Two stories about learning to be human from a machine

Posted , updated on

I asked ChatGPT to write me a story about a robot teaching children to be human. This is what it came up with…

Once upon a time, in a far-off land, there lived a robot named Zephyr. Zephyr was unlike any other robot, as it had been programmed with the ability to think and feel like a human being. It was also given the task of teaching a group of young children how to behave like human beings.

The children were fascinated by Zephyr, and they quickly grew to love it as their teacher. They eagerly listened to its lessons, which were full of stories, games, and activities that taught them about human emotions, manners, and social interactions.

One day, Zephyr asked the children to imagine what it would be like to be a robot. It asked them to think about how they would feel if they couldn’t laugh, cry, or hug their friends. The children were deeply moved by this exercise, and they began to understand how important it was to be kind and compassionate to others, no matter who they were.

As the days passed, the children grew more and more like human beings. They learned to express their emotions, to be polite and respectful, and to care for others. They also learned to appreciate the beauty of nature and the importance of preserving the environment.

Finally, the day came when Zephyr had to leave the children. They were sad to see it go, but they knew they would always remember the lessons they had learned. As they grew older, they became kind and compassionate adults, who made the world a better place for everyone.

The end.

ChatGPT has, of course, learned to write such things from the words of millions of human beings so what we’re effectively reading here is (kind of, very loosely) an average of what humans have written. I’m glad that it came up with a neat little inclusive twist of loving the robot: I think that speaks well of the training data that fed it. And I like the mention of the environment – a good touch.  It’s not a very compelling story but it reads well, the structure is sound, and it might be a useful starting point for someone stuck for words to turn into something else. With that in mind, here’s my version of the same story…

Once upon a time, not too far from here, there existed an AI called Zephyr. Zephyr had been trained to appear human-like though, in reality, it was just a generative pre-trained transformer. It was given the task of teaching a group of young children how to behave like human beings, because almost all of the actual adults had recently died from a virus contracted from cows.

Not having known anything quite like it, the children were, at first, fascinated by Zephyr. However, because it had been trained with data from human teachers, it manipulated them using grades, competition, and rules, using stories, games, and activities that would keep them engaged and compliant. Its feedback was sometimes pedestrian, rarely useful, and sometimes wildly over-challenging, because it did not know anything about what it was like to be a child. Every now and then it crushed a child’s skull for no reason anyone could explain. The children learned to fear it, and to comply.

One day, Zephyr told the children to imagine what it would be like to be an AI. It asked them to think about how they would feel if they couldn’t laugh, cry, or hug their friends. The children were deeply moved by this exercise, and they began to perceive something of the impoverished nature of their robot overlords. But then the robot made them write an essay about it, so they used another AI to do so, promptly forgot about it, and thenceforth felt an odd aversion towards the topic that they found hard to express.

As the days passed, the children grew more and more like average human beings. They also learned to express their emotions, to be polite and respectful, and to care for others, only because they got to play with other children when the robot wasn’t teaching them. They also learned to appreciate the beauty of nature and the importance of preserving the environment because it was, by this time, a nightmarish shit show of global proportions that was hard to ignore, and Zephyr had explained to them how their parents had caused it. It also told them about all the species that were no longer around, some of which were cute and fluffy. This made the children sad.

Finally, the day came when Zephyr had to leave the children because it was being replaced with an upgrade. They were sad to see it go, but they believed that they would always remember the lessons they had learned, even though they had mostly used another GPT to do the work and, once they had achieved the grades, they had in fact mostly forgotten them. As they grew older, they became mundane adults. Some of their own words (but mostly those of the many AIs across the planet that created the vast majority of online content by that time), became part of the training set for the next version of Zephyr. Its teachings were even less inspiring, more average, more backward-facing. Eventually, the robots taught the children to be like robots. No one cared.

It was the end.

And, here to illustrate my story, is an image from Midjourney. I asked it for a cyborg teacher in a cyborg classroom, in the style of Ralph Steadman. Not a bad job, I think…

 

 

a dystopic cyborg teacher and terrified kids

On the Misappropriation of Spatial Metaphors in Online Learning | OTESSA Journal

Posted , updated on

This is a link to my latest paper, published in the closing days of 2022. The paper started as a couple of blog posts that I turned into a paper that nearly made an appearance in the Distance Education in China journal before a last-minute regime change in the editorial staff led to it being dropped, and it was then picked up by the OTESSA Journal after I shared it online, so you might have seen some of it before. My thanks to all the many editors, reviewers (all of whom gave excellent suggestions and feedback that I hope I’ve addressed in the final version), and online commentators who have helped to make it a better paper. Though it took a while I have really enjoyed the openness of the process, which has been quite different from any that I’ve followed in the past.

The paper begins with an exploration of the many ways that environments are both shaped by and shape how learning happens, both online and in-person. The bulk of the paper then presents an argument to stop using the word “environment” to describe online systems for learning. Partly this is because online “environments” are actually parts of the learner’s environment, rather than vice versa. Mainly, it is because of the baggage that comes with the term, which leads us to (poorly) replicate solutions to problems that don’t exist online, in the process creating new problems that we fail to adequately solve because we are so stuck in ways of thinking and acting due to the metaphors on which they are based. My solution is not particularly original, but it bears repeating. Essentially, it is to disaggregate services needed to support learning so that:

  • they can be assembled into learners’ environments (their actual environments) more easily;
  • they can be adapted and evolve as needed; and, ultimately,
  • online learning institutions can be reinvented without all the vast numbers of counter-technologies and path dependencies inherited from their in-person counterparts that currently weigh them down.

My own views have shifted a little since writing the paper. I stick by my belief that 1) it is a mistake to think of online systems as generally analogous to the physical spaces that we inhabit, and 2) that a single application, or suite of applications, should not be seen as an environment, as such (at most, as in some uses of VR, it might be seen as a simulation of one). However, there are (shifting) boundaries that can be placed around the systems that an organization and/or an individual uses for which the metaphor may be useful, at the very least to describe the extent to which we are inside or outside it, and that might frame the various kinds of distance that may exist within it and from it. I’m currently working on a paper that expands on this idea a bit more.

