← Edtech Insiders
Edtech Insiders
What Should Universities Teach When AI Knows Everything? with George Siemens of SNHU
00:56:44
Loading summary
Transcript39 lines
- [00:00]
Alex
This season of EdTech Insiders is brought to you by Cooley LLP. Cooley is the go to law firm for education and ed tech innovators offering industry informed counsel across the pre K to gray spectrum. With a multidisciplinary approach and a powerful edtech ecosystem, Cooley helps shape the future of education. This season of EdTech Insiders is brought to you by Starbridge. Every year, K12 districts and higher ed institutions spend over half a trillion dollars, but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts and strategic plans and then helps you turn them into personalized outreach fast. Win the deal before it hits the RFP stage. That's how top edtech teams stay ahead.
- [00:53]
George Siemens
If we're less so about hey, we took you through in course in our LMS and now you got your grade. After a year or so you'll lose access to that LMS environment. We're doing them a bit of a disservice because what we need to enable them to do is to get clear around what they know and what they've acquired and what they've done in a way that can then be parsable, accessible and engageable with a range of AI agents to help them be more productive in their personal lives in in their work life. Our need is to not just say teach you something now, but to teach a way of being that will persist outside of your engagement with the institutions.
- [01:33]
Alex
Welcome to EdTech Insiders, the top podcast covering the education technology industry. From funding rounds to impact to AI developments across early childhood, K12, higher ed and work, you'll find it all here at EdTech Insiders.
- [01:49]
George Siemens
Remember to subscribe to the Pod Pod, check out our newsletter and also our event calendar. And to go deeper, check out EdTech Insiders plus where you can get premium content, access to our WhatsApp channel, early access to events and back channel insights from Alex and Ben. Hope you enjoyed today's pod.
- [02:13]
Alex
We have an absolutely incredible guest this week on EdTech Insiders. We are speaking to one of the prominent learning theorists of the Internet age. Truly a pioneer in everything research and learning related and edtech. You know Hero George Siemens is a prominent learning theorist and digital education researcher serving as the Chief AI and Innovation Officer at Southern New Hampshire University. We love Southern New Hampshire University. He co founded SNHU's AI learning platform and he's widely known as the pioneer of the Learning Theory Connectivism which we'll talk about today and for creating the first massive open online course. George Siemens welcome to EDTECH insiders.
- [02:54]
George Siemens
Alex, thanks. I've been long tracking your work and exciting to spend some time with you today.
- [02:59]
Alex
I am truly honored to be with you here today because I think you've done incredible work for a long time and what you're doing right now is just right at the cutting edge. You've basically spent decades studying how technology changes, learning how connections between people and current events change, learning. As we're sort of in this crazy moment right now where AI is changing education yet again, what do you see as the most important shifts coming now from your vantage point as a longtime researcher of education and technology from the AI revolution?
- [03:31]
George Siemens
I think that's a great question, Alex, in terms of wanting to get at what is the substance that's in front of us and how do we begin to align ourselves in a way that allows us to make sure that we're doing what's right for our learners and for our learners success? And there's really a twofold angle here, which is on the one hand, the question is how does AI influence and impact the learning process as we know it? We're seeing a number of reports that come out that there's, you know, improved productivity, there's some equalizing capability. Let's say someone who uses English as a second language may now have a bit of a leg up to be able to perform better in areas that they couldn't do before. We're also hearing concerns like what is this idea of cognitive surrender? Not just cognitive offloading. Recently seeing a port that talked about how there is a diminished impact on creativity the more you use AI. And so there's that bundle that is so ambiguous and entangled that says AI is good, AI is bad, AI shouldn't be used for this. So that's that bucket that says AI in relation to our existing system of education. Then there's another bucket over here that still doesn't have clear boundaries. And that's one that's emerging that says, what if AI changes the core infrastructure of learning and enables entirely new capabilities that we don't yet realize? And what if it opens doors for us that changes how and what we learn, but actually makes us more capable, more impactful, and more effective in the kinds of work that we're doing? That's the challenge because so much of the dialogue right now is on AI's impact on the education system as we know it. All the values that are coming out, you know, you need to have this knowledge in your head. If you outsource it to an LLM, we call it Surrender and so on. I think I'm interested to see how those two will resolve. Which systems are capable of imagining a new relationship between humans and information and knowledge and AI as a partner, and which systems double down and want a traditional assessment model and a traditional learning model. So I think there's going to be some tremendous friction between those two visions. But I think both need to coexist short term. No university is going to blow up their entire model to just go out and be frivolous, running through meadows of innovation without regard for the practical realities of job market, labor market, and actual learning or development. So I don't know, it's sitting a little uncertain in that regard. Yeah.
