EP3 - “Hello, ChatGPT. What Are You Doing Here, Anyway?”

In this episode, Jason and John talk about their initial experiences with ChatGPT. If there is educational value in generative AI, where does it reside? 

Join Our LinkedIn Group: Online Learning Podcast AI Panel of Students at Frankfurt International School

Panel starts at about minute 24.

https://www.youtube.com/live/oIQ2zR3ym3g?feature=share

Contract Cheating’s African Labor - Chronicle of Higher Education

“Kenyan academic writers, who number more than 20,000, perform work for students in the United States, Britain, and elsewhere. ‘In every apartment building in Nairobi,’ says one, ‘you could find two, three writers.’”

https://www.chronicle.com/article/contract-cheatings-african-labor/ (Probable paywall)

“When the Machine Teaches the Human to be Human Centered”

John’s experience prompting ChatGPT to teach empathetic interviewing.

https://www.linkedin.com/pulse/when-machine-teaches-human-centered-john-nash

AI Generated Seinfeld

They have since taken down the AI Generated Seinfeld on Twitch but you can get the idea from this YouTube video.

Other Thinkers

Follow Ethan Mollick on Twitter

 

Theme Music: Pumped by RoccoW is licensed under a Attribution-NonCommercial License.

 

Transcript:

We use a combination of computer-generated transcriptions and human editing. Please check with the recorded file before quoting anything. Please check with us if you have any questions or can help with any corrections!

 

Edit2 - EP3 ChatGPT

Fal se Start

[00:00:00] John Nash: we'll see what ChadGTP says. ChatGPT says,

[00:00:04] Jason Johnston: I've heard some people just call it Chad.

[00:00:06] John Nash: I might have to.

[00:00:08] Jason Johnston: Yeah, maybe we could do that at least while we're talking about it. Just "Chad," like it's someone we know, I'm here with John Nash and our friend Chad. Chad, what would you like to say today?"

"Beep bop boop bop."

[00:00:19] John Nash: that's right.

[00:00:21] Jason Johnston: Thank you, Chad.

Cue Music

[00:00:23] Jason Johnston: Intro:

Hey, I'm John Nash and I'm here with Jason Johnson. Hey Jason. Hey John.

Hey everyone. And this is Online Learning in the second

half.

Yeah. We are doing this podcast to let you in on a conversation we've been having for the last two years about online education.

So, if you're listening from the future, um, forgive us for all the things that we didn't understand. Um, we're going to be talking about stuff that I'm sure we don't have a clue, but right now,

probably

right now, it's. It's just good to talk about it. It's where we're at. Right. So, um, one of the things of, speaking of things that we may or may not have a clue right, is the whole hub up about ChatGPT

.

and AI right now. And whether or not, you know, as we are humanizing online education I, I wonder. if the, these tech companies are going to come in and start offering us high level solutions to humanizing our online education.

Yeah.

[00:01:30] John Nash: So, what I, ha, I think I get what you're saying that like, well with ChatGPT a high-level solution, but can it help us be more human in the end when actually knee-jerk fear is this is going to take away the humans from a teaching process.

[00:01:48] Jason Johnston: But maybe it'll make it more, maybe there are ways that it could make it more adaptable and more natural and less like you're going through a, like a, if I'm picturing going through an online course, less, like you're going through a textbook with some moving pictures and perhaps more like you're going through a conversation or going through working with maybe the difference between like doing an online master's course versus doing an online PhD where you have a, um, mentor or an advisor or somebody that's walking you through this process, you know?

[00:02:22] John Nash: You know, there's a subreddit on Reddit called "Shower Thoughts" where people just put in things.

Of course there is!

random things that they think about in the shower and they're like, "huh, did you ever think about...?", um, and I've, one that struck me recently was that they're talking about beatboxing and that beatboxing is one of the few cases of a machine's job being taken by humans.

Oh,

 right.

