Saturday, March 12, 2016

Will Amazon be the `Netflix’ of learning?


Amazon often surprises. When it moved out of books only into anything and everything, we were surprised. When it started to deliver cloud services, we were surprised. When it said it wanted to open bookstores, we were surprised. But the idea of Amazon being a global education provider – that’s a shocker.
Secretive
Notoriously secretive, we can only guess what they’re up to. But this much we know. They’ve got a ‘wait list’ for their new educational service. That’s an interesting little marketing play. Keep it secret, keep it scarce – then launch. They bought TenMarks a couple of years ago, use predictive analytics to sell stuff and have the ability to deliver a super-massive global service if they so desire.
Amazon have been playing around with TV, with Amazon Prime. But they produced Alpha Four about four Senators - a dud. Netflix used data, but much more fine-grained, and produced House of Cards. Data analysis, in itself, is not enough. You need data plus experts. That's why Amazon made their mistake - they were too cocky about the data alone. See this TED talk for more. 
Role of AI?
I’ve written tons on the future role of AI in teaching and learning. I’ve invested in it, am building my own company in the area, talk on the subject, write on the subject, so I’m a convert. But I’m not Jeff Bezos and I don’t have a global platform that is as good as anyone at delivering stuff with consummate ease to the entire planet. Jeff does.
Knowing Amazon, there will be some predictive, recommendation engine, review, ratings and an interface that works. They are the masters of ‘ease of use’. They’re bandying about the word ‘open’, which is heartening but could mean anything. An open publishing platform could be interesting but the OER world is full of teacher-created content that lies dead in the unloved repositories of reusable content. If that is their strategy – a sort of share and swap service for resources, with ratings, - it will fail. Delivering smart, interactive e-books could be interesting. Add the magic dust of AI, it has a real chance.
Textbook wipeout?
The textbook market is ripe for a Wikipedia-like cleanup. They’re often poorly written, linear, text-heavy, media unfriendly, quickly out of date and far too expensive. If they have a pop at this market, I for one, will cheer them on. The very concept of a textbook is under attack and it is well on its way to becoming obsolete.
Polish experiment
There's already been nationwide work done in Poland on OER textbooks, the first country to politically support an open-textbook strategy. The government funded Creative Commons Licensed textbooks that can be translated, reused and adapted in primary and secondary schools. The huge savings for both parents and government are obvious, running to around €200 million. They plan to continue the program until 2020. Other places to watch are S Africa and Brazil. The question is whether the clout of a global brand, like Amazon, will help. The evidence suggests that private sector delivery does help. Most OER initiatives fail through lack of business and marketing skills, and remain unloved and unused. Amazon may just provide the infrastructure, marketing and skills to turn this into a global phenomenon.
Conclusion
I wrote some time back about the possibility of a Netflix in education. I feel that we’re moving closer to this, with the rise of AI and adaptive learning. What’s missing is the organization with the chops to pull this off. There are a few around but it really comes down to the big five – Apple, Facebook, Microsoft, Amazon and Google. It is often claimed that IBM’s Tom Watson, who sold a LMS to Hitler, said that there world would only need five computers. He said no such thing. Like most quotes from Einstein and others at educational conferences it’s bullshit. Yet it may, despite its false providence, turn out to be true. These guys do have a grip on the market, and enough cash, to make them almost invincible. As they say, watch this webspace.


