Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
We Get Requests
225 Plays4 months ago
Title reference: https://youtu.be/C51CVmjkAyo
In this episode of "Online Education Across the Atlantic," Morgan, Neil, and Phil address listener questions for the first time and discuss Coursera's recent introduction of AI-based tools for plagiarism detection, proctoring, and grading in their pilot program in India. The team also explores Queen Mary University's move to end several online courses after their OPM partnership with CEG Digital, while still offering other online options. Later, the hosts tackle the future of OPMs in the US, the challenges and potential of AI in education, the necessity of innovation in online learning, and whether tech giants or edtech companies will drive AI advancements in education. Join us for an insightful and engaging episode!
00:00 Online education resumes with methodical reader engagement.
04:51 Indian higher education facing challenges in integrity.
08:20 Generative AI enables new approach to plagiarism.
11:13 Presentation on consolidated educational tech with AI.
13:46 Enthusiastic about personalized homework, data democratization concerns.
18:11 Queen Mary University ends online course partnership.
20:24 Investors learned the dangers of excessive enthusiasm.
23:50 Innovative university investment and Dean's gone wild.
30:01 Online programs need niche strategy to succeed.
32:26 Doubtful about AI progress in education products.
34:57 Productivity distinction in certain areas has impact.
39:38 Edtech vendors should be better equipped.
41:17 ASU's partnership with OpenAI and potential impact.
Recommended
Transcript
Welcome Back and Reader Interaction
00:00:07
Speaker
Hello and welcome to Online Education Across the Atlantic. After a two-week break, that's what we call vacation in America. Nothing more than that. But it's ah great to be here with my colleagues, Neil Mosley and Glenda Morgan again. Got Morgan in a Las Vegas hotel about to go to the Infocon conference, but ah a trooper to be online again, working from the hotel. Today, what we're going to be doing is actually, we want to start being more methodical about addressing reader questions, because we've had them come in in various formats, occasionally deal with them, occasionally not, or occasionally just respond in email. But what we're going to try to do with the podcast is solicit more direct emails to email at oeaa.fm.
00:00:56
Speaker
And we're starting to collect these, and we're going to address some of our initial questions coming through that way. But before we get to the reader request, I guess we should start out with some of
Coursera's AI Initiatives
00:01:08
Speaker
the news. And at least two of the items are AI-based, if you will. No no big surprise there. This is the second year of AI being dominating the news cycle quite often. But yesterday it came out and Coursera is introducing some AI based plagiarism detection tools and proctoring tools and grading tools. And that was a tough one to figure out partially because it came out as a very weird announcement. Somebody broke the embargo.
00:01:41
Speaker
CXOToday.com broke the embargoes that they released early. And the news was all around doing this in India, I guess as a pilot program. And then when I saw that, and then I had an interview with Inside Higher Ed set up, so I was doing some research and you start looking and you could not find any news anywhere else. And in fact, the blog post that Coursera put out was buried deep and very difficult to find. There was nothing on their news section. They didn't share it from their ah Twitter handle. It was almost like a hidden news announcement of releasing these tools, and it was a suite of tools.
00:02:22
Speaker
And one other thing that I had noted, although this didn't make it an inside higher ed, is what's interesting is it almost appears like this is using generative AI to cut out former partners. So in 2021, there was a partnership and integration point would be between Turnitin and Coursera, where they would use the plagiarism detection from the originality score from turn it in and display it and use it within Coursera. And they, you know, that was just three years ago. Now with generative AI Coursera is saying, we're doing that ourselves. Now that's not how they emphasize the news, but that's what I read into it.
00:03:01
Speaker
Likewise, they had an integration partnership point with Examity and ProctorU for Proctoring. Now Coursera is saying, we're doing our own Proctoring using generative AI. So it's a weird bit of news, in my opinion. um What was your reaction as you two saw the news about the Coursera AI? yeah I was trying to figure it out, you know, as a way, I mean, on on on one level, AI has been used for proctoring for a long time. So I wondered if there was something else going on, because in many ways, you know, ProctorU bought a bunch of companies, including Examity. and And at one point, they had said that they were going to pivot to fully in person proctoring. And that certainly seems to be the model that is mostly used, for example, at Western Governors University. And that would be horrendously expensive.