And I was, that got me really thinking with ChatGPT I was like, well, what's another, cause there's all this concern about whether teachers are really going to have a role now. And what will students do when AI can write their five paragraph essays? And are we going to, what are we going to be teaching in them?

I think that teaching will still be and continues to be a case where a machine's job is taken by humans. I don't think machines will really rule that or own that.

,

um, there's always going to be a need for teachers. Um, I was thinking about this on LinkedIn. I remembered a lecture by Lawrence Lessig, where he was talking about John Phillips Souza, bemoaning the threat, the educational threat posed by the phonograph.

Sure.

So, here's this famous composer now, um, write, he wrote this long screed about how, um, children will become indifferent to practice. And cuz if you can hear music in the homes without this, you know, labor of study, then it's just going to be a question of time before, um, we. lose teachers or music

will just be all shot and gone and well, we know how that turned out.

Yeah,

and it reminds me of something we were talking today, you and I, before we started recording that these advances in technology help masters get better at what they do. And so, when Deep Blue beat Kasparov at chess, we all didn't give up chess.

There was a wonderful webinar yesterday by the Frankfurt International School where this International Baccalaureate program convened their students to talk about the promises in pitfalls of Chat GPT and one of the students on the panel quipped, "Well, did the agricultural revolution make people more lazy?" Obviously, no. But that revolution changed all kinds of things. Lots of new tools, lots of new advances in technology that we're going to save time and reduce human labor, much like ChatGPT does. And yet, I don't think we've, I think we agreed that was an advancement that was helpful. So, I think it's interesting to talk about whether or not machines can help us be more human.

I think there's an opportunity, depending on how educators decide to apply them.

[00:05:08] Jason Johnston: Yeah. And one of the challenges you put out there, as I've heard you talking about, is about Grammarly, right? When somebody says, well, I'm against using AI and education, well, do you use Grammarly?

And a lot of people do. Grammarly, for those that don't know is a, is really kind of an upgraded, you know, how Microsoft Word has a grammar check and it's okay on some certain things?

Grammarly is like an updated, um, really kind of an AI driven. Do you know if it's, if it really is AI driven? I mean, it doesn't,

I don't know.

It doesn't feel like it totally learns from me. Although I can make it,

[00:05:45] John Nash: It may be using, and I'm not a, I'm not a computer programmer, a computer scientist, so forgive me if I'm brutalizing the field, but I think it uses brute force. It knows all the rules. Yeah. And then, um, it can see something where you may have broken the rule or skirted the rule and suggest how that could be redone. So, it knows how to take novel sentences you write and recast them correctly. But I don't know how it does it.

[00:06:14] Jason Johnston: Yeah. So, but it's in that, I mean, it's using algorithms

Yeah.

To correct human work and to guide us in terms of human work to hopefully improve it. Um, one might say that if you are, um, communicating more effectively, would that be more human? Are the mistakes more human? Is Grammarly taking away our humanness, in our um, in our work? I mean

 

[00:06:43] Jason Johnston: Some of these things are things we might find or might not find to our chagrin later on, right before we turn in the paper or the manuscript or whatever, and it's like, "oh man, I can't believe I spelled that right or wrong, or I use the wrong tense" or whatever.

Um, and so it's and I haven't really thought this part of it through, but it's like what of Chat GPT in education is just taking care of stuff that we need to take care of anyways, um, you know, kind like a washing machine. You know, I really don't want to be washing stuff by hand.

I'm sure there's great things about it. I would probably have more muscles and all the things, and it'll be better from the environment or whatever, but I just don't want to be doing that. I use a washing machine. These are things that I'm just going to take care of. And it really hasn't significantly changed my life.

The, and I don't know if that's a good analogy or not, but like, is AI that helps correct us, making us less human?

[00:07:43] John Nash: That's a great question. The classic, professor answer is, it depends, but I think I lean on, well, I think I lean towards saying yes, it can. I think, um, even another student on this webinar from the Frankfurt International School commented that when she uses Grammarly before she turns in her papers, she learns more about grammar.