- [05:50]
Alex
So let's take those one at a time because I think they're both incredibly interesting. Let's start by talking about the existing mode. You know, before we get to sort of how AI might change our relationship to learning, might create entirely new infrastructures and ways of learning, let's look at how it works with our existing system. One of the things you mentioned, cognitive surrender and a lot of the edtech companies, both in higher ed and in K12, what they've been talking about a lot of the time up till now is, hey, AI can take the load off of instructors. It can do a lot of the bureaucratic tasks, it can speed up lesson planning, can make things more efficient, make study more efficient, but they sort of are skirting around a little bit the core of learning itself. And you know, what AI's relationship is. And the research is really still out on it right now. I'm curious how you see AI going far beyond hopefully just efficiency gains and really start to change even in the current system learning outcomes, you know, so
- [06:42]
George Siemens
I'm going to take a slightly peripheral path to get to the outcomes question. So one of the things I did when I joined SNHU last year and took on, you know, the role that I currently have and you know, the goal is, if so much of the dialogue is around AI, when you don't necessarily have it used in your process of daily work and daily activity. And so my argument was to say, how do we open the aperture of what's possible with AI on the part of leaders and have them engage with tools and technologies that actually allows them to do more in ways that they aren't aware of. Because so much of the dialogue still is the LLM versions we had at the start, namely, this paper was cheating. And I'm like, we are so far past this paper is cheating when you're starting to look at managed agents and multi agentic systems that are producing variable outcomes for different reasons. And so if we're still locked into this framing, it's about plagiarism and cheating, we're missing it. So the first thing I did was I created Executive Council has the C suite at the university. And so I put together what I called architect your work. And it was a boot camp, for lack of a better word, that I ran for six weeks with a senior team. And it started off with just the basic. We use Claude, you know, Codex with a current iteration, but we use Claude. And we just said, hey, you know, my argument was don't worry about literacy, don't worry about any particular pre existing framing. You have to align to just worry about developing a relationship to when this is good and when this is useful. And we started with the basic, okay, here's your little chatbot does its thing the way we all did with ChatGPT, and then we progressively move through. Now you're working with projects, okay, now you're going to build a skill. And what was interesting, you know, when we landed ultimately with what does it mean to create a second brain, so to speak, where you've got sort of obsidian or you know, markdown files, where you're managing the ways that the tools are being used and you're starting to amplify your capability by better structuring your own information that you've generated or that you've created. And so we moved through that whole process and the core argument was the literacy conversation is so fluid because it can literally change overnight. Something gets dropped. There was a time we were doing prompt engineering as a primary vocation with AI. Nobody really talks about that anymore because we've shifted. So my argument continues to be the way to understand AI is to become familiar with what it does well. And the team I'd always articulate, you have to have that blow your mind experience that says, oh my goodness, this changes how I work. And then you have to have a bitter disappointment where it embarrasses you because you forgot to check a citation before you share it with colleagues. And somebody's like, this is AI swap. And you're like, ooh. So let's say at a leadership level, you have to understand the attributes of this technology that we say is going to transform our society. So dealt with the executive council and one individual came up to me and she said, I've been in higher education for like 35 years and this is the most significant learning experience of my life. This has transformed how I work from there. We moved it down to our senior leadership council and we just did our wrap up just prior to getting on the call with you. And at the end, I asked the group that had gone through it, I said, if we took clawed away from you, would you be disappointed? And it was like a resounding oh my goodness. Like my workflow would change, my productivity would drop, and so on. So finally, to get around to your outcomes, question, that experience of doing more and being more capable, no one is dumbing down. They're doing bigger things and having bigger conversations and solving routine, mundane problems with AI so they can do more substantive work that supports student success. So I would say with students, if we get that right, if we change how we do assessment so that it actually unlocks creativity on the part of learners, if it gives them the capability to create and to build. And I keep pushing internally, you know, we want to build culture where you're creating and developing artifacts and outputs. With AI, I think that's where you're going to start to see outcomes improve and learner capability and confidence advance as well. So that would be my vision within the system.
- [10:31]
Alex
Do you feel like that downstream of that, as people expand their capabilities as they build, you know, what you're sort of calling a second brain? It seems like having a second brain would probably have beneficial outcomes on your learning process. Do you see that coming or is that even beside the point when we talk about the current system?
- [10:48]
George Siemens
Yes, I do. So let's just say briefly, if you were to give. One of the things that's most important I found, at least in interacting with AI, to make it useful to you, is you have to bring intentionality and you have to bring your resources, if you will. And if you've clearly articulated what you want and what you need from a particular set of agents, then you now need quality data, which is could be meeting notes, reflections. So I spend quite a bit of time curating my personal data reflections at the end of a day that just, you know, use whisper flow and, you know, run my daily note in Obsidian and sort of detail this and this and this and observations, right. Once it's there, it can be processed and engaged and connected to the models you're working with. So I think if we did the same for learners, if we're less so about, hey, we took you through a course in our LMS and now you got your grade after a year or so, you'll lose access to that LMS environment or at least that particular course. Even if you're Continuing to take courses, we're doing them a bit of a disservice because what we need to enable them to do is to get clear around what they know and what they've acquired and what they've done in a way that can then be parsable, accessible and engageable with a range of AI agents to help them be more productive in their personal lives, in their work life. A lot of SNHU students are professionals, they're mid career or they're in a career setting. And so as a result of that, our need is to not just say teach you something now, but to teach a way of being that will persist outside of your engagement with the institution. So you'll hear me skirting the direct answer to your question on the outcomes from that lens, you know, because the outcomes are often assessed and terms of grades and so on. But the real outcome should be five years after you completed a program, were your earnings, did they increase? Did you have sort of a transformation in your life, did you get a promotion in your work? And so it's our, I think obligation to learners who engage with institutions that want to be AI capable is to give them that bridge to that kind of goal or desire that they have that's life changing. Not just oh, they moved from, you know, a B to an A across the university sector. That's a short term metric. The big one is did you help them acquire goals and objectives in their lives that they were unable to achieve before they started with you?
- [13:07]
Alex
We'll be right back. Innovation in pre K to grade learning is powered by exceptional people. For over 15 years, edtech companies of all sizes and stages have trusted higher education to find the talent that drives impact. When specific skills and experiences are mission critical. Higher education is a partner that delivers offering permanent, fractional and executive recruitment. Higher education knows the go to market talent you need. Learn more@higheredu.com that's H I R E edu.com that's a terrific answer. And I don't think you're squirting the question as much as sort of expanding in the definition of learning outcomes in a very southern New Hampshire style way and saying what are your, what are the students actually looking for in their lives? What are the changes? Yet a short term metric like an assessment or a final exam or a grade is really irrelevant compared to the long term learning outcomes of, you know, how to live with your second brain and how to have a better life and how to have make more money and how to have the job you want and how to have workflows that are more fulfilling. I think that's actually a really. A really interesting way to look at it. You sort of are jumping to learning outcomes, but in a much expanded definition, which I think is where the whole field is headed. I want to move to the second part of your definition about what I might do to sort of change the paradigm of what learning looks like. And I think it's a natural segue to talk a little bit about your theory of connectivism, which you've pioneered, you know, about 20 years ago now, for people who don't know about the connectivist, you know, theory and philosophy. Tell us a little bit about the core aspects of it. And in a lot of ways, they really precede the AI revolution in a very specific way in terms of how connections between people, between people and information is where knowledge is actually built. But you could say it much more than better than I can tell us about connectivism and how it's relevant to this moment. Sure.
