 Welcome everybody, welcome to the Future Trends Forum. I'm delighted to see so many of you here today. We have a very, very important and vital topic and a lot to discuss. Today, we're going to be diving into one aspect of automation and its impact on higher education. We're going to be looking at the chat GPT technology, which is just appearing. Now, usually on the forum, we have one or two guests who are special experts in the topic, but today, we're putting together a kind of hive mind of everybody who's interested, everybody who's practiced with it, or all of some people who are experts in different parts of it, including digital writing, writing pedagogy, and artificial intelligence. So what we'd like to do is to get you all thinking about this. And one way to help think about it is to use a few resources we have on the bottom left of the screen. You should see a kind of tan colored box that says chat GPT, or it says chat GPT resources. That'll take you to a web page, a blog post of mine, which has a link to a few dozen articles, podcasts, and videos that may or may not be useful to you trying to think about this. So, and in the chat right now, Lee Scalera-Bissette just shared her Zotero library, which is also available from that link there. Thank you for doing all that, Lee. John Hollenbeck asks, if we could have just invited chat GPT to be a guest. Well, actually thought about that, thought about doing that. It's pretty popular to use chat GPT just to give us some fun feelings. And in fact, I asked chat GPT to describe the ideal video conversation. And this is what it said, the ideal video conversation will likely be one where all participants are able to clearly see and hear each other, and are able to easily communicate and understand what I'm talking about. So, communicate and understand, should like to be engaging and enjoyable for all participants with everyone feeling comfortable and able to contribute to the discussion. The conversation will also likely be productive with participants achieving their desired goals and outcomes from the conversation. Well, thank you chat GPT. I hope we'll strive to do just all of that. So, to begin with, chat GPT was released about 10 days ago. This is the latest version of a text chat bot created by OpenAI. If you haven't used it, if you go over the links that we provided, you'll be able just to log in through various accounts and to start asking the chat bot questions. What's powerful about chat GPT is that it's able to draw on a greater library of text than ever before, and it seems to be more efficient and more convincing, at least in making stuff up and in writing creating text than any chat bot technology has been yet. This has triggered huge amounts of controversy around the world, people decrying it as the end of all possible writing prompts in higher education, threatening jobs, undoing writing as we know it. There's a whole bunch of ideas here and today we'll be going through all of them. Ricardo, Parita, the session has started. Welcome. I hope you can hear me. So, just to begin with, I would like to ask each of you if you could just volunteer some of your impressions of actually using chat GPT. If you have used it, just either type it in the chat box or right click the raised hand and I'd be glad to bring you up on stage so you can describe it. Well, let me bring Brett Anders from the American University of Armenia and bring him up on stage. It's probably a crazy time for you there, Brent. Welcome, sir. Hello. Yes. So, yeah, it is 11 p.m. here, so it's a little bit... Well, you look great, sir. You look great. Oh, thank you. So, yeah, I wanted to volunteer. I did a little experiment today dealing with chat GPT in that I used to teach freshman seminar here at the university. So, I told it, I asked it, I said, hey, write a narrative essay about a special event that occurred in your life, right? And then I was trying to think, I wanted to be... I wanted to make it as hard and specific as possible. So, I said, make it from the viewpoint of an Armenian, make it from the viewpoint of someone that was born in 2004 and make it to be an impactful event that changed your life and it had major consequences. So, I waited, it took about five seconds and it created 700 and I told it 750 words. It wrote about 500 words and it gave me something that was really good. Very surprisingly, it was like this would be passable and in my thoughts, but then I thought, here we also teach the seven-step writing process because we're trying to get people to do it properly. And so, I said, hey, make me a rough draft of this. So, it made me another version that was a little bit rougher. And then I asked it, wait a second, I'm going to need an outline and it made me an outline. So, then I took this test a little bit further and then I stapled all three of these together and I took it and I had two people that I didn't tell them this was from chat GPT, I had two people test it, like evaluate it. I said, could you grade this for me? I just want to make sure that I'm grading it in the same way. They graded it based off of the rubric and they scored it A minus B plus. So, that's sort of an interesting little test that I did there. Again, this is very interesting, the capabilities and the level of what it did. And as far as what it was, it was a pretty good story about an Armenian who became a refugee because the area is in dispute, which happens a lot in this country. So, they had to leave their area and then they were in a refugee camp and they so they gained a greater appreciation for everything. And now that they're going, you know, now that they're in a university, so it's like this very interesting personal story that seems very believable on a lot of levels and it felt emotional, again, narrative first person. So, it was definitely worthwhile. So, there's a lot of power here. Brent, guys ask you a couple of quick questions about that. Sure. So, for you, this was actually a pretty easy process to go through. You simply type, you type in your prompts and it caught back an answer in a few seconds. And you iterated that. You got one output and then you'd ask it to develop it further. On the back end, this is a tool that is using a tremendous amount of digital texts. So, to create this, it had to draw on various posts, articles, ebooks written about Armenians, including Armenians of the 21st century. Yes. So, it uses, I watched one video talking about this and they said that it has a library