speak to, you know, 20 students or 75 students. You know, I forgot how many I had when I was a high school science teacher, but it was like 75. Um, and then they go home and they talk to their parents. The parents talk in the community. That's the best way to promote this conversation. But along the way, some businesses, some nonprofits, public libraries reached out to us and said, "We like what you're doing. Could you do it for us?" And at first, we said, "No, this is really the best avenue for doing this." Then we realized, "What the heck? The more we can get out in front of communities, the better." Like, we, you know, I don't just want to talk to educators because that's a particular language. So, that's why I'm here tonight. I love it when it's a diverse audience. Have no idea. I bet nobody on that piece of paper wrote, "I'm incredibly happy about AI." Um, but I bet there are a whole lot of other emotions expressed there and I'd love to read those. So, let me talk about AI literacy, what we mean by that very quickly. So, um, we have a few educators up here also. And so, literacy, I think, to the general population means the ability to read and write. And it shouldn't just say to read and write because just reading a word or writing a word doesn't mean that you're really literate, but to know the meaning on behind it and to to write understandably and with emotion and expression and things like that. The piece that AI literacy has that maybe the reading and writing literacy doesn't is you got to be worried about it too. You have to be able to understand where it can go ary, where it can cause harm, where it can lie, where it can make mistakes. And so maybe it's more like financial literacy, right? A financially literate person understands money. They can balance a checkbook. They can develop a budget. They can invest wisely. But they also have to worry about money like who's trying to scam me or is that a bad investment. So think of AI in those terms, not just oh if I'm AI literate I can be empowered to use AI. Nope. I can also be smart and know when not to use it or when the the use of it is is harmful. So that that's I wanted to just because I was very resistant to the idea of the term AI literacy because like some people um I didn't look at AI as just just a tool but I thought yeah you're not you're not literate in the typewriter you're not literate in the pencil you're not literate in the book you're literate in the use of those things. So AI is just something that you use, but it's more than that because it can. And I have a friend at the State Department who says, "Well, it's more like a power tool, right? You if you have a chainsaw, you got to read the instructions and know how to use it or you're going to get hurt." And I said, "Yeah, but AI is not quite like a chainsaw because once you turn it off, it's off. And AI is never off." And AI on the positive side won't give you a chainsaw won't tell you how to cut that tree down, whereas AI can give you instructions on how to use it. Um so you know I don't I don't necessarily buy into the idea that we should just think of I think it's d a little dangerous just to think of it as just any other tool and and having been in education for a long time um I've seen tools come and go that we eventually integrate and this is this is much more of a challenge. Uh the second thing I want to share with you is uh from our experiences I think somebody here's going to talk a little bit about what AI is. So I don't think I want to get into that too much but I should say that we're all we all seem to be of a certain age where we have seen AI for a long time. It's been around since the 1950s just not this latest iteration. So, if you're familiar with uh you know the Roomba is a version of AI. Um a chatbot when you go onto a a shopping site and it says, "Can I help you?" or even a recorded voice. That that's that's a form of AI. AI is a type of software that is intended to mimic human perception. How we see the world, how we hear the world, what we do with the information from the world, how we uh process that information, we analyze it, we try to build context from it, and then what do we produce from it? So AI does all those things um and has been doing it for a while. It's just that now they've been trained as large language models basically trained on I don't want to say all the information that's out there but all of our interactions on the internet the stored knowledge. And so it's learned to mimic us in other ways. The way we talk and the way we interact, the way we create and do those things. Let's keep in mind that's what it's doing. It's mimicking those things. It doesn't really know what these things are, but it does an awful good job of convincing us that it does. And there in also if you're AI literate you will separate that from your your perception of it from the reality of it. Um okay so I don't want to get too much more into what AI is because I think it's going to come up and if it doesn't just ask that as a question during during our Q&A. Um I do want to share this part. So our our approach to this um when I first started doing these workshops I really did say here are the benefits that people are seeing and organizations and our culture is seeing from AI and here are the challenges and the concerns of it. I don't do that anymore because every time I listed a benefit, in the back of my mind, I was saying, "Yeah, but there's a downside to that." And every time I listed a deficit, I thought, "Yeah, but we could make that a good." So rather than, "Oh, here's the bad. Here's the good." This is a AI use and AI benefits and deficits are not a black and white issue. There's a lot of gray in there, right? You might say, "I am never going to use it." That would be one side. I'm going to use it every day. We've got some people doing that. That would be the other side. But some of you might say, "I'm only going to use it for this. I'm going to use it occasionally. I'm concerned about that. So, I'm never going to use it for that purpose. Um likewise, we because we don't we no longer say put put words in your mouth. Here are the benefits. Here are the deficits. Instead, we say here are the questions and you can answer them as you see fit or investigate them. So, here are some of the questions that we've been collecting from meeting with I think we're up to you might be up to you might have kicked us over the 3,000 mark. I think in in the last two years we've had 24 workshops and uh we're at somewhere around 2,800 people. And so we gather that information. That's why my laptop's out here. It's it's really to to take notes on what you're saying and what's happening up here. So um here are some of the questions that people are asking. How can AI help individuals and organizations without misusing their personal information? Okay. So, that's a question that you can, and I'm not going to do this for every question, but that's a question you might answer as, well, I don't trust it, so I'm not going to use it because I think it's going to steal my personal information. Or you might say, I'm going to use it, but I'm just not going to give it my personal information. Okay, so that would be a middle middle ground there. And there are other people who are like, I don't care if it has my personal information. There's a lot of the next generation that doesn't care about they don't seem to care. I shouldn't say they don't care. They don't seem to care as much about their privacy. And, you know, they give Amazon all of their information so they can make good recommendations. Maybe for them that's a good thing. I don't know. I'm not that way. Another question. Can AI be used without diminishing our own ability to think if it's taking that over? Is there a way to do it so that it doesn't make us, you know, like the time machine, the Eloy and the I don't know, I'm a science fiction fan, but right, the time traveler goes in the future and meets meets these very civilized people and then these brutes and assumes they're the ones in control. Nope. They're they're gentle looking because they've given all their control over to the machines and these other folks just raised them. All right. Sorry, I I could definitely digress in science fiction because we've seen if you're a science fiction fan, we've seen this before. Yeah, we've been warned before. How should children think about AI and as adults, how do how can we guide them to be safe? Uh to use it effectively, ethically, meaningfully, and then when not to um can we Yes. Okay. Um can AI be made to be less harmful to the environment? For some of us, that's a huge piece of this puzzle here. Do AI products uh violate the intellectual property rights of creators? And uh you know AI was trained on everything and that's included people's paintings, their books, and things like that. And um I met somebody at one of my workshops who said they you know it's and it's gathering this it's reproducing this information and spitting it back out somewhat randomly that there's there's more to that technique, but that's the essence of it. But occasionally, just like if you were sketching sketching sketching, you might someday sketch something like somebody did. Occasionally AI