- [14:59]
George Siemens
And it's worth emphasizing. You know, I. The original article was me struggling with. I was having this experience where I was teaching at Red River College at the time, and it was pre PhD. I started my PhD after that, that paper. But the goal in that original paper was to say, you know, wow, I'm. I'm sitting here. At that point, we had blogs prominent, not pre Facebook still. I mean, Facebook was there, but not the behemoth it is now. I was engaging with how do we build knowledge in an environment where we have multiple voices, we don't have a textbook exclusively, or we don't just have an academic paper and we don't have the single instructor. So I was involved in online environments talking to people like, you know, Alan Levine. At that time, I think he was with Maricopa and, you know, Stephen Downs, he was out in Canada and a group of colleagues out in Europe and across the US and there was, you know, a number of us that were all just trying to riff on what was happening in our world because it was now easy for us to just post our thoughts, reflections, comments, and opinions in a blog format pre Twitter. So people listening may not quite be aware. Like, there was a time where it was actually kind of hard to get your ideas out into the world and to find ideas. Now we're at the overwhelmed stage. Stage. There's too much. So as I was doing this, and there was one simple little mechanism that tied a lot of this together. So I didn't go to 50 blogs every morning. You know, that time I had RSS reader. It eventually ended up being Google, which was probably the most disappointing product Google ever nerfed. But it was just. It allowed me in one space to collect the voices. At that point, it might have been 50 or 100 people writing globally about, you know, ed tech at the time. And it was such an unbelievably rich learning experience. I. I would. I would get ideas from how to implement a site in Blackboard or web CT at the time where we were at, or Moodle or other tools. And it was just this frenzy of connections, and you never knew what you were going to learn. It was serendipitous. There was no path A to path B. The way learning outcomes through to teaching, through to assessment are structured in the university. That was just this Precambrian explosive period of experiencing an alternative model. And I sat down, I was in the middle of a course. I just read Driscoll's Psychology of Learning, which remains one of my favorite accessible texts on. On learning as I was going through it. And she does a great job just unpacking different theorists and different frameworks. And I realized, like, none of this captures what I'm experiencing. I'm experiencing this enormously chaotic, randomly motivating process of just growing and learning. And so I tried to capture that with connectivism that said, what if our framing of the central expert, the structured, rigid process of learning, what if that actually isn't how it works when you have sort of this cacophony of connections that are present constantly and everything is accessible to a degree? And so that was what I was trying to write. And so my argument with connectivism was that the pedagogical model that defined the university experience that I taught in at the time was not quite representative of the way that the architecture of information was emerging. And I started to hypothesize that if you want to understand the future of education, you need to understand the architecture of information creation, generation, absorption, validation. And as that core architecture changes, our learning needs change because they're built off of an information architecture. So that was the attempt with connectivism and to carry it over a little bit more to what you're saying with AI. AI yet again changes the architecture of information. And so I would argue, and I keep meaning to sort of put, you know, words to page, so to speak, around how does artificial intelligence restructure the idea of learning and learning theory? Because at the core of it, we have more autonomy, we are less reliant on the central expert. We have more capability than we possibly ever imagine. And I did have a statement, I think, in one of my principles at the time, which was, you know, that knowledge can reside in non human appliances or some word like that, you know, so I was already thinking it doesn't just have to be humans connecting, it can be AI or agentic AI or related tools that could be part of our knowledge ecosystem. But ultimately we have high degree of information and data sovereignty ourselves, much higher agency the university enables, and we can identify and solve our own knowledge needs without going through a mediating agent like a faculty member or even an institution. So that was roughly the core ideas around connectivism when I first posited and the context where it arose.
- [19:24]
Alex
Yeah, I mean, I have the principles of connectivism up here right, right now and I. Exactly. Learning may reside in non human appliances. That feels like a good prediction of this as well. But you also see, you know, learning is a process of connecting specialized nodes or information sources, or, you know, nurturing and maintaining connections as needed to facilitate continual learning. When I look at, you know, learning from Reddit, learning from Wikipedia, learning from YouTube, learning from networks of people on Discord, where they're all trying to wrestle, especially with this fast changing world of AI, it feels like this connectivist philosophy is so core and it's so. I mean, my follow up question for you on this is really, you know, I can play sort of two very different roles in this structure, right? AI could be the central expert yet again. It could be like, oh, you know what, when people talk about AI companions or AI tutors or the idea of just, you know, learning back and forth with an AI and the AI itself, right, becomes the expert because it strained on so much of the world's knowledge that sort of recentralizes it. But I think realistically what AI is really going to be doing is actually becoming, like you said, like the RSS feed, right? It's pulling together all the information, all the newest information, and being able to sort of speak with you, but also connect you into the networks that are most exciting in whatever space you're trying to learn. I'm curious how you resolve those two sort of roles that AI might play. Is it the ultimate sage on the stage or is it the node that connects to all the other nodes and just creates a network that is unbelievable for you among other human experts.