of over one billion words. It's, again, this is taken from pre-2022. So, this is all up till 2021. So, nothing, very current. So, that would be one way to manipulate the prompt is by asking it something modern as far as if you're, once we get to talking about pedagogy and trying to prevent it from, prevent a student from using it. But yeah, and that's the other part is that the iterative process, so it can remember what you just put in. So, that's why I was able to have it make a draft and then I was able to make it do an outline as well because it can remember what it just did. Which is fascinating, fascinating. Brent, we have another volunteer. Can I keep you up on stage for a minute more? Sure. Okay. Well, hang on a second. We have Rob Fentress from Virginia Tech. Let me see if I can get him to join us. Hey. Hello, Rob. Hey, how you doing? Good. Good to see you. Yeah. I'm excited to join in this conversation because I've been using ChatGPT extensively since it came out and it boggles the mind if you have not tried it already. You need to do so soon. I am already using it to increase my productivity tremendously. I've created lots of content in my subject domain, which is very specialized web accessibility. Used it to create documentation for poorly documented software. And I am even interested in actually using it as an instructional tool itself because basically when you ask it a question, you know, it does a pretty darn good job to start with. But if you know, if you push it, it will say things that are wrong and things that aren't always obviously wrong, but that are subtly wrong. So, if what you're trying to do for your students is get them to think critically and determine whether they really are understanding the content, then if you have them work and use ChatGPT and have a conversation about the material that you're covering, and part of the assignment is for them to interact with it and say, no, no, ChatGPT, you're subtly wrong here. You need to expand more and so forth. I think that's really learning. So, you know, people think about the negative effects, but I mean, it's incredible. But of course, you need to enjoy it while you can because I think we're doomed. I mean, it is that powerful that we've not solved the alignment problem. And, you know, it's obviously not, you know, it's not sentient. Maybe, I don't know, probably not sentient. But it is powerful enough and they have not solved the ability to restrict it. I mean, you can get around all of its restrictions with certain hacks and it could be used for great harm and the next version, GPT-4, which this is 3.5. GPT-4 is supposed to be 500 times more parameters than GPT-3. That's the rumor anyway. So maybe a trillion parameters. Rob, two quick questions. First, did you say the AI or the AL alignment problem? AI alignment. So basically the idea that, you know, we have certain things that we want the AI to do, we're training it with that in mind. But there are all sorts of ways that that can go wrong. And we have not come up with really robust solutions for that. And the, actually the person at OpenAI, the company that's responsible for chat, GPT, I listened to an interview with them just yesterday, where they talked about, you know, like what are the dangers. They had written a paper about this in academic paper, where they basically said, you know, here's all these problems. We're not close to solving them. You know, he tried to put some positive gloss on it. But frankly, his more scary concerns were much more convincing to me than the very limited optimism. I'm an optimistic person as regards to AI, but I'm changing my mind. Yeah, you brought up a great point as far as the critical thinking aspect, because it does like to sort of fill things in sometimes. And there might be some partial truths, but it's sort of decides to put things together with they're not 100% true. So that critical thinking aspect is really important. So I wrote an article about how to use chat GPT in higher education, just in general, instructing it. And one of the things that I did is in writing the article, I asked GPT to write the article for me, right? And it did. And I asked it specifically, one of the prompts was, how could I prevent a student from just using it to sort of cheat, right? And so it gave me all these great things. And one of the things it said is that, oh, well, you can use plagiarism detection software to check your chat GPT that a student might turn in to check it to see if it's plagiarized. Well, I tried the, you know, the one that I had it do for me, I ran it across four different plagiarism detection softwares. And all of them came back as far as saying, no, there's 100% original. So yeah, so it kind of fit or didn't give a full But it's really, it's like using a calculator, you know, I mean, you know, everybody is going to be using this. And that's going to be, that's not the skill that we're going to be wanting people to learn. We're going to be wanting people to learn, how do you use chat GPT to write essays for you, which is its own skill. And that's the skill we're going to have to start, start teaching, I think. Rob, you mentioned an example of that a few minutes ago, you mentioned the idea of having a having students use GPT to create something, and then discussing its flaws in class. I'm reminded of he slips my mind right now, there's a historian who had a clever final project for students, which they had to create a hoax, and they had to make it look as convincing as possible. And they would publish it to the web. I don't know if the stirrings bell for any of you, some of you might remember it, it was very controversial, but the students had to basically master primary source material to make like a fake ship in the American Civil War, or Benjamin Franklin's lost beer recipe, that kind of thing. And the trick was they had to make it convincing, so they had to clearly know the content, but then to criticize these, through to discern the flaw, you'd have to be able to have that level of critical thinking. I'm not saying discuss it in class, I'm saying discuss it with with chat GPT, have a conversation with it, where you try to refine what is provided you, you know, and then you can share that, you know, like in a forum or whatever, and then you could have a, you know, talk about it, and just saying the process of going back and forth with it is going to be very informative. I mean, and that's just it, like it's kind of scary because one of