Chris: 0:13

Welcome back everyone to another episode of Shifted Podcast. Um our school year is uh still kicking. Uh we're just getting into uh our September, October. Well, still mid-September, getting into October. Probably by the time you hear this, it will be October. Um and I have a great guest, Stephen Downs, coming in, um, who works for the National Research Council of Canada with a specialization in digital tech, which you know that's our jive here at Shift Podcast. We love all this stuff. And Steven's such a thinker, a philosopher, and um just a great mind to kind of tap into about education, the digital landscape and uh connectedness. One of his big things is connectivity. That how do we stay connected through the X's and O's of digital technology? Steven, thanks so much for joining me today.

Stephen: 1:10

Thanks. Pleasure to be here.

Chris: 1:13

So, Steven, can we uh rewind time a little bit? And could you kind of tell us a little bit about how you kind of came to what you're doing today? Um, what were some of the experiences that brought you to the job you do today um at the council?

Stephen: 1:31

Um well pretty much happenstance uh I was studying philosophy in university and fell into a job teaching philosophy by distance education with Athabasca University.

Speaker 01: 1:47

Okay.

Stephen: 1:48

I did that for seven years with them and um always had an interest in computers. Uh I was actually enrolled in a computer certificate course even before my philosophy degree. And um so it was a natural to try to adapt distance education to computers. I created my own bulletin board service way back when. So um, while I was with Athabasca, I got hired by uh a Cineboy and Community College to develop their online presence, uh, everything from their website to their distance learning. Uh and I built uh a learning management system while I was with them, and that's basically where it took off from there. I went to the University of Alberta and then in 2001 came to the NRC.

Chris: 2:40

Awesome, awesome. Great experiences, and I love happenstances. I mean, those are sometimes we get to where we're at just by like a conversation or replying to an email. Who knows, right? It's just so uh amazing how that happens. I kind of want to drill down a little bit on the MOOCs. Now, MOOCs are online learning environments for whoever, right? I mean, you can have them in a plethora of different varieties. It could be on philosophy, it could be on you know car mechanics, whatever you want, right? You can probably find one online. How did you get started with that? How did you I know that you started with George uh Siemens, right? Siemens?

Stephen: 3:21

George Siemens, yeah. Um, so I've always been on about networks, and that goes back to my days studying philosophy. And back then I I represented uh knowledge as being based in similarity and association, and that led me to something called connectionism, which was a philosophy or approach in computer science, uh, which was based on artificial neural networks, which totally meshed with what I was thinking at the time. And so uh, as I developed my own thinking about online learning and online communication generally, networks fell naturally into place because, of course, we were on the internet works, right?

Speaker 00: 4:12

Right, right.

Stephen: 4:13

Um so George Siemens came along with an article titled Connectivism, where he drew out some uh features of networks and applied them to learning, uh, which totally meshed with what I was thinking. And um, so consequently, George and I got together. He held a conference called Connectivism. It was an online conference hugely popular. So we're sitting in a bar one day in Memphis, believe it or not, and we decided to offer an online course about connectivism. Um and we decided that the course shouldn't just try to explain connectivism, it should try to model connectivism. Um, and that was a good thing because we had 2200 people show up. And because we designed the course as a network and not one of your standard, you know, uh people sign up for a discussion list or whatever, the course worked. We could accommodate that many people, and that prompted Dave Cormier to call it a massive open online course, and thus the MOOC was born.