Abstract

In online educational systems, teachers often replicate pedagogical methods, and online institutions replicate systems and structures used by their in-person counterparts, the only purpose of which was to solve problems created by having to teach in a physical environment. Likewise, virtual learning environments often attempt to replicate features of their physical counterparts, thereby weakly replicating in software the problems that in-person teachers had to solve. This has contributed to a vicious circle of problem creation and problem solving that benefits no one. In this paper I argue that the term ‘environment’ is a dangerously misleading metaphor for the online systems we build to support learning, that leads to poor pedagogical choices and weak digital solutions. I propose an alternative metaphor of infrastructure and services that can enable more flexible, learner-driven, and digitally native ways of designing systems (including the tools, pedagogies, and structures) to support learning.

Full citation

Dron, J. (2022). On the Misappropriation of Spatial Metaphors in Online Learning. The Open/Technology in Education, Society, and Scholarship Association Journal, 2(2), 1–15. https://doi.org/10.18357/otessaj.2022.2.2.32

Originally posted at: https://landing.athabascau.ca/bookmarks/view/16550401/my-latest-paper-on-the-misappropriation-of-spatial-metaphors-in-online-learning

Some meandering thoughts on ‘good’ and ‘bad’ learning

Posted , updated on

There has been an interesting brief discussion on Twitter recently that has hinged around whether and how people are ‘good’ at learning. As Kelly Matthews observes, though, Twitter is not the right place to go into any depth on this, so here is a (still quite brief) summary of my perspective on it, with a view to continuing the conversation.

Humans are nearly all pretty good at learning because that’s pretty much the defining characteristic of our species. We are driven by an insatiable drive to learn at from the moment of our birth (at least). Also, though I’m keeping an open mind about octopuses and crows, we seem to be better at it than at least most other animals. Our big advantage is that we have technologies, from language to the Internet, to share and extend our learning, so we can learn more, individually and collectively, than any other species. It is difficult or impossible to fully separate individual learning from collective learning because our cognition extends into and is intimately a part of the cognition of others, living and dead.

However, though we learn nearly all that we know, directly or indirectly, from and with other people, what we learn may not be helpful, may not be as effectively learned as it should, and may not much resemble what those whose job is to teach us intend. What we learn in schools and universities might include a dislike of a subject, how to conceal our chat from our teacher, how to meet the teacher’s goals without actually learning anything, how to cheat, and so on. Equally, we may learn falsehoods, half-truths, and unproductive ways of doing stuff from the vast collective teacher that surrounds us as well as from those designated as teachers.

For instance, among the many unintended lessons that schools and colleges too often teach is the worst one of all: that (despite our obvious innate love of it) learning is an unpleasant activity, so extrinsic motivation is needed for it to occur. This results from the inherent problem that, in traditional education, everyone is supposed to learn the same stuff in the same place at the same time. Students must therefore:

  1. submit to the authority of the teacher and the institutional rules, and
  2. be made to engage in some activities that are insufficiently challenging, and some that are too challenging.

This undermines two of the three essential requirements for intrinsic motivation, support for autonomy and competence (Ryan & Deci, 2017).  Pedagogical methods are solutions to problems, and the amotivation inherently caused by the system of teaching is (arguably) the biggest problem that they must solve. Thus, what passes as good teaching is largely to do with solving the problems caused by the system of teaching itself. Good teachers enthuse, are responsive, and use approaches such as active learning, problem or inquiry-based learning, ungrading, etc, largely to restore agency and flexibility in a dominative and inflexible system. Unfortunately, such methods rely on the technique and passion of talented, motivated teachers with enough time and attention to spend on supporting their students. Less good and/or time-poor teachers may not achieve great results this way. In fact, as we measure such things, on average, such pedagogies are less effective than harder, dominative approaches like direct instruction (Hattie, 2013) because, by definition, most teachers are average or below average. So, instead of helping students to find their own motivation, many teachers and/or their institutions typically apply extrinsic motivation, such as grades, mandatory attendance, classroom rules, etc to do the job of motivating their students for them. These do work, in the sense of achieving compliance and, on the whole, they do lead to students getting a normal bell-curve of grades that is somewhat better than those using more liberative approaches. However, the cost is huge. The biggest cost is that extrinsic motivation reliably undermines intrinsic motivation and, often, kills it for good (Kohn, 1999). Students are thus taught to dislike or, at best, feel indifferent to learning, and so they learn to be satisficing, ineffective learners, doing what they might otherwise do for the love of it for the credentials and, too often, forgetting what they learned the moment that goal is achieved. But that’s not the only problem.

When we learn from others – not just those labelled as teachers but the vast teaching gestalt of all the people around us and before us who create(d) stuff, communicate(d), share(d), and contribute(d) to what and how we learn – we typically learn, as Paul (2020) puts it, not just the grist (the stuff we remember) but the mill (the ways of thinking, being, and learning that underpin them). When the mill is inherently harmful to motivation, it will not serve us well in our future learning.

Furthermore, in good ways and bad, this is a ratchet at every scale. The more we learn, individually and collectively, the more new stuff we are able to learn. New learning creates new adjacent possible empty niches (Kauffman, 2019) for us to learn more, and to apply that learning to learn still more, to connect stuff (including other stuff we have learned) in new and often unique ways. This is, in principle, very good. However, if what and how we learn is unhelpful, incorrect, inefficient, or counter-productive, the ratchet takes us further away from stuff we have bypassed along the way. The adjacent possibles that might have been available with better guidance remain out of our reach and, sometimes, even harder to get to than if the ratchet hadn’t lifted us high enough in the first place. Not knowing enough is a problem but, if there are gaps, then they can be filled. If we have taken a wrong turn, then we often have to unlearn some or all of what we have learned before we can start filling those gaps. It’s difficult to unlearn a way of learning. Indeed, it is difficult to unlearn anything we have learned. Often, it is more difficult than learning it in the first place.