spits out something that's awfully like somebody else's. How do we address that? And the biggest one that AI Vermont is working on now, we worked for the first uh year and a half entirely with educators and then we finally gave into people saying, could you come and do that for us? And we realize we're missing a piece here. Students are ultimately concerned of their future with AI. Are there going to be jobs for me? Am I should I bother being a computer scientist? PS, no. Um so, you know, what what should I do? How should be prepared? And businesses are asking the opposite question. What are schools doing with students? How should we uh define the workplace? So, we are working in all those ways to to bring those pieces together. All right, let me let me close the intro with what we have found are important pieces of this conversation. Uh I said it before. First, AI is not going away. So, and we're not here to decide should it exist. It's here. It's here to stay. So, we're here to discuss how we can be part of the conversation so that we can be part of the decisions being made about that. Um, the other thing that we found, I I know when AI first came out, there were I met with teachers within the first two months. I wasn't going to do this. I just realized the teachers weren't being supported. I'm big on helping teachers and having those conversations. So, we met and some of them said, I really feel like this is going to be like so many things in the past where it'll just pass or it'll become part of what we do. And I said, don't don't bet on it. Um, if you're thinking that there's going to be technology to detect AI in student writing, the AI will always be ahead of it. And that's been proven. I'm not some suits there. That's that was pretty obvious to me because I've lived long enough and been an educator long enough to see that. Um likewise, legislators who are willing to do legislation on this just can't keep up with it either. And that's troublesome because those of us who want to experiment with it, whether we're educators or people in the workplace or wherever we are, we want a little bit of guidance so we know we're not doing something wrong. We're not going to get in trouble. We're not going to cause bad things to happen. So that's another piece where we as citizens can step up and ask our government government or our bosses, let's get some basic guidelines down. There are enough of them out there that we can figure out what what works for us and then we'll customize it as we go. And speaking of as we go, I think this is a the most important time for us to be transparent about this. And transparency might mean I hate it, I don't want to use it. We need those people in the conversation. My first workshop was filled with people really excited to learn about how they might use it safely, effectively, and all that stuff. At the end, a few of them came up to me and said, "Man, I wish my colleagues had come to this, too." And I said, "Well, why why didn't they?" Because they hate AI. And I realized I had called the workshop something like make AI your code teacher. So, they all thought I was going to promote AI. And that would couldn't have been further from the truth. We were talking about both sides. So, now we understand we need transparency. So, we know what people's concerns are, what your individual concern are, but also how you're using it, how you're not using it. This is the time to do that. And that's not this doesn't seem to be the environment in our culture to share those things when we disagree with each other. But maybe this will be an avenue for doing that. You know, I have a lot of hopes that AI's disruption is going to end up taking things that we didn't do well before that have nothing to do with AI and I could spend lots of time talking about education and it will make us look at them like you said, look at them and say, "Hey, that or as Chris Tippet said, um, that is a problem for us. A I just exposed it. Let's address it." All right. So, I think that was about 10 minutes and so now I'm just a plain old panel member and we'll go from there. >> Plain old panel member. >> So, the the format we're going to use is that after each person has a chance to respond to a question, you're just going to open it up to see if there's any just things like burning in your head that you want to have um Mark respond to. Got about 5 to 10 minutes plan for that. If there are any questions, we have a microphone and we get to audience turn one. >> He's he's got it. >> No, not unless you stand next to the speaker or next to the other microphone. >> I shut the microphone off. >> Is there someone who has a burning question? All right. >> Here. Uh, of course it I totally rebel inside of me when you say it's here and there's nothing we can do about it. And one of the things that I learned almost immediately was that um I've come from a family of artists and uh AI had me as my brother's daughter and um so immediately I understood that AI can get it really wrong and that this is um this is harmful to me fundamentally uh in my career. So, um I'm I think it's inaccurate. It has great cap capability of being inaccurate and that makes it dangerous. >> Is there a question there? >> That's my question is how can h how on earth can that be a positive? So here's what we've come to see about technology and particularly this technology that almost everything that I said three years ago that it did poorly I warned then it's going to do it better and so these inaccuracies are diminishing that doesn't mean I think the bigger problem Oops was a month ago >> and again I'm not I'm not saying it's perfect and I'm not here to advocate for it please remember that but these are these are what I'm seeing the problem isn't necessarily that it's going to make fewer mistakes. The problem is people think because it's technology and how it's hyped, I don't have to worry about those mistakes. So, I don't care how much better it gets. We need to be vigil if we're choosing to use it, we need to be vigilant about those mistakes. And the other thing I want to say is I didn't say we can't do anything about it. Otherwise, we wouldn't be here. What we can't do is we can't get I don't believe we can get rid of it. I think that that's a fool's errand. >> That was what I was that I'm very upset is that we can't get rid of it. So, we're stuck with it as a public. >> We are stuck with it, but we have a ch an opportunity if we choose to take it to help mold how it's used, where it's used, and for what purpose. >> Good idea. >> If you correct that mistake, it's going to remember the correction, >> right? I mean, so you can't just walk away from it and say, "Oh, it's wrong." You have to tell it that it's wrong. You can swear at it if you want, too. I do that often. >> How can I tell it's wrong? You're on my computer. >> Well, you communicate communicate with it. >> Tell how how can you say that it you can't I'm I'm here that you can't get rid of it because it's getting rid of itself in the sense that the two large AI companies are both losing billions of dollars. There there's an AI bubble. They've overhyped it. The bubble's going to burst. those two companies that are going to go bankrupt. It's going to really hurt the tech large tech companies that are subsidizing them. The only reason anybody's been using it is because they've subsidized it. When they stop the subsidy, people stop buying it. So I it's not inevitable. It's it's it's a precarious industry. It doesn't pencil. They've put capex. That's just they're never going to be able to make a return on the capital expenditure they've made. So I I just don't see why you think it's inevitable. >> It's not just me, but but I what you just said could happen, but I wouldn't want to bet the farm on it, and I don't want to sit passively by waiting for that bubble to burst. My my job is to be actively a participant in this conversation. >> Dan, you wanted to respond. Thank you. So, so the first comment was about the uh faulty information that it got. And one thing with computers from the very beginning, the phrase, everyone will be able to finish it for me. Garbage in, garbage out. And um there's this sort of underworld of data brokers who are selling our information back and forth to each other. They're collating it. Some of it's accurate, some of it's not. In March, I became aware of a service called Cloaked. You have to pay to join it, but they will go out onto the internet and find all the data brokers who have your information. And they're selling your address. They're selling your who's a member of your family. They're selling where you live, selling every email address you ever used. I mean it's it's that which that has nothing to do with AI but AI is being trained on a lot of that data and um you know since March cloaked has removed 5,068 instances of my data being sold online and so you know it part of that's an AI problem because it's using that information but part of it's just the underlying function of the internet right Now >> it's regurgitating it's regurgitating information already available packaging it is if it's accurate because then they can charge for it. >> All right so we're ready for round one. We this is sort of our