Chris: 5:31

Wow, wow. And there are many forms, as I alluded to before. What are effective MOOCs for these online? Which ones do you find personally grab not only you know the content that you want to, you know, eat, consume, but also have that personal connective um aspect to them? What are some of the components that you would look for in a MOOC that would allow that connectivity to happen? Uh these days, um, and you have to keep in mind, right? We're we're almost 20 years from the first MOOC. That was in 2008. So these days, I would say it needs to be in some way federated. Back then, I would have said distributed, and I still mean the same thing when I use the two different terms. And what I mean by that is it should not be based in a single place. Um, and just as an aside, that's what the commercial MOOCs all did is they were one big website with a whole bunch of Amazon uh AWS backing for storage and stuff and AI to do the marking and that, but they were all just centralized, and none of those models survived. One by one, they all ended up selling out one way or another and becoming closed products. But a massive open online course really ought to be a network, of course, right? Not one site, but many connected sites. And the properties of a MOOC follow from that thinking, right? Massive because networks can support mass much better than single centralized services. Open because it's not a network if it's closed, right? Open means it can join and leave, that content can come in and come out and flow freely throughout, right? Online, because well, that's the easiest place for us to have networks. And the only thing that's different about a MOOC, um and just uh a network generally is in the word course, and here course implies kind of two things. First of all, a fixed start date and fixed end date, so it's it's a moment in time, and second, some kind of well, I used to talk about going back to the original Oxford-Cambridge model where the learning was in fact all organized by the students who would gather around the professors, and each student would have a professor as a mentor. But what they would do is they would convince, and they had that this they actually had to do this, they had to convince a professor to offer a course of lectures, where course in this case means series of lectures. And so that that would form the core of the MOOC, right? There'd be a course of discussions on a sim on a single topic, uh, you know, hosted by one of us and bringing in people, and then the whole network would surround that and discuss that, which is what the students back in the Oxford-Cambridge model did. So it's taking that model and applying it to the digital age. That's interesting. Wow. And do you feel that when we're looking at these, like how the technologies, the ethics, the privacy of all of this, like, because you do say it's open, right? It's an open um network, you can come and go. Like more and more we're concerned at schools anyway, and like school boards, and you know, they're worried about the ethical use and the privacy use. Like, how does how would that contour or cause a problem with it being used more prolifically throughout, example, our youth sector education, like for high school students, example?

Stephen: 10:02

Well, yeah, I mean, it depends on what your view of education is. Um I'm just trying to think about how to word this diplomatically, because uh, you know, a lot a lot of people are very concerned about controlling education, controlling the content of education, um, and they're very concerned about education imparting specific knowledge and specific content and even more specific values, especially cultural values, right, on students. Um and that's not really the model that a MOOC follows. Um a MOOC is much more about the students themselves determining for themselves where they want their learning to go. Um, George and I used to say, and I still say, you know, the content, quote unquote, is a MacGuffin. Right? Uh it's the thing that we're talking about, but it's not the thing that we expect people to learn. Uh you know, it doesn't matter if they learn the content. The content is just the starting point. It's the seed, it's the catalyst, it's the thing that gets people going. Um, and what they're actually doing in our course and in any course is exercising their mind, exercising themselves, thinking about these topics, working through these topics, hopefully, and we really encouraged this, creating and writing about these topics. Right? Um so it didn't matter whether they learned the topic, what they what they were learning were the skills needed to deal with kind of things like this. So uh you know, it it's a different way of thinking about learning. Um coming into a massive online course, you know, we we talk about things like privacy and things like that. Um my courses, um, I don't even ask for registration. I don't want to I mean I guess I'm kind of curious, but it doesn't matter to me whether or not I know who or how many people have signed up. Right? I mean this this is you know the the the need for registration isn't the same as the need for learning. There's incredible for you. Um, you know, so uh if I wanted to sign up for a mailing list, then I'd need their email address, otherwise I wouldn't know to send where to send the newsletter, but they could just as easily use the RSS feed and access to that as anonymous or as anonymous as it gets, right?

Speaker 01: 13:11

Right, right, right.

Stephen: 13:12

So there's much less concerns about privacy and surveillance. A person's contributions to a MOOC are are not required by the curriculum, right? They're voluntary. Each person contributes whatever they feel comfortable contributing.

Speaker 01: 13:31

Okay.

Stephen: 13:32

So if the person comes into the course having a basic understanding of you know not sharing too much on the internet, then you know there is a prerequisite there. Um then from privacy considerations they should be fine. Um so you know these kind of questions don't arise nearly as much in student, I don't want to say student managed because that's not quite the right word, right? But and and and not even student-led, perhaps student-driven might be a better word, but even that's not quite the right word. But you get the idea, right? Uh these questions don't arise the way they do in a course with the prescribed curriculum, specific content that a student must learn, specific activities that a student must undertake, specific rules of participation that a student must follow, right? If you remove the word must from learning, uh a lot of the questions about privacy, security, et cetera, disappear.