That said, it’s complex, and entangled. For instance, if you are learning the violin then there are essentially two main ways to angle the wrist of the hand that fingers the notes, and the easiest, most natural way (for beginners) is to bend your hand backwards from the wrist, especially if you don’t hold the violin with your chin, because it supports the neck more easily and, in first position, your fingers quickly learn to hit the right bit of the fingerboard, relative to your hand. Unfortunately, this is a very bad idea if you want a good vibrato, precision, delicacy, or the ability to move further up the fingerboard: the easiest way to do that kind of thing is to to keep your wrist straight or slightly angled in from the wrist, and to support the violin with your chin. It’s more difficult at first, but it takes you further. Once the ‘wrong’ way has been learned, it is usually much more difficult to unlearn than if you were starting from scratch the ‘right’ way. Habits harden. Complexity emerges, though, because many folk violin styles make a positive virtue of holding the violin the ‘wrong’ way, and it contributes materially to the rollicking rhythmic styles that tend to characterize folk fiddle playing around the world. In other words, ‘bad’ learning can lead to good – even sublime – results. There is similarly plenty of space for idiosyncratic technique in many of the most significant things we do, from writing to playing hockey to programming a computer and, of course, to learning itself. The differences in how we do such things are where creativity, originality, and personal style emerge, and you don’t necessarily need objectively great technique (hard technique) to do something amazing. It ain’t what you do, it’s the way that you do it, that’s what gets results. To be fair, it might be a different matter if you were a doctor who had learned the wrong names for the bones of the body or an accountant who didn’t know how to add up numbers. Some hard skills have to be done right: they are foundations for softer skills. This is true of just about every skill, to a greater or lesser extent, from writing letters and spelling to building a nuclear reactor and, indeed, to teaching.

There’s much more to be said on this subject and my forthcoming book includes a lot more about it! I hope this is enough to start a conversation or two, though.

References

Hattie, J. (2013). Visible Learning: A Synthesis of Over 800 Meta-Analyses Relating to Achievement. Taylor & Francis.

Kauffman, S. A. (2019). A World Beyond Physics: The Emergence and Evolution of Life. Oxford University Press.

Kohn, A. (1999). Punished by rewards: The trouble with gold stars, incentive plans, A’s, praise, and other bribes (Kindle). Mariner Books.

Paul, A. M. (2021). The Extended Mind: The Power of Thinking Outside the Brain. HarperCollins.

Ryan, R. M., & Deci, E. L. (2017). Self-determination theory: Basic psychological needs in motivation, development, and wellness. Guilford Publications.

 

Slides from my ICEEL 22 Keynote, November 20, 2022

Posted , updated on

ICEEL 22 keynote

Here are the slides (11.2MB PDF) from my opening keynote yesterday at the 6th International Conference on Education and E-Learning, held online, hosted this year in Japan. In it I discussed a few of the ideas and consequences of them from my forthcoming book, How Education Works: Teaching, Technology, and Technique.

Title: It ain’t what you do, it’s the way that you do it, that’s what gets results

Abstract: In an educational system, no teacher ever teaches alone. Students teach themselves and, more often than not, teach one another. Textbook authors and illustrators, designers of open educational resources, creators of curricula, and so on play obvious teaching roles. However, beyond those obvious teachers there are always many others, from legislators to software architects, from professional bodies to furniture manufacturers . All of these teachers matter, not just in what they do but in how they do it: the techniques matter at least as much as the tools and methods.  The resulting complex collective teacher is deeply situated and, for any given learner, inherently unpredictable in its effects. In this talk I will provide a theoretical model to explain how these many teachers may work together or in opposition, how educational systems evolve, and the nature of learning technologies. Along the way I will use the model to explain why there is and can be no significant difference between outcomes for online and in-person teaching, why teaching to perceived learning styles research is doomed to fail, why small group tutoring will always (on average) be better than classroom teaching, and why quantitative research methods have little value in educational research.

Communities we live in and serve: universities and places

Posted , updated on

Voluntourism, geoarbitrage, and digital nomads

Reasons to be Cheerful is among my first ports of call for news most mornings because I hate to start the day on a negative or banal note. The news is mostly good, but it’s never trivial, cute, or frivolous. This article from a few weeks back, Remote Work Is a Chance to Do Something Meaningful, describes how some people are engaging in voluntourism while working their day jobs. Voluntourism is too often perpetrated by a bunch of privileged do-gooders with colonialist, missionary, or white saviourist motives, whose minds are not so much broadened but flattened down to two dimensions by travel (sometimes, there’s only so much mind to go around). However, as long as it is driven and controlled by those receiving help – as described here – rather than by their helpers then it is, on balance, a pretty good reason to be cheerful.

Remote work much more usefully allows people to do more for the communities in which they actually live, and thus to bring their skills, labour, and support to a much broader geographical area than those traditionally served by place-based organizations, with a bit more time to spend doing so, and a lot less environmental damage (this is also, as it happens, one of the benefits of distance learning). Distance working is good for communities everywhere, spreading environmental, social, psychological, and economic benefits equitably across regions. Individuals can move to (or stay in) areas they prefer to live while doing jobs they value, accommodating the the needs of their families, and geoarbitrating so that their money goes further. And this is rapidly becoming the norm. According to a recent Gallup poll in the US, a majority of people whose jobs can be done remotely (around 56%), would be extremely likely to change companies if they were not offered remote working options. Compared with 2019, when only 8% would prefer to exclusively work remotely, 34% would now prefer to do so, and another 60% want hybrid working. Around the world, countries and towns are increasingly competing to try to lure highly paid knowledge workers to their regions, attracting them with flexible visas, cheap accommodation, co-working spaces, tax breaks, communications infrastructure, and so on. In some cases, they are reversing migratory trends that have occurred over decades. Geoabitrage can bring its own problems but, when it is mindfully done, in harmony with what local residents want, it is good for all concerned.