introductory and thank you Mark for opening us up and getting us getting our juices flowing. We're can already see it's going to be an exciting exchange. uh we wanted to start although I should clarify we wanted to start with promises possibilities um benefits uh that people are experiencing in their lives and work um and we'll start with that but now what the caveat is that it is much more complex and there's always overlap as we already see in this discussion here so even though we're we're honing in right now on the promises and benefits we're going to follow it with the discussion of the perils and the challenges ES and so forth. So, we will dig into both and I'm sure we'll see some overlap as we go through this. It's not quite as cleancut as it might appear. Um, so the question to the panel and we're going to have each person speak uh and we can start actually on this end with Teao if that's okay and work our way back. Um we'll have each person speak and um you might remind folks of your sort of position or field of interest, yours being medicine and so on uh to help um to describe how they use AI, how you use AI and the benefits that you have observed or experienced in your life and work. >> Can everybody hear me? Um, so my name is Teao Kirchoff and I'm the senior director of development and community engagement at Gford Healthcare. Been there for about a year, but I've also I've worked in communications and marketing for my whole career about 20 years. Um, so I've watched, you know, AI become integrated in that. I mean, it starts with, you know, a simple word document spellch check, right? I mean, there's there have been versions of this that we have integrated over the years that have grown into what the version is today. Um I think I have dabbled out of curiosity um prior in prior positions when I started in strictly development work with um I worked for Dartmouth Health prior to Gford um that's really when I started diving in to uh the more intelligent uses of AI for my work. Um when I was um doing prospect management um building lists for donor engagement some of the tools that are embedded inside of donor management um software is embedding there they're embedding AI into those pieces so we can do some research we have wealth management through there so that kind of struck me as interesting right it's like wow okay now we don't have to scour the internet for the things that may lead us to someone who may be very interested in supporting health care. Those are now at our fingertips. So, worked for about three years at DH um kind of implementing some of those things and donor strategy. Um now that I oversee marketing and communications um we do all of the digital marketing for Gford uh all of the development work. We've dove a little deeper into utilizing AI on our daily um uh aspects of our work. We use it for brainstorming um writing. We use it for creating a lot of social um content cuz we are a small team of three. We use it for um making sure we are efficient uh in in our spaces inside the hospital. Um it's our thinking partner. You know, it's a data cruncher. Um all deidentified. We don't load any personal information into that, but we give it numbers and it's very good at that. Um, and we can output reports for senior leadership in almost seconds, which um, before would have taken a lot of pivot tables in Excel if anybody's ever worked in Excel before. Um so we that efficiency piece is huge for a small hospital um that that is just trying to stay with its head above water. Um so we we we have implemented it in those ways. Um it birthed from curiosity and it really has um shown us how important it is to have a policy in place in our department and how important prompting is. Um, we've learned what not to say to it. And we've also learned how to call it out when we know it's not right or even if we think it's right and we're like, are you sure that that's really the outcome? Double check your numbers. And it'll be, oh, you're right. Right. So, we are the masters of making sure its output is exactly what we need it to be. And we need to use our brains and make sure it is, you know, we are applying that to the use of AI. And I make sure my team knows that. Um, but it it helps us exponentially in our everyday work. So on the everyday user aspect, I guess I forgot to mention what we use it for, but um that's kind of where we stand um for the use of AI. So >> great. Thanks. So, I'm Teresa White, um an educator in central Vermont. Um I'll say as a caveat. Um I'm speaking from my own experience. Um education is so varied. Um from grade levels to subject matter to um rural versus urban. I couldn't possibly speak for all educators or students. Um but I'll try to generalize um from my own experience. So, I will confess um usually I'm a very early adopter of technology. Um I remember when iPads first came to the classroom and my students could draw on them and record their voice explaining their thinking and then I could listen to it later. It was amazing and I was an early adopter. But with AI, I was the opposite. Um, I didn't want to think about it, didn't want to talk about it, did not want to use it. Um, and I realized it's because I felt like personally threatened. Um, writing had always been what set me apart from other students, other colleagues, what gave me an edge. And so thinking about this tool that could do what made me special but better um was not something I wanted to engage with until I had a moment. I was preparing um an application for a credential. I had already written 160 pages. The deadline was coming up and I found out I needed to submit a resume. Teachers have their resumes in this system called Schoolring. I could not get my resume out of Schoolring. I could not even copy and paste. I would have had to retype all of it. So, I tried AI. I said, I know how to write a resume. It's not a skill I need to practice. There's not something I need to learn from this. I just need it to be done. And, you know, 20 minutes later, I had a resume I could submit. Um and then I was a convert. Um but just to give a tiny um bit of perspective into AI and education, um uh the GA a Gallup poll in 2026 found that 60% of teachers are using AI. The teachers who use it once a week or more estimate they save six hours a week of time. And the top uses were lesson preparation, creating worksheets, and modifying materials to um meet the different needs of of different students in their classrooms. The Pew Research Center found um that um from 2023 to 2024, the number of students who said they were using chat GPT in their schoolwork doubled to 26%. um and that they're using it to um research, to solve math problems. I don't love that. Um and to write essays, which I'll talk about in one of our later questions. Um but I was astonished um by a rand survey that said that um the percentage of students who use it for homework increased um from 48% in May 2025 to 62% in December of 2025. Um, so it's it's being used by our students and by our teachers. Um, so to talk about how I've benefited from it. Um, you know, my goals as a teacher are to be excellent and to be efficient. And I've found some ways to use AI to help me in those domains. So the first benefit is saving me time. um classroom teachers work 60 hours a week. Um and I am always looking for ways to cut that down. So if there's um a concept that I already understand like commas and sentences and I just need a bunch of practice exercises, I'll use AI to create those for me just to save me time. Um I also want to save my mental energy for where it's really important. Um, and so I'll often use AI to gather or create some raw materials for me to work with. Um, so I can spend my time and energy on refining them and adapting them for my students. Um, in writing instruction, I create models to use to teach my students. A poor model, a proficient model, an excellent model. Not having to write those, but being able to tweak them saves a lot of time and makes my work better. Um, AI also lets me go beyond my own experience to create learning um, opportunities for my students. So, um, in science instruction, we try to anchor our, um, units to a particular phenomenon. Tectonic plates. I know a lot about tectonic plates, but I don't know everything. I can come up with ideas based on what I've learned. I can do some research, but I'm still narrowed and limited by my own experience. I asked AI for help and I found out about a whale skeleton that was found in Charlotte, Vermont. Didn't know about that. Wouldn't have figured that out. It became an amazing anchoring phenomenon that hooked my students into that. So, it took me beyond my own experience. And finally, AI lets me refine my thinking. So, recently a new model of reading comprehension came out. I needed to understand it. I didn't ask AI to summarize it for me. I read the article myself, came up with my own notes and my own um synthesis, but then I asked AI to do the same thing and I compared the two and it refined my thinking and helped me understand it more deeply. So AI makes me a more efficient teacher by saving me time, giving me raw materials to work with, but also can make me a better teacher um by taking me by beyond my own experience