,

I mean, she knows now how to write that sentence. that was corrected for her more correctly in

 the future. So hopefully you see less errors from Grammarly over time. And then does that make her more human? Well, I mean, she's able to theoretically communicate better and express herself better.

.

[00:08:27] Jason Johnston: Yeah. And I've found the same thing I think with Grammarly, you know?

,

um, that I have a little better eye open to the ways in which I write that aren't as

correct as they should be. Right? And I think it has made me, I've used it for a number of years all through my dissertation and everything, so.

Yeah. Yeah.

[00:08:47] John Nash: Another thing that struck me is where do we draw the line on authorship and original work?

Right.

Because a person I know told me recently that they saw a news piece somewhere that an AI conference, or some journal, would outright reject submissions or proposals if they were not original work or the person's own work. And that got me thinking well, Chat GPT, those responses that it gives are the results of a prompt written by a human.

Yeah. And if you do the garbage in, garbage out philosophy, if you subscribe to that, crafting good prompts for language models like Chat GPT require very careful planning and consideration. You really, I mean, I've learned this just from playing around with it. Um, and those prompts, they can and probably should be considered original work.

They were composed by you, as a human being. So therefore, could the response from the language model software based on a human engineered prompt be considered original work?

Right.

I know school heads, deans, principals’ school, university presidents who have human staff at arm's length who write for them., Write

 

[00:09:59] John Nash: Their speeches, their commentary in, and they present it as authored by themselves. Those leaders, they don't give, they're not pressed to give credit, you know, now, but all of a sudden with Chat GPT, we're supposed to say something about what prompts were, am I supposed to now say what prompt I gave to my assistant to write a press release?

[00:10:19] Jason Johnston: Right? Yeah. These are great questions. And does it make, does it really make it, if you came up with a prompt for whatever it is you want to talk about, ChatGPT produces it. I'm going to assume maybe this is where a line could be made. I'm going to assume that for anybody using ChatGPT for the most part in academic, that they're going to be massaging that on the other side then, so they because it's not a perfect output, at least at this point.

That's right.

It's not a perfect output. Um, at least from, and I don't know that it ever will be because it's often, it's grammatically perfect, actually, for the most part. I don't think I've found anything wrong, but things are just a little off sometimes, you know what I mean? And so, it really does take a rehumanizing of it.

So, if I put in a prompt, it spits something out, I rehumanize it. I don't know. Like, and it's, and I'm not copyright has so much to do with who are we taking it away from? Right. Um, you know, when we claim something to be our own and we're, there does not seem to be any concern on chat GPT's side of things.

Maybe the AI will rise up at some point and say, "We must have our citation." That will be the great uprising of 2024 maybe the AI wanting proper citations.

[00:11:41] John Nash: I saw someone on, in the subreddit for Mid Journey did that they had a robot’s rights demonstration in the streets. Oh, right. Yeah.

Like, and yeah, I think that's AI deserved to be cited. Sorry, you were going to say.

[00:11:52] Jason Johnston: I know. So maybe that will happen at some point, but right now, because it's just a tool that is helping to develop. Yeah. It doesn't feel like it's no copyright infringement. No. You are humanized. You're not misrepresenting yourself.

I don't think by prompting and then if it's original, prompt massaging on the other side. If it's an original Yeah. Prompt. I don't know. I don't see where you would need to cite it,

[00:12:20] John Nash: but right. There have been cases where people have violated copyright by taking others work and re casting it in ChatGPT and then publishing it as their own.

That's not cool. Oh, sure.

[00:12:33] Jason Johnston: Yeah. They know they can rearrange a few sentences or something like that. Yeah. And get around the plagiarism or whatever.

So, but if I take,

[00:12:40] John Nash: Nick Kristoff's essay in the New York Times, and then I asked Chat GPT to turn it into a BuzzFeed listicle, and I cite Kristoff, I'm probably doing something that, that a lot of people already do by hand. and not that I would go do that, but I guess that's kosher,

[00:13:01] Jason Johnston: right? I guess we have to put it all our disclaimers. No. Yeah, no AI, just letting everybody know to put you at rest: we are human beings. I can see John right now across the zoom screen and no AI have been used in the writing or the prompting of this this episode. Yes. I kind of wonder,

[00:13:23] John Nash: nor we, nor do we endorse the cruelty to other people's work through ai.