Chris: 14:45

Interesting. Interesting. And I mean, one of the things too that kind of struck me when you're when you're speaking was that this need for content that we have in structured schooling, right? It's all driven by content. We test it, uh, you know, we have goals you have to set. And we often forget about the person behind all of that and those skills that we need to develop in school and not just content. I mean, I think about school, I mean, uh what I remember is my personal interactions and plays and sports I did. I don't really remember what happened in C4Math at all.

Stephen: 15:27

Well, we don't know how to add, but that's basic numerous, you know.

Chris: 15:32

I mean, they say grade eight is probably about as much as you need to survive in this world. Where do where does where do we connect those two from the content to the person? And also help in that development not only of the knowledge of this of whoever is interacting with the MOOC, but also how do we develop them personally and their skills and their critical thinking, their you know, creativity, etc. Yeah, I mean, I know those are important questions. Uh the last thing we want uh is to have a generation of students who are incapable of functioning, especially in a modern industrial information age economy. Um, we we can look at any country where the education system has collapsed or been compromised to see the results of that. Um we we don't want that to happen, obviously. Um since the very beginning of my work in educational, I've urged caution. And I still urge caution, right? You don't jump into something whole hog before you know what's gonna happen out the other end.

unknown: 16:55

Right.

Stephen: 16:56

Um which is unfortunately what I think we do sometimes. Uh I tend to have two things that are are my my guiding points here when I when I think about this question. Um the first guiding point is if a thing is fundamental, it's gonna show up as soon as you try to do something. Right? Think about that. Yeah, uh think about anything simple, like we're having a conversation, right? The need for language is gonna come out, right? Um, you know, the better we are with language, the better the conversations we're gonna have. And that that's going to be true with a lot of things. If if a if a person wants to, I don't know, um sew dresses, like pull that out of the air, right? They're gonna have to learn to read. They're gonna have to learn to work with patterns and directions. The more they get into it, the more these fundamental skills will become important. Measurement, mathematics, area, geometry, uh, 3D geometry if they're doing fashion, right? Uh you know, all of these will come into play. And and a whole bunch of other subjects, right?

Chris: 18:29

But you know, so the fundamentals your your starting point is the hook. Where where are they gonna find entry into whatever you're asking them to do? Right, right.

Stephen: 18:42

The fundamentals will emerge, right? Right. So it's it's our role as educators to make sure that the resources are there in place for them to learn these when they need them, but you know, forcing them on them, probably not the best way. And that's leads obviously, and you've already alluded to it. The second point is what is it the person wants to do? Right. So a couple, there's there's a couple of caveats there, right? Because you know, uh people say, well, all this all the students, all the kids want to do is play. Well, strictly speaking, that might be true, but if you look at what play amounts to, in many cases it amounts to imitating, uh imitating what they see around them. Uh so you see them playing a lot of sports. Why? Well, they see sports on TV, uh, you know, uh, or they see the older kids playing sports or whatever. You know, so uh you kind of need an environment that's inspiring that will give kids ideas what they want to do, right? Um and you need to provide the role models for them. Um, and and and then the access point into whatever it is they want to do. Some kids will want to play basketball all day, every day. Um should this be encouraged to a certain point, sure, why not? Um, you know, they they will be developing themselves as athletes. And you know, again, the better at being an athlete, you're going to need to learn a bunch of stuff, right? It's not just gonna be about the physical performance. You're gonna want to get better, you're gonna want to learn about arts and and and physics, uh, and tactics, uh, and physical fitness and all of that. Right. Is it is it okay for a person to focus on that in their life for the rest of their life? Sure, why not? They don't need to be a worker in a factory these days. Nobody needs to be a worker in a factory, uh, it's all potentially automated. So that that's really gonna take, you know, and and uh so other kids will want to make robots. I would have been one of those kids, except my hands were not very good. But you know, maybe, maybe well no, I played with Mechano and things like that, and I was all so yeah, that that never really worked, but but computers would have been a natural for me, and were a natural once I had access to them. Right, right. And so but the thing is, right, to a large degree, kids can follow their passions and it will lead to a good outcome. And then then we can deal with the exceptions. There is there will probably be exceptions, but if you think about it, dealing with the exceptions and guiding them so that you know they actually do develop worthwhile skills and and abilities is better than stamping the same quote unquote foundational knowledge on all of them, in my opinion.