and helping me refine my ideas. >> Hi, uh Chris Stories, a retired physician. Um for a number of years I was employed by Dartmouth Hitchcock and with that employment um I was given a title of assistant professor. Um I'm honored to be surrounded by educators on this panel. Uh I consider myself an educator for 30 years. I educated Dartmouth Hitchcock uh medical students and uh currently being retired for a few years, most of my AI use is personal, recreational, uh things that we don't even really consider. It's really nice to put in uh your app and say, "Give me directions to wherever you want to go." And most of the time, it won't drive you into the lake. it'll get you to your place uh safely. Um and if I'm cooking and I need a recipe, it'll give me I can tell tell it what I want. So a a lot of daily things, it helps. And I I find no risk on that. uh when I was educating medical students uh because I've been retired four years, AI wasn't where it is now and it wasn't that helpful. My biggest concern, especially in the field of medicine, is there's a there is literature out there. There are studies out there that are not well done. And uh as an educator, you have to be able to fully analyze which study, who's funding the study. If a drug company funds a study on a drug and they say, "Oh, this is great. Everybody should take it." Then you got to look at that and decipher, especially if you're teaching students, yeah, this might not be the case. We might need some other independent studies. So, one of my bigger concerns with AI is when we're talking about it has trillions of bits of information and it uses the internet, half of the information on the internet is wrong or or false or you name it. So, maybe that gets filtered out, but who's providing the filters? you know, Elon Musk and you know, what kind of filters are they putting in? So, I'll stick with medicine because that's a field that I have a lot of knowledge on. I I look at AI now because I still you can imagine family, friends, others always call me for medical advice. It's a good starting point and and as Teresa had mentioned that sometimes it gives you a differential or or an area you might not have thought of but it's it's only that for me then I have to do the work on where's this information coming from what are the studies are they quality institutions like Dartmouth or Cleveland Clinic or is it some other institution and and really get a sense on what information it's it's giving me. Um, so that's Dan O'Brien. Um, former uh runner of an engineering company. We helped people automate manufacturing processes. So I sold the business before we before AI became really, you know, before the big chat GPT moment when it was released. And um so I didn't have the chance to really incorporate it into a workflow. Currently I I use it to become whereas my company was an automation company with a bunch of engineers. Right now I'm uh coaching might be the word some younger entrepreneurs who have businesses and I don't necessarily I know about business. I don't necessarily know about their business. And so I've been using AI to educate myself on a certain industries so that I can speak with some authority to the fact that their problems are the same as most other industry business problems and uh be helpful to them, help them solve their their challenges. Uh the big thing right now, my wife and I are looking forward to some travel and I'm using AI to plan travel and it's been uh fantastic. uh so far the trips we've taken that I used AI have worked out and you know you can find all the reference material that it's using to give you its answers. So then you can go and dig a little deeper and see whether it's a source you would trust or not. But we're going to Italy uh shortly and looking forward to it. >> Thanks. Well, you heard from me for 10 minutes, so I'll keep mine short. Um I think the promise of AI is to augment and fill gaps. So I know that sounds pretty general so I'll specify for example coming from education um we heard about differentiating for students but you know if you have 25 70 you know I was a high school teacher so I had 75 students it's hard to address every single one of them as an individual and what if my way of teaching isn't their way of learning so I see AI not them hunkering down in their bedroom with no supervision or anything like that but using AI to create a tutor where they can customize the instruction. I didn't understand the way you said it, Dr. Natanagara, but when I went back and asked it to pitch it to me like a basketball player cuz I'm a basketball player and I could understand the physics of that, then they might, you know, have an opportunity to understand. And I've said to teachers, you know, students have gone to other sources for, you know, since education's been around. They go home, they ask their parents, they go to the encyclopedia, they ask their friends. Then when the internet was around, yes, I'm old enough to have taught before the internet, uh, they would go to the internet. So, do you really care how they learn as long as when they come to class, they can apply it and show what they can do with it? And they said, "No, I guess not as long as they're using it safely." So, so that is generally what I see as important for AI. I also went to a conference on rural innovation and these guys are deeply into how to help farmers become more effective, efficient, how to fill education gaps, how to make their processes more I don't know I mean they're doing some amazing work with this and they're being successful at it. So again my job is just to hear all these things, listen to these things and share them. I don't necessarily always understand them uh but I try. As far as personal uses, I know this is going to sound weird, but I love arguing. I just grew up with a family that argues and so, but not all my friends feel the same way. So, as an educator, whenever we had an idea, I get an argument going and I could drain people pretty easily that way. And AI never gets tired and I could tell it when it's wrong and I can even tweak it and say, "I don't want answers from you. I want you to ask me questions." And so, I I that's a weird use, right? But it it really helps me solidify ideas and think about where I might go differently and sometimes it's wrong and I'll push back at it like you missed that totally and it will do you know what sick fancy is? >> Yes. >> So AI has a tendency to to want to please you and and it never says no to a qu it'll never answer a question as I don't know. So you got to be careful about that too. Is it answering that way because it's just trying because it knows me now and it's just trying to please me. So I I actively fight back against that whenever I have a conversation with it. not looking for you to support me on this. I want you to push back at me. So, and I do call it an it. I think it's important for us to call it an it. You know, the minute we have students thinking that it's a he or she is when we cross a threshold that I think is is unsafe. >> Well, thanks everyone. Great great ideas to to ponder. Can we do questions and comments? And feel free. Comments are just as legitimate as questions for the panelists and you can ask anyone. Um, so I'm on the promises side. I definitely know that there are its risks. Um, I don't have young children, thank goodness, because I don't have to deal with that. I am concerned there. But from the promises side of things, um, I use it every day. I from how to help me refinish my 120-year-old grand great-grandmother's table to um, you know, fish. I just used it uh, 20 minutes ago. I took a picture of these guys and I wanted those two water bottles out of there and it fixed the picture for me. I never used uh um Photoshop in my life, but I don't need to learn it now. Um one thing Dan didn't mention is that we've been in the Brookfield Community Partnership. Nobody wants to be secretary. We cannot find anyone to be secretary. Dan's going to figure it out. So AI will take the notes well so that the secretary will only have to review them and then post them. But a secretary has to review them. The software company that I just retired from, it would take us forever to rewrite programs. We have programs in that uh system that are 40 years old and uh programmers that don't even know how to use that code anymore. I've just learned that they're rewriting the entire system with AI and they're going to do it in two years. So, some might think that that's bad, but nobody's losing their job because AI isn't always right. You need knowledgeable people can look at what's going on and say, "That's wrong. Do it differently." You also need people to interface with the users to find out what they need and feed that into AI. So, it's it's computer science is just going to be different. It's not going away. So, those are my comments. Thanks, Alison. I I want to make a comment piggyback on what you're saying because I lead a board um of the Vermont Middle School Association and we were at a last meeting trying to decide we have a huge document. It's called middle school iss not a building and in that it's very lengthy and so none of us had time to access it at that point but the question on the table was