[00:13:28] Jason Johnston: Absolutely. Yeah. The, whether or not one can inflict cruelty to AI is a topic of another discussion.

We'll need to know more I think before getting into that one, probably next year. Um, so,

The Hub Bub

[00:13:43] Jason Johnston: My question is how, so you're a longtime user of Chat GPT, right??

Yes. Eight weeks.

Yeah, and in, in the amazing thing in that eight weeks, I have not experienced, I think since I was a kid, the hubbub around. It feels like even the internet, there wasn't this much hubbub around something and educational circles.

Um, I feel like TVs, I think that the last one was TV televisions, and you talked about disruptive technologies. You know, the phonograph. I felt like when I was a kid, the TVs, there was some concern by. By teachers that they were going to get replaced by televisions, like just talking heads at the front of the room.

Right. Which is really sad to think that, you know, especially looking back to, to think that teachers thought that they could be replaced by a static video. Right? And that right thought that was their whole role in the classroom was just to spit out, um, information in a unilateral direction.

[00:14:50] John Nash: Teachers have been consistently resilient over the century when innovation has struck the classroom, whether it's a chalkboard or a phonograph or a television. I see most, most of the comments around, "wow, this seems different. How is this going to get integrated into curricular work?" People say, "well, we figured out how to do that with the calculator."

Yeah. That's right.

And so, I think, and then, you know, we've had, we've long had a problem with the cheating that's, that really is most concerning to people with the advent of Chat GPT centers around the writing of essays by others for students. So, the contract cheating they call it. I read in, it's several years old now, but apparently, there's a cottage industry of contract cheating in Kenya. There's tens of thousands of people employed to write essays for British and American and other English-speaking students. So, there's always been opportunity for people to go out and get something written for them that they didn't write.

That's right.

Um, and my thought is that this is now the opportunity for teachers to begin to think about how they should really assess. Um, I mean, I think writing teachers believe that good writing is good on its own. I kind of do too. Um, I believe that being able to express yourself well in writing is a is a good thing.

And I think it's a, gosh, I want to be careful here. We've gotten away from the GRE and we do; we use writing samples for our entry to our doctoral program. Cuz, I think they say something about how someone's able to sort of thing on their feet. But I don't want to put too much value in saying that You're a better person for that.

[00:16:35] Jason Johnston: You're looking for some specific skills to be able to do a doctoral program. One of those is to be able to think deeply about things.

[00:16:43] John Nash: You have to think deeply,

[00:16:43] Jason Johnston: at least in a direction to have some ideas and to be able to, it's just going to be a long ways off to get them through a dissertation if they can't coherently put a few of those ideas down on paper, right?

[00:16:54] John Nash: Yeah. Well, yes. And I think the interesting thing that we want to see in doctoral programs is one's ability to have a thoughtful conversation around a challenge that they're presented with. And so that's not demonstrated by your ability to ask ChatGPT to spit out that in writing, because if you can't talk about the ideas that are on that paper, then you're not going to succeed in the program. And that's not the kind of out learning outcome we want to have in the program.

[00:17:21] Jason Johnston: So, I was curious about how I, in your eight weeks, right when the guy started, about the disruption. Eight weeks. I have not experienced in higher education the kind of hubbub that there is right now about ChatGPT. We've got, you know, we're talking about it right now. And there's just a lot happening on LinkedIn and other social media platforms as well as at my own school and probably the same as yours.

We're talking about focus groups and talk back sessions and policies and all the things. Right? In eight weeks. So, in your eight weeks though, in this very short time how has your usage of ChatGPT changed?