00:03:52
Speaker
for Coursera kind of a model. and and And it doesn't seem to function really well if you hang out in some of the Western governors forums, as I do sometimes at the risk of appearing creepy, that that that is number one as a complaint. ah So, you know, it it made sense to go to a more AI kind of model, which is this this this human and in the loop, but it is weird that it's so buried.
Strategic Implications for Coursera in India
00:04:19
Speaker
Yeah. yeah nothing And using a country as your pilot, you know, we're going to release this in India and see what happens and try to keep it very quiet and then maybe come out with more. And immediately what my mind went to also, though, was, um you know, the Indian government has been quite strict in how it's recognized online things and has been cutting down on those things, particularly at the degree level. So that was immediately where my mind
00:04:45
Speaker
When was its party partly to get acceptance in a more official Indian kind of sense? I think the Indian dimension is an interesting one. I don't i don't claim to be an expert on Indian higher education, but I know I've had conversations with people who are, and you know they have said to me in the past that you know there has been an expansion of online learning in the country. but that integrity piece has been a particular kind of challenge. So, you know, ah maybe if I was more acquainted with the kind of system over there, that could be one of the drivers behind it. I can't say definitively, but i certainly I've certainly heard that like kind of combination of expansion, but problems around the kind of perceived integrity of student work and things like that. So that might be one aspect of it. I mean,
00:05:34
Speaker
I was just interested in the kind of sheer array of features that they've announced on their blog that was really kind of quite the splurge yeah of stuff there and it's kind of the sort of things that I to a certain extent I associate more with a an LMS or a VLE or a different type of player in the space, not to say that you know obviously they't they have a product and so they have to have features like this, but um it kind of read more like something from from that realm than Coursera Realm.
00:06:10
Speaker
And yeah, it just struck me that there were so many things. there I mean, it's interesting that things like the AI-based Viva is an interesting approach to tackling you you know some of the issues that AI presents where you know they but I think they they question students, don't they? They generate questions on you know their approaches and that feeds into instructors. I think that's an interesting development and idea You know, but there's things like time and attempt limits, which don't feel like they're particularly new and locked down stuff and plagiarism detection. So yeah, strange, strange for the sheer array. And the fact that some of the stuff is interesting and some of the stuff seems more run in the mail, I think to me.
00:07:00
Speaker
Well, the proctoring comment ah that you mentioned, Morgan, or I hadn't thought about that, it that almost might be it's replacing a partnership because that partnership wasn't working out, was infeasible, as opposed to the turn it in feels a little bit different. Like it's directly getting into plagiarism detection. Why are you replacing that? I'm not saying everybody should use turn it in, but you have to ask the question, what's the purpose of getting into this space within there? And it just wasn't clear. one thing sort of covering turn it in on some level for a number of years was the steady march of cost. You know, the prices kept going up and up and up and up. And you know, there are there are tests out there are done that actually just pre chat GPT using Google was just as efficient as using turn it in. um Or you know, most of the other copyright kinds of things. So especially now i I wonder whether there is a way to just
00:08:00
Speaker
short circuit some of that, you know, you don't get access to a lot of the the journal articles, but in many ways, probably that's not what people are using. So I think just in terms of scale, the cost was getting unsustainable. And you know, now there are more and more other kinds of solutions probably that you can that you can get out there. Yeah, it does ah generative AI does enable things. And I would note the Coursera is taking a different approach to plagiarism detection. And it's a little bit, I wouldn't call it formative necessarily, but it's a little bit more of trying to help the student along. And that's you know what questions to ask the student as they go through the process. So it's interesting.
00:08:44
Speaker
And I would say generative AI does enable you to take a different approach. But the other question I would have is, who is the audience for this announcement? Now, they didn't make it a big announcement. But if it works in India, they're I would assume they're going to expand this. And so on one hand, it almost feels like investors were the focus of this news release. And that's part of the reason why they gave a plethora of uses. you know We're doing this, this, this, and this. And I wonder if it's somewhat to overwhelm investors and say, hey, we're with it. We're, we've got AI just like you're demanding. We're ahead of the curve. yeah It doesn't feel, if that's true, it doesn't feel very strategic, but I do question who the audience is. I guess it's interesting because I feel like they have really, you know, not necessarily put all their eggs into the AI basket, but they've certainly, you know, they certainly really kind of pushing that forward. You know, when we've talked in the past around,
00:09:41
Speaker
you know, LMS vendors and the extent to which they're investing in AI. And so I wonder if it's it's something that's congruent with their strategy of really you know going deeply into AI and maybe they see an opportunity to go a bit further than others have been prepared to to do and and an opportunity in that, possibly possibly. So in other words, this ties to their strategy of emphasizing AI content within their courses and the AI programs that they're trying to support.