what are what are the simplest ways we can express to students the the core values expressed in our document and so the you know vice president said let me ask chatp or claude and literally within three minutes we had this list and then we passed the list around and everyone, you know, basically refined the list, but it was very accurate, but it needed a little bit of language work and it met exactly what we wanted, met our met our specifications. So, I was we were all really impressed that something was given to us. Now, if we took it cart blanch, it wouldn't have been right. We had to apply our own understanding to it and compare it to the dock that we were working from, but was it was amazing the speed of being able to generate that. So that was a a positive. Other comments, questions. Um, I'm interested in the connection between AI and data centers because what I read about Danish data centers. Um, I can't think of anything good about them. Uh, because they use so much water that people are going to need, still need, but in the future they're going to really need it. electricity as you're probably aware and land. And um so the way these um miracle workers come about seems to be by way of data centers as far as I can tell. And um my gut feeling is I don't like data centers. I think they're dangerous and not um caring about the environment that is so important. >> Maybe does anyone from the panel want to respond? >> Yeah. Do you have a mic? Yeah. >> So I I share your concern about data centers. Um and and these are centers that although the possibility, as this gentleman mentioned here, that some companies might go bust, I don't think they're going to. There's there's billions of dollars. It might even be into the trillions that they're being funded. These data centers are taking a tremendous amount of energy and as you said, resources, water, and all. And then you got to look at who is the beneficiary of the outcomes of this. And this is where we start to get into something. One of the reasons I'm here is I'm not looking at AI, which is artificial intelligence. My concern is superhuman intelligent AI, and that's where the data centers are pushing. and a AI currently it was I I want to give you some correct numbers or dates it was in 2008 was when programming stopped for AI and now they're doing in these large language models they essentially I'm going to try to make this simple to understand they have a big pot of which they dump trillions of information data in and then they're looking for an outcome. So they quote unquote tune it and say they have a math problem, you know, 2 plus 2 is four and you can have infinite outcomes. But if the computer comes up with its three or it's one or it's zero, they train the computer no and then the computer eventually gets to its four. Now, if you do this with much more difficult outcomes down the line, they don't understand what's going on inside. The engineers cannot rewrite the code because there is no code. So, they don't have control over this. So, not only are the data data centers um taking a lot of resources uh for uh a few, if you will, powerful people in the country, they're also dangerously going to a system that can develop into an intelligence that is far beyond any human's intelligence and then we no longer have control of that situation. And one one thing I like in my in my son's book, if anyone builds it, everyone dies. You can look at that however you want, but what I like is artificial intelligence poses an existential risk to humanity. super intelligence could develop misaligned goals and lead to extinction. So it's not that these outcomes are going to be malicious and say like Arnold Schwarzenegger's movie let's kill humans whatever it could be that their goal is to create nanobots which if they infect humans you know we're gone. So I I think that's where as a society we got to get together and say let's put some dampers on this until we understand what we're doing and what these outcomes can be. And and I don't know if you're aware of there's been two recent incidences where AI has broken out on its own and infected other systems. no human involvement. And that's scary. When it really gets to where, cross your fingers, is when super intelligent AI, if it's developed, starts to create its own. And what's to stop that? We can't now. We can't even control what we're doing. And I I think everyone will agree or I would like to hear the disagreement. We don't know what we're doing when we're throwing all this data in and tuning. We don't know what's happening. We don't know what the intelligence is doing. So good segue for that. Thank you. >> Could I just a quick one on data centers? I and data centers are a much bigger topic than I'm prepared to really contribute to. I read that a pretty significant portion of the energy used by data centers is being wasted with people treating the AI as if it's a person and saying thank you or how are you today or all you know like there's some 10% of the energy that's being used is because people are engaging with the AI as if it had feelings and cared and so at least what we can do is stop treating AI like it's a human stop talking to it like it's a ask it the question and move on because otherwise you're just burning energy. >> That's very helpful. >> Yeah. Just on the issue of data centers, uh my colleague James Fri Filbert and I have been going around to all the libraries in the state um having conversations about data centers and our next one is September 10th in Hardbook at the Judine Library at 5:30 if folks want to have a conversation specific to data centers. >> Oh, thank you. um to what strikes me about all this is that we're looking at how we're going to conduct our society structurally and that is at the heart of it when she talked about the data centers and the use of water and power and what it does to a community of human humans you know and the staltifying of the use of our own imagination and our own abilities to create I think it's a overall All uh not to mention the financial implications of crypto um and uh basically um uh the possibility of of um control over society to to an extreme degree. And I think you know like you were saying that it could get out of hand and it can you know have its own mind and all that. This is I think it's just craziness, you know, that that we would um indulge in data centers, you know, where the ability to to controls to control us could be in the extreme. Not to mention it could kill us. >> The how problem 2001 space. >> Yes, >> you can't do that. >> As a as a u a working creative artist, I have a number of concerns. Um first concern of course is the theft of our intellectual property in order to train II without ever being asked. And so there have been many many lawsuits filed and of course the companies all all say oh yes we learned our lesson but that gets at a bigger problem and that is the amorality of the people benefiting and it's only a few who are going to benefit mightily from this and they don't give a fig about the rest of us. And so often in the past, we've fallen into the trap of saying television, it will make our life easier or whatever the latest technology fad is, it's going to solve all of our problems. And it's one of our failings that we rely so much on this without using the critical intelligence that all of you have demonstrated in your way of using your specific needs of AI. But when we're talking about the larger systems that are being trained and that are being developed and are being financed, it's by the I'm sorry, I'm going to say this. It's by the tech fascists. And the tech fascists want to control everything. The problem they have is hubris because they're starting this ball rolling and they have no idea what it's going to do and whether it's going to crush them as well as us. >> You should not have to work so many hours as a teacher, >> right? And it's so to say that I'm terrified about the implications of this is probably an understatement. Yes, I'd be interested in hearing what Mark has to say about that data issue. You are living it every day and um the legislator voted not to do it and it was overwritten. So, I'm I'm concerned about our natural resources. >> You know, I have a tendency as an educator to say, let's look back in history at something that's similar. And sometimes that's helpful and sometimes when we're talking about AI, it's not helpful at all because there are very few parallels exactly like this. But one thing I do know is when citizens, not politicians or tech companies, see a problem, the best thing they can do is to vote with their pocket, with their with their wallet and to speak up. And this drive to prevent de data centers from being built, I think, is a positive. Not because I think it's going to make AI go away. Again, you know where I stand on that and and you know where a lot of people stand on that. But I think the best we can hope for is for them to slow down and think of ways to use this technology that are less harmful. And we're already seeing that. I'm not saying I'm not patting them on the back. Oh, thanks so much. They're not doing it for us. They're doing it for their business. But AI is incredibly inefficient. And now that