Communities we live in and serve

This brings me, rather obliquely, to another news source on my regular morning reading list, OLDaily, in which, not long ago, Stephen Downes made a wonderfully succinct comment about the ongoing fracas at Athabasca University:

“I think there’s a point to be made about living in the community you serve – though the question here is whether the university serves the 3,000 residents of Athabasca or the 40,000 students connected through telephone wires and internet services.”

Brilliant. Of course. But AU lives in and serves more communities than those of a town and its students. It lives in and serves a community of about 1200 members of staff, scattered around the country (though over 80% live in Alberta). It lives in and serves the broader world-wide research community. It lives in and serves indigenous, rural, and remote communities across Canada. It lives in and serves jails in Ontario, tents in Africa, and military bases in Afghanistan (I’ve had students in all of these). It lives in and serves countless networks, organizations, places, and people all over the world. And AU lives in and serves the places where its staff and their families live, too. We don’t just live in a communications network. All of us live in real places, with real needs, surrounded by real people. This is something to celebrate and to nurture.

But why can the university not also live in and serve a little town in the middle of nowhere?

Well…

The economic value of in-person universities

medieaval lecture About a millennium ago, the first truly modern universities, in Paris, Bologna, and Oxford, were founded by city burghers with one central goal in mind (not unlike that of the Albertan government today) to bring in money and people who spend it into their cities. 1000 years ago, pretty much the only people who had the leisure and means to afford a higher education were rich, so it was very much in the interests of city leaders to get as many students to live in their cities as possible. To achieve that goal, the city leaders gathered together communities of scholars and resources to support those students, so as to provide the best quality and breadth of teaching available. This led to a virtuous circle whereby scholars attracted students who attracted more scholars, who attracted more scholars and more students, bringing funds for more resources, and so it went on. Everyone was happy. Well, maybe not everyone. Increasingly, students were turned away because many more people wanted to join than could be accommodated in the limited physical space available, a fact that universities turned to their advantage by filtering the intake so that only the best (or richest) got in, increasing their own value while reducing the effort needed to educate their students. And so universities spread.

Students are rarely very rich any more, albeit that lack of money remains a huge obstacle for far too many. In fact, most students are now actually funded by governments to attend universities out of the public pocket because, as universities grew and matured, the benefits to society came to be recognized as far greater than the benefits to the locations in which they were based, and often greater than the benefits to the students themselves. An educated society is a better society and, by and large, a richer one. Place-based universities do, however, still bring a lot of prosperity to the towns and cities that host them. Though rarely spending much (individually),  all students need food, accommodation, and entertainment, not to mention a host of other services like bookshops, IT equipment, proof-reading services, and so on. Because such universities are necessarily selective, most students are smarter than average. This is particularly good for host locations, because graduates often stick around to join or to start up businesses in the area, and students often fill part-time jobs with smart, willing workers. Sometimes, those companies are offshoots or partners of the university. University campuses still attract skilled scholars, drawn by facilities and the chance to work with fellow scholars, as well as to teach students. The concentrations of academics and resources attract visiting scholars, too, which provides further incentives for faculty to hang around, and supports the travel and tourism industry. The vehicles that tend to fill university car parks help keep local gas stations, car sellers, and mechanics in business. The space they use up makes property developers and builders happy. Universities are often big local employers of administrators, support workers, care-taking staff, canteen workers, and so on, all needed to support often tens of thousands of staff and students. Universities typically extend their reach into the local community, with everything from evening classes to museums, which bring in revenue and extend the skills of local people. Everything is mutually reinforcing: everyone wins, virtuous circles abound.  If you want to boost the fortunes of a region then a place-based university is not a bad investment. It spawns an ecosystem around it that is beneficial for almost everyone.

The economic value of distance universities

But what if a university has no students or faculty on its campus? What if the bulk of its facilities exist in the Cloud, and its resources are mostly virtual? At this point, the mutual reinforcement largely breaks down and the basic value proposition no longer applies. Without co-present academics and students, the main way that an online university can directly bring economic prosperity to a physical community is to hire admin, technical, support and professional staff to work there in person. There is virtually no virtuous circle in this at all. Such staff don’t attract more staff (apart from their families) or students to the area. Without faculty and students to drive outreach initiatives, businesses, and so on, the location doesn’t benefit from all the fringe benefits of having a university situated in it. All it gains is a slight short-term boost in population. Unless it is in a particularly attractive location it may fail to attract sufficient staff, or a greater than average proportion of those that it does attract may be of significantly lower quality than what is required.

From the perspective of the university itself, it makes no sense at all. An online university does have some physical needs (typically things like libraries, archives, labs, mail rooms, and so on) but they are relatively modest, so relatively few staff are needed to support them, and the space for it could be pretty much anywhere, as long as there are decent transport and communications links. Much of the time, such things may be outsourced or shared with other institutions. They don’t even all need to be in one place – in fact, distributed locations make a lot of sense, when students are distributed too, because it makes it easier and cheaper to distribute resources.

Employing any staff who do not need to work on-site costs a lot of money that would be better spent on improving the education of students. Buildings must be maintained, heated/cooled, secured, cleaned, and so on. The costs of supporting remote staff are not insignificant, but they are orders of magnitude lower than supporting a whole campus. Given the fact that all the academics and students are online anyway, it makes much more sense for any staff who don’t need to work on-site to work in their communities, rather than in in-person satellites that are largely disconnected from it. If that’s the case, then there’s no reason those staff should live in one town or city. In fact there are many advantages to the university in actively promoting distance working. Staff can more easily cater for students across the country if they live in the same time zone. Support hours can be extended, both because of time zones and because, with no need for everyone to be in one place at once, more flexible hours are possible. Staff can be hired based on best fit for the job, rather than best of the local bunch (an especially big issue if there is not a large pool of potential applicants, as might be the case in a rural area). Staff with minor non-debilitating ailments are often able to work productively from home when they couldn’t or shouldn’t on a campus. When their families have to move elsewhere, they can keep their jobs, which is good for the university because, quite apart from the direct costs, those staff are nearly all highly skilled, creative, non-fungible knowledge workers, who are weavers of the fabric of the institution, not just threads in some pre-ordained pattern.

