The automated teaching assistant aka ‘teacherbot’

The developing potential in it was clear to see in Sian Bayne’s paper (Bayne S. (2015) Teacherbot: interventions in automated teaching. Teaching in Higher Education. 20(4):455-467) on the ‘teacherbot’ developed by a team at the University of Edinburgh, and used in an earlier MOOC.

The twitterbot examples we’d found as a collective were by their very nature designed for consumption by large audiences, and in most cases offered little interaction. They use the power of algorithms to source, rework and mix up varieties of content.

Given Twitter’s primary use as a platform for interaction, it was really intriguing to find how the teacherbot had been designed for use for a particular set of people (namely the MOOC participants) but also provided value in several ways.

It seemed like the bot offered some immediate efficiency gains, for example, automated reminders of assessment deadlines, based on keywords. But it seemed to me in the paper that one of Sian’s key arguments was that automation should not just be viewed as simply a way of gaining efficiencies. Indeed exploration in this field should be framed differently, to provide a wider field to test its potential.

So the real excitement comes from the examples of interaction between the teacherbot and MOOC student, even if it was “slightly ‘clunky’ and often rather wide-of-the-mark”. It seems in some cases it provoked useful reflection and discussion, which would add value to the student’s experience.

The paper did make me ponder on a few things: It’d be fascinating to see

  • It’d be fascinating to see teacherbot v2. Given its first iteration was understandably clunky at times, and that an algorithm/bot improves when it is receptive to feedback, it’d be useful to view how much interaction happens the second time around, and the value it brings to the student experience. Given subsequent ‘polishing’ of the teacherbot interactions, I wonder if it would become increasingly difficult to spot it as a bot. (Obviously, there are the ethical discussions to be had here about not raising awareness of this with students, but putting this to one side for now…).
  • Taking this further, and should bots become more commonplace with these environments, would students increase or decrease their interaction with the bot as a result, and ultimately would they even care if it was human or non-human? Is that relevant to the experience, and is this just a personal choice?
  • In some of the forum postings, there has been some discussion about the procurement of technology within education, and how this has in several instances been sold into the institution or organisation gain as an efficiency gain (primarily). The pedagogical advantages have been at times been a secondary consideration. Given the speed of development or development of any technology is often tied in with the adoption rate (more usage brings more development), should we think more pragmatically about this, and ensure that any technology we wish to bring into education has an efficiency element to appeal to certain stakeholders within the decision-making process?


Having come towards the end of week 2, it’s been interesting to see what twitterbots have been highlighted by the other course participants. Forgetting the bots that simply provide alerts on new events or information for now, it’s evident there’s a real richness in some of them.

My own particular favourite is @the_ephemerides. Harking back to the recommended blog reading on bots I’ve already touched on in a previous post, I think this twitter bot certainly is an excellent example of creating:

“something that is actually greater than the sum of its parts”.

Here’s an example:

The creator of the bot provided some useful background on the development of the bot:

“One of the affordances of generative text in general is satire and humor, and I have made my share of satirical and humorous (and borderline mean-spirited) bots. In making the Ephemerides, I was trying to expand my aesthetic range a little bit, to make something a bit more lyrical and evocative. The poetry is made by remixing two 19th-century texts, one on astrology and the other on oceanography, which combined have the feeling (or are intended to have the feeling) of otherworldly poems about the exotic icy landscapes of the solar system.”  (Source, 30 September 2017)

Having read through chapter 6 of Selwyn (Selwyn, N. (2011), Education and Technology: key issues and debates.), there’s a discussion about the art and science of teaching, and the role that technologies play within this. While not directly relevant, these discussions did get me thinking about the role of technology and ‘art’, particularly in light of these bots.

In my opinion, bots like these allow us to explore new areas beyond previous thresholds, simply due to the scale and reach technology like this allows. In the example above, technology has been already used to create the various pieces (camera’s/telescopes for the photos, and the pencil to record the written word), and now the technology is providing us with the opportunity to mesh these in weird and wonderful ways. If art is ultimately about “producing works to be appreciated primarily for their beauty or emotional power” (Source, 30 September 2017) then surely this ticks the box?

Twitterbots – something more than just scale

I’ve got to admit, I’d always assumed twitterbots were just the platform’s version of spam. I’ve used twitter fairly regularly over the past 2-3 years so, while no expert, would assume I’m one of their more savvy users. Sure, I’d had the occasional tweet from a ‘bimbot’, but assumed bots were being used for little else.