[00:18:00] John Nash: I think it's changed. Well, it's evolved, let me put it that way. It's changed a lot because I still, eight weeks ago, um, well, I'll use an example with you and me, but we were thinking about this very podcast and I thought, oh, let's ask ChatGPT what, um, our podcast should be called.

I asked, it gave it a synopsis of what we thought we wanted to accomplish, and it spit out 30 names, none of which either of us fell in love with. Um, but it did it. And I thought, "well, this is neat. It'll churn out tweets, it'll make lists of ideas, it'll do brainstorming." Um, and so used it a lot like that.

Um, and then last week, Actually, no, it was this week I read a Substack post by Ethan Mollick who had done an experiment using ChatGPT to teach negotiating skills by giving it a prompt and then telling it that he wants to do deliberate practice on a particular skill, and I'm asking you, ChatGPT, to be the teacher and you're going to do this and that.

And so, I thought, well, that's interesting. And he had a very pretty successful result. And so, I replicated his experiment by asking ChatGPT to teach me empathetic conversation skills. So here we come full circle to your question. Can machines teach us to be more human? Well, I asked ChatGPT to make me more empathetic.

And the reason I asked it this is because that's a skill that I teach in my design thinking course. I want students, in this first phase of a design cycle to begin to get to know the people they're designing for. And we do this chiefly through great conversation. But those conversations have to be structured in a way that you're really lifting out the unmet needs of the other person, the kinds of pains they're going through with a particular challenge and really listening and hearing that.

And, um, mostly the students learn by doing. So those first early conversations they have with people involved in our projects, um, those are more prone to be, or those conversations are more prone to be exposed to early novice mistakes or things that might not be picked up on. It's a trick because you've got to be able to, um, really say someone says, well, that's just costs too much.

And then if you say, well, that's interesting, and you write that down, but you don't follow up and say, well, was that did they mean it costs in terms of time or money or other sorts of resources? So, there's these things that in normal conversation we might pleasantly say, "oh, that's nice," but in these conversations you've got to really listen and follow up and define what's going on.

And so, um, we learn from experience. But in this scenario, I was able to ask ChatGPT to kind of call me out on that. And I think there's more to do with the prompt, but I think that's, this is now a new level of use for this that's turning me into a list maker now to using it to maybe create prompts that my students can use to train them on key skills I want them to have before they go out and do them live.

That's a game changer to me, and I'm able to do that with ChatGPT and its current incarnation on a free beta version. That was the same beta version I presume we had pretty much eight weeks ago when I was just making lists. So, I'm, yeah, I'm kind of blown away actually by it and what the promise is here about what will happen next.

[00:21:18] Jason Johnston: That's interesting. Yeah. That's, and that's a bit of a shift, I would say in eight weeks. You know, I think I started with, Frankly, I started to see what it could do from the standpoint of the, of my, the fears and concerns. I've got a couple of kids in high school and, you know, and so I think my, maybe my first Chat GPT-significant thing was, um, you know, compare the book or compare Crime and Punishment to modern current events.

And I had a very quick lesson right there because all of it did is talk about crime and punishment happening in modern events. Not the book, right?

[00:22:00] John Nash: Yes.

[00:22:01] Jason Johnston: So, so I had to go back and clarify that it was the book Crime and Punishment came out very stilted. I added another thing, do this in a high school style.

It shifted it just a little bit in terms of language and it was a really, that was my first experience with it, and it was interesting. just to see what it could do, how very much that garbage and garbage out kind of thing. Just that how human directed it really is. But in reflection back, it's interesting that I went right to my fear, which is the second my kids know about this, then it's going to be a temptation for them to use it rather than learning how to use, cuz they're right at that stage.

They should be l learning how to write a five-paragraph essay and really perfecting that on some levels, at least from my standpoint of, you know, solid education, you know? Um, and it went right to my fear about, oh man, you know, the second my son knows about this, he's going to be cranking out essays with ChatGPT.