Remember this date – 12 02 2012 - the day our species lost to AI - but won

This is worried look of Lee Sedol and he was in the match of his life - he lost. But it wasn't Lee who lost, it was us all or is it really a win for us, the certaors of AlphaGo? The Human v machine sparring that has been going on for some time with checkers, Chess and Jeopardy, were featherweight contests. This was the Big Fight and we, as a species, got thumped. There’s a new breed of champion in the ring and it’s not just smart, it’s a superfast learner, even its own teacher. It’s eats up human expertise for breakfast, then the real game begins, as it uses this experience to play itself, as it’s the only opponent worth playing. Having learned from us, it sucks our experts dry, then transcends their abilities to boldly go where no brain has gone before.
Momentous moment
This is a momentous moment. In less than 2.5 years, since Google snapped up Deepmind for a mere half billion dollars, 2500 years of human experience and expertise at GO has been trounced. But this is only the start. Software that learns is exponentially more powerful than software that has to be written by humans. Given the huge processing power of Google Cloud Services, AlphaGo has one of the greatest engines on the planet under its hood. It also has some of the best algorithms and that’s what matters. Machine learning algorithms are like small Gods. Free from the tyranny of time and space, speed is no limit. They can learn faster than any of us. These algorithms are the new DNA of progress. This machine moves beyond teaching and teaches itself. That’s essentially what humans do as they become expert learners, few in the later years reply any longer on teachers, as we’ve learned to learn for ourselves. AI just moves to abandoning the teacher faster.
A bit of history
Throughout our history as a species we have always benefited from the delegation of the mundane. This has largely been achieved through technology. We conquered the planet through technology. First through stone then metal tools, needles for clothing, tools for agriculture and so on. Then we invented machines that to did the manual work and we moved from the fields to factories. Then we mechanised the factories and moved towards mental work. Now we’re delegating the drudgery of some of that mental work to machines or, more accurately – AI, even more accurately to machine learning.
Teaching and learning
Amid all the hubris that surrounds education and teaching, there’s a deeper problem. Parents know it, learners know it, even teachers, lecturers and trainers know it. Performance has plateaued and everyone is getting a little fraught. Politicians, driven on by the poor foundations, and therefore learning tower of PISA results, demand more testing. Parents, the most conservative of lobbyists, demand more schooling. Teachers scream ‘enough already – we’re exhausted’. Well, isn’t it about time we looked for the sort of solutions that gave us the industrial and information revolutions of the past? Can’t machines solve the problem of teaching?
Teaching trumped by learning?
Could teaching be trumped by a learning machine? Are we beginning to glimpse the possibility of machines that teach themselves to teach? They learn what works, what doesn’t and deliver ever better performance. We see the embryonic evidence for this in adaptive learning systems, that are truly algorithmic, and do use machine learning, to improve as they deliver. The more students they teach, the better they get. They even tech themselves. This is not science fiction. This is real AI, in real software, delivering real courses, in real institutions. The future has been here for some time it’s just not distributed.
Teaching free like search?
Imagine what will happen when these super-teachers are commoditised, delivered from super fast cloud-services and let loose on the web? Teaching and learning will be as free and accessible as Google search. You will not only be able to find things with ease, you’ll be able to learn them with ease. We may see dramatic rises in performance among learners, right across the board, as such systems will be far more sensitive to individual needs, even learning difficulties. Who is likely to deliver such as service? Well Amazon are on the march, Gates has been seriously funding this stuff but Google is the front runner.
Future without teachers?
This may see hopelessly utopian. But could we have a future without teachers? Why not? Teaching is essentially being a conduit. It is a means to an end, not an end in itself. Wouldn’t academics really prefer to do pure research and not teach? Wouldn’t most teachers prefer not to have to mark anything and avoid the stress of the classroom? Couldn’t we dispense with teaching and just have learning?
Probable, improbable or impossible?
Agricultural workers were largely mechanised out of the process by machines. factory workers by robots, secretaries by word processors and It looks likely that we will see the obliteration of drivers, cabbies and truck drivers, through driverless cars. No one predicted that! There’s a lot of evidence to suggest that many professions, even white collar, middle-class professions, may be replaced by smart AI. So what’s so special about teaching? If we can teach millions, of not hundreds of millions at cents per learner, isn’t that desirable?
Remember this date

So remember this date – 12 02 2012 – it sounds almost providential. It may go down in history as the day we lost our several million year long reign as Champions of the World, not to the super-smart Frankenstein we created, but to the machine teachers who help us learn to be better humans.