So I was a little surprised to see this introduced in week 2 of the IDEL course, and it’s been a very pleasant surprise! Of course, having now read a little more about them, I sense the deep relevance in the discussion about technology and the role of ‘teacher’.

The recommend blog post provides a great introduction to twitterbots, and I was particularly struck by the description:

What these bots show is that the future of automation—whether of work, errands, or other routine tasks—is about the combination of human creativity and the raw processing power of machines.

And boy, when the balance is right it can be a beautiful thing. What could be suggested here is it combining technology with human creativity can help scale this – to reach a wider audience, have more impact, produce more, and even produce

“something that is actually greater than the sum of its parts”.

The ‘infinite monkey theorem came to mind when I read this. I wonder if technology has previously just been assumed as a way to scale (an industrial version of the monkeys), to aid efficiency and deliver increased permutations. Perhaps what we are now beginning to explore is something much deeper – how it can elevate our creative output to a higher level, and help us create something much bigger than ourselves. Ultimately our creative output is the centrepiece, but technology can help us mould this is new and exciting ways.

I’m looking forward to spending more time this week looking into twitterbots, and how they can help us in learning and education. My favourite so far? The ‘here’s your reminder’ bot, simply because I think it has very sarcastic overtones…

Algorithms in learning

I was lucky enough to listen to a talk from Chris Littlewood from Filtered in Manchester recently. His business is exploring algorithms as a way of providing a scalable solution to the idea of content creation.

Ben Betts, from HT2 labs, suggested at the same event that a transition may be taking place within L&D functions – moving away from the creation of elearning content, to simply curating the best of what’s already out there.

The technology that Chris’ team are developing takes this one step further – so rather than a person deciding what is relevant and scouring the web for the best content (and by ‘best’ I mean most likely to add value), algorithms are used to scale this up, gain feedback from students on the choices made, and use this to improve the algorithm itself.

With algorithms and technology helping education and learning become more ‘intelligent’, it feels like we could be on the cusp of something really exciting.

Week 1 takeaways

So week 1 is already over! Wow that went quick. The case studies on ‘constructing community’ have provoked some real food for thought, particularly given the polarised situations in the examples.

A few key points and thoughts from the first week:

  • Whether the student environment is online or offline, the complexity of community doesn’t really change. It may be different in format, and with different challenges, but ultimately all individuals within a community are unique with their own needs, personalities and capabilities. In short, ‘digital’ doesn’t seem to dumb things down, or make it more straightforward. Instead it may heighten certain issues (e.g. the feeling of being swamped), reduce others (e.g. physical impairment, and unconscious bias towards this), and bring new challenges to the table (e.g. increased diversity of nationalities could bring cultural clashes to the fore)
  • Tutors or leaders within a community have a tough balance to strike! Be too interventionist and the community could lose some of its as yet unexplored potential, – the value be ‘throttled’. Keep things too unregulated and focus could be lost, and the community could become a real turn-off for some.
  • How someone designs an online learning environment has a major bearing on the nature of the community within it. It seemed to me that in several of the examples, the structure of the course influenced a student’s reaction and ability to ‘learn’ from it. Perhaps in the past I’ve underestimated this. However, I also think that building community is not something that can be overtly designed (see my first point about the complexities of individuals). Perhaps it is also about ongoing refinement, and being observant of any improvements that could be made from one iteration to the next.
  • Digital literacy is likely to play a part in the participation levels within a community. It seemed to me that in some of the examples, some of the causes of the challenges could be placed with the individual, rather than the learning experience itself. Again the tutor has a tough job to identify these and support when it makes sense to do so, particularly when in an online environment it may be more difficult to spot those struggling (they may just go quiet) than compared with the more traditional classroom. And how do we avoid losing their contribution in the community discussions?

And finally, one rather pertinent point:

  • There is a huge amount of value to gain from a community within a learning experience! May sound like a rather odd thing to be explicit on, but I’d previously mentioned that ‘community’ has not been a common element in many of the online learning experiences I’d had (or created – for a variety of reasons). I’ve been massively impressed (but not surprised by this!) by the other individuals in the groups and really benefitted from their expertise and opinions. Being able to see the group interact with each other, and challenge each other (it’d be great to see more of this though), really enhances your own thinking, rather than relying on one source of expertise or content. It’s going to be interesting to see how the community develops over the next few weeks, particularly when initial novelty value and enthusiasm may wane?

Feeling swamped – MOOC perspectives

We’ve been posed some very interesting scenarios this week on the topic of ‘constructing community’, and it’s been fascinating to read the different takes from fellow students on these on the forums.