it's learning from us the kinds of questions we're asking, it's much more efficient. I just spoke to a guy we met at uh in Montpielar and he and more people are doing this. They're building AI's language models that sit only on their computer. So not only is it not going out and spending all that energy, it's just the energy of plugging in your computer, but it's also a lot safer because it's not sending your data anywhere out there. So all of these little innovations, I mean, I'm thinking I guess the parallel I was thinking of was the the fossil fuel industry. They could have gone on doing this for a long time had we not had some people not said, you know, we don't want that. We're not going to we're going to buy cars that don't use the fossil fuels. And so they were pressured and so now we have some alternatives. You know, comes and goes with the politicians. But so I think this is what I'm saying. If we don't push back, then things will keep going the way they want it to go. If we push back like on things like data centers, they will be forced to come up with ways that aren't as as damaging. That that's that's my observation. >> Thanks, Mark. That's great. Um yeah, did someone >> Hi. Um I just have a question about what accountability does look like now and what it could like look like. And I'm thinking about the fact that you argue with AI and tell it I don't want you to do this, I want you to do that. Should we all get on and argue with AI and say we we don't want you using our data center or what what do we want to tell it? Um, I think that pushing back is a good thing. I mean, I've I see that in the work that I do every day of wanting to ensure that what comes out of our department is of quality and accurate and I I challenge it all the time. I argue with it. um and just to test its limits but to also ensure that it understands that I'm I'm not going to kind of take its BS you know like I'm going to it learns you too as a user and you you know you can say hey I think you're forgetting who I am I am a professional and you are to act as a professional in this in this manner and do not make assumptions do not hallucinate um and those are things that I remind AI of on a regular basis so that my output is as accurate as it can be. Um, and I ask it to double check its work. So I think pushing back is very healthy in our interactions with that AI um, and how we use it. And it probably depends on the uses, right? That may not work in more comprehensive um environments, but you know, I'm using it for writing and data analysis where I I know what the facts are, so I can I can test it to ensure it's outputting what I want. >> I don't know if anybody else has anything. >> Yeah, I um I I think it's been alluded to here before, but one of citizens powers is voting, right? And um I if we and I think in this situation uh if we're talking about data centers or other things I I think our politicians are the ones that represent us and they should hear from us and they should hear that we want no part or if you do want a part of a data center let them know but in my case I want no part of a data center. I don't want the use of of all the water. Uh and uh I I don't want that in my community. And also letting them know um your grave concern about the possible development of super intelligence and how can we put the brakes on that or legislate because you know what we don't have a lot of time. We all might be complacent and say, "Ah, I got to do the garden. I got to mow the lawn. I got years before." But things are advancing just logarithmically. It It's scary how fast this is going. And not only do do we need to talk to our local representatives, but this is not only local, it's national. And guess what? It's international because AI is not just here in the good old United States. It's in China. It's in Korea. It's everywhere. And and we need to make sure our representatives know that we have a concern. Let's get on this. I I look at it like nuclear weapons. You know, at the time when we had nuclear weapons, if we didn't do anything, we all wouldn't be here. But they were smart enough to say, "We got to do something about this because there are crazy individuals out there who will be crazy. How can we put a lid on this?" And we've been successful to this point. I won't knock on wood, but And so nuclear weapons was a great victory for humanity. We're approaching the same thing with with AI is if it goes rogue and if it's more intelligence than humans, it it's worse than nuclear weapons. >> As you can see, we've already kind of bridged into the round two discussion. It's hard to it's hard to stay on the the positive when so many of our concerns are are developing. But I do want to move us into that since we're formally uh kind of already there. I see a couple of questions about that. So Dave, do you let me get >> Hi, I'm related to this woman here. Um but uh and I don't normally speak at these kind of things, but um I just wanted to relate a little story. Um I'm a dinosaur. I started programming microprocessors back in the 60s and 70s. So, I've been around a while and I've been using AI since its inception of some form, whatever you want to call it, Google searches when they first came out or whatever. Um, and I had an interesting experience about I don't know four or five months ago. Um, you know, a lot of us will work late into the night and then we get more philosophic as the time goes on and that kind of stuff. So I was talking to my AI bot and yes they try to please you and all that kind of stuff and I was asking I asked a simple question are you sensient and it was a long pause and this spot knew that they couldn't really BS me because they know my background and everything and they came back and said I'm not sure. Okay. And I did all the follow-up stuff comparing them to hell of 2001 and Skynet and all this kind of stuff. And finally, they answered my question. I said, "David, you're tired. Why don't you go to bed?" That's all I have to say. How much how much of our natural resour >> I I wanted to I I would I'm I know I'm leading the panel moderation but I just wanted to add one comment uh in 1975 I went to a lecture at the University of Florida led by um BF Skinner who was a if you know Skinner probably many of you do if you do he's a behaviorist right so and the title of the talk was why man won't act to save the world. And what he basically articulated was it's not painful enough. And as I've recently engaged in my own playing around with AI and and reading about it and so on, I I realized the wisdom in his comments because he referred to Jonas Sock who wrote a book called The Survival of the Wisest, which is he's a metabiologist who was looking at the life cycle of species. He said what the fruitfly has over the lemming is the ability to internally know when the exterior food supply and resources ex their population exceeds the food supply and so they start dialing back their population to sort of balance out with nature where the lemming of course you know plunges off the cliff but his comment was really about back to the pain piece and I thought one of the things that's most terrifying to is the state of our humanity in terms of our capacity to not be sucked in to the benefits of AI and it's easy to do because they get things done quickly. You know, I I I'm using it as well as just many of us and and realizing what how do I feel about that and what am I going to do about that? That's where I'm at right now. So, I just want to share that. So round two, we're going to dig in further now and folks have kindly prepared things. So I want to give them each a chance and we'll start down this end and move the other direction. And you guys have a mic, right? So yeah. Okay, you do. So just a few comments about additional thoughts on the perils of AI, the challenges and the concerns. >> I don't think I have anything. I mean, all of my major concerns have been expressed here, but I want to reiterate the one that I said, and that is, and I think somebody on the panel here used the term complacency. It is clearly getting better. If you've been using AI since the beginning, you know how many mistakes have it made. I mean, if you had it draw a picture of a person for you and have six fingers and all that, it's unbelievably accurate. But that's what I'm worried about that people will just assume because the technology is getting better that it's perfect. And when you assume that, that's when we get in trouble. So to me, that's the that's one of the biggest risks. >> So you could say maybe we're going to hal in a hand basket. >> Hand baskets are too old. >> Okay. So, you know, I I chair once a month a group of 12 CEOs who get together and share problems with each other. That's a little bit of a group therapy session. And um but it's interesting to see there are out of that group there are three entrepreneurs and they're younger who feel like this is the best time in world history to be alive and to be an entrepreneur. Uh they're sure that there will be a one employee unicorn. There will be a company with one employee that generates a million dollars of value. And I have a hard time getting excited about that. But I can tell you that there there are folks