Thursday, March 10, 2016

7 questions you need to know answers to on legalities of ‘data’ in education

Everyone is in a tiz about data in education. Yet students don't seem to care and, if truth be told, there's not a great deal to worry about. Most of the fuss is from older folk who get all angsty about 'privacy'. Nevertheless, it's worth knowing where we lie on the issue and the law.
1. Are students queuing round the block for their data?
No. Students are not that interested, as they are generally brought up in a world where they know that letting people hold and use your data is the price you pay for free stuff. Don’t imagine for one minute that educational institutions have students queuing around the block demanding to see their data. Students are remarkably blasé about all this. That’s not to say it is not an issue or that it will not become an issue in your organization. Sure, there will be a small army of people wanting to deliver 'digital literacy' courses. That's always the solution to a problem in education. But there's no need for panic here, as students probably know as much about this as anyone. They just don't care.
2. Should we promise all data on demand?
So, if students demand it, should an institution promise to give them all data? No. In the UK, we have the Data Protection Act, which is a legal right. However, this is not a blanket rule and there are lots of exemptions, as well as practical issues. So don’t promise the world, as you may neither need to nor  be able to deliver.
3. Are we legally bound to show them?
No. An important point, that is often missed, is that if they haven't asked, you don't have to provide the data.
4. What form of data?
Even if you were to be asked for data, it is likely, that in its raw form, it would be meaningless. Data is rarely useful unless it is analysed and then visualized. So don't give yourself a technical rod for your own (and not the students) back. Data is usually dead, inert stuff.
5. What about predictive models using analytics and adaptive learning?
Should we provide that data? First, you only have to provide 'stored' data, not data used on the fly, which is common in these adaptive systems. You also have the argument that much of this is not stored data but algorithms with inputs and outputs. It is not so much isolated data, as statistical inference or probability. Although try explaining that to a student.
6. Is all data declarable?
No. There are exemptions. You do not have to provide data where there have been infringements of IP by students, data to do with a crime/investigation or third party data provision.
7. Risk of prosecution?
Remember that the risks of prosecution are relatively low. The prosecution has to show that there has been monetary damage and that you have acted unlawfully. This is a quite remote possibility. Nevertheless, as we’ve seen in the US, in these days of onerous student loans, angry, unemployed students may have axes to grind.
Conclusion
Remember also, that if you’re thinking of loosening up on data, making it available to students, you have also to be very careful with access. Students are smart, savvy and skilled. Many have the ability look for access to change, say grades.
(Found JISC podcast very useful on this subject) 