I was particularly intrigued by some of the points raised about the sense of feeling ‘swamped’ in a MOOC (Massive Open Online Course) environment, and thought it would be worthwhile diving into this a little deeper.

In the example provided, there seemed to a variety of factors contributing to this feeling:

  • the sheer volume of students on the course
  • the volume of content created by the students on the course
  • engaging with content outside of the main environment (e.g. additional reading)

It got me thinking about the nature of MOOCs, and strategies that could be put in place to help manage these challenges.

(Before looking at this, I think it’s important to note from the outset that this could be seen as a ‘nice’ problem to have! The fact that the MOOC has attracted successfully attracted thousands of students obviously taps into a subject that people are keen to develop their understanding in!).


Although the example doesn’t give details, I’ve found several MOOCs I’ve participated in to use ‘engagement’ as a metric, or contributor of success.

The majority of MOOCs I’ve come across have primarily been a) free of charge and b) require no prior knowledge or experience. While this is one of the fantastic aspects of MOOCs – being able to provide a learning experience for anyone, anywhere. But this could be a double-edged sword, with a bigger and more diverse audience, perhaps that sense of community and connection is more difficult to bring together, simply because there may be more that makes people different, than similar.

Therefore it’s difficult to base success of a MOOC simply based on an end assessment. Given a person could potentially complete an assessment without actually touching the course content or interacting with other users, how to do you evaluate success this way?

(And given many MOOCs are used as a brand awareness activity by universities, and could be considered a ‘marketing’ activity, engagement (however this is defined) is probably a better metric internally within an institution).

Given engagement is a key driver then, perhaps it creates an onus on students to contribute, even if it offers little value (to themselves, or the other students). As the example itself states:

“What was the point of adding another blog post when there were so many floating around already?”

Perhaps then a strategy for the course designers is to think carefully about the metrics related to the course, both from an academic, student and institutional perspective. As ultimately what may be quite subtle measurements could extrapolate to fairly profound impact on the overall course experience. (I think of Ken Robinson’s quote here, “If you design a system to do something specific, don’t be surprised if it does it”.)

Group size

I wonder if there was a way the course tutors on the MOOC in the example could have split the group size into smaller cohorts. It seems there has been plenty of discussion over ideal group sizes within MOOCs elsewhere, with understandably little agreement (given the nature of MOOCs are quite diverse).

One interesting question to pose is how to create cohorts, and what to base the cohort decision on. Registration date is the obvious one, simply batching them based on chronological enrolment. But could geographic location but one option, and perhaps levels of engagement itself? Ultimately this would need some consideration and piloting to make the MOOC experience for students more enjoyable, without losing any of the diversity and breadth of opinion that could occur outside of the student’s specific cohort.

Managing expectations

On further reflection, I wonder if the feelings expressed by the MOOC student were representative of the wider group. It strikes me that this student could be trying to view everything created by the course and its students, which in this scenario seems unrealistic.

Hajira Khan, a fellow student on this IDEL course, suggested moderators have a key part to play in this process too:

“The tutor for the course should be an active moderator so that discussions are moderated and are kept relevant and concise.”

That role of tutor could be key to keeping the discussion on track, but also managed what is expected.

The feelings are obviously genuine and not to be dismissed, but changes should be considered for future MOOCs in line with wider feedback too, this student’s experience could be an outlier.

One consideration for tutors is to ensure there is the facility to capture feedback like this early on, and where possible have the flexibility to adapt the MOOC where possible to get it back on track.

Slightly off-piste I know, but I also wondered if that feeling of being swamped had been cited as one of the reasons for the notorious MOOC drop-off rates that occur. After looking at several sources, it isn’t named explicity (although it could contribute to the wider tag of ‘bad experience’).

Daniel Onah, of Warwick University, discusses the following contributors to MOOC drop-off rates here:

  • No real intention to complete
  • Lack of time
  • Course difficulty and lack of support
  • Lack of digital skills or learning skills
  • Bad experiences
  • Expectations
  • Starting late
  • Peer review

So while feeling swamped could be a contributor, it seems the reasons for drop-off could be quite wide-ranging!

“The ledger”

So we aren’t even at week 1 of the IDEL course, and the University of Edinburgh team are already serving up some real food for thought.

We were given some links to video content to peruse – I’m guessing as a way of getting the juices flowing and priming us for some of the content we’re likely to be discussing.

One in particular peaked my interest – ‘Learning is Earning 2026 by the Institute for the Future (IFTF)’.