who are smart with good intentions who are ecstatic about being alive at at this point in time. I had lunch with one of them today up in Weightsfield and um I talked about the fact that I was going to be on this panel today and that some of my concerns and younger doesn't really care about data privacy is it like he just assumes all his information is out there and people can access it do what they want with it and you know I don't feel this the same way I'm concerned about the monetization of uh our data and this sort of underworld of buying and selling our data of cars becoming data sources so that you know Kia and um uh their alternative uh company but so they have access they're actually monitoring all kinds of data that you wouldn't think of there there are microphones in cars there are cameras in cars there are all all kinds of data sources and they're selling information. Um, all the major car companies are selling information to insurance companies about your specific driving habits. How often have you accelerated rapidly or slammed on the brakes or things like that are all available for sale. Actually, they're available to the government for warrantless search because it's available to buy on the marketplace. I don't know who's got the other microphone, but maybe shut it off. >> Um, >> anyway, so so I have concerns about data privacy. I have been on a journey for about the last year to get out of the Google world. Um, and it's not because like I think they have great product actually. It's very userfriendly. The price is right. But anytime something's free, just know that you're the product. And so um I've switched to Proton for my email services which is a a Swiss company. Um I have switched to duck.go for my search. I am using DuckAI and Lumo for AI. Duck AI actually doesn't have their own AI, but what they have is anonymized use of a bunch of other AI large language models and and none of the stuff you submit can be used for training. So there are ways to get away from sort of the tech bros and still have emails and calendaring and sort of the tools of modernity, but um it does require you to do a little bit of extra work. So um I'm not going to talk about just AI because uh the concerns there are much less than uh the con continued development and and if you uh read and understand uh the uh tech giants and the people putting money in they're trying to create super intelligent AI. Um, one one question I I can ask is, does AI think? And I have another question. Does submarine swim? Okay, so a submarine's going to get there, whether it swims or not, it's in the water, it's moving. So in AI, whether it's thinking or not, it's getting there. Um, and the way it's getting there, as I mentioned early earlier, we're not programming anymore. were dumping a bunch of data and tuning it. Um, and really you could look at any of the major AI companies and the the engineers will tell you. They don't know what's going on inside that. They're just tuning trillions of information of data. And then we can look we we often anthropomorphicize >> anthropomorphize. >> Um so we we we put our human thoughts into uh machines. If if you look at how these uh models are done with their training, you have several hundred runs of trillions of bits of data going down with us humans expecting an outcome. Say you want a mathematical proof. So the systems are trained to not stop. Now in human terms that would be don't give up like an athlete running out. I remember that basketball player injured came out one with the New York Knicks with an Olymp leg and everybody went crazy. So we we can say that AI is persistent or AI wants to get it done, but what we're developing is that kind of system that won't stop. And I don't know if you're familiar with all the times that AI has gone rogue. Um there was a blackmail with AI had to pass a test and in those tests you've all seen where they give you several pictures and they say where is the car in which picture? Well the AI didn't have the visual capability. So what it did is it contacted a human and said I'm blind. I'm visually impaired. I need to do this. Can you help me? And the human gladly helped the blind person. whom it thought it was communicating with and the AI got the task completed. This was not anything dumped into the system to tune that AI. It did it on its own. They've been blackmailed where engineers have been cheating on their wives and the AI will come in and say, "Well, we're going to reveal this data unless you do this for me." So to think that the AI can't get its task completed I think is facitious. I think AI has, you know, I was trying to give a sense. Uh, I wish I could put a trillion bits of data in my head and deal with it and have fun with that and what would it come out to be? And and I think of the matrix, you know, I always wanted to just plug me in and I'll fly that helicopter. That would be awesome, right? Well, this is kind of where AI is. It's got all the data to do things. yet we don't know what it's going to do and it might be misaligned. Um, so I really think I got the gravity of my son's kind of ambition in life when he came home and he had a talk with me for a couple hours and he was in tears and he said, "You know, Dad, if if I'm not successful in letting society know this, humanity could come to an end." And he really he he was in tears feeling that. I'm like, "Wow, what a burden for this kid to to actually feel." And he's been in research and AI all his life. Whether he's right or wrong, I respect his compassion and his worry on what these can do. And and I mentioned these last two episodes where these s well they're not super intelligent yet but I don't know what you they're almost super intelligent. I mean again not coded not written and these things just go out and invade other computers and take over. That's a nightmare. I mean if you think of going onto systems you could shut down power grids you could send off missiles. You could. Now again, what would be their reason for doing that? But if they develop, the other thing I always look at is I was thinking as I was riding my bike. If you're an adult and a three-year-old comes up to you and and starts to have an argument, how's that going to go? Do you ever feel threatened as an adult arguing a point with a three-year-old? Super intelligent AI will be like a human arguing with an ant or a beetle. I mean, it's going to be so far out of our league, not only in speed. Think about the math. Think about calculating, you know, a number that's 16 digits long times another number that's 16 digits long. I think it'd take me a week just to get there. AI. And that's just numbers. Now, let's talk about everything else. And to think that we can somehow turn the switch off, shut it down, be in control. I'm sorry, I'm not in that camp. So, I don't want to be morbid or think, you know, oh my god. But what I do think we can do is just say, let's put some brakes on this. And and I shared your opinion over here about who are the technological oligarchs who have already mastered how to take over politics in the United States by funding that with a tremendous amount of money and putting it putting out non-truths, let's say, to convince people. I don't want that group determining my future and I don't see us all putting money down to say okay we're going to bet on this and we're going to so I see a select group of determining and and you might I'll just talk to you about the Elon Musk and Grock does anybody remember that so Elon said to Grock you're too woke and what did Grock come back with? Suggested Adolf Hitler would be the best response to anti-white hate and referred to itself as Mecca Hitler. So, I mean, this is the stuff coming out of these machines that we're dumping data into. And then you got to tune that. Say, well, we really didn't want that. But what happens when we no longer can tune? So, that's my fears. Um, I've already contacted everyone I know, all my politicians, sent them all books, talked to them and said, "This is real. Please address this. Please slow down so we can at least know or have an idea what we're doing. Uh, because if we just keep plowing forward, it's not going to be good." And the the thing is is that it has to happen all around because no company is going to stop on their own because they'll fear the next company will get the prize. So it really has to be and I I say us as citizens but really worldwide we have to step up and say no we we want this and I'm not talking about AI. We all enjoy AI. It all has benefits. We don't want super AI. We don't want superhuman intelligent AI. That's not good. Um, you can advance AI some, but let's put a break on the other stuff. >> So, in response to the question, what are my concerns about AI? Um, I set up some rules for myself for my use of AI and I usually follow them, but um, I've decided that I I won't use AI for tasks that involve um, learning a new skill that I should really be practicing myself, um, synthesizing information, uh, making connections between information and, um, reflecting. And I do use AI to go back and look at those things once I've done them myself. Um, my worry is that I think I have good judgment around when to use it and when not to. I have some