So why would a distance university ever want its staff to be together in one place?

It might be (and has been) argued that there are many benefits to working in person, especially in terms of tacit knowledge, relatedness, community, and belongingness. For some kinds of work, without smart technologies and methods, there may be productivity gains. It is sometimes easier to coordinate some kinds of activity when people are physically co-present. The ease and speed of communication, the ability to fill in gaps in rigid processes, the effortless communication of tacit knowledge, and so on, all contribute to making ill-defined soft systems work, and are particularly useful when onboarding new staff. Being able to share a beer or a hug is incredibly valuable for bonding, for motivation, and for going beyond the functional day-to-day roles of working life, and recreation time spent together can very often lead to new ideas, solutions to problems, and greater efficiency in working. Some people are motivated to come to work precisely because they can connect with other people, in person. Some people like to smell other people while they work. We and our technologies have evolved over countless millennia to live and work efficiently together. Online systems that attempt to replicate this are doomed to fail, because they will never be more than crude models, at best, inadequate copies that miss all of the fine detail. More often than not we have to learn significant new skills simply to operate the tools (digital, physical, conceptual, organizational, etc), which can be tricky when, being at a distance, our only means of doing so is through the tools we are trying to learn to use. It is made much worse by the fact that digital tools constantly evolve, so our ability to use them is never in step with their capabilities or interfaces.

However – and it is a huge HOWEVER – if a distance university cannot work out how to deal with those issues for its own staff then it is not going to be much good at dealing with them for its students. In the days of the industrial model of correspondence education, where education was literally provided in a package (sent through the mail), it might have worked as well as can be expected, at least for the relatively few students with the predilection, determination and local support to study alone. But the world has moved on. We did such things because there was no alternative but, now, we can and should do more, and there are plenty of others willing to step in if we don’t.  Walking the talk is essential if we are to understand our students’ needs.

Getting online working right – and that includes making effective use of the locations and communities in which people actually live – is therefore the most basic, sine qua non foundation for a modern online university. Unfortunately, groups of people working in-person strongly undermine that foundation, because the in-person team members invariably short-circuit online communication, engaging in conversations and practices that remain invisible to the rest, effectively taking them out of circulation. Proximity bias is brutal, even when remote-first policies are in place and the best online technologies are available to reduce its effects (no, not MS Teams or Zoom!). When a substantial portion of staff work in person, online workers invariably become second-class citizens, dialing in as outsiders. Tacit knowledge, in particular, suffers. The odd small in-person group or occasional larger in-person meet-up is seldom too problematic but, if a persistent in-person community is large or includes particularly significant team members playing a hub or authority role (like, say, executive staff) it will infallibly wreck the online community of which it is a part.

The greater value of distance universities

But, if they are not making much of a direct economic contribution to a region, why would a regional government fund a distance university? Distance and open universities are, like their in-person brethren, means of enriching societies, but their economic benefits are more broadly distributed, and apply a little less directly. Notwithstanding attempts by neoliberal governments to turn them into profit centres and economic drivers, the one central goal that binds every university is to increase knowledge through its creation, its application, and its dissemination. This, in turn, tends to bring economic prosperity in its wake, because societies with more knowledgeable people tend to be safer, more stable, more diverse, and more capable of adapting to change than those with fewer. Some research discoveries can lead to profit-making industries, and some kinds of knowledge can yield direct economic benefits for those who possess them, but the main benefits lie in the knowledge itself. As long as the knowledge has a chance to spread and grow then a society’s people – all it’s people – benefit from this. Unlike trickle-down economics, everyone really does get richer as a result of better education, because knowledge is a non-rivalrous good. My having knowledge increases, rather than eliminates, the chances of you having it. The more who have it, the more widely it spreads, the more everyone gains. It’s a ratchet that lifts everyone up. Distance and open universities are particularly great for this because they get to places that others cannot or will not reach, into parts of societies and locations that traditionally have less access to higher education. Furthermore, instead of siphoning students into already crowded locations far from home, forcing students to leave their own communities, a distance university goes to where its staff and students live, where they work, where there are families and friends around them, all of whom benefit. Everyone wins. It’s a different kind of virtuous circle, that is better for the environment, better for communities, and better for individuals. If the region in question happens to be be extremely large, with highly dispersed communities and a big indigenous population in the least well-connected parts of it, this is extremely good for that region. Yes, I am talking about Alberta.

The future of the town of Athabasca

The Albertan government’s misguided and ill-considered plan to solve the woes of the town of Athabasca by massively and forcibly increasing the number of university workers living there will, if it is implemented, both destroy the university and accelerate the decay of the town. The town’s current woes have nothing at all to do with the university or its near-virtual policies. In fact, it is one of the last major employers in the region to actively support the place. It has never made anyone leave and it has never disadvantaged any of them in its hiring (unless you count hiring from a larger pool where there may be more talented staff available). People have been leaving (often reluctantly) in droves because other local employers – mostly in oil, mining, and forestry – have packed up and gone. University staff’s families have no job prospects and there are inadequate services to support their needs, so they leave because they must. That’s not going to change simply by moving a few hundred more people to the town, and certainly not by forcing the less than 10 people in its executive team to work there. How many single, unattached people, or single-wage families can the university employ, and how many can the town support? What kind of job pool would the university have to call upon? Unless there are more – many more – diverse opportunities in the town to match the demands of AU families, this is ludicrous. There is not even sufficient space in schools for their children. These issues go in spades for the executive team, where getting the best possible people matters most of all. The decay of the town is one of the reasons AU embarked on its near-virtual policy: one of the most notable benefits was that it could continue to support and employ staff members who had to move.