It’s a video produced by the Institute for the Future. I’ve never come across them before, but am really glad to have done so. As an organisation it seems to me that they speculate on what the future could look like, by combining several recent developments or trends. In this example they’re exploring a combination of block-chaining, digital credentials and a flat, non-hierarchical form of learning from your peers.

It’s already got the cogs whirring, but some initial thoughts came to mind:

  • The digital credentials mooted in the video are not something for the future, indeed they are already here. From recent conversations I’ve had with my own clients, this is starting to trickle into the mainstream. Whereas in the past they’ve seen quite abstract (I always felt Mozilla’s backpack was a great initiative but had some flaws), we’re starting to see commercial competitors, such as credly and accredible come to the fore. I see great potential in the digitisation of credentials, especially when it comes to providing granular information around the ‘achievement’, real scale (from quick wins to full degrees), and also providing rigour around the attainment of the credential.
  • I’ve also previously come across xAPI, which is using of technology to capture ‘learning moments’ outside of any formal programme and recognising these. Again this is a theme raised in the video. It’s already a tangible form, but it’s still in it’s infancy.

(Both points remind me of a quote by writer William Gibson – “The future has arrived — it’s just not evenly distributed yet.” source)

Some other thoughts came to mind:

  • I’m really intrigued by the concept of peer-to-peer teaching – and particularly how this can be tied into a form of digital credential. Is a digital credential only of value if it comes from a ‘reputable’ source, and the output of the learning can be assessed and qualified? I’m sure this is an area we’ll be digging into over the next few weeks.
  • Following on from this, the video raises the incentivisation of peer-to-peer learning through digital credentials (although it doesn’t have to be limited to this). This could be a direction that commoditises, and ultimately weakens the value of the credentials, so if I was the CEO of ‘Ledger LTD’, I’d be wary of the approach to this.
  • It strikes me that using ‘ledgers’ as a basis for granting work, as described by the HR director, can empower hirers to make better decisions. But unless I’m missing something, it should be used in conjunction with the more traditional methods of hiring, to assess overall culture fit and motivations. But this is probably a subject for another time!

Now, back to that list of videos…

A step into the unknown

Well some 13 years after graduating from an undergraduate degree in Economics, here I am back in the realm of formal education! Yet the motivations for doing so could hardly be more different.

It’s a real maelstrom of feelings, particularly in the last week when it’s moved from being just an exciting project in the future to a sense of ‘wow what have I taken on!’. I’m slowly working my way through all the introductory information and it’s a lot to take in – but as with all these things once you have a bit of momentum it all starts to fall into place (or so I hope).

So what am I looking forward to in the next few weeks? There’s quite a lot to be excited about:

  • I think there is so much yet untapped in the area of digital education and learning, we’re still just exploring the potential. I’m looking forward to learning about where we’re at, where things are headed, and ultimately what I can help shape.
  • Connecting with others to share, discuss, argue, disagree (!) in an online setting is something I can’t wait to get stuck into. As a father of three young kids, working from home, life can be quite isolated at times so extending my network, both professionally and personally is a real motivator for me.
  • Having worked in a commercial environment for the entirety of my working life, I think it’s going to be quite a culture shift to get back into ‘education’ and learning alongside people of different, and perhaps more academic, perspectives.
  • Having dipped my toe into other online offerings in the past few years, the only real experience that stood out as ‘breaking the mould’ was the University of Edinburgh MOOC on digital cultures. Given the format and thought that’d gone into this, I can only be enthusiastic about what lies ahead. In my opinion ‘elearning’ has quite a negative reputation, and that’s largely down to some of ‘click next’ compliance courses we’ve all experienced in the past at some point. I’m looking forward to being challenged about what a digital learning experience could be, and how this can be facilitated.

So what about the fears?

  • I’m sure every student has the same doubts, but the anticipated time commitment is on my mind. I want to contribute, and ultimately gain full value from the course but am a tad nervous about baking this into my week. But I’m sure once I’ve got into a rhythm it’ll be fine – once the conversations start flowing they’ll be fuel for my learning fire!
  • It seems the real framework of the IDEL course is the personal blog. Having always battled to articulate my thoughts from many words into few, I think this’ll be a big challenge at the start. But I see the opportunity to practice, and ultimately work on this soft skill as a real fringe benefit of taking the course.

So what lies ahead over the next few days? Well, I’m travelling back to the UK tomorrow (I’ve left the family in Holland with the in-laws for another week), and am downloading what I can now to get up to speed on ‘Week 0’ while in transit. Then come week 1, it’s all systems go!