00:10:14
Speaker
Yeah, but even wider than that, that they're, you know, that they're sort of practicing what they what they preach around, you know, AIs. primacy I suppose and they're kind of reflecting that in terms of some of the kind of learning based features within there within within their platform and maybe there's a sense in which they're going a bit further than some of the other platforms and vendors are doing things slightly differently and they see an opportunity in that, hard to say but that's just one thought that came into my mind.
Power School's AI Strategy
00:10:47
Speaker
I'd read it as the primacy of India, like it's such a big market, you know, um true and and the need to sort of crack that and and make that work in a way. um But maybe I've been ah reading into it because in some ways, some of the other vendors that I think we have some news items are are also sort of heavily influenced and and talking a lot about India. Did you have this thought while watching a cricket match that involved India? Maybe. That might be it as well. Well, speaking of what at least I consider more of a strategic usage, Morgan and I got to get a presentation and quasi demo from power school, the K-12 provider, you know, that center, they have much more of a consolidated approach than most of the ed tech vendors we deal with because the K-12 space allows that student information system, CRM, LMS through Schoology.
00:11:41
Speaker
a number of systems and then they've developed this data lake to combine data from multiple sources and are putting AI on top of that and they call it Power Buddy. And as we're watching it, I mean, of course, I had to have the reaction and tweak Morgan because they start out with the sort of chatbot tutor type of example that she is so fond of. But I have to say, it was an maybe I give them too much credit, particularly because Shivani is such a good presenter and so knowledgeable that if you push back, she understands your question. But they seem to have a much better thought out AI strategy. It's based on their strengths.
00:12:22
Speaker
which is combining multiple systems into one consolidated approach, then they're putting AI on top of all of that consolidated data. And part of what was surprising to me is they're at least claiming to be system agnostic. So this can still work in the data lake, even if you use Canvas or Moodle or you know some other LMS instead of Schoology. It can even work if you use a different student information system. And they're really pushing this as a consolidated approach. and And it gives you some good advice things, like some of the examples they're showing is proactively prompting a student, hey, we notice you have a 3.7 grade point average. you could be ah You could qualify for this scholarship, or you might be able to get into this school. Do you want to check it out? And it just had some really tangible use cases.
00:13:20
Speaker
that take advantage of power school strengths, but also speak to what could be useful to K-12 students and parents. So my, as I watched that, I had a very different reaction. I get what you're trying to do. There's real value there and it's building on your strength. I had a very different reaction than I did in reading the Coursera news. Morgan, what was your reaction while watching this? i I also really think that that they have thought it through, and I think Shivani is awesome. So you know a lot of my enthusiasm comes from her. They do lose me a bit, not so much in the demo we saw, but in other things I've been watching. um you know they're They're sort of pushing the idea of personalized homework, particularly. And I want to know what that looks like you know in terms of some of the stuff behind that.
00:14:13
Speaker
you know, I love what they're doing with data, ah in terms of being able to interrogate the data and say, what is it telling me and and really democratizing that as you sort of pointed out, when they start getting into the personalized homework, I start having questions about what that actually looks like and, and how that is. I also wonder about if they in using generative AI have stumbled into the early alert problem, and how much they have red teamed about that. You know, because in, in higher ed, particularly, you know, people are gung ho will wear gung ho about, about early alerts, but it turns out that they can have some perverse effects. For example, you know, Hey, you're, you're in danger of, of not passing this course or something, at which point people that, that are less well-prepared say, yeah, i'm I'm not cut out for this. I'm just going to give up now.
00:15:06
Speaker
and and And that sort of thing. So you have to be really careful about some of those early alerts and and their use in higher ed was very hyped and not always supported by data um in terms of actual success. So I worry a bit about that as well. i'm I'm looking forward to learning more about those parts, but certainly I think, you know, there does seem to be a more thought out strategy and ah how these things are linked together and they certainly have the data chops to do it. I don't know if you've, Neil, since I doubt you've been looking at Power School, but have you seen any vendor-based solutions that strike you in a similar way in terms of really strategically thought out? Like, I get where you're going, and you're building off of strengths?