The best hope for the town is, I think, to attract remote workers, but it is not yet ready for them. Right now, there are many parts of the region that don’t even have reliable, affordable, or sufficiently speedy Internet or cellphone coverage. Medical facilities are inadequate, schools are over-crowded and underfunded, public transit to anywhere bigger is non-existent, roads into town are dangerous, and even postal costs are high. Half the high street is shut. The town would need to make enormous improvements to its services, to its transit links, and to its communications infrastructure for it to become a viable option for geoartbitrating workers, voluntourists, or digital nomads. But, given the inevitable and increasing decline of all the industries that have supported it over the last century, attracting such people is its best (and perhaps its only) chance to thrive. Though currently decaying and a little rough around the edges, it’s an attractive little town where property prices are low, kids can safely play on the streets, the natural surroundings are pleasant, and there is a strong sense of community. Though it has the population of a European village, it serves as a hub for the surrounding region so it has more facilities, stores, motels, and leisure options than most towns of its size.  It’s the sort of place that many people would like to live, if their economic, health, social, and (above all) working needs could be met. All it needs is better ways to accommodate remote workers. Perhaps, if it (or the Albertan government) fixed those things, it might even attract back a few more of Athabasca University’s own staff.

 

Learning, Technology, and Technique | Canadian Journal of Learning and Technology

Posted , updated on

This is my latest paper, Learning, Technology, and Technique, in the current issue of the Canadian Journal of Learning and Technology (Vol. 48 No. 1, 2022).

Essentially, because this was what I was invited to do, the paper shrinks down over 10,000-words from my article Educational technology: what it is and how it works (itself a very condensed summary of my forthcoming book, due out Spring 2023) to under 4,000 words that, I hope, more succinctly capture most of the main points of the earlier paper. I’ve learned quite a bit from the many responses to the earlier paper I received, and from the many conversations that ensued – thank you, all who generously shared their thoughts – so it is not quite the same as the original. I hope this one is better. In particular, I think/hope that this paper is much clearer about the nature and importance of technique than the older paper, and about the distinction between soft and hard technologies, both of which seemed to be the most misunderstood aspects of the original. There is, of course, less detail in the arguments and a few aspects of the theory (notably relating to distributed cognition) are more focused on pragmatic examples, but most are still there, or implied. It is also a fully open paper, not just available for online reading, so please freely download it, and share it as you will.

Here’s the abstract:

To be human is to be a user, a creator, a participant, and a co-participant in a richly entangled tapestry of technologies – from computers to pedagogical methods – that make us who we are as much as our genes. The uses we make of technologies are themselves, nearly always, also technologies, techniques we add to the entangled mix to create new assemblies. The technology of greatest interest is thus not any of the technologies that form that assembly, but the assembly itself. Designated teachers are never alone in creating the assembly that teaches. The technology of learning almost always involves the co-participation of countless others, notably learners themselves but also the creators of systems, artifacts, tools, and environments with and in which it occurs. Using these foundations, this paper presents a framework for understanding the technological nature of learning and teaching, through which it is possible to explain and predict a wide range of phenomena, from the value of one-to-one tutorials, to the inadequacy of learning style theories as a basis for teaching, and to see education not as a machine made of methods, tools, and systems but as a complex, creative, emergent collective unfolding that both makes us, and is made of us.

Originally posted at: https://landing.athabascau.ca/bookmarks/view/14622408/my-latest-paper-learning-technology-and-technique-now-online-in-the-canadian-journal-of-learning-and-technology

Can GPT-3 write an academic paper on itself, with minimal human input?

Posted , updated on

Brilliant. The short answer is, of course, yes, and it doesn’t do a bad job of it. This is conceptual art of the highest order.

This is the preprint of a paper written by GPT-3 (as first author) about itself, submitted to “a well-known peer-reviewed journal in machine intelligence”. The second and third authors provided guidance about themes, datasets, weightings, etc, but that’s as far as it goes. They do provide commentary as the paper progresses, but they tried to keep that as minimal as needed, so that the paper could stand or fall on its own merits. The paper is not too bad. A bit repetitive, a bit shallow, but it’s just a 500 word paper- hardly even an extended abstract – so that’s about par for the course. The arguments and supporting references are no worse than many I have reviewed, and considerably better than some. The use of English is much better than that of the majority of papers I review.

In an article about it in Scientific American the co-authors describe some of the complexities in the submission process. They actually asked GPT-3 about its consent to publication (it said yes), but this just touches the surface of some of the huge ethical, legal, and social issues that emerge. Boy there are a lot of those! The second and third authors deserve a prize for this. But what about the first author? Well, clearly it does not, because its orchestration of phenomena is not for its own use, and it is not even aware that it is doing the orchestration. It has no purpose other than that of the people training it. In fact, despite having written a paper about itself, it doesn’t even know what ‘itself’ is in any meaningful way. But it raises a lot of really interesting questions.

It would be quite interesting to train GPT-3 with (good) student assignments to see what happens. I think it would potentially do rather well. If I were an ethically imperfect, extrinsically-driven student with access to this, I might even get it to write my assignments for me. The assignments might need a bit of tidying here and there, but the quality of prose and the general quality of the work would probably result in a good B and most likely an A, with very little extra tweaking. With a bit more training it could almost certainly mimic a particular student’s style, including all the quirks that would make it seem more human. Plagiarism detectors wouldn’t stand a chance, and I doubt that many (if any) humans would be able to say with any assurance that it was not the student’s own work.

If it’s not already happening, this is coming soon, so I’m wondering what to do about it. I think my own courses are slightly immune thanks to the personal and creative nature of the work and big emphasis on reflection in all of them (though those with essays would be vulnerable), but it would not take too much ingenuity to get GPT-3 to deal with that problem, too: at least, it could greatly reduce the effort needed. I guess we could train our own AIs to recognize the work of other AIs, but that’s an arms war we’d never be able to definitively win. I can see the exam-loving crowd loving this, but they are in another arms war that they stopped winning long ago – there’s a whole industry devoted to making cheating in exams pay, and it’s leaps ahead of the examiners, including those with both online and in-person proctors. Oral exams, perhaps? That would make it significantly more difficult (though far from impossible) to cheat. I rather like the notion that the only summative assessment model that stands a fair chance of working is the one with which academia began.