Chris: 22:20

Right, yeah, totally, totally. Right, right. But still kind of satisfying the role that they are they their interest guiding them to, and then kind of pulling in all this stuff as you're going discovering, you have no choice, right? Yeah, that's really fascinating. I love, I love there's some great, great quote quotes in there, Steven, for sure.

Stephen: 22:53

And and just as yeah, this is not none of this is my unique idea, right? There's there's a lot of people who have said this before me. Um everyone from Frieri to Illic to uh uh well John Holt would have mentioned it. Um other names. Um there's there's one person in specific, I can't remember his name, but oh uh dang, it's right on the took of my tongue. I have Alvin Toffler, but that's not it.

Chris: 23:29

Well, maybe it will by the end of the day. It'll come back. Boom!

Speaker 02: 23:34

It'll come back.

Chris: 23:35

Steven in walks AI, right? Now we know since November 2022, things have dramatically changed, not only in education, but the world is feeling these waves coming through. We know it's not going anywhere. How do you see AI supporting you know online learning connectivity? Like, is it gonna be an asset to us? And what do we have to get our heads around or beyond? Because I know a lot of the times right now in in education is kind of seen as well, it's a cheating machine, you know, it's like a quick way to get around doing the work where there's no critical thinking. And where do you stand on AI? Like, what what do you what are your hopes and also kind of your cautions about you know going all in with AI in in in in in learning?

Stephen: 24:29

Yeah. So well, again, first of all, be cautious and conservative, you know, you know, like don't just go all in on AI because you don't know what you're going all in on. But that's it. Okay, that's the first thing. Second thing is AI is not magic, it's math. Okay, that's why it's not going away. So if you know, you you can criticize open AI and meta and anthropic all you want, and there's lots of reasons to criticize those companies. The fact itself of the math is not going away because the math is networks. The networks is the stuff that I've been talking about my entire life, right? Uh that's not going away. AI is based on connections, it's based on strength of connections between individual units in a network. Um, and in large language models, it's based on strength of connections between words or perhaps parts of words. And in some of the more uh recent AR systems, strength of connections between words and phrases and longer sentences, increased attention, if you will. So none of that's going away, none of that is magic. Okay. Uh it does point to um how irrelevant the content is. Right.

Chris: 26:05

Um I think that's the big exposure too, right? That yeah, I mean, we thought when when kids had handhelds in Google, yeah, forget about it. I could go and like ask the teacher a question now, and they're suddenly the rules reverse. I think AI like trug like triples that.

Stephen: 26:24

Yeah, yeah. So now, interestingly, um, because you know, artificial intelligence, properly so-called, originated in a research project to try to emulate human reason using computers, right? Hence artificial intelligence, right? Um and it has been more or less successful. The part that hasn't really changed is emulating humans, and many of the human foibles are found in artificial intelligence. Um, for example, confidently asserting something that is not true as being true. Humans do that all the time, right? Sometimes on purpose, which we call lying, sometimes by accident, um, which we call, I don't, I don't think we even have a name for it. Um, you know, um depending on how they're set up, an AI system might be a sycophant, might might say, oh yes, you're so correct, and all of that. So you you can't just accept what an AI says as necessarily true. Um now that should have been the case all along, um, but especially well, I was gonna say especially recently, but I suppose it's something that's always been the case through human history. Uh, we've been taught to accept what the authority says as true and objectively true, um, especially in the age of science. Um, but you know, even in the age of faith, I guess. Um and of course, that should never have been the case. We should always have been critically reflective of what anyone tells us. And uh, you know, the the scientific method, properly so-called, basically is a set of mechanisms that enables a person to determine whether or not something that is said to be true is in fact true. And it involves um, you know, trying it for yourself, getting the same information from multiple sources, seeing if it stays true over time, um, imagining counterfactuals and falsification and testing and you know, those sorts of principles, right? And we can be really loose and fuzzy about what those principles are, and they'll still more or less work. So there's a strong correlation between what we think of as scientific method and what we think of as critical thinking, although, bracketed aside, there's a whole industry of false critical thinking. Um beginning with De Bono and continuing on. Not to say that everything de Bono says is false. There are ways of broadening your creative and imaginative powers, not determining whether or not something somebody tells you ought to be believed. See the distinction, right?