10 PERSONALISED learning ideas

This brain the most complex thing we know of in the universe. We all have one that is absolutely unique, that doesn't like impersonal, long linear, flat experiences. Yet, in learning we feed it exactly that - long lectures, flat pages of text, one-size-fits-all learning experiences. Learning is a deeply personal experience and most learning theory, in terms of attention, deep processing, self-generation, practice and so on, pointstowards the need for individual needs.
Personalisation is the one great gift that online learning offers teachers and learners. But it comes in many practices, not all of them obvious. In classrooms teachers talk about differentiation then struggle to differentiate, with 30 plus kids. In Universities, the anonymous mass that turns up to mass lectures (or not), is the opposite of personal learning. In the workplace, the compulsory compliance course is far from ideal. But ‘personal’ can mean relevant, timely, location-based, self-determined, self-generated, favourites, targeted, private, even intimate. Learning is rarely any of these. Set-meal courses are the norm. What to do? Personalise.
1. Timely
Learning delivered when you need it is called performance support. That specific thing that you need at that timely moment, is often something precise and special. This is what predictive agents and AI will deliver for learning. That precise answer, the one you need right there and then, the time predicted by software.
2. Location-based
Plenty of apps are now location-based and learning can be delivered tat is sensitive to your personal position. It may be in an art gallery in front of a painting or when operating a piece of electronics. I know of one programme that locked people out of the central area of a nuclear power station if they failed their risk assessment course. It disabled their electronic badge. Well see more of this I’m sure.
3. Real time adaptive
Personalised delivery can be in realtime, as you move through a learning experience. The software decides what you need next, based on predictive analytics, a bit like a SatNav, getting you back on course. This approach is sensitive to personal needs in terms of attainment, misconceptions, and so, preventing failure. The personalisation depends on both the individual, aggregated data and the algorithms used to determine such paths. There will be no linear course, only vectors through content.
4. Porosity
With AI driven online learning tools you can create courses that are ‘automatically’ porous, with links out to automatically linked content. This can be done with YouTube, Wikipedia and other text sources in WildFire. This allows the learner to move out of the more structured online e-learning into more detail, where they want to dive deeper. It allows a more personalised set of routes through the learning experience driven by curiosity, not deterministic paths. One such tool is WildFire.
5. Generative
Self-generated content has always been a powerful way to learn. I’ve produced a revision tool, where you simply type in your notes – from a lecture, class, course, revision – whatever. The software then automatically creates active online learning, where you answer automatically generated questions on your own notes. This moves the learner from reading/underlining to active, recall and remembering.
6. Playlist
Jeff Staes put me on to this – the idea that efficient learning has less to do with ‘courses’ than personal playlists, like our music playlists. We all have different needs that vary over time but these needs are more like personal playlists of skills, that vary over time as our needs change. Most of our behaviour is personal and so learning should be individualised. We have preferences, needs, likes, dislikes, bad habits, good habits – these can be reflected in our playlists.
7. Personal assistant
What we may need is a behind the scenes is a personal service that does the messy stuff for us. A good example is the Personal Assistant. I had a PA for years and found it difficult when I stopped running a large company to go it alone. But try AMY the AI driven personal assistant. It’s uncannily real. We are now seeing agents enter the learning arena, to help us navigate our way through what we need to know and learn.
8. Social
Social media is highly personal, with AI dealing with the delivery of Facebook timelines, Tweets and other social stuff. The billions that use social media clearly like the fact that it is a personal experience, not only in terms of their preferences and likes but also personal in terms of people, who you follow and so on. This is why social learning matters, It is a set of highly evolved, even intimate, habitats in which one can learn from friends, relatives, colleagues, experts and others.
9. Assessment
There are all sorts of ways that assessment can be personal in online learning, especially using contemporary AI techniques. I can identify you digitally for an online exam using facial recognition, typing patterns and other forms of digital identification. I can deliver randomised, adaptive test items to avoid cheating. I can automatically grade and provide feedback on you essay through automated essay marking. I can spot cheating patterns by teachers and students through statistical techniques to spot deliberately de-personalised assessment.
10. Spaced-practice
Now that we have the ability to deliver learning via mobile devices, spaced-practice is coming of age. Having blown through the 100 million users mark last year Duolingo is a good example, there are many others. These learning experiences are finely tuned to your personal reinforcement and needs in terms of remedial practice. Most AI adaptive systems use spaced-practice in one form or another. You receive practice items, not only finely tuned to your performance but also to your availability and pattern of use.
Conclusion

All of the above are examples of the personalisation of learning but there is one more dimension that takes the personal to another level and that’s scale. The trick in learning is to personalise, at the same time as scaling up. It’s pointless personalising for the few when personalisation means a personal experience for the many. In my view, it is AI that will accelerate the personalisation agenda by delivering more than just the illusion of personalisation but the reality.