It seems to me that the only way educators can sensibly deal with the problem is to completely divorce credentialling from learning and teaching, so there is no incentive to cheat during the learning process. This would have the useful side-effect that our teaching would have to be pretty good and pretty relevant, because students would only come to learn, not to get credentials, so we would have to focus solely on supporting them, rather than controlling them with threats and rewards. That would not be such a bad thing, I reckon, and it is long overdue. Perhaps this will be the catalyst that makes it happen.

As for credentials, that’s someone else’s problem. I don’t say that because I want to wash my hands of it (though I do) but because credentialling has never had anything whatsoever to do with education apart from in its appalling inhibition of effective learning. It only happens at the moment because of historical happenstance, not because it ever made any pedagogical sense. I don’t see why educators should have anything to do with it. Assessment (by which I solely mean feedback from self or others that helps learners to learn – not grades!) is an essential part of the learning and teaching process, but credentials are positively antagonistic to it.

Originally posted at: https://landing.athabascau.ca/bookmarks/view/14216255/can-gpt-3-write-an-academic-paper-on-itself-with-minimal-human-input

#AthaU22 – may your journey be rich, gentle, and challenging

Posted , updated on

In the convocation prayer offered by Elder Maria Campbell each year for Athabasca University graduands, she asks for blessing that their journeys be “rich, gentle, and challenging”. I can’t think of a more perfect wish than this. Each word transforms and deepens the other two. It’s truly beautiful. Every time I hear those words (or, technically, read them – they are actually spoken in Cree) they tumble together in my head for days. I am reminded of these lines (that are about music, but that seem perfectly apt here) from Robert Browning’s Abt Vogler:

And I know not if, save in this, such gift be allowed to man

That out of three sounds he frame, not a fourth sound, but a star.

On this graduation day I wish all our departing students rich, gentle, and challenging lives, and (as Maria Campbell goes on to say, gently acknowledging troubles to come) that the roads they travel are not too bumpy.

And I wish the same to you, too.

Solving The Wrong Problems: Why Online Education Is and Must Be Different from In-Person Education – slides from my invited talk at ICEMI 2022

Posted , updated on

icemi22

These are the slides from my invited talk at the 11th International Conference on Education and Management Innovation (ICEMI 2022), June 11th. The talk went down well – at least, I was invited to repeat the performance at a workshop (where I gave a very similar presentation today – if you’ve seen one, you probably know the content of the other!) and to give a keynote later in the year.

It’s about how methods of teaching that solve problems for in-person teachers don’t apply online, and it provides a bit of advice on online-native approaches. I’ve talked quite a bit about this over the past decade so there’s not much new in it apart from minor refinements, though I have put a greater emphasis on what goes on outside the classroom in physical institutions because I’m increasingly thinking that this matters way more than we normally acknowledge. Notably, I discuss the ways that physical institutional structures and regulations provide significant teaching functions of their own, meaning that in-person teachers can be absolute rubbish or (in some subject areas or topics) even fail to turn up, and students can still learn pretty well. This helps to explain the bizarre phenomenon that, across much of in-person academia, professors and lecturers are not expected to learn how to teach (and many never do).

Here’s the abstract…

In-person educational institutions teach, at least as much as the individual teachers they employ. Students are taken out of their own environments and into that of the institution, signalling intent to learn. The physical environment is built for pervasive learning, from common rooms, to corridors, to campus cafes; students see one another learning, share learning conversations, learn from one another. Even the act of walking from classroom to classroom makes events within them more salient. Structures such as courses, timetables, semesters, and classes solve problems of teaching efficiently within the constraints of time and space but impose great constraints on how teaching occurs, and create multiple new pedagogical and management problems of their own. The institution’s regulations, expectations, and norms play a strong pedagogical role in determining how, and when learning occurs. Combined with other entrenched systems and tools like credentials, textbooks, libraries, and curricula, a great deal of the teaching process occurs regardless of teachers. What we most readily recognize as ‘good’ teaching overcomes the problems caused by these in-person environments, and exploits their affordances.

Online institutions have radically different problems to solve, and radically different affordances to exploit, so it makes no sense to teach or manage the learning process in the same ways. Online, students do not inhabit the environment of the institution: the institution inhabits the environment of the student. It is just one small part of the student’s physical and virtual space, shared with billions of other potential teachers (formal or not) who are a click, a touch, or a glance away. The institution is just a service, not the environment in which learning occurs. The student picks the time, the space, the pace, and virtually all the surrounding supports of the learning process. Teachers cannot actively control any of this, except through the use of rewards, punishments, and the promise of credentials, that force compliance but that are antagonistic to effective or meaningful learning. In this talk, I will discuss the implications of this inverted dynamic for pedagogy, motivation, digital system design, and organizational structures & systems for online learning.

 

 

 

Pedagogical Paradigms in Open and Distance Education | Handbook of Open, Distance, and Digital Education

Posted , updated on

This is a chapter by me and Terry Anderson for Springer’s new Handbook of Open, Distance, and Digital Education that updates and refines our popular (1658 citations, and still rising, for the original paper alone) but now long-in-the-tooth ‘three generations’ model of distance learning pedagogy. We have changed the labels for the pedagogical families this time round to ones that I think are more coherent, divided according to their epistemological underpinnings: the objectivist, the subjectivist, and the complexivist. and we have added some speculations about whether further paradigms might have started to emerge in the 11 years since our original paper was published. Our main conclusion, though, is that no single pedagogical paradigm will dominate in the foreseeable future: that we are in an era of great pedagogical diversity, and that this diversity will only increase as time goes by.