- [20:54]
George Siemens
I mean, what an excellent question. And I think it's, it's a very pertinent one to process at the end of the day, you know, an LLM or any of the generative AI tools we're Using now and keeping in mind, AI as a toolset is much broader than just the transformer LLM slash model. There's much more to it. But in that regard, I think it is a network fundamentally. I mean, it's neural network actually in terms of how it's trained, and it does collect and bring in a series of that kind of knowledge. I think the one difference though is that unlike a faculty member in front of and I've taught cognitive processes, you know, 150 students at university of Texas, and you know, you're starting, I think it was a third year course at the time, and you start in September or August and you've got this wall of people in front of you. I knew my content and my curriculum. I didn't know anything about the learners or the intention of each learner because the education system can't afford that kind of insight in person unless we have just a totally different economic model of how universities operate. What's interesting with AI, where I think it may not become that centralized expert, is we bring our intentions, our questions, our context, our second brain to that relationship, which means it is a rapidly amplifying connector. And so it's more of a connector than a node in some ways now, it can certainly play an active relationship. So for example, you know, I've been playing with Hermes and you know, everybody's kind of mucking around with openclaw at various times. And the idea that you can have a series of agents parsing off parts of your tasks and parts of the work that you're doing, or automating things like your morning news scan of what happened in AI and education overnight. What's interesting is it's more the RSS feed that AI in that regard than it is kind of the central node. So I go to it to go to the real knowledge that I want, and if it doesn't provide me sort of the guidance to the answers I want, you know, I'll seek and go in other directions. But I think you're raising an important point that we don't want to fetishize AI as the central node, and we want to instead pedagogically implement it as an enabling capability that allows you to do more, access more, understand more. And a key component of it is if you come my experience, this will be different, differently. But if you come to AI without answers, without hypothesis, you're miss the real opportunity. Like the chatbot stage was a short stage with generative AI. I mean, there's a lot of people still use it for that, but as a functional tool, that wasn't the big one. Like my stage right now is I keep emphasizing to peers, you know, in SNHU is one of the critical capabilities of AI is information transmutation, that it's, I know what I want. Like I wrote a paper on something, but now I gotta make a PowerPoint out of it. And normally to do that, if I'm doing a presentation, you know, I go, okay, you know, I spent however many hours putting a presentation together and, and now I can take a paper that I've written or ideas that I've generated already, or, you know, emails I've sent to my team to say, hey, this, these are our priorities and I can just turn my ideas into a format that's in different space. So to me, AI in that instance is enabling my capabilities. It's not inhibiting by being the central conductor through which my ideas and concepts have to, to flow. And I think it'll be more and more of that, which is why I keep emphasizing what matters most to individuals today is get your knowledge architecture, your personal knowledge architecture in order. Get into a habit of daily summary notes, get into a habit of your most brilliant emails that you sent to your team or staff. Get them into a structured format. I mean, it could be markdown files or HTML, it really doesn't matter for now because you can flip one to the next as AI gets more capable as you go. But get that brain worked out so that you are the central node and you choose when and where to connect and engage with others, other tools.
- [24:39]
Alex
I love that your personal knowledge architecture is a really interesting way to look at it. And I, I love how you're playing with the idea is, is it a node, is it a hub, a node? Or is it actually the connective tissue between things? And it can sort of play different roles in different contexts, but the idea of the person, of the human as the central node connecting to other people, many other different ideas, and AI being sort of the pipe with the infrastructure, the piping between things. It's a really interesting metaphor if I'm following it properly, but one of the things that I think is a really interesting through line of your career, and I'd love to hear you talk about it here is how traditional university systems, and I say traditional, you can define that however you'd like, but you can sort of imagine what, what I'm trying to refer to here, tend to be a lot more centralized. They don't always rely on networks, even though there are networks, right? There are departments There are lots of networks within a university. A student is usually engaging with a single professor who is sort of a central node. And as we know, especially with things like, like AI or modern technology. The day that Mythos and, and Fable launched yesterday. And if you want to learn about Mythos and Fable, the answer is not going to your computer science professor and saying, what is Fable? The answer is getting on the Internet, right, and seeing What a hundred YouTubers and a hundred people on Reddit, a thousand people on Reddit and all the people who write those AI newsletters are saying about it and then piecing it all together. It feels so connectivist to me. And I guess my question is, how should universities square the sort of traditional model of internal expertise, very high level internal expertise, with that sort of knowledge of the network? How must a university evolve, like southern New Hampshire to allow its students to be parts of that network and point them back to the Internet if they need to know what's happening on a daily basis?
- [26:22]
George Siemens
Yeah, fantastic. Because Fable, Mythos is just a great example as we're, we're daily. I'm daily spending more of my time in what I'll call sense making or meaning making practices than I am in sort of core learning tasks that I experience in university. And this is probably about almost a decade ago, but Kaggle did a paper or report where they said 59% of the people at that time that had a job in data science acquired their skills and capabilities through at that point, YouTube or MOOCs or open online courses or other avenues because universities don't have a production pipeline that is fast enough to move. And at that point it was, it wasn't today's fast. It wasn't like Mythos or Fable dropped yesterday fast. It was like, oh, data science takes a year to do. But I developed a Master of Science in learning analytics at UTA. It took me five years. And the subsequent document was like 500 pages. One of the most substantive artifacts I've produced in my life. And just to get somebody to say, yeah, you can teach our students about learning analytics. So I think this gets back to the university systems learning systems track to the architecture of information in a particular era. And right now, you know, we need some mechanism that is immediate and update that could be skills or literacies you've acquired yourselves. But it also gets that to a degree, the core part of much of the university experience today is directly in the line of path for the steamroller of AI. Right. The idea of taking something from textbooks, something from a faculty member's head and then putting it into the brain of a student is that's AI's territory. EFF used to have what they called an AI progress index that I used to use in presentations that would talk about in which domains like games or reading comprehension or some rudimentary math work, character recognition and so on. Did AI exceed human capability? And this goes back like 20 years. This is pre generative AI. And in numerous cognitive domains, AI had the capability or in some cases able to exceed beyond what human beings could do. And so much of our educational experience today needs to be restructured to account for the fact that it's roadkill for the steamroller of AI. And if you want to prepare learners for the future, you need to begin focusing on process capabilities, on sort of metacognitive attributes, judgment and wisdom and insight. But more than that, you don't need to just blindly adopt AI. Critical reflections on what are the environmental effects of AI? Why are data centers disproportionately early indication at least being built in poorer, lower income regions rather than sort of wealthy? We need to grapple with the totality of AI, not just what it does well and why we love it and why it gives us some help. But there is an uncomfortable plagiarism aspect, to put it mildly, to how AI models got to their capability now. So we need to be very blunt in looking at the totality of AI, what enables and what it provides to us, and then also look at and say, where does that tweak that relationship? And I'm going to argue, and I've said various venues, there's this idea of cognitive escalation where we move from. This is the work we used to do. And now if I can do that, we're not going to sit here and twiddle our thumbs and, you know, watch reality tv. We're going to now move up and do bigger things, more complex things, more integrative things, and quite possibly more human things as well. So I think I'm optimistic that we will make that part of the right judgment. But I'm worried about the large swaths of society which you're already seeing some of the negative effects of AI having sort of a toxic effect relationally and so on, so covering maybe more territory than you intended. But I think the core argument I would make is what AI does best is kind of exactly what our universities try and replicate in our students. And that means that we need to broaden the canvas of human capability and human potential to get at fuzzy words like meaning making and wayfinding and sense making and these cognitive processes that are more complex, that allow you to wake up in the morning and say, oh, fable dropped. What does that mean? What are examples? How do I apply to my job? How can I take this and make it a practical impact that serves me and my personal life or family life or in my career? So you're no longer just looking at do I know a certain mathematical formula or do I know a certain theorist that has created in psych some contribution to psychology, but more can I use this to improve my life, improve the life of my family, improve my contributions to my community? So you're shifting from this sort of transactional relationship between learning and knowledge to, to this transformative transaction around it. And I think I'm optimistic that AI will help us get there, but we'll obviously see. But that's the direction that I'd love to see us move as university system.