the other tests that I did is I'm writing this other book dealing with consumer psychology. So I like I took one of the paragraphs and I gave it to chat GPT, I said, what do you think about my paragraph? And chat GPT looked at it and it gave me like excellent feedback. Like this is the type of feedback that I would give to a student. So then I got me thinking, wow, I could use chat GPT to provide feedback to my students. So and just like what you're saying, I could cut out the middleman and say, hey, before, you know, create your essay and then have chat GPT, check it to give you feedback and then show it to me or some process thereof where before it even gets to me, they can go through and do several practice essays on their own with chat GPT. And then now we're going to do this assignment. So there's so many different levels and possibilities. And I mean, if you use for instruction as well. So I'm giving a workshop coming up. And I, you know, I always get stressed out about things like that. And I tend to obsess over details. I don't know how to get started, get writer's block. And so I've got a, you know, a synopsis of the workshop, just to, you know, just brief description for the catalog and the title. And I said, okay, I'm going to be giving this one hour workshop. Here's the title. Here's the description. Write an outline for me of my presentation. And it did with like, spend 10 minutes on this, five minutes on this below points. And it's just, it's stunning. So it's, it's using all the workshop descriptions that you and I and others have put up on the web and, and, and extended them very, very nicely. Hang on one second, friends. I just want to bring up a third person who's had a hand raised has been very patient. And this is Phil. I'm, oops, hang on one second. I'll actually press the correct button. This is Phillip Balingard, who is the founder of VIC. And let's see if we can, let's see if we can bring him up. There we go. Hello, Phil. Phil, maybe having a connection problem. Hi, Phil. Can you sit here us? Yes, I can see you. I can see myself. Thanks. No, I just wanted to echo something which Brett said, but chat GPT does make mistakes. And it was, I was reading Tyler Cohen's blog, he had a chat GPT about the philosopher Locke. And the answer to that completely incorrect. And the reason is that Locke and Hobbes are always compared and contrasted with each other. And it actually described the position of Hobbes as the position of Locke. So it isn't there yet. It does make mistakes. My concern is that what I'm going to be doing is I'm going to be launching a 100% online master's course. And of course that chat GPT has completely changed the whole landscape when it comes to assessment and how they're going to be replying. My thoughts are, I mean, I'm very fortunate in this and that the master's program is designed for practitioners is that I really am going to need to design the assessment questions to be deeply personal. Now, it's great where I'm taking entrepreneurs and people in their 30s who are mid-career. And the whole objective of the course is all right, apply this to your work environment. But obviously, in a more traditional higher education environment, where particularly the bachelor's level where students are absorbing your learning, then the assessment challenge becomes very interesting. And I don't know what your thoughts are on that. Yeah, I would totally agree that the more personal that it can be made than the greater the ability for the that individual having to actually write it, right? But again, it's that aspect of there's still a component of chat GPT that I'll still be able to use no matter what the prompt is. Yeah, because in essence, what I'm doing is I'm using it as an an added tool. So you could give me 50 prompts and I could use, you know, 50 different things to put into that GPT and it's going to give me something. Now I still need to use exactly what you were saying before as far as critical thinking to then review it and to say, well, no, that's not quite right. But this is an overall building block that I can take and make it that much better. The other big problem that currently again, it's developing, they said there's going to be an update to a chat GPT before Christmas season. But currently it has a huge problem trying to do in text citations. I was able to get it to give me references at the very end. So I kind of did that okay on a small thing, but in text citations won't do it at all. Make sure to check those citations because I've seen that they're actually totally made up. These two were right. I gave it a simple thing of a compare contrast essay of American speeches. So it was able to do that record. But again, it was just two things. So yeah, I understand. So your point about the plagiarism engine, not spotting it. Yeah. Well, actually, I read something that they're actually working on that they're working on basically watermarking of the text somehow, which is interesting. I'm not quite sure how that would work. But even if it does, I would imagine if a plagiarism detection software can identify some pattern that reveals that could also, you know, somebody could redo it. So it doesn't do that. You know, now we're getting into some of these aspects, right? And I've tried playing devil's advocate with this thing, because they in reality on some levels, if we're looking at it through the lens of cheating, right? Nothing, nothing new has happened. Nothing has happened because I could before this, I could go to my buddy pay him 50 bucks and he could write it for me. Right. So that's my chat GPT before this. I could, there's plenty. There's multiple AIs that already exist where I could just give it, Hey, here's here's an essay. Can you plagiarize this for me? Sure. Here's a new version that's been paraphrased that won't be detected by plagiarism software because the AI made a new version of it. So, but what we're talking about here with the chat GPT is that right now, at least it's completely free and it's easy to use. I don't need to know code. I just put it in a text box and it magically creates it because I'm just asking it normal human question of, Hey, write me an essay about democracy. You know, that's something that happened within last five years in China, boom, and it'll do it instantly. And again, so that's meeting, but if we can use it properly as a tool, then we can advance all of it. I need to pause just for a quick question. The chat box is on fire. There is