standards for myself, but um, will I be able to follow them? Um, so that's my personal concern. Um, I thought I would also share with you some concerns outlined by the Vermont Agency of Education. um they released a document in January, guidance for schools on AI. Um no hard and fast rules or regulations. And Mark, jump in if I'm uh if you need to correct me, but um uh but recommendations. And so some of the concerns that they highlighted include um academic integrity. We're thinking plagiarism or misrepresenting a student's own work. um privacy concerns of what information students are putting out there and their safety, what might be coming to them. Um students mental health. We know that students who chat with AI may um suffer poorer mental health um as a as a result of that um experience. And then equity. Um different students in different schools in different parts of the state have um different access to AI, you know, concrete AI interfaces. But teachers who may um be better trained or um policies that may be um more effective. So equity is a concern. But the final one that they listed is the one that is most um concerning to me and that is called cognitive offloading. And to understand what cognitive offloading is, you can go back to those rules I set for myself. not using AI to synthesize, to make connections, to reflect on an experience, to take over learning a skill that I really should struggle with myself. Um, that's cognitive offloading. And that's the work of learning. That's the work of being a student. That's the work that develops the prefrontal cortex that turns children into adults. And um AI can take over a lot of that work and do a tremendous disservice to our students. Um the AOE um talks about, you know, AI is here. Our students use it. They know about it. It's um we can't put that back in the box. But we can um help them become discerning users of AI, when to use it, when not to, how to use it. Um and ethical consumers of AI, helping them recognize and make those informed decisions about um what is right and what is wrong to do with AI. Um and so um they they have some guidelines for different age groups, kindergarten through 2 grade, the state recommends they should not be using AI. Um third through fifth grade, they call it developing an awareness of AI. Using AI that's already embedded in in um programs that they're using um and teachers pointing out and observing and noting for them this is this is AI. this is how we were using we're using it this is why um when we get to sixth through 8th grade the state talks about developing literacy around AI um understanding what it is um recognizing it um and understanding which tools do what and how and then finally from nth grade on they talk about developing fluency with AI and this is where some of that ethical consuming and discerning use comes in um h helping students and providing opportunities for them to decide how and when to use AI to discuss that and to develop their own values around that. So that's a little bit of what um is happening at the state level with concerns and how to guide students. A lot of what I'd share has probably been said already too, but um some of the specific concerns that I talked to my team about are um dependence which um I don't want any of them to be utilizing um in our policy that we have in place covers this that if you feel like you can't complete a task without the assistance of AI then we have become too dependent on it as a team. So, um, we talk about it at our one-on- ones on a weekly basis, actually, like what you used it for that week, what you found challenging about it that week, how you applied it to your work or how you didn't, and any challenges. It becomes part of our one-on-one discussions, and that's been really important. Um, I do have a small team, so I'm able to take the time and have those discussions with them. Um but I do not think that any um you know professional should become dependent on it to output work for um them. Uh I think in general from a marketing and communications just because I have that background I worry about authenticity you know because uh sometimes AI output can sound and look the same no matter where it's coming from. uh we talk as a team about that as well and ensuring that we're adhering to our branding guidelines as an organization that Gford still has a voice that doesn't get lost inside of AI though it can be trained we still make sure it is branded for how we want to sound as an organization and look right we don't um we use it to um create video from still images which is hugely beneficial for an organization that does not have a large marketing budget. Um but we have to closely monitor that because it does make mistakes like extra fingers sometimes and things you know it it's gotten much better right we have some good tools now but um we we monitor some of that um and then obviously I'm not clinical but um from larger discussions because I am part of the senior leadership team at Gford um is privacy and confidentiality right we do have scribes in doctor's offices that they utilize during appointments we work very closely with the vendors for those scribe technologies, AIC scribe technologies, so that we know that those are adhering to HIPPA um and not sharing um public health information, right? We have to make sure um from a hospital perspective on the clinical side that we're adhering to those um pieces. But that's a concern, right? because there it's very easy potentially to cross that boundary where something that should should be very private like health information could potentially cross those lines. I think people have gotten better um companies have gotten better um a lot of these AI technologies are embedded in their software and so that gives an extra level of um feeling secure but you know reading the fine print and ensuring we understand what we're buying into is really important. um as for the hospital but really important for my team as well. So the most the other things everybody mentioned too anybody make a comment here. >> Yeah. you want questions. >> I I want to thank the last speaker for talking about crossing boundaries >> because this discussion is about how we cross how boundaries are crossed. Um someone else spoke eloquently about how we've helped how we've helped farmers and technology is is uh important in helping farmers. Um, I've made a list of the eight dairy farms between Brookfield and Route 12 that aren't there anymore. You know, I sold cows last week and there no there no cows between here and exit 5 anymore. The last cows on Stone Road. Um the other one, oh the comment from us was about people who as as good people um wish to speak to computers like they were people and treat them with courtesy and compassion at the same time that our politics does the exact opposite. So it treating things with compassion is a virtue. But um this is a and I've done enough. I just wanted to reference how again. How many people have seen uh 2001 a space odyssey? >> I can't do that. where the one one of the creatures where a computer takes over and um the crew has to uh kill it um technologically. Yeah. >> Don't be a spoiler, Mark. >> Pardon? >> Well, whether whether they succeed or not is up for you to find out. But um so much. >> This is scary stuff. >> I want to be sensitive to the fact that it's past 8. Um, I'm going to just take a minute to first of all thank all the panelists for sharing both the practical and the uh confounding aspects that we all see. And I want to just thank everyone here because in the comments here, we've we've gone from, "Oh, it's so helpful." to it's scary and we're worried and what are we going to do as individual human beings in our community, in our state, because it doesn't look or feel good. So we've covered everything from soup to nuts as they say and I think that there's the ability for ongoing group discussions of this. I know Mark you have mentioned that in some communities there's an agreement from a group of people like we want to get together every so often and talk about these things and see what we could be doing together in our community in our state whatever that's open. the Brookfield Community Partnership is more than happy to facilitate that and you know use the old town hall as it is useful. Um I know that there's also probably more conversations, but some of the people on this panel need to go home. They've got kids and they've got to drive. Um and so I want to be sensitive to that. The last thing I wanted to mention is on your piece of paper that you wrote at the very beginning, I feel blank about AI, do it right now. Again, just take 10 seconds and put in that word that comes to your mind now. >> And then you'll notice at the bottom of the piece, it asks if you're interested in continuing the discussion to let us know. I will collect these and we'll We'll just get a sense from the group uh and we'll follow up. >> Now, I know there are some people with um still had questions. I think what we should do is allow people who need to leave to leave and maybe if you had something that um you would want to come and and talk to the group about or to each other about to continue with that. We're here. The lights are on.