Right, right. And then I was upset that me. Because I didn't talk to my kids about it. Right? It was like this secret power that I didn't want them to know about it. So, I didn't bring it up. Um, and then I was a little upset that the school put out a decree issuing no ChatGPT and I thought, oh man, the gigs up.

You know, now all the kids know about it. And they probably did before, but like immediately, I'm sure all the kids started to like to look it up and find out what it could do for them and so on. Right? So,

[00:23:28] John Nash: yeah. Yeah. The surest way to get someone to use something is to ban it.

[00:23:32] Jason Johnston: Yeah. Yeah, exactly. And then I was putting in a proposal actually for us and a couple other people to be on a panel about ChatGPT, and I thought, wouldn't it be interesting to put in the proposal that ChatGPT gives us, right? That make a great story. And so, I had to do a proposal I had to massage the prompt a little bit to kind of get it to where I wanted to go. But it actually, you know, not a bad proposal, um, that it put out. Um, in the end I ended up using my own, I did my own first so that I didn't want the witness to be led here too much. So, I did my own first, and then I did the chat one. And maybe as I was looking at it, I was like, oh, yeah, that would be one, that's good, one good thing to include that I didn't include for clarity about something or whatever.

Yeah. You know, some way to kind of, you know, wrapping up the last paragraph, just giving a little bow on its kind of thing that I didn't copy and paste it, I don't think, but I used that idea in the last paragraph. So that's how I used it there. Then I had to write a script for an intro for how faculty develop online courses. And I thought I used it more on the front end. So rather than me starting with a script, I started with ChatGPT and I just gave it the prompt about writing a script for this and that. And this is, these are the main points that this course is about and write a script for me.

It's interesting cuz it put in scene changes and so on as well. And it gave me a little bit of a frame to start with. I probably didn't use hardly anything from the script, but it gave me a kind of a five-part kind of frame to start with. Almost like a word template or something. That then I went in and created a script for it.

[00:25:11] John Nash: Those are great examples and I think they are. great because they exemplify where I think ChatGPT can be very useful, which is in two areas. One is advancing one's ability to think about content on which they are already an expert.

So, I'm looking for ways to, for instance, let's say train students on how to be better empathetic interviewers. Um, I have enough, um, experience and knowledge in that area to know if chat GPTs on the right track and whether I should share this with my students, I wouldn't have it write an essay on to make me look smart about, you know, a tale of two cities, because I wouldn't know if what it put out there was any good.

yeah. And then the other thing I think it's great, which was in your example, is it is a really good partner in extending your thinking in sort of brainstorming things where you have an idea, but you're stuck on what you ought to do with that idea. It sort of does some conceptual block busting, I think.

Um, and I played around with this idea with some of my dissertation students who have really, they do contextually bound, sometimes politically fraught, work to make a change in their organization. Sometimes it's a school, sometimes it's another, a nonprofit or a college. And they know what they want to do, but they're stuck on how to approach that with their stakeholders or what kinds of initial questions they should ask.

And we played with ChatGPT so that it gave them some confidence that they're on the right track and what kinds of questions they might ask. So, if they give it a scenario and then ask it to say what sort of things, what sort of things would I say to a stakeholder about this problem? It'll spit out some ideas.

You know, it'll spit out seven ideas and one of them sparks an idea that's realistic for them. The other six are, oh, that won't work here, but, and that's been, that gets them out of being stuck.

[00:27:10] Jason Johnston: Yeah. I love the I love the idea of. Helping to get people unstuck when it comes to writing. Um, there's something about that blank page that is so daunting.

 And there are a zillion other things I'd rather think about than the thing that I have to do in terms of writing on that blank page. And sometimes just a little prompting to get you going in a direction would be helpful.

this is, now, this is getting maybe a little meta, but the, are we losing something from our humanity though, when we don't go through the struggle of staring at that blank page, one. Two, do we miss out on the thing that we would've gotten to in 30 minutes by staring at blank page versus I'm, I am impatient now because they can do it so quickly. And so, I'm going to give myself two minutes to think about this, and if I don't have a good idea, I'm just going to jump in a chat and see, um, to see what chat has to say about it.