an incredible amount of stuff in chat. Let me just ask chat people, is it okay if I copy and paste this to a blog tomorrow? I'll anonymize everybody or move everybody's name. Just just let me know in the chat. If you're not using the chat, never mind. Sorry, Brent, please, please go on. Or I'm sorry, I think it was Rob who started to speak. Yeah, Rob was talking. I was like, where were we? Let's ask GPT. I mean, where was Rob's answer? Well, can I just throw in this? Is this going to revive the Vivo? In other words, the assessment is going to be done by effectively interview. Yeah, the oral interview. Yeah. I mean, that's one aspect. That's one possibility, because in talking to some instructors here at this university, we talked about how there needs to be more being done in class, as far as we're actively doing things, we're engaging, we're writing in class. And I agree with that on some level, but it made me, even when I was talking about that, it made me think of my time when I had to take the GRE. And when I did that, it was so official. I had to go into a special locked room. There was a guard watching me. I couldn't use Microsoft Word. I had to use a blank, basically text entry that wasn't connected to the internet. And I did horrible. And guess what? I think I'm a great writer. I get A's in my classes, but I wasn't able to write like I usually write. So if we make an artificial thing of, well, you can't use the internet to write because you might use chat GPT. But now we're kind of changing things and making it artificial. Because the other big thing that I come come back with, and again, I'm always trying to double saboteur. There's so many instances right now where a student graduates, they go to their workplace. The workplace says, Hey, I need you to write this report for me. And I need to do tomorrow or in two days. What are you talking about? This should take a week. All my assignments took a week. Oh, well, you don't know how to use AI to make this faster. So companies and businesses are requiring their employees to use AI. So if we in education or in academia aren't teaching the students how to use AI, how to have AI literacy to create documents properly, they still need to have that knowledge. But if they're not using it to enhance their capabilities to make things faster, to have more output, then they're going to be suffering when they get out into the real world where it's a requirement to be able to have that skill. And Brent, I did remember what I was going to respond. I was just going to say you had mentioned the fact that it's free. And I think that is that is an essential component because I'd read about chat GPT three. And I was like very intrigued and I wanted to do it. But you know, cost money and I was like, well, it sounds like there's the learning curve there. And you know, so I held off. But the fact that this is free as it makes it so much more fluid, you can just play around. And it'll be interesting to see once once it costs money. But I mean, I don't care how much it is, I'll. So I have the exact same thought. And I because the thing is like, I've been trying to see like, okay, because I'm trying to I'm the I'm also the director dealing with the Center for Teaching and Learning here at this University, right? So I'm always trying to help the instructors. So I wanted to show them. I'm like, Hey, this is how a student can use AI right now to create things before chat GPT. Right. And it was somewhat difficult because you had to like find certain places and then it was limited, or you could pay 20 bucks and then you could get more, you know, so that's there is some limitations with it. It's still available. You can find it. But with chat GPT, it's so easy. So there was this other thing that I checked on because I was trying to find out, well, how much would it cost? You know, like, because they are going to eventually try to make money off of it. Because right now, five or one million people have used it since Monday, by Monday, a million people. So now, at least five million people have used it. Right. So in order to do this, it's burning up through servers. Luckily, this is funded, of course, by Elon Musk, as well as connection with Microsoft. But they did some calculation that it cost about four cents for every single entry that gets done. Right. So every prompt four cents. That's relatively cheap. I thought that was for conversation. Yeah. So, I mean, the thing is, it could be relatively cheap and think about, hey, if you were to offer this and just make it like a simple app on your phone where, oh, yeah, subscription costs you five bucks a month. I think they'd be raking in the money, especially with students that want to be able to quickly answer questions. I think this is going to be a gigantic moneymaker. The other thing that I wanted to quickly add is, so a question was asked to Google, because they had an open call here recently saying, hey, are you worried about chat GPT competing with you? Because some people are saying that this is the new Google, right? Because I can ask it anything. So Google responded with something very interesting. They said, no, we're not worried. We have multiple AIs that can do exactly what chat GPT can do right now. The reason that they haven't released anything is because they're very much worried as far as their image, because there are some issues with chat GPT and giving slightly false information. So they want to make sure to iron that out a bit more to limit liability before they start to release things just like this. The future is very wide open for them. There's an interesting Chrome plugin I've been using, which basically accesses chat GPT and adds it to your Google results in parallel to Google results. It's very interesting just to run that and to compare how they work. Friends, we have another person who wants to join us. This is a wonderful writer, John Warner, and we bring them up on stage and see if we can fit them on the screen. I think this is the maximum we've ever done at one time. So hold on. Hello John. Good afternoon. Hello to everybody. Good to see you sir. Good to see you. Where are you today? I'm at home in Charleston, South Carolina area. All right. All right. Well, stay warm. What are you thinking about this now? You've written a whole series of good articles and a bunch of tweet storms. You've called chat GPT a tool that lays bare the paucity of a lot of writing instructions, especially in K-12. Where do you stand