00:15:52
Speaker
Not really, I mean I actually think in a weird way there's even in the short period there's a kind of a degree of orthodoxy that's been built up around the way in which like features are addressed with AI and it's kind of rubrics and quizzes and a virtual assistant. Construct the syllabus. Yeah, yeah, and so I I i mean the Corsair of Viber thing was interesting because I haven't seen an Application where it's more focused at asking questions. So that was a point of difference for me Which was interesting that that doesn't speak to their wider strategy But just as a feature is interesting and I think you know I think your point around the early alert Morgan I think is a really interesting thing about this because often
00:16:40
Speaker
there is a kind of particular realm of learning science or something like that that kind of gets jumped on. And there is a sense in which when features are applied in a kind of universal manner, they may work for some but not for others. And I think that's one of the challenges around AI and personalization because you know if you know a student say you might know how they might handle an earlier warning system you know it might actually crush their self-efficacy and their confidence you know it people can take things and universalize them and they lose a bit of the context and in in the kind of context of discussing AI and personalization that is
00:17:27
Speaker
that is an aspect of it, knowing knowing the individuals a little bit too. so um Short answer is not not really. um and i think yeah I still think there's a it's bit of orthodoxy around all of the feature stuff that people are developing. I think Neil and I need to start a break off podcast or newsletter called Dumb Uses of Learning Science.
00:17:51
Speaker
Put it right in the title, what you think. Yeah, with with with merch like posters, you know, captions on books. Yeah. Yeah. Tell me about your learning style.
The Evolution of OPMs in Education
00:18:03
Speaker
Well, let's get into one other quick news item ah that, Neil, you were going to share from the UK and then we'll get into the reader questions. Yeah, it was an interesting article that was published this week which basically talked about the fact that Queen Mary University over here, that article was kind of headlined as that that that they were stopping a bunch of online courses. You know, really the crux of this matter is that their OPM partnership with CEG Digital has kind of come to an end. But it was a strange piece because it it was very much framed as in, it was a piece that was written
00:18:37
Speaker
to report this news without necessarily understanding what had happened. That was my reading of it. and I think this has more to do with the viability of that partnership for the company, then it was about a university pulling away from offering online courses. And actually this university offers a whole bunch of online courses that weren't in partnership with an OPM. So it served as useful confirm confirmation of actually something I knew already, ah but it was weird in that it didn't seem to understand
00:19:10
Speaker
what had actually happened and so you had kind of students saying I don't know what's going on and yeah so yeah that was interesting interesting news and and not the sort of news that often comes to the surface because usually when these happen these things happen they they're not really reported and they kind of fade into the the distance but yeah that was one piece of news this week that kind of popped up. Well let's use that as a segue and to reader questions and Morgan had one reader question that directly touches on this in terms of the OPM market. And ah for this one, as we move along, by the way, for readers, we'll start being explicit about whether you want to us to reference your name within the questions. Today, we're going to just treat everything as anonymous. But Morgan, if you can share that question. Sure. So broadly, the question was asking about what comes next after OPMs in the US, you know because they've been under such stress in terms of regulations and and things like that. So what comes next? But more specifically,
00:20:09
Speaker
I think the interesting question is what have we learned from watching the rise and fall of the whole OPM regime as we build the next chapter? I think that's a really pretty interesting question. So what have we learned? Well, one thing I would argue that we learned is that you don't want to, it's not quite Icarus, but you know, I think there's a lot of risk of flying too close to the sun. You go look back at the euphoric days of 2015 through 17, at least in the US, I think it's later in the UK.
00:20:44
Speaker
where you had too much enthusiasm and two both from a this changes everything point of view but also from a what's the value of a company. So like the company valuations back then when 2U at one point was worth over five billion dollars, I forget the exact amount, and everybody was growing and And by the way, this is I've said this before, this was the origin of the Mad Max diagram. I was getting tired of so many news articles making it sound like OPM market was just a run for the riches and everybody's doing great. And I think that success in investment markets and in the mind share that it created with universities on what you can do with an OPM is part of the reason they're suffering so much right now.