Speaker 01: 30:01

So yeah.

Chris: 30:02

Um, and so you have you need to be kind of careful about that when we talk about what what we mean by critical thinking. I've I've seen that misunderstanding play out.

unknown: 30:13

Okay.

Speaker 02: 30:14

So, um, so how do you teach critical thinking? Well, not as a subject.

Speaker 00: 30:21

Right. It's again kind of a lived experience. Like you get better at it by doing it.

Speaker 02: 30:27

You get better at it by doing it, and you do it, especially when you're a kid, when you see other people doing it.

Speaker 01: 30:37

Right.

Stephen: 30:37

Right. A big part, a big part of my work over the years has been to try to offer a model of what I think critical thinking is. And I have my daily newsletter. Yes. And each article that I review, I give it a little bit of critical thinking, right? So you read my newsletter, you see uh four, five, six, eight items of me doing the critical thinking thing on an article. Um, you know, and in my longer work, I try to do longer instances of critical thinking and develop the whole thing. Um that's part of the answer. The other part of the answer is let's think about how how do I want to put this? Um let's think about what thinking is. Let's think about what even critical thinking is, not in terms of a process or a method, but in terms of how we know what's true. No, no, I don't even want to say that because truth is an attitude we have toward a proposition. It's what we call a propositional attitude. Truth is a label that we give to a sentence in a language, but our our thinking doesn't work in languages, right? Uh we hear languages when we think, but that's our perception of ourselves thinking. What a mess.

Chris: 32:17

What a mess. What we do is what the computers do. We recognize patterns. Right.

Stephen: 32:27

All right. So what we want to model is pattern recognition, pattern testing, all of that. Um there's a whole story I would tell. And you know, if we we think about the different types of uh logic and critical thinking, these are different forms of pattern recognition. That's why Sesame Street got it right when they did this whole one of these things does not belong. They're after pattern recognition.

Chris: 33:01

Great. Could that be like that's a part of computational thinking, right? Like where you have certain pattern recognition, like there's certain mathematical concepts that are more philosophical in a sense, but they're hemmed to you know, math and science and like reality. Um I I've always been really fascinated with that mindset instead of a math thinking. Like, I wish instead of math, you would have taught me more about computational thinking, so that I start to understand patterns and connections between things and sequencing, and which which, like you said, is running not only our technologies, but AI is um I mean it's a prediction machine in in essence.

Stephen: 33:51

Well, yeah, uh absolutely so is the human brain. It's a prediction machine. Uh well the but it's also like a super sensitive uh content receiver, uh signal receiver, you know. Like uh, you know, we have we have the various senses in that, but it's like our head is our antenna, right? It's well, our whole body is the antenna, right? So we're very sensitive to the environment and we and and signals, causes, sensations, whatever, come in, and then we're very sensitive to what they are, and then we look at the patterns in those sensations. Um, and some of the patterns are sounds, and some of the patterns are words, and so on.

Chris: 34:53

Um fascinating, Steven. That's like you're making me think of things in different ways here. I love this.

Stephen: 35:00

And we go back to computational thinking. Uh we think about what computational thinking is, and it's fascinating if you think about it. Uh, what is it? Right. So what what's a computer program? Computer program is a set of declarative statements, perhaps. Uh, but really the the magic in a computer program works um in the form of conditional statements. Right. Okay. If this then that. And and the rest of computer science, um, as any logician will tell you, is logically equivalent to a set of conditional statements, right? A loop is a type of conditional statement, right? Uh a series is a type of conditional statement, right? If you've done this, then do this. If you've done this, then do this. What is a conditional statement? Conditional statement is a fascinating thing. I spent a long time studying it in my youth from a philosophical perspective. My first published paper ever was in fact called Conditional Variability. Um and what makes a statement if X then Y true? Right, and we think it's just your standard truth table, but there's so much around a whole context around that. Ultimately, what a conditional statement is is a pattern recognizer, right? If something is the case, right? So, but pattern recognizers don't, and this is the this is where neural networks were so brilliant. Pattern recognizers don't have to recognize patterns exactly.