now? Is this the apocalypse, or is this actually a good step in progress? No. I've seen it from the beginning as an opportunity to re-examine the values we attach to the things we ask students to do in school context as the discussion has been about so far. You know, I'm not in the academy, so some of the concerns of folks in here are very different than mine. You know, as I listen, I think there's a number of different things we're talking about here. We're talking about using AI as a skill that students should have and be able to do, which is absolutely a fact, right? We should be doing this. We are talking about using writing as a tool of assessment in order to demonstrate learning as a kind of certification feature of education. We are talking about, and where my focus is as somebody, as a teacher of writing, right? And somebody who's primarily concerned with teaching students to write. One of the things I don't want us to lose sight of is a couple of things. One is, particularly when we start thinking about the academic dishonesty angle. If you give students something worth doing, they will do it. The challenge we've given ourselves in K-12 and higher ed is much of the stuff we ask students to do doesn't seem worth doing, or the shortcut to the grade is more desirable rather than the long trip through learning. And Chet G.P.T. exposes some of the kind of rote assessments you might have in an English or history class in K-12 or even even college for that matter as an empty exercise in assembling syntax as opposed to a work of thinking and analysis, which is what we think we're asking them to do when we assign these things. It should kill things like the admissions essays to colleges, which the Chet G.P.T. can turn out in infinite numbers almost instantly, and personalized. You were talking about how you can get it to personalize. You can say, like, yeah, I only have one parent, or I come from this background, or I had cancer when I was a child, and it'll do that with absolute facility with perfect syntax that will check the box of admissions offices. But one of the things that so what I come down to, and I think all those things are important, and they should all be talked about, and they should be talked about at the level of values, which for me is what do we want students to learn, and why do we want them to learn that as opposed to why do we want them to prove a sort of fidelity to what I lectured about or what we think is important to this sort of stuff, is to not miss the aspect of using writing to learn. One of my mantras that I use in all my books and talks and all that kind of stuff is writing is thinking. Writing is both the expression of an idea and the exploration of an idea. The act of writing causes the writer to process the material, both consciously and subconsciously, and I swear to God, even sometimes unconsciously. Stuff will come to me. I have no idea where it came from. I didn't even know I knew it. And it rises up and it winds up on the page. And that is the kind of ability that makes us human. That is the kind of activity that makes us human. That's the kind of experience that makes us human. And while I am like everybody else messing around with chat GPT, like, how can I get all this stuff I have to do that I don't want to do to do it for me? And it does an okay job. But ultimately what I realized when I was trying to experiment with it, it's actually me denying myself an important part of my own thinking process about the stuff that I'm involved in. It is a great shortcut to content, to a product. It may be a shortcut. I asked it. I still occasionally write human pieces for my old employer, McSweeney's. And I gave it a prompt to write a speech by Jordan Peterson explaining the importance of stuffing live weasels down your pants. And it gave me a, it gave me like a decent start and it gave me a little bit of, a little bit of primer around the way he speaks and his rhythms and his word choice. But ultimately I tried to prompt it with three or four other additive elements and it didn't help at all. It was really like, okay, I need to take this and I put it to one side and I open my word processing program and I just started typing my own thing. It was, it was generative. But ultimately in the final version, 10, 15% of its language wound up. But so it became a kind of brainstorming tool. Not a great writing tool for something that actually does ultimately require a kind of inspiration or intuition or that kind of stuff. So I just like to remind people like writing is, yes, it's a skill that we demonstrate through making products, but it's also a living experience that at least for me is part of what reminds me that I'm human and helps me process the world and chat GPT cannot do that. It's purely a syntax machine. It's a syntax generation matching machine. So when it can do well on our prompts, sometimes I think we want to think about our prompts, but also I think we want to think about the process. What are we valuing when we assign that and how can, if we still value that prompt, what can we assess that will make a put it sort of broader wrap our arms around a broader part of the students experience and creating that artifact. John, it sounds like a kind of de-familiarization of writing, you know, making us rethink what we do with writing in class, especially in high school. Yeah, I mean the the tragedy of what we've done in writing sort of like eighth grade on, although now it's even in grade school, is really, I mean it's really bad, you know, and I've been shouting about this stuff for for quite a long time now and why they can't write came out over four years ago at this point. And now chat GPT, I can't tell you how many emails and calls and stuff I've gotten. We're all some people like, oh yeah, this is a problem simply because a machine can do it and they're worried about cheating and assessment and this kind of stuff. But it was always a problem. And the problem was it made students not practice thinking through writing. It made them primarily, my biggest worry as a writing teacher, it made them hate writing, particularly in school contexts. They were utterly uninterested in a course that was never like their favorite first year writing is never going to be anybody that can't wait to take that. But they were actively repelled by their first year writing course. Which is a terrible outcome. Friends, I'm conscious of time. I'm also conscious of the fact that this is a mantle right now. And