00:21:34
Speaker
I think investors are saying, okay, we bought into that. It's not happening and things are getting worse. We're going to give you almost no value now. And I think a lot of schools are questioning it and it just provided red meat for all of the activists who want to shoot down the OPM market. So I'll throw that in as one lesson about the rise and fall is don't, don't let yourself get too big because it can come back to bite you. And I think we're seeing that right now. Yeah, I think all of this debate is such a muddle because there's a whole bunch of other factors that lay into it. i I don't think in the UK we are remotely where you are in terms of the type of debate that's being had about OPMs, but I think you know that to kind of pick up on what Phil said, you know um I'm reading a book at the moment called The Company of One and and it talks about this idea of just
00:22:28
Speaker
some of the pitfalls of just continually growing or having the perception of a need to kind of continually grow and go forward and I think you know OPMs have kind of been hampered by that but that's happened in a climate where you know as we said before there's kind of been three money and so you know I think there's a degree of yeah hoop hubris and arrogance and greed that have influenced you know the OPM world, just as it has influenced you know many other kind of worlds in terms of companies. I personally think in the UK there's still an appetite for someone external to come in and help people do online well. But yeah, I guess the kind of lesson really is
00:23:14
Speaker
for for OPMs is kind of how they can do that sustainably and profitably and not get, you know, I think in online education in general there's still a sense in which is the coming thing even if it's not within your national geography but kind of in the world and it's easy to just kind of get wrapped up into this kind of hubris and hype that the fields are kind of ripe to be harvest with kind of students and there's lots of kind of money out there. So I think Yeah, it probably should teach us that there needs to be less hubris and a little bit more focus on profitability. As you were talking, I was thinking there's a line in in Pride and Prejudice where where Mr. Bennett says to Elizabeth about Darcy, he can't look at anybody without seeing a blemish. And I think he probably was describing me. I always seem to see that the worst in everything. But the the thing that immediately sprung to my mind was one of the things that that
00:24:11
Speaker
this whole OPM chapter has taught us is how willing, how unwilling universities are to invest in innovation, you know, because 80% plus of them were going into revenue share models in part because they weren't willing to to upfront the money to actually invest in building out online infrastructure, you know, that would would support the university Also, I think it it points to the fact that universities actually have to figure out their online policy because one of the things that OPM
00:24:44
Speaker
OPMs enabled was the sort of Dean's Gone Wild scenario where each each college ran off and and did their own thing. and and And I curse Phil and Michael for coming up with ah the Dean's Gone Wild thing, because whenever I say it or hear it, I immediately imagine lots of college deans on Fort Lauderdale Beach in Speedos. And then I need to bleach my mind, which To be fair, I didn't come up with that actually came ah that actually came from Steve flowering from key path, who was the founder of infinite compass. He just at the time, he said he did not want to be credited with it. So we asked me not to use his name. So I don't know if that's ah
00:25:25
Speaker
That's a good excuse or not, but that's why you have that visual in your mind. It's according to Steve. and So, you know, they they they weren't willing to sort of do the upfront work to actually make it happen. So I wonder what we're going to see going forward in terms of places, you know, making that investment and, you know, and actually having to figure out, okay, we're going to do this centrally and we're not going to have a system where most of the money flows to the edges because it creates these perverse incentives. Yeah, I think that's a really interesting point. I think it's like the the the willingness for universities to make the hard change is part of it as well. you know There is an element of which you know I know kind of speaking to
00:26:05
Speaker
universities over here, some of the challenges that they face to kind of create the conditions to offer online learning well involve like serious change in investment within their institutions and you know the conditions and the appetite to deal with those things to put themselves on a better longer term footing I think is you know I can forgive them a little bit if they're presented with a and an option that circumvents that in a way but I agree with your the essence of your point, Morgan, around the the need for them to do that. In the US, too much of the conversation when you get to this, it gets down to post-OPM. We should do it internal OPM. We can do it ourselves. And that does get to the heart of we need to invest in our own capabilities. At least ah at least it throws it a bone. But what it doesn't get to is, well, there's a reason
00:27:01
Speaker
that so many schools have gone for rev share OPM deals and taking the financial risk off of them. And part of that is, you know, the rule of thumb is most online programs, at least for OPMs, don't hit break even for at least four years. And that's the reason they have long term contracts is the vendor is paying for the program up front. So if you go back to the, we need to invest in our own capabilities, what we don't talk about enough is you need to have a realistic timeframe yes for payoff as, okay, maybe it does make sense to invest in your own capabilities, but you you can't also keep to the same aggressive timeline expectations from what you did.