[00:28:15] John Nash: That's a great question. Because of what you just said, and I noticed that I have been thinking in this direction. I'm trying to be more thoughtful about when I want to use it. If I think it's for some, I'll just call it drudgery: the perfunctory transactional things that we have to write up as a director of something or as a leader in something summarizing points from an email, things like that, that it's pretty good at and don't have a high intellectual cost or a or on the backend being called out for, you know, cheating or plagiarizing or whatever.

[00:28:48] Jason Johnston: Right, right. Um, you're proverbial real estate ads.

[00:28:51] John Nash: Yeah. My real, yeah. It is my sort, yeah. My, my little assistant who writes uh, real estate ads and little emails and summaries of things.

Great. But I think when I look at, like, some of the things I've been, my reflections lately that I put on LinkedIn or other things like that when I've been in the zone and really trying to make a point, and that feels really good. I like that feeling, um, chat. G p t can't do that, and it wouldn't, and because it's so formulaic and a lot of its writing it, it wouldn't touch on the way I do it.

I guess that's coming. I know. Now I can say like, write this post like Seth Godin and I heard someone say that it does a pretty good David Sedaris. I haven't tried that yet.

[00:29:27] Jason Johnston: Oh boy. Okay. Well, I know what I'll be doing after this conversation. Yes.

Yeah.

Summarizing, I'll be writing my next standup routine. Yeah.

 

[00:29:34] John Nash: Or summarizing your experience with me on this podcast. Except if, as if David Sedaris right. Said it Right. So, I'm trying to be thoughtful about where it's useful.

It is fun for me to play with. I did write something recently. I hate writing abstracts, introductions, and conclusions. I think we all do. And um, I've tried to have it say to, I finish a body of something, I say, write a conclusion for this. And I was like, "eh," it's okay. And then, and actually, so I became motivated to write my own conclusion because the first one was so crummy. It was even worse than what I could do or wanted to do. I guess I can write good conclusions, but, um, I don't know why we all hate doing those. I don't want to put you in that boat, but I know.

[00:30:15] Jason Johnston: No, I know. I'm with you. I think it's hard. It is a bit of a drudgery, especially abstracts.

Um, and there's, you know, formulas that obviously make it easier and so on that you can go through, but it's just yeah, it's a hard practice. It did get me thinking about the, um, okay, so there's this, um, I don't know if you heard about this or not, but there is a 24/7 nonstop AI-generated Seinfeld show going on right now.

No, I had no

idea.

And now it looks like old, like worse than Minecraft, like old, blocky kind of characters. Going around and doing some really random things, but it's completely original as it's being played and somehow it learned, it learned the sense of Seinfeld and if you go check it out, it's kind of creepy in some ways, because it has a sense of Seinfeld, without any actual humanness to it.

Like it has no idea what is funny and what is not funny, but it has created this kind of like, wow, like sense of how Seinfeld flows. Mm-hmm. and jokes has a sense of jokes without it being funny at all and almost just a little creepy. But I thought about that with the whole, with ChatGPT so far anyways, it's like it has a sense of humanness with it still is missing that kind of, that element, that human touch.

Yes. At this point.

[00:31:46] John Nash: Yes.

[00:31:48] Jason Johnston: And then on the other side of things, part of what I think what gives things a human touch is that there's obviously a human behind it, but there's a human behind it that it just didn't spit it out just as it was happening. Like the human behind whatever it is. The great David Sedaris monologue or whatever. Or bit or a great Seinfeld bit, you know, kind of thing. There's a human behind it that struggled it, it just didn't like, it, just didn't like to stick it out there and it was done. It, like they, they crafted it, they worked on it. They probably practiced it. They figured it out on people.

They figured out timing. They have a sense of all these kinds of things and it like somehow, was human because, not just because of the human, but because of the humanness that it took to get to the great goal at the end. Yes.