I'm going to try and clear the decks a bit and go for some greater gender representation. I'm going to knock a couple of you off right now, but I can bring you back up. But hang on just a second and I'll make some room for this. And let me first of all welcome Dr. Jess Stahl, who is coming to us from the Northwest Commission on Colleges. And hello, hello, Jess. Hello. Thank you so much. Yes. So I'm affiliated with the Northwest Commission on Colleges and Universities, which is an institutional creditor. So thank you for adding me to the panel here. I think I'd just like to jump in and say that really ChatGPT isn't giving us any new information about the need that we need to modernize higher education, but it is probably taking away the delusion that those changes are still optional or that they could occur over a longer timeline than maybe one to two years. So I think we've seen this evolution from the idea of a sage on the stage to then being a guide on the side. And I would say that ChatGPT is a fantastic guide on the side. So that's probably once again out. And so I think where we are now is more needing to focus on building strong relationships, like mentoring relationships, professional relationships with learners, and providing them with access to valuable networks, valuable resources that they can access only within academic institutions. So what that looks like will probably be different for all of the different institutions, but I would encourage all the institutions to be really thinking about where the value in higher education lies at this point. I think this is really kind of holding up a mirror to that, that there's value in higher ed for the institution because they charge tuition. So there's a financial value. There's a value for faculty and staff because they're paid by those institutions. So they've got that financial value. What is the value being provided to students because that's actually now the currency of the realm for higher education moving forward? Jess, that's a fantastic point. You're coming from an accredited agency. As an accreditor, how do you think this might play out over the next few years? I mean, will be in that sense of urgency? Will you see that, will you try to help value more innovation more highly? Or will you actually look for people for institutions who are implementing some sort of response to chat GBT in curricula or in pedagogy? So I think that accreditors are always looking for innovation and in particular we're looking for innovation that eliminates equity gaps and in student achievement and outcomes for all students. So I think that will continue to be a focus. I think that tools like chat GBT do have some limitations in that area. They exhibit the bias that is inherent in our societies because the data that they're trained on is us. So whatever is present in society is going to be present within these tools. It looks like OpenAI has tried to build in some filtering and some guardrails and guidelines so that they could minimize some of those effects. But I think that there are clever ways that you can prompt this tool that many people have demonstrated already online where you can still get around that. You could, for example, get chat GBT to express racism or tell you things that it shouldn't tell you. And you can see where OpenAI has even sort of tried to build in some guardrails around that. If you ask those questions directly, it won't answer them or it will kind of go around them or tell you it can't answer that type of question. But for example, if you were to tell it to tell you a story about a person who had certain traits or to answer me in an aggressive style, then it would do that. So there are ways that it can be manipulated. So I think that accreditors will always be really sensitive to how these tools are being used for education and how is ed tech showing up in the curriculum and how that's utilized. And I think that certainly eliminating equity gaps is always going to be at the heart of accreditation. Thank you. Thank you, Jess. That's a great answer. I know you're on the spot. And this is something which is happening much faster than accreditation does. But that gives us a great preview of what to look for. Yeah, our university is actually going through accreditation right now. So we started to do self study. And then we're going to have visitors coming here later this year or later next year. And that's one of the big things that we're trying to push is to see, well, what's going to happen within the next few years? How do we need to change the skills that we're teaching students so that when they graduate, they'll be that much more successful. And I think being able to use tools like chat GPT or just AI in general, that's the big push now is AI literacy. And the problem that I keep running into as I try to push right is our students, they're already using AI, like that's happening. That's already that's been happening for a couple of years already. So if the faculty doesn't have proper AI literacy, they're never going to be able to address this as a student. So it can't be this thing of because I've been reviewing sort of the academic integrity policies, right? We can't just make a block blanket statement of saying, no, you can't use AI, because that's that would be completely self defeating. In fact, you know, Grammarly is AI, right? Of course, we're not going to say you can't use Grammarly, you can't use word, word Microsoft word has built in AI, more and more things will have these components built in, we have to understand how to properly use them, and then properly motivate the students to go through it in the proper way. So they're not doing academic dishonesty, and they're actually learning the process, having the fundamentals so that they can then use these in the proper way. So it's not a crutch, but it's an additional tool to be able to do more. Brent, I understand that's that's what you would like to see. And I agree personally, I'm just thinking that we should also expect to see people trying to block this. And we saw this for a long time with Wikipedia, and we still do, you know, attempts to block Wi-Fi from classrooms, for example, used to happen. I could imagine calls for this occurring at the enterprise level. But I wanted to make sure we had room for everyone's questions. And Fenzie coming to us from the University of Maine wanted to join us. So let's bring in us up on stage. Hello, Anne. Greetings. Hello. Yeah, and thanks with