Speaker 01: 36:52

Okay.

Chris: 36:53

Um and and you know is that some of the bit of like with AI, like with hallucinations sometimes, like they'll want to satisfy, but it might not fit exactly to that, but they'll work their way around it somehow.

Stephen: 37:12

Have you ever walked in a crowd and thought you recognized someone and then it turned out it wasn't the person you thought you recognized?

Chris: 37:20

Totally, totally, yeah. So that's that's a good, good connect. That's a great example, Steve. Wow, yeah, totally.

Stephen: 37:29

That's exactly what's happening with a hallucination, right? Um, you and this is a really important point. You're predisposed to recognize a certain set of things, right? Right? People in your family, uh, famous people, um, people who look different, uh whatever, right? I mean it's the set is different for everyone, right? Um and I'm pretty sensitive to this because my eyes are so bad that I'm really bad at recognizing people, horrible at it. And I've insulted so many people over the years as a result. I'm just but you know, in the absence of something, you feel its presence keenly. Um, but you can recognize on a partial pattern, right? That is, in other words, you can recognize things based on similarity rather than identity. But what makes similarity works is this precondition, or as we say these days, context. Right? So what is computational thinking? It is recognition by similarity, what counts as similarity, context, and so now you're into topics, you know, how do you do pattern matching? Um, how do you do classification, categorization, but not in terms of necessary and sufficient conditions, but in terms of similarities or partial representations. What is a representation? What counts as a representation? If I told you this is uh Karim Abdul Jabbar, first of all, let's check. Steven's holding up a he's holding up a spoon.

Chris: 39:29

Yeah, you recognize the name, right?

Speaker 00: 39:32

Yeah.

Speaker 02: 39:32

Okay, good. Just checking.

Chris: 39:34

Totally, totally. I'm just because this is just gonna be audio. And so Steven was holding up a spoon and asked me that question.

Stephen: 39:43

And now if I do this, yeah, I just conveyed a signal to you that you can probably interpret. And again, nobody on the audio saw that. I bounced the spoon a bit and made it leap up toward a basket.

Speaker 01: 40:02

Right.

Stephen: 40:03

There was no actual basket, but you would have inferred that there was a basket there. Now, you know, and I know there wasn't really a basket there, but that was never the issue. The issue was did you correctly infer that I intended you to infer that there was a basket there? That's computational thinking. So it's not about following rules and principles and constructing algorithms. It's about pattern matching, data awareness, um, classification, categorization, representation, those kinds of topics. And those are the kinds of skills that are fundamental in an information age, not content.

Chris: 41:01

Right, right. Where we are now, which is um I I I love I love your words, Stephen. And I mean, the reflections that you're you're making me and I know the listeners have is just out of this world. I really, really appreciate the time. I know that we're running a bit long, but I just this is so fascinating to me. I mean, we could just keep going and going about this, but I do want to respect our time. I'm gonna have you back, Steven. It would be great to just have a follow-up on this because it's so um, I love how philosophy ideas connect with tech. I mean, I love that marriage or that that relationship that they have. Um, so I'd love to dive deeper into that. If you want more of Steven, he does, as he alluded to, OL Daily. It's a great newsletter. You can find it on LinkedIn or on his site, which I'll put in the descriptor, um, so that you have access to all these great um ebooks he has, and also, like he said, his reflections, his critical thinking on articles, etc. And he's he's prolific in getting these out, and they're really good. So I recommend those to uh you listeners as well. Steven, this has been an absolute pleasure. Uh it's been such um I I I need to sit and just think a little bit with everything that we've exchanged today. Um, but I would really love to uh continue this one day.

Stephen: 42:30

Sure, it'd be my pleasure. Happy to do it.

Chris: 42:32

Amazing. Well, Steven, you have yourself a great day. And again, thanks for joining us and and sharing some of your knowledge with us. I think we're all that much smarter today, now after this.

Stephen: 42:42

So thank you. Oh, you're very welcome.