- [31:12]
Alex
I love that. And my reaction to that when I hear, hear it is, you know, cognitive escalation. That idea that as AI take so many of the tasks we do on a daily basis in a personal way and makes them doable, you can outsource them to an agent, you can outsource them to a, to an LLM or you know, a system, and then it, we escalate. And you mentioned words like, you know, wisdom, sense making, meaning, making judgment, you sort of go higher and obviously intentionality, as you mentioned earlier, the, it's like if you can do almost anything, what do you want to do? What do you want your effect to be on the world? That feels like a phenomenal direction for universities to go in. I guess the, the follow up question for me is if university system had to cognitively escalate its own purpose. Purpose, right. What it's been doing for a long time is, you know, conducting research. It's been obviously giving credentials and certificates to people for a long time, but it's also been instilling knowledge, often to certain kinds of jobs. Right. Trains people for business jobs and law and science. If that whole system actually has to escalate and say, okay, maybe we're now about wisdom, maybe we're now about ethical questions. Like your question about data centers. What does it mean for society? How might the university says, I know this is like the trillion dollar question, but how might the university system uplevel itself to higher order thinking?
- [32:32]
George Siemens
Oh, that's, I mean, that's our work. That's what we're in the middle of answering that. I don't, I think one of the Things I really want to emphasize sometimes AI is cast as a kind of an inevitability that, you know, it will run this course and you know, somewhere between Altman and Dario and Zuckerberg, they're going to run the world. But anytime we use a tool or a technology, you know, we're voting in a sense. And when we begin to vote intentionally, when we begin to do things like create our personal knowledge architecture, when we begin by choosing community over sort of automated tasks and processes, I think we're more in control. And I did a TEDX talk, an SNHU TEDX talk a couple years ago and my argument, quite naive, but I'm still optimistic, is that what if AI does more and more of these kinds of tasks in this kind of work and then it allows us to become more present in sort of our daily lives and family and other areas. So as an illustration, we should acknowledge there is a dehumanizing aspect to the way that our school systems are sometimes structured. It is rote learning. It reflects the industrialization mindset of schooling as a factory system model. So there are ways that we could make that entire system more human and more focused on connectivity and, and so on. So within one of the platforms that we were building that was brought into SNHU last year, you know, Tanya Gambi leads, the scope of that work focuses on this idea of whole person learning, the human element, the connection to one another, the care economy, like pick your term. But we know something is there. We don't know what it'll look like, we don't know how. You know, you're probably not going to get hired at Google if you come in and say, I don't know anything about Python, I know nothing about neural networks, but man am I a delight to work with. Like, you're probably not going to get hired on that. So we can't naively say if you have wisdom and judgment and great capabilities for sense making that you're going to have a good job. You need to couple that with technical skills and self regulatory capabilities so that you can be a fantastic teammate while being, you know, okay yourself. So I think as a university system in the future is going to have various in, you know, that, that trio of being, well, you know, having durable, capable skills and having core technical skills. That entanglement of those three core concepts are going to be vital going forward for sort of whole person engagement. But everything from research, I look at what we're seeing come now with some of the models and obviously DeepMind has made that Research, especially the medical field, a big priority. We're going to, I think, start to see improvements in quality of research and research output directly, those that impact the human condition. And so we'll see a shift from this grappling with core knowledge information that the university does right now to more the transformative aspect of that, more the implications of it. And I don't know if I've seen a model yet that lets me say this is a model that's replicable across higher education globally. I haven't seen that yet, but I'm excited to meet and connect with people who are at least trying to answer that question. Because if you're grappling with that question, you're much closer to something. Asking the question like you did is a step toward the right direction in what it might be. And we'll experiment and play and try and figure it out collaboratively and together. And we'll learn from other systems as they're equally engaged in it as well.