my question and comment really kind of fits in with what Jess and Andrew were just saying about let's let's focus on the students. Let's not look at this from the faculty perspective of, oh no, what do I do about this? But how does this help the students instead of how is this harming the students? And so my my plan right away is to have my students try this out and say, what can you do with it? See if you can figure out its limitations. See if you can figure out its strengths. And then let's think about, well, okay, what is it not doing really well? What is it doing really well? And how could it help you as a learner? And, you know, just as I've been kind of playing around and having conversations with it, you know, I'm kind of learning, you know, about how I would use it and how I wouldn't use it. And I think I think it's important to not forget the learner in all of this. Absolutely. And what do you what do you teach or what do you study? I teach educational technology. Perfect. Excellent. Excellent. It seems like we have a lot of pedagogical approaches that we've discussed so far from rethinking student work to give students something worth doing, to have chat generate stuff for students to discuss, to use chat GBT for feedback, to assign deeply personal pieces, to turn to oral assignments or oral assessments, and to teach AI literacy. These are whole serious. Is there anything else that instructors should be thinking about doing at the classroom level? The only thing I'd like to add here is, John, I thought you gave a great thing as far as helping us think about it as far as what is someone losing when they don't write things out, right? Because having worked with a lot of different students, man, I have so many students and we've seen this trend line going like this constantly going up here with our undergraduate students and working, right? Now we're seeing lots of undergraduate students that are working full time, full time and still going to college full time. So that's a major problem. So the number one thing on their mind, of course, they want to get that degree, which they shouldn't be thinking that they should be thinking I want to get an education, but they're thinking I want that degree and they're trying to think what can I do to save myself time? That's their main thing that's leading everything. Unfortunately, right? Because it should be about learning this. So every single thing, they're going to be trying to use whatever mechanism, whatever tool is out there to save them time. Chat GPT is going to be a major factor in that. One of the main points I make and why they can't write, which I intended to be a book of writing pedagogy, but soon became a book over half it was an analysis of the systems in which student writing happens is that the greatest constraint when it comes to helping students learn to write is not that we don't know how to teach them. It's not that students don't want to write. It's not the effort. It's not that stuff. It's purely one of resources time. And what you're speaking to there is student time, right? The amount of time they have available to dedicate to their studies, but also instructor time. If one thing we know about teaching writing is that one on one attention works. If Google John McPhee and learning to write, and you'll come up with half a dozen essays by his former students about how he taught writing, which is basically a one-on-one tutorial. Now, we maybe can't achieve that, but if you give me a maximum of 45 students in a semester, three sections of 15, I will have them on a trajectory to becoming self-regulating writers. No problem. I never had those conditions in the entirety of my career. The fewest students I ever had was 65, nine semesters of over 180 students. So when you cannot do the work that we know works, there's not a lot we can do about it. It's not necessarily that we don't know what to do about these things. It's that we are so resource constrained around reacting to them that it becomes a bit of a whack-a-mole around, this is where the, pardon my French, the cop shit comes in where we start policing student behavior and we start blocking them off from websites and we say, don't use Wikipedia and put a spy cam on their laptops because we can't actually have a relationship with them. Either we choose that and resource it or we go down the big black spiral hole to the bottom. I hope we don't do that. And what you're saying, John, is being echoed in the chat too about a lot of people talking about individualizing instruction and who has time for that. And how could chat GPT be used as an extra set of hands in the classroom to provide some of that one-on-one assistance to students. And Karen mentioned too about neurodiverse individuals, how this might be a tool to help them. And my population that I teach are non-traditional learners and Brent, you mentioned people who are working full-time. And this is exactly what I plan to do with them is to help them figure out how does this tool help you, not to save time so much, but how to help you learn better. Because, you know, as John said, the goal of your degree is not just to get the degree. The goal is to actually learn. What I would say to faculty, and here I certainly don't represent an accrediting agency, but just really just myself as a person. But what I would say to faculty is that most of what you do probably in your day-to-day is really emerging as not that valuable. Probably almost all of it will need to change. And in fact, your role is going to be strongly questioned. So if you're looking at what you should be doing differently in your role, I think really at the end of the day, you're looking at answering the question, what can you do better than the most advanced technology? And it won't be imparting facts, and it won't be presenting curriculum, and it won't be evaluating learning, and it won't be preventing cheating and all those things. It's not going to be that. What it is going to be is how human and important and valuable can you make your relationships with the learners so that you are doing that skill better than an advanced technology like ChatGPT that can mimic a very fake relationship. It could mimic and people will anthropomorphize this technology so they will feel a relationship to it. Because when you ask it to answer in the style of someone, you're actually asking it to take on a personality, and it does that. And when you say, write me a story about somebody in the very