00:27:47
Speaker
So I look at the project Kitty Hawk. That was an internal OPM for the University of North Carolina system. We're going to invest $97 million and within three years we're going to stand up 120 programs and we're going to do it ourselves. So high marks for investment and of $97 million dollars internally, but they still had ridiculous expectations of the payoff. And And give me a good example where an internal investment did have a payoff within five years or even 10 years. We don't talk about that aspect of it.
00:28:25
Speaker
Yeah, and also on that, the the internal investment, although for some is presented as, you know, the best option or the wisest option or whatever, you know, there there are plenty of examples of fairly disastrous internal or, you know, not very successful internal investments. So, you know, it's it's not just what you do, it's how you do it. Yeah, you got to know what you're doing. Yeah, something i I often share with people is, you know, I've got a chart that shows the growth at at UCF over time in their online program. And it is a very long chart and even ASU, similar chart, you know, and they were using an OPM. And, but you know, it still took some considerable time to get to where they're going. And that's even with an OPM, let alone without one. So places, you know,
00:29:13
Speaker
I mentioned in ah in a post recently that I'm on a one-woman task to educate senior leadership about you know the the costs and the time involved in building online programs because they seem to be dubiously believing that it's short and they'll they'll earn billions and billions from it. um would I'll add one more lesson, then we'll get to another question that we have. One is the fact that the environment's changed. It wasn't just the Dean's gone wild, now that's changed. It was also to use another movie metaphor or video metaphor. it was ah If you build it, they were come back during the 2010s. There wasn't a lot of competition.
00:29:54
Speaker
particularly for name-brand schools who had online programs. So there was this mentality that was partially
AI's Role in Future Online Programs
00:30:00
Speaker
backed up by reality. If you build it, you're going to get a lot of people coming to it. Well, in 2024, we're saturated almost with online programs, certainly in the US. So it's a highly competitive environment. which means that typically you need a niche strategy to design a program um and you can only expect it to get so big without a massive investment. So the it's not like the underlying environment has stayed static as the OPM market has risen and fallen. I think that people need to understand part of the falling is the competition, academic program to academic program. And that changes a lot of assumptions.
00:30:43
Speaker
Alright, so let's move on. This gets the award for the longest question that we've seen over since we've been doing the podcast. So you guys can have some coffee while I read it. and First off, thank you for your podcast. It's so refreshing to hear each week thoughtful commentary that comes from a highly informed and independent perspective. Too many podcasts push their own barrow. Either they're self-promotional if they're on the tech side of the equation, or they're everything's fine, move along, nothing to see here if they're from the ed or institutional side of the equation. Obviously not a question with that first paragraph of his, but I really enjoyed that.
00:31:22
Speaker
So going back to the email, my question, and it's really a topic for conversation, is an extension of Phil's recent comment that it won't be Microsoft, Google, or the other tech giants that are going to solve the challenge of the educational application of AI. This task is going to fall to ed techs that are steeped in pedagogical and learning processes and that engage closely with educators to solve the challenge of finding ways to solve particular challenges. And then he added in one other aspect of making a distinction between AI as a productivity tool and AI as an educational tool and the difference between those two. So let's just ah start out with sort of that setup
00:32:08
Speaker
Do you agree with sort of the premise, which he ah he seems to agree with, that it's not going to be the tech giants who really are the ones to figure out how to use AI directly in an educational and not just a productivity but an educational context? So what are your thoughts on that? I would be reluctant to be so confident about that because you could you You can question the extent to which, you know, whether it's tech companies or at tech companies, the extent to which they actually have landed on a solution that makes things better always. Like, ah maybe I'm reading too much into the question, but the sense in which we'll get to this great destination in terms of AI and tech.