Saturday, March 05, 2016

The LMS is dead, long live the LMS! (10 pros & 10 cons)




The LMS is dead, long live the LMS! (Swap out LMS with VLE if you wish.) Some love them, some hate them. Some love to hate them. Me? I see downsides and upsides. They work for some, not for others. So here’s my 10 pros and cons.
1. Zombie LMS
Some organisations have a Zombie LMS. At the very mention of its name, managers and learners roll their eyes. Organisations can get locked into LMS contracts that limit their ability and agility to adopt innovations. Many an LMS lies like an old fossil, buried in the enterprise software stack, churning away like an old heating system – slow, inefficient and in constant need of repair. Long term licences, inertia and the cost of change, see the organisation locked into a barely functional world of half-dead software and courses.
2. Functional creep
Our LMS does everything. “Social?” “Yes, that as well”. Once the LMS folk get their hooks into you, they extend their reach into all sorts of areas where they don’t belong. Suddenly they have a ‘chat’ offer, that is truly awful – but part of the ‘complete LMS solution’. For a few extra bucks they solve all of your performance support, corporate comms, HR and talent management problems, locking you bit by bit into the deep dungeon they’ve built for your learners, never to see the light again.
3. Courses, of course
The LMS encourages an obsession with courses. I’m no fan of Maslow’s clichéd pyramid of needs but he did come up with a great line ”If you only have a hammer, you tend to see every problem as a nail.” That’s precisely the problem with the LMS - give an organisation an LMS and every problem is solved by a ‘course’. This has led to a culture of over-engineered, expensive and long-winded course production that aligned with the use of the LMS and not with organisational or business needs.
4. Cripples content
Throw stuff into Blackboard and it spits out some really awful looking stuff. Encouraged to load up half-baked course notes, teachers and trainers knock out stuff that conforms solidly to that great law of content production – GIGO – garbage in garbage out.  Graphic, text, graphic, text, multiple-choice question….. repeat. Out goes simulations and anything that doesn’t conform to the simple, flat, linear content that your LMS can deliver, like WildFire, where course use AI to create high retention learning in minutes, not months.
5. One size fits all
With the rise of AI, adaptive and personalised learning, the LMS becomes an irritation. They don’t cope well with systems that deliver smart, personalised learning pathways. The sophisticated higher-level learning experiences are locked out by the limited ability of the LMS to cope with such innovation. The LMS becomes a sort of cardboard template through which all content must fit. There are the formal courses that most organisations need to some degree, informal learning (often with a social dimension) and performance support (rarely done well from a LMS). But it’s the ‘learn by doing’, or experiential learning, that most LMSs really squeeze out of the mix. It’s often disguised within an LMS delivery as ‘workforce planning’ but that’s a sop.
6. Compliance hell
We all know what happened in compliance training. L and D used the fallacious argument that the law and regulators demand oodles of long courses. In fact, no law and very few regulators demand long, bad, largely useless courses. This doesn’t work. In fact, it is counterproductive, often creating a dismissive reaction among learners. Yet the LMS encourages this glib solutionism.
7. Completion cul-de-sacs
With the LMS, along came SCORM, a ‘standard’ that in one move pushed everyone towards ‘course completion’. Learning via an LMS was no longer a joyous thing. It became an endless chore, slogging through course after course until complete. Gone is the idea that learning journeys can be interesting, personal affairs. SCORM is a completion whip that is used to march learners in lock-step towards completion.
8. Privacy
I once questioned the surveillance role of LMSs at a Masie conference, pointing out that the first LMS had been built and sold to Hitler by the then IBM CEO Tom Watson. He went apeshit (I later realised that IBM was a sponsor). But I’ve always been wary of the privacy issues through an LMS. There is general unease among employees about being measured in this way. Now, of course, there are data and privacy legal issues. These vary from country to country, Germany being particularly fierce. This remains an interesting and contentious area.
9. Surveillance
Explicitly gathering data through an LMS may also have a deleterious effect on learning, making people more nervous than they should be about who is watching their behaviour and why. There is the nagging worry that such data may be used to determine their future in a deterministic and unfair fashion. They’re not far wrong. Take Myers-Briggs, HRs favourite Ponzi scheme, that has been shown to be wrong but is still used, shamefully, to determine recruitment and promotion decisions.