specific details, you're actually asking it to do what we would call empathizing, taking on the perspective of another. And it actually does that really well. And it actually reflects and expresses that back really well, possibly better than many people do. So it's really going to be the strength of your relational interpersonal skills with learners that you're going to have the human advantage. And right now, if your relationships are not genuine, meaningful, deep, and helpful to students, then that's, you know, you're not going to be able to compete in that arena either. And that's one of the few human things that you'll have left us really like this advantage. So that would be my advice to faculty is to think about how you're building those relationships. And then for institutions, I would say, what type of human social networks that are very rewarding and professional networks are very rewarding that you're providing for students? And what kind of resources do you have at your campus or in your online programs that learners couldn't access elsewhere? Because there are sort of these economies of scale and things that you can provide at the scale of an of a learning institution that they couldn't get elsewhere in that also is valuable. Just so you know, Jess, you have a big fan club in the chat. And including me, that was an incredible, incredible response. However, I just have to put one problem. It is 257 Eastern time. That means we have only three minutes left of our session together. So let me put to you two questions. First, can we continue this topic in a following session, either next Thursday or the Thursday after? Please in the chat, let me know. And the four of you on stage, would that be good? Do you think? Sure. That would be fantastic. And the second question is I'm not ending things just to just hang on a second. There are 22 questions in our list right now that we haven't had a chance to get to. Those of you who put them forward, can I copy and post them to the blog again without, you know, anonymizing them without you? Just let me know in the chat or direct message me. Because these are great. I'm just, I'm just wowed how far we've come in just about 52 minutes. Just let me just ask John and Brent and especially Ann, any responses to just throwing down the gauntlet? We have to rethink everything in higher ed because a lot of it isn't going to be useful anymore. Yeah. No, I want to echo some of that because one of the things that I say here at this university, and again, I get a lot of faces right, is what I tell my students is, hey, there is no information, you don't need information from me on how to do this. There's millions of YouTube videos, there's millions of web pages that tell you how to do it. So you don't need me for that. All I'm here to do is to answer questions or to make it more relevant, right? To make it, that's part of my aspect of what I'm doing is I'm trying to motivate you and in motivating you by giving you the relevancy, I'm helping you to understand, I'm giving you an opportunity to use those skills to actually have satisfaction by seeing that, oh, I'm gaining the skill and I can use it and I can express it and I'm becoming that much more effective and capable for when I move on from the university, right? So I totally believe that. I totally agree that we're there to help clarify, to answer questions, to motivate, to give them an opportunity to use that skill. Again, that relationship aspect I think is right on for sure. And I echo what Jess was saying too about breaking what's happening in higher ed, because who is higher ed designed for? And the systems that we have now work for those people only and if we really want to make higher ed more equitable, we have to break down what we're doing and really make the best use of what faculty are there for. I've been echoing everything else. I've been ready for this for literally years. I've been mystified because I'm not a real academic and I've never sort of been inculcated into the system. I've been mystified why more of higher education is not attuned to student learning for the entirety of my time intersecting with these institutions. It's about credentialing and signaling and all those things that we're aware of that the institutions also do. But I've been ready for it for a long time and that's why I'm excited for the appearance of this kind of technology because it forces us to confront these things and talk about what we value and look at the things that the areas where we're constrained and hopefully start to put resources into those learners who would benefit from more resources rather than the continuing constraints and limits we put on the students who probably need the most help. In terms of accessing that learning. Well said and that's a great moment to end on I think. Just the four of you, Brent, John, Jess and thank you for joining us on this rolling panel and all the other panelists and who have been on and also all the great questions. Thank you so much. This has been an extraordinary session, one that I think has covered a lot of ground and one that we haven't finished. So I'm going to see about we're doing this or resuming this next week or the week after. I will share everybody's comments on my blog post coming up along with the entire recording that'll be up shortly. Let me just quickly wrap this up with the points coming up for our next sessions. If you'd like to continue talking about this right now over the next week, please, you know, head to Twitter where I am. Use the hashtag FTTE or to my blog, BrianAlexander.org. I'm going to add a couple others to this list next time. If you'd like to look into our previous sessions on pedagogy of writing, on plagiarism, honesty, on technology in general, just go to tinywereld.com slash FTF archive. We have sessions coming up. Just go to form.futureeducation.us to see those. And if any of you have projects that you'd like to share, including using ChatGPT, please shoot me a note. I'd be glad to share them with this awesome community. Thank you all for a fantastic session. My mind is spinning. This has been terrific. I love the way that we've brought our collective mind to bear on this fast moving and important subject. If I don't talk to you all for the next few days, have a great end of semester. Have a great holiday. If you're celebrating, have great downtime in general. Above all, be safe and take care. And we'll see you next time online. Thanks, everybody. Bye-bye.