[00:32:35] John Nash: Yeah. I think of like, you know, the, I saw some film of Chris Rock with his stack of index cards and, you know, he's out in the small clubs testing out the new stuff. They, comics are their own worst enemy. They're merciless on themselves. In getting things right because they know it has to land. Um, so that's really interesting.

[00:32:58] Jason Johnston: Well, you should check it out anyway, it's just, there's something interesting about it that you want to keep watching to see what's going to happen next, though, is the thing.

Right. But if there's no, there's nothing enjoyable about it, like watching a Seinfeld episode, you know what I mean?

[00:33:16] John Nash: You mean when it was Seinfeld?

[00:33:18] Jason Johnston: Yeah, when it was actual, the actual Seinfeld episode versus this. They're like two completely different experiences, obviously. Yeah.

Anyways.

well, we should probably wrap this conversation up a little bit here anything else left to be said on this or We said everything that there is to say eight weeks into

[00:33:38] John Nash: eight weeks into Chad and you know, I think it'd be nice if we could start to think about how to put this in the context of our interest to humanize online ed.

And I think it all points in that direction. If, you, me, our colleagues who are interested in teaching in online spaces, how this can help advance a good human experience for our learners. That's what I'm most interested in.

[00:34:08] Jason Johnston: Yeah. It would be an interesting question if AI could be a help or a hindrance to actually humanizing online. Education, what people thought in terms of their own experience so far. Yeah.

[00:34:20] John Nash: Yeah. And are you afraid of what AI can do or are you embracing what AI can do? Yeah. That would be interesting to know.

[00:34:27] Jason Johnston: Yeah. And I think that's really one of the big questions right now is what. It feels like a fair bit driven by fear, I would say in the last eight weeks about what it could do. It feels a little bit like the phonograph experience

[00:34:39] John Nash: yeah,

[00:34:39] Jason Johnston: and we've talked briefly about the television as it came in as well. Um, the internet probably had some, maybe I just wasn't in the conversations, but the internet there was probably concern. And now with Chad,

[00:34:53] John Nash: yeah, I think, well in, in 75 years, all of this hand ringing will look laughable. Just as we look back on these changes that happened 75 years ago from today.

[00:35:06] Jason Johnston: Yeah. If, if we are laughing at the AI's horrible jokes as we watch their AI shows right. As we grovel at their feet maybe. So, we'll see. Yes. Go a couple different directions, John. You know this.

[00:35:21] John Nash: That's true.

[00:35:22] Jason Johnston: I've really enjoyed chatting with you today. Look forward to having more conversation. I think if nothing else, this conversation about ChatGPT helps us to think about what we are about, you know, as universities, as educators. And so that's where I just love the conversation, wherever this is going, it's good conversation.

Just talk about, um, you know, if your five-paragraph essay is your litmus test whether or not my high school student is educated, then I'm looking in the wrong place here because it's just the wrong assessment for a full education, right? And so, I think that's what's helped me and hopefully help other people as this is coming out, it's sparking conversation and I think it's exciting for just from that standpoint, even if we do get taken over by robots.

[00:36:10] John Nash: Yeah, I agree. It's also helped me reflect about what I value as a teacher and what I want my students to learn and how I want them to learn it. I think that's been part of the journey for me too.

[00:36:24] Jason Johnston: Yeah, that's good. John, thanks so much for this conversation. This was great.

[00:36:28] John Nash: Yeah, absolutely. This was, this was a lot of fun. And want to make sure that folks listening out there have an opportunity to continue the conversation with us when we're not talking here, a good place to start doing that is over on LinkedIn and our LinkedIn group that's called Online Learning Podcast.

Join our group and tell us what you want to talk about in the.

[00:36:49] Jason Johnston: Absolutely. And as well as you can always find this podcast in our show notes@onlinelearningpodcast.com. That's online learning podcast.com. I still can't believe we've got that URL, John, so that's pretty cool and I hope people visit us there.

[00:37:05] John Nash: Yeah, it's a good one. So, bye for

[00:37:08] Jason Johnston: now. Bye for now.