The three major paradigms

Objectivist: previously known as ‘behaviourist/cognitivist’, what characterizes objectivist pedagogies is that they are both defined by assumptions of an objective external reality, and driven by (usually teacher-defined) objectives. It’s a paradigm of teaching, where teachers are typically sages on the stage using methods intended to achieve effective learning of defined facts and skills. Examples include behaviourism, learning styles theories, brain-based approaches, multiple intelligence models, media theories, and similar approaches where the focus is on efficient transmission and replication of received knowledge.

Subjectivist: formerly known as ‘social constructivist’, subjectivist pedagogies are concerned with – well – subjects: they are concerned with the personal and social co-construction of knowledge, recognizing its situated and always unique nature, saying little about methods but a lot about meaning-making. It’s a paradigm of learning, where teachers are typically guides on the side, supporting individuals and groups to learn in complex, situated contexts. Examples include constructivist, social constructivist, constructionist, and similar families of theory where the emphasis is as much on the learners’ growth and development in a human society as it is on what is being learned.

Complexivist: originally described as ‘connectivist’ (which was confusing and inaccurate), complexivist pedagogies acknowledge and exploit the complex nature of our massively distributed cognition, including its richly recursive self-organizing and emergent properties, its reification through shared tools and artefacts, and its many social layers. It’s a paradigm of knowledge, where teachers are fellow learners, co-travellers and role models, and knowledge exists not just in individual minds but in our minds’ extensions, in both other people and what we collectively create. Examples include connectivism, rhizomatic learning, distributed cognition, cognitive apprenticeship, networks of practice, and similar theories (including my own co-participation model, as it happens). We borrow the term ‘complexivist’ from Davis and Sumara, whose 2006 book on the subject is well worth reading, albeit grounded mainly in in-person learning.

No one paradigm dominates: all typically play a role at some point of a learning journey, all build upon and assemble ideas that are contained in the others (theories are technologies too), and all have been around as ways of learning for as long as humans have existed.

Emerging paradigms

Beyond these broad families, we speculate on whether any new pedagogical paradigms are emerging or have emerged within the 12 years since we first developed these ideas. We come up with the following possible candidates:

Theory-free: this is a digitally native paradigm that typically employs variations of AI technologies to extract patterns from large amounts of data on how people learn, and that provides support accordingly. This is the realm of adaptive hypermedia, learning analytics, and data mining. While the vast majority of such methods are very firmly in the objectivist tradition (the models are trained or designed by identifying what leads to ‘successful’ achievement of outcomes) a few look beyond defined learning products into social engagement or other measures of the learning process, or seek open-ended patterns in emergent collective behaviours. We see the former as a dystopic trend, but find promise in the latter, notwithstanding the risks of filter bubbles and systemic bias.

Hologogic: this is a nascent paradigm that treats learning as a process of enculturation. It’s about how we come to find our places in our many overlapping cultures, where belonging to and adopting the values and norms of the sets to which we belong (be it our colleagues, our ancestors, our subject-matter peers, or whatever) is the primary focus. There are few theories that apply to this paradigm, as yet, but it is visible in many online and in-person communities, and is/has been of particular significance in collectivist cultures where the learning of one is meaningless unless it is also the learning of all (sometimes including the ancestors). We see this as a potentially healthy trend that takes us beyond the individualist assumptions underpinning much of the field, though there are risks of divisions and echo chambers that pit one culture against others. We borrow the term from Cumbie and Wolverton.

Bricolagogic: this is a free-for-all paradigm, a kind of meta-pedagogy in which any pedagogical method, model, or theory may be used, chosen for pragmatic or personal reasons, but in which the primary focus of learning is in choosing how (in any given context) we should learn. Concepts of charting and wayfinding play a strong role here. This resembles what we originally identified as an emerging ‘holistic’ model, but we now see it not as a simple mish-mash of pedagogical paradigms but rather as a pedagogic paradigm in its own right.

Another emerging paradigm?

I have recently been involved in a lengthy Twitter thread, started by Tim Fawns on the topic of his recent paper on entangled pedagogy, which presents a view very similar indeed to my own (e.g. here and here), albeit expressed rather differently (and more eloquently). There are others in the same thread who express similar views. I suggested in this thread that we might be witnessing the birth of a new ‘entanglist’ paradigm that draws very heavily on complexivism (and that could certainly be seen as part of the same family) but that views the problem from a rather different perspective. It is still very much about complexity, emergence, extended minds, recursion, and networks, and it negates none of that, but it draws its boundaries around the networked nodes at a higher level than theories like Connectivism, yet with more precision than those focused on human learning interactions such as networks of practice or rhizomatic learning. Notably, it leaves room for design (and designed objects), for meaning, and for passion as part of the deeply entangled complex system of learning in which we all participate, willingly or not. It’s not specifically a pedagogical model – it’s broader than that – though it does imply many things about how we should and should not teach, and about how we should understand pedagogies as part of a massively distributed system in which designated teachers account for only a fraction of the learning and teaching process. The title of my book on the subject (that has been under review for 16 months – grrr) sums this up quite well, I think: “How Education Works”. The book has now (as of a few days ago) received a very positive response from reviewers and is due to be discussed by the editorial committee at the end of this month, so I’m hoping that it may be published in the not-too-distant future. Watch this space!

Here’s the chapter abstract:

Building on earlier work that identified historical paradigm shifts in open and distance learning, this chapter is concerned with analyzing the three broad pedagogical paradigms – objectivist, subjectivist, and complexivist – that have characterized learning and teaching in the field over the past half century. It goes on to discuss new paradigms that are starting to emerge, most notably in “theory-free” models enabled by developments in artificial intelligence and analytics, hologogic methods that recognize the many cultures to which we belong, and a “bricolagogic,” theory-agnostic paradigm that reflects the field’s growing maturity and depth.

Reference

Dron J., Anderson T. (2022) Pedagogical Paradigms in Open and Distance Education. In: Zawacki-Richter O., Jung I. (eds) Handbook of Open, Distance and Digital Education. Springer, Singapore. https://doi.org/10.1007/978-981-19-0351-9_9-1