10. Limits data
Given the constraints of most LMSs, there is the illusion that valuable data is being gathered, when in fact, it’s merely who does what course, when, and did they complete. As the world gets more data hungry, the LMS may be the very thing that stops valuable data from being gathered, managed and used.
Now the pros….
1. Migration from classroom
It’s often forgotten that the LMS was, in the early days, the prime mover for shifting people away from pure classroom delivery. This is still an issue in many organisations but at least they effected a move, at the enterprise level, away from often lacklustre and expensive classroom courses. In fact, with blended learning, you can manage your pantheon of delivery channels, including classroom delivery, through your LMS (classroom planning is often included).
2. Scalable
Scale has many meanings. Do you want 24/7, on-demand, self-paced, secure, location flexible, responsive, multi-language, global and local, affordable, value-added learning? Without technology this is undoable. An LMS not only gives you scalability, it makes you think on scale and solve the problems that scalable solutions bring. This has been made much easier by metered, cloud delivery.
3. Controls chaos
There are arguments for letting a thousand flowers bloom but this can turn into a nightmare if everyone starts to become auteurs. Amateurism can turn learning into a cottage industry with lots of duplication of content, poor quality resources, unnecessary licence costs and an unmanageable mess. An LMS can bring order to potential chaos.
4. Consistency
Large organisations, especially global organisations, need to have some level of consistency in terms of strategy, brand and messaging - this affects learning. Consistent rules about design, development and delivery are not always bad. It can lead, if managed properly, to a rising tide of relevance and quality. An LMS can help an organisation be consistent. There's a reason organisations have marketing, sales and finance departments - they deliver strategic intent. This can be true of an LMS. 
5. Integration
Whatever way you cut it, an organisation will have a load of systems that need integrated. However you identify your people, store stuff, deliver stuff and manage data, there will be integration issues. An LMS is simply a single integrated piece of software. It’s that simple. You may want to do without one but you’ll end up integrating the other things you use – and that will be, a sort of LMS (thanks to Andy Wooler for this interesting observation).
6. Manages people
Large organisations need to manage large numbers of people, especially as they come and go. The LMS, linked as it should be to HR data, can ensure that the right people get the right learning. It also allows the organisations to guide learners forward, with essential, even legally required learning and desirable options for induction, relevant management training and their personal development. It gives people choices.
7. Manages stuff
Organisations have a lot of stuff to handle; induction materials, compliance, product knowledge, management needs, practical skills…. To operate in a complex environment you have to be as good as, then smarter than others. This level of complexity needs management. An LMS will manage, not only learners but what has to be learned. With newer authoring tools, such as ADAPT, and WildFire content can be made to look contemporary, with a move away from the linear page-turning paradigm.
8. Costs
Enterprise software may seem expensive but at the ‘cost-per-user’ or ‘cost-per-learner’ level, an LMS often makes sense. Newer approaches to the LMS, like the open source Totara or VLE Moodle, have changed the landscape, offering a lower-cost and more agile approach to the management and delivery of learning.
9. Security
There’s a good reason for having an IT department. They worry so that you don’t have to worry. There’s a whole load of scary problems around bandwidth, loading, legals and security that most people don’t fully understand. In this age of DoS attacks, phishing, hacking and malware, we should be grateful that these people are looking after our interests. An LMS is a controlled environment that can save us from ourselves.
10. Manages data
All organisations need some level of management data. Let’s not forget that in the supposedly good ol’ days, the only data you got from classroom trainers was dumb-ass, happy sheets. Data, with analysis, gives you insights. Insights are what managers need to make decisions and innovate. A good LMS spits out data and that is useful.
Conclusion

The LMS market has moved on with new players, open source options, xAPI and more data sensitive delivery. The danger lies with those vendors who just see more and more control as desirable, as opposed to a degree of looseness, if not chaos. The arguments between the LMS is dead and the LMS fanboys is often one between realists and idealists. In practice, we need