- [35:53]
Alex
Yeah, and it is such an interesting way to look at how the education system works, to sort of think about that trio. As you mentioned, there'd be these technical skills, this idea of, okay, there are certain things and the technical skills may, you know, we're seeing this coding revolution happen right now with cloud code and codecs and everything. And people are starting to say computer science may not be the hot major anymore in a few years because those technical skills, which have been on the upswing at the expense of the humanities for decades now, suddenly, maybe that's what will be, you know, outsourced with these technical skills combined with human skills. Maybe there's creativity in there, maybe there's ethics and morality and wisdom and sort of whole human dimension. If you think of that as the nature of the school system of the future, I guess the question would be how might we, as an ed tech ecosystem right now, help engender that future? You know, it's tempting, I think, for many edtech companies in higher ed and K12 to feed into the existing system. Right. How do we make things work more easily? Because that's where the buyers are right now, that's where the students are, that's where the teachers or parents are or professors are. But if you're trying to sort of build that future, do you think that the there's a role that EdTech can play there? Or is that going to come more top down from, from Frontier Labs, or is that going to come from universities or nonprofit? Like who's going to help build that model? Will sort of Universities come together and have a come to Jesus moment about, hey, this is our new goal, or is that going to be built from some other part of society?
- [37:21]
George Siemens
Yeah, you're asking all the right questions,
- [37:23]
Alex
Alex, all the big ones. You've thought about this more than most people in the world. I really, I really believe that. So I'd love to hear how you think even nobody knows the answer to these.
- [37:31]
George Siemens
Yeah, yeah. And so I just want to articulate that, you know, exactly that. Like, I think it's a little uncertain on where we'll go, but I think directionally there's a few angles I'm interested in. So one of the things that we did recently, so I'm going to take you down a bit of a meandering path and try and get there. So one of the reasons to your question, around the Frontier Labs universities have been a bit slow to respond to AI at a structural or systems level. I think most people would acknowledge that universities often are intentionally slow because a big move in a university sector can have long term impacts in the lives of an individual. So you can't be reckless with that. So we have an obligation to serve our learners well. That helps them prepare for the lives that they want. So that may be a part of it. A part of it I think is until you've played with AI, until you've done stuff with it actively, and until you've had that blow your mind and be horribly embarrassed experience, I don't think you understand what's capable. And I think in many instances university leadership hasn't done that with AI. They've been sitting to the side, they're not using it actively and you can't govern direct something that you just don't understand or don't have an experiential relationship with. So I think in that regard, within snhu, we recently held an event in Boston where we had a large number of universities come together. It's called the Chief AI Officer Network. And the goal of this meeting was to bring together senior leadership, a different tier one, R1, you know, all the elite systems down through to the community college systems to HBCUs and HSIs. And it was a really eclectic mix of people who are all grappling with what do we do with AI in our institutions, how do we respond to what's happening with Frontier Labs? So we had three goals. Is one, to engage directly. We had all, you know, the three primary labs, you know, OpenAI, Anthropic and Google, you know, come in and speak to us and about their vision for the education sector. So we want to be able to speak to the product offerings from that sector, at least have an opinion on it, rather than just be a net buyer. Second thing we wanted to do as a group is we wanted to get a better sense of how to learn from one another because there's people around the country and around the world really, because we have upcoming events in Adelaide and in Mexico and the next event will be at ASU in October. And so we wanted people to get together to learn because if you're grappling with, gee, how do you guys orchestrate multiple agents? And how do you guys reduce token costs by compressing context as you hand an output over to a new agent? Like, we're all trying to figure this out. And so we wanted as a secondary learning experience. And the third component was to contribute to policy that relates to regulatory guidance that makes sure our learners are centered, not the labs are centered. So those are our three goals I'm giving. That's one group where we're trying to drive the conversation in the system. A second initiative we start is the Institute for Practical AI, where we brought together a handful of universities, you know, Stanford, Michigan, Minnesota, Georgetown and Adelaide University to try and walk through how can we practically assess the impacts of AI. And that practical impact assessment are, you know, pretty critical because we need to decide if we're using a tool, is it actually doing what we think it'll do, or are we at the end of the day just the target, a vendor target on the back of, of big tech. So that's, I think some examples of at least where I'm involved in what we're trying to do to get clarity around where and how can we innovate around the AI response. I think the startup ecosystem and the edtech ecosystem needs to be an active driver of ideas and capabilities because in many instances universities don't have that capability in house. Big Tech is right now and Frontier Labs are concerned with other multi billion dollar operations. And that means that the startup ecosystem is the one that's going to likely make the fastest imprint on universities for accelerating capabilities, increasing capabilities. So yes, I'll just briefly call out and I'll say those would be three domains where I would say, yes, Frontier Labs will shape us, but we ourselves have to dive in and start creating products, being AI builders as a university sector, commenting on the conversations happening broadly, and then of course relying on the edtech or the AI ed tech ecosystem and offerings from that ecosystem that kind of help shape and drive the kinds of work that we're doing broadly. So it's a way to say we need to. We don't know where the river is going, but we're sure not going to figure out by standing on the banks. We need to build to get in the canoe and see where the river's going. Because once you're in it, once you're in the water, this, the water raises, your canoe raises and you're now part of it. As Mark Carney has said, that, you know, if you're not at the table, you're on the menu. And I think there's a responsibility for universities to drive that, but to be very active partners with a number of organizations that serve the university sector and collectively, we at least have a chance of getting to a better answer than we do sort of on an individualized, disconnected way.
- [42:16]
Alex
That was a phenomenal answer. I did not realize that the AI officers from all these universities were coming together and influencing policy and thinking about how to, how to raise the awareness of what's needed in the university sector. That's incredible work. I would have loved to have been a fly on the wall in that meeting. It sounds like, just fascinating, especially at this moment where like, nobody, including the Frontier Labs truly knows where the river is headed, to use your metaphor. I mean, they don't know either. So it's wild to see how all of these very smart people right at the center of this are trying to work together and, and figure out where it's all going. That was a terrific answer because I think that probably is exactly where it's happening at the moment. And there's probably some governmental, different countries, governments around the world are discussing it in different ways, but that's really, really fascinating. I know we're almost at our time here. I can't let you go without talking about the MOOC world because I know we mentioned connectivism. The original paper was from 2005. Now we're going to flash forward to 2020 12, 2013, where you were a pioneer. You basically coined the term. I mean, you did, you coined the term massive open online courses, but very much in the context of connectivism, of people working together, of it being highly connected, activist. The movement of Coursera and edx and udacity@ the time was sort of branded as ex moocs, a sort of different branch of the MOOC world. I know this is maybe starting to feel like ancient history in ed tech world, which is crazy to think about, but I would love to hear your short take on that whole world and then sort of bring us up to the present. Where has that movement gone? What is the future of massive online learning, even if it doesn't maybe look exactly like the MOOCs of 2012?
