So, FutureLearn, the new MOOC platform begotten of the Open University, has ‘gone live’ and opened the doors for people to sign up for beta courses.*
I’m impressed: they have gone from ‘good idea’ to ‘people can actually use it’ in just 9 months. That was by no means certain: it’s not unknown for high-stakes IT projects like this to run a teensy bit late, or even crash and burn completely before delivering anything. (Also in the news today is the prospect that the final bill for the disastrous NHS National Programme for IT will top £10bn – with nothing substantial to show for it.**)
More impressively, they have lined up contributions from 20 universities at launch, with more in the pipeline. The work involved in getting academics to do anything is considerable, and getting them to do something they’ve never done before is even harder. The partners have worked really hard to get some great-looking courses up there. Most of the feedback I’ve heard so far (noon on launch day, 18 September 2013) has been interest in the topics of the courses – which is as it should be.
I’m going to ignore the sexy content and look at the underlying stuff.
They’ve jumped two of the biggest hurdles: getting something live, and getting some decent stuff there. When it was announced, I said it was well worth a try. They’ve tried. In fact, it looks like they’ve tried really hard. They’ve delivered something. But how good is it?
I think it’s too early to tell.***
The most important question isn’t whether what they’ve done so far is good (though it has to be good enough): it’s whether what they’ve done so far can be developed to be really really good.
There’s an obvious sense in which it’s too early to tell whether it’s good: none of the courses have started yet, and that’s when we’ll really start to see how the platform works and whether the courses are as good as they look.
There are some really interesting features. Mike Sharples, the academic lead for FutureLearn, and the rest of the Learner Forum, have clearly done some very good work: Mike’s set out some of their thinking about massive social learning. But we know from the alpha experience that there are some real issues to address, and there are sure to be more flushed out in this beta phase.
The next stage is to test out empirically whether they have enough for people to learn with, and whether they can rapidly build on the feedback from those learners to make it much better than it is now. Their approach – as documented on their blog – is very much the rapid iteration model, so I’m fairly hopeful (see e.g. Learnings from our alpha trial).
Now the courses are nearly starting, we can start to see what FutureLearn looks like. Ok, so it looks like Rhubarb and Custard, or Fruit Salad chews – but we knew that. I mean, we can start to see what it can do.
But the even more important question is what it will be able to do, and that’s a question about the team’s approach, and also fundamentally about their code.
Imagine you’re the FutureLearn dev team, newly appointed in tearing haste back in January. You have to decide what to do for the platform. All of the options are wrong in at least one important way:
- Use an existing platform as-is: You’re a pointless clone and are not offering anything new.
- Modify an existing platform: You’re a bit of a pointless clone, and you will probably struggle to make it do something different (and you may not be able to).
- Write from scratch: You will have to run very, very fast just to have something to show people, and will lack lots of features that existing platforms have.
This is a fundamental problem for any moderate-sized software project, and is a really interesting question.
With iSpot, we made the decision to go with option 2, using Drupal to get something up and running quickly. We really needed the feedback from actual users: nobody had done something exactly like this before, so we needed to see if our ideas about what would work with actual, real users. The downside is that as the site has matured and grown, the overhead for using Drupal has grown huge – so we’re looking to migrate over time to written-ourselves as we add new components. On Cloudworks they used the same ‘use Drupal to get something up fast’ approach, but then later rewrote it from scratch: something Joel Spolsky says you should never do; Jeff Atwood finesses that as ‘understanding means rewriting‘. You can read a bit about how Juliette pulled it off.
FutureLearn went with the courageous strategy of writing a platform from scratch.
This was high-risk, but arguably the best decision of a bad lot. (I’ll leave aside political considerations, except to note that they were also crucial to the decision, and probably point the same way. I’m also not going to get in to a relative review of MOOC platforms and capabilities, fascinating as that would be – that’s a job for later and/or someone else.)
In technical functional terms, the fact that other platforms exist and are used heavily and to some measure of success gives you a baseline of what sort of thing can work. We didn’t have that for iSpot and Cloudworks. So for FutureLearn there was not quite so much learning-by-doing required merely to get off the ground: they could repurpose/steal the general approach from things that are already known to work. Also, the high profile publicity they can generate (and kudos to the people doing that job, which I certainly couldn’t) means they were likely to be able to get a mass general audience through the door when they wanted one. iSpot and Cloudworks were much more community-building exercises and so had to start small and grow with those communities.
In launching at all, FutureLearn have successfully dodged the main risk of this approach: of not delivering at all. Which is a huge achievement. They’ve talked about their ‘Minimum Viable Product’ (which to my mind sits slightly uneasily as a phrase when compared to the high-stakes publicity claims.) They’ve certainly hit Product – they’ve shipped, as the quaint skeuomorphic term has it.
What we’ll see over the next few weeks as courses start is whether it is in fact Viable.
And after that is when we’ll get to see whether the built-from-scratch platform and the team around it are good enough to develop in to something that really is an exciting new learning experience.
And then after that we might get to understanding a bit more about whether this approach to MOOCs is great, or not. Or – most likely – somewhere in between.
Fundamentally, though, the launch of FutureLearn looks – so far – like an exciting set of learning opportunities made available to anyone who wants to see them. Some of them are even better and are proper open educational resources with CC licences and everything. (Well done to the Nottingham team who put together the ‘Sustainability, society and you‘ course and gave it a CC licence.) I can see that there might be an argument that the time and effort might have been better spent on something else (though I’d disagree), but I find it hard to see that this is a Bad Thing.
Is FutureLearn any good? It’s too early to tell. But they’ve leaped over some major hurdles already. More importantly, can it develop in to something really, really good? I’m optimistic – on balance and very cautiously optimistic, with many caveats and all that. We’ll see.
* Disclosure: I work for the OU, and many of the OU people involved in FutureLearn are my colleagues and friends. However, I’m not directly involved in the project, and in practice have very little access to it beyond what anyone can see from the outside.
** Back in about 2000 or so I was chairing a course on information in health and social care, and some people commented that the new NHS Programme for IT would render the course obsolete before it went live. I disagreed, partly because we were writing it to focus on principles rather than details, but mainly because I just did not believe that it would deliver in the timescales promised. The course ran for many years and was retired before the plug was finally pulled on the NHS PfIT.
*** This is an excellent thing for academics to say about almost anything that hasn’t (yet) sunk without trace. The urban legend is that “it’s too early to tell” was Zhou Enlai’s response to being asked about the impact of the French Revolution, which sounds fantastic, but real life is a let down and it seems he was being asked in the early 70s about the events of May 1968, not those of 1789. I’m almost tempted to launch in to one of my favourite anecdotes about long-term thinking in the Academy but I’ll resist.****
**** Actually, I see I’ve already posted it at the end of this post back in 2008.
This work by Doug Clow is copyright but licensed under a Creative Commons BY Licence.
No further permission needed to reuse or remix (with attribution), but it’s nice to be notified if you do use it.