00:32:52
Speaker
I'm not not so sure, and I would like to leave scope open for companies that fall outside of those to actually maybe potentially come up with something. Because you know as I said before, I think there is a sort of degree of orthodoxy around the ways in which AI is conceived of in terms of product for ed tech at the moment and where people are focusing on. And it actually might it might lend itself to someone outside of that bubble. to to identify something or to stimulate an area that that might be might be of use. um So I don't feel confident across the board that we're necessarily going to see AI
00:33:38
Speaker
being developed within products around education that that you know is going to kind of move things massively forward. But I'd like to leave it open that that there may be other players outside of ed tech companies that that might actually provide something um of of value in that space. So I'm disagreeing with you, I think. That's refreshing, isn't it? I like that. I want to question the premise of the of the sharp divide between productivity applications and learning applications. I think there's a ah way that they flow into one another, or you know they can lead to one another. So they're not as unrelated as as one might seem. But this is sort of not answering the core of the question. but i
00:34:26
Speaker
A lot of people, at least in K-12 are upset this week because mike Microsoft is doing away with Flip, which used to be Flipgrid, and they're rolling that into Teams. So it it seems like nothing that Microsoft will do will we benefit learning because it's it's it's all Teams all the time, maybe. Although in defense of of ah before Phil goes after Teams as he likes to do, I just want to say that He's just so critical of teams because he never had to use WebEx as I did for eight and a half years. It's all relative. Yes. I think the productivity distinction is, does play out in some areas. I think I talked about this in relation to kind of like video conferencing where you can have recaps and you know, that's a, that's a kind of feature we can kind of recap or.
00:35:14
Speaker
ah Even summarizing documents, there was an article I read recently around, I think it was titled, you know, we're not talking about AI's impact on reading. And so you could, you could say maybe the productivity features for more of a corporate environment are like, we'll, we'll, we'll provide a recap of this. video conferencing meeting or provide a summary of this document and in a business context the efficiencies that you might gain from that are appealing but the efficiencies that you might gain from that as a learner might actually be detrimental because you're kind of investing less effort in something that would you know
00:35:51
Speaker
help you in some way, or will help you retain it or help you digest it or help you kind of ruminate on something. So ah like I, I agree with you that there's a danger that you can overstate that distinction. But I think that distinction does play out in certain, in certain ways. And I think there's just a couple of ones that come to mind for me. Can I interrupt with a funny story that I heard yesterday? So you know, the whole story about the the bad summaries that have been coming out of the new Google AI enabled search and the glue on the on the
Humor and Pedagogy in AI Development
00:36:24
Speaker
pizza. So Google essentially Google bombed itself, because now if you use regular Google, apparently to search about about pizza, keeping cheese on the pizza, it comes up with that reference to, to read it. And I won't mention the name of the poster, because it's a family show. That is the number one search because so many people have searched it. Oh, they want to go see that. Yes.
00:36:46
Speaker
I have missed this glue on pizza, but I didn't think that there would be anything more offensive than pineapple on pizza, but I was there. All right. That'll be another episode topic. Yeah. Well, I have to say I disagree with myself somewhat as well, or at least I disagree with myself as presented. and First of all, it's a great question and a great topic of conversation and one that I think is going to be relevant for the next several years. My point was more about the tech giant approaches and that it became quite clear, particularly even if you read that Google paper that they put out around AI, is that it was very much a tech solution and then they're trying to find education use cases and then they're trying to figure out how to fit with an education.
00:37:36
Speaker
But it's not my point wasn't, hey, leave it to the edtech vendors. They're the ones who know this. My point was, you've got to have people involved who understand education, involved in the beginning, so that they understand what might work, how to avoid a rabbit hole. And so it's you need knowledgeable people in education. And yes, a lot of them work at the EdTech vendors, but that's not my point. I'm not saying the EdTech vendors will do it. I'm saying you've got to have those people involved earlier on. And one example is what we already talked about.
00:38:11
Speaker
the problem with early alerts, somebody with enough knowledge to say, hey, this makes sense on paper, but you'll end up having this perverse effect. So if you want to avoid that, learn from past implementations and deploy it the right way. And Google doesn't, at least right now, they don't have those people in open AI. They don't have those people. And It's good that they're focusing on education, but they're doing it almost as an afterthought. Now that we've created something, we're going to help you implement this. And there are going to be some aspects that really take a lot of knowledge.
00:38:48
Speaker
Now, Neil has mentioned the downside of this. If you get people who are steeped in education and history and ed tech vendors, then there's a tendency to say yeah the orthodoxy question. There is a tendency for those people to say, oh, we're going to do it this way because that's what we know and we're all thinking alike and we have such a history. And they're not thinking outside the box. They're not thinking of how Generative AI might help you rethink a problem instead of just doing the same thing better So as presented and as I probably as I originally said it I need to go look at the transcript
00:39:27
Speaker
I'm not saying EdTech vendors are the solution. What I am saying is we need more pedagogically aware solutions before we should expect dramatic outcomes to come from this. Yeah, and I think the way I think about is it is you'd like to think it would be the EdTech vendors, but there's but plenty of legacies of them not.
EdTech Innovation and Incremental Change
00:39:49
Speaker
taking the approaches that you're advising around being well informed about the way that they develop products. I guess the other complicating factor is there's and there's a kind of a sort of a assumption that there is a consensus in education about the way that we do things and teach. And so even if you did adopt an approach where you're consulting with educators or people on the ground, if we just say it like that,
00:40:17
Speaker
It depends who you're engaging with because there's contentions of different approaches and in education. So that complicates the picture even further. But I think the way I would describe it is that you'd like to think that EdTech vendors would be better armed to develop solutions than the big tech companies. And i think I think part of the problem, both with EdTech vendors and with the big tech vendors in general, is that you know, to borrow an analogy from James Lang's small teaching, they all played big ball, you know, they're all trying to find the the home run, instead of trying to think about small ball, you know, little buns and sacrifice flies, or your little ways of of advancing in in small ways. And I think they need to think about AI in that in that sense, because they're all trying to like,
00:41:06
Speaker
use AI to to to help you do your syllabus or something like that instead of some very small thing that really is impactful, but is perhaps something that they're not thinking about. Yeah. What we're not talking about in this conversation are schools, people who know education, but from the ah college or university side. So it makes me also wonder the ASU, Arizona State University partnership, ah with OpenAI and where they're essentially enabling campus-wide programs, pilots, to try out generative AI, but it's really academic-led pilots. I wonder if that's going to help lead to any kind of transformational thinking and development that might then spread to the EdTech community.
00:41:58
Speaker
ASU is a university of one and you could almost question, yeah, but Phil, are they really a university or a vendor at this point? But there's the big question about the role of schools. And right now, Southern New Hampshire, where Paul LeBlanc is changing his role in leading an effort to rethink AI, Arizona State, it's the same players ah quite often, but there are a handful that are saying we're going to think deep and long about AI and education. So it makes me wonder, is that going to be the source as opposed to EdTech vendors that really come up with much more beneficial AI usage that might be not following the orthodoxy, but also something that can really be useful with educational outcomes?
Universities and AI Leadership
00:42:47
Speaker
Yeah, and the real onus there is on the universities to be proactive in that domain. I remember hearing Paul Lebronc speak and he was talking about ways in which they've created, you know, games using AI for SEO. um And I know that's not exactly what we're talking about here, but you know there's that's an example of a university being proactive. around the ways in which they might use AI, you know, maybe to Morgan's point and kind of slightly smaller or less big ball kind of ways to improve what they're doing. But I guess it depends what their disposition is, whether they they're waiting for someone to serve them a good AI solution, or whether they're proactively looking at ways in which they can use some of those tools to to to do what they you know, to make improvements. So I'm in I'm in Las Vegas at the moment about to go to infocom, you know, the big classroom technology thing and
00:43:35
Speaker
I don't know that this is true, but I've always heard that Pyre Red borrowed the overhead projector from bowling alleys. And I think that's the number one edtech technology in the world. like That had the biggest impact. And it came from bowling alleys. Can I just also add that for regular listeners, we we might want to develop a Morgan bingo. And if we did, overhead projectors would be on the list, I think.
00:44:07
Speaker
remember them before. I think you're you're brewing a kind of a history of overhead projectors book, I think. Oh, maybe I'm just scarred from the impact of trying to take them out of classrooms and having death threats made against me by faculty members.
Listener Engagement and Future Topics
00:44:22
Speaker
Well, there we have our use of AI to look at the at the articles and podcasts musings from Morgan and come up with a bingo card out of that. Hey, well, listen, the great conversation. I really appreciate the feedback that we've had from listeners. And as you can tell, part of what we're looking for are not just quick question and answers, but really topics for conversation. So I want to thank our initial emails that came in and keep them coming. We'd love to address them in the future. But Morgan, go off and enjoy Vegas, or at least enjoy the conference. And Neil, and and enjoy Cardiff and your lovely summer that must be emerging at this point.