- [43:51]
George Siemens
I think the part that I found most interesting in that process and where we are now, I think generally is that let's step back. Open education resources. You know, David Wiley and MIT and others were driving in that space. They taught us you can scale content without significantly additional input costs. You break the relationship between one piece and the cost. Right. Once it's digital, it accelerates. The second aspect that is I think key is that there's a lot of attention paid to the development of open online courses because it allowed us to scale instruction. And now you could have content and now you could have someone directly instructing someone and that would scale as well. Now with AI, just to bring that in, there's a capability that we can comparably scale parts of the engagement, assessment and tutoring of those learners. So in each of these innovations, they scale some part of the education system. But I will say AI is one scaling mechanism. But so social connectedness, like you can have a thousand people teaching one another different parts of it if you have the right configuration of social relationship. So you know, if you're on Twitter, for example, which remains probably the single best source for AI news and AI drama for that matter, you've got Reddit where you have large clusters of people engaging or teaching one another, the comments section, some YouTube spaces and so on. So I'm just trying to articulate what is it, you know, that we scale with our big education innovations. And I know scale isn't always a good thing. Sometimes it's way better to have a little boutique idea and like four people having a fantastic philosophical discussion. So scale isn't the ultimate goal and target, but we be given the economic reward infrastructure of the university model. We are scaling something with each of our innovations. And so initially it was a little bit like all these X MOOCs, they're not what we wanted. Stephen downside when we ran the first course, it was about connection and you know, taking RSS feed model of many ideas blooming and we just centralize the voices in a daily email or something to pull them together. I think what started to happen as I saw more people around the world taking the Coursera and the EDX moocs, you realize like, hey, it's changing their lives, it's giving them new opportunities they didn't have before. And that's. Then you've got to take Your ultimately, we're here for the learners, not for our own personal philosophical mantras that we like to adhere to too. And if it's really centered on the learners, when you see data and evidence that students are developing capabilities and skills even, it might not be the pedagogical model that I prefer, which would be higher social, higher connected, but if they're finding it through a technology layer where they didn't, they, you know, they could CS50, you know, out of Harvard, and now you've got students in other parts of the world taking it and they had no access to that before. I'd be an idiot to not give credit to the fact that that is giving people opportunities. Did they find a community in a social space for dialogue? Perhaps not, but they achieved a capability. So I would love to see multiple innovations of this type exist. I would love to see AI supporting and guiding and directing that same learning architecture as well, that that continues to advance. You know, at the end of the day, it's really about you let a thousand flowers bloom, but you only pot two or three and keep them in your house and maintain them and take care of them. In the AI conversation as it relates to MOOCs or any other educational innovation, we're in the thousand flowers blooming stage. In the near future, university leadership and sector level leadership is going to need to start deciding which ones they want to pot. But for the time being at least, that's roughly where we are.
- [47:24]
Alex
Yeah, that's a very enlightening answer. And it makes me wonder about the sort of complementarity of if OER scaled content. It was the first real mass movement to scale content. And then MOOC scaled instruction, but it tended to, you know, the original live MOOCs sort of had some of that connectivist tissue a little bit, but then as they went to on demand and asynchronous, it became a lot more scaling of centralized instruction, not as much connectivity. But then maybe, as you say, maybe those students, that student in Malaysia and the student in Madagascar who are both taking CS50, maybe the place they're connecting is on Reddit, right? Maybe they're connecting on a totally separate platform and they're asking each other questions and it's like, maybe it's the social media networks that scale the connectivity. And then as you say, AI, we don't know what it's going to scale yet. But I love that idea of scaling, tutoring, maybe scaling differentiation or personalization as we call it, or even, you know, knowledge. As you say, information transmutation is that, did I get that right?
- [48:19]
George Siemens
That's the one.
- [48:20]
Alex
Yeah. It's like the idea of like taking all of these sources. If you start adding these to your own personal knowledge architecture, look, I'm picking up all the vocabulary. If you start adding all of this information to your own personal architecture, then it helps the AI make sense of what you're trying to accomplish in the world. And you're adding, you know, world class knowledge and world class instruction to your own second brain. I love that idea of separating them. I've never thought of it that way. That's really fascinating. And I feel like maybe it's the pieces that maybe those are sort of the meta nodes of the future of instruction, of online instruction. It's really interesting.
- [48:53]
George Siemens
Yeah. And there's a final point. You know, we're wrapping up here, but final point, I would say as well that, you know, everything is something in relation to something else. Like few things stand purely on their own. So I might say, hey, it's really good. If you want to become aware of your community, if you want the, you know, a care economy and a sense of connectedness, you need to learn in social spaces, shared accountability to one another, you know, look people in the eye rather than yell at somebody through social media. I might say that, yeah. And I would prefer that, however, to use your example, you know, about, you know, a student in Madagascar, that conversation may not be accessible to them until they've, say, taken some baseline courses. And that may be an ex mooc that's async, but it levels them up. And I think as they level up, they become more capable of participating in what I might feel would be a more desirable pedagogy on a few fronts because I'm trying to drive that shared commitment to one another, that shared connected nature in our engagement. But anyone who's learning, I think it's a little arrogant to presume that it needs to be our way and on our condition. And anytime someone sits down and wants to make sense of their world and engages in any environment to do that, I mean, of course I'm going to sign up for it and I'm going to in the long run advocate for, hey, we need social, connected, engaged, collaborative, community oriented, wisdom, judgment, personal self regulation, personal control. I'm going to advocate for that. But you don't necessarily need to only be that or only start there.
- [50:19]
Alex
I like that a lot. They feed off of each other and the idea that, you know, maybe somebody in a forum trying to learn data science, unless they've gone and taken An Intro to Data Science course from Johns Hopkins on Coursera or from Harvard and edX. They don't know what the terminology means, they don't know what a skew is. And just being able to enter that conversation, that is worth its weight in gold. And previously that was only accessible through a university, often a university in a very developed country. So fascinating conversation. I feel like we could talk for hours about this. I'd love to go deeper. And boy, next time you have those AI officers come together, shoot us a message. I am so curious about the future of that. I think that sounds like the most cutting edge conversations you can, you can think of in AI and education, especially higher ed. Thank you so much for being here. We, we sometimes end the podcast with two questions. I'd love to hear your answers to them. This one could be a very open ended one, but it's probably something we've already talked about. What is the most exciting trend you see right now in the landscape and particularly we're looking for here? Something that might be around the corner. You've been doing so much with AI at a very cutting edge level. What's around the corner? What's something that people aren't yet talking enough about that you think is going to come into our edtech world in the next 6 to 12 months?
- [51:29]
George Siemens
My core argument is with our team is that what you're doing today digitally, you will likely be observing an agent doing in the next six to 12 months. And I think knowledge work will track what happened with software engineering as it relates to AI. Not 100% because software engineering is a bounded, structured, rule based domain with clear indicators of success or not at compile or at, you know, delivery. Knowledge works a little more complex. So I'd say that's one point I would say is that what you're doing now, you'll probably be observing an agent doing in the next 12 months at least, growing segments of it. But I do want to emphasize something I have already talked about. I don't hear it enough, but the single most important thing for people to be doing right now is to get their personal knowledge figured out. So I think have a space where your brain exists that you can point AI to in the future because that'll make you far more impactful in the kind of actions that you engage in. So that'd be two things that I would call out as we should pay more attention to that.
- [52:25]
Alex
Yeah, that's amazing. I've tried sometimes to point AI to all the transcripts of these podcasts where we've done over 500 at this point. And I'm like, well, this is my second brain. It's all the amazing people I've gotten to speak to over the last four years and, you know, still, still getting that in place. But it's been really interesting when, when you sort of have that as a corpus of knowledge and then ask questions, a lot comes out of it. It's really, it's, it's intriguing. Our last question is what is a resource you would recommend to somebody who wants to dive deeper into the topics discussed today? Are there papers, are there people on Twitter or, you know, out there that you think, oh, this person really knows what they're talking about that other people may not have heard of yet? However you want to take that question.
- [53:04]
George Siemens
One of the longest voices I've listened to is Canadian Stephen Downs. So I would say, you know, he has a very particular bent on a number of fronts, which is why it's always a good reader. I've definitely enjoyed. I'm sure you're well aware of this. And I think I can declare EdTech insiders another one that I definitely follow. You know, I'm finding Twitter from the ed tech space a little different than what it was, as I'm sure you've heard from others, you know, some of the other distributed media, whether it's, you know, Mastodon or the more alternative media with threads or Blue Sky, I haven't quite found the same vigorous discussion that we used to have through social media. So that's a little bit hard. But I'm seeing a few people like, you know, there's a, I think a colleague that Hollis Robbins, I've been finding some of, you know, her substack and some of the things that she's sharing interesting. And you know, and I think, but as a rule, my core argument is we're always, as the seasons change, so do the voices. And I'm always, you know, as if you're still listening to only the people you listen to five years or 10 or 15 years ago. You're missing a little bit of the. Because the wonderful part of it is, and this has been a definition of my career, I like to focus on something for about five years and then just drop it and start with a whole new network and a whole new community and hope, because otherwise you spend your time defending previous ideas.
- [54:18]
Alex
It's true.
- [54:18]
George Siemens
It reduces the aperture of curiosity and suddenly you're rehashing conversations. So I love having completely new ground zero conversations. And I'm seeing that with some people it seems like there's a little bit more momentum on some social media and share out, you know, and others that have started to do more, you know, Michael Horn and others that I'm sure you're well familiar with that share. So those are a few that I would say they're tracking what's happening and they're interesting. But I'm always looking for the person that is not in my network because you get an amplification effect when you or what others call structural holes and networks. You have a network component here, you add another network here. When you get that bridge that connects the two, it's like your whole knowledge system lights up.
- [54:57]
Alex
Totally fantastic recommendations. And yes, I'm always looking to fill those holes as well. It is, it is interesting how much, especially in this era, people are going so deep on all these different aspects of agentic architecture and of education and of health and there's so much we could learn from each other, but it's getting complicated and they're also separated. It feels like connecting them up is a fantastic idea. This has been an amazing conversation as I knew it would be. George Siemens is the Chief AI and Innovation Officer at Southern New Hampshire University snhu. He co founded their AI learning platform. He's also the pioneer of the Learning Theory connectivism for creating the first massive open online course and for just being an incredibly keen observer of the EdTech landscape for decades and the education and technology and learning landscape writ large. Thanks so much for being here with us on EdTech Insiders.
- [55:46]
George Siemens
Thanks Alex. Fantastic. I appreciate the work you're doing and a delight to spend time with you.
- [55:50]
Alex
Thanks for listening to this episode of EdTech Insiders. If you like the podcast, remember to rate it and share it with others in the EdTech community. For those who want even more EdTech Insider, subscribe to the free EdTech Insiders newsletter on substack. Tuck Advisors was founded by entrepreneurs who built and sold their own companies. Frustrated by other M and A firms, they created the one they wished they could have hired but couldn't. Find one who understands what matters to founders and whose North Star KPI is the percentage of deals closed. If you're thinking of selling your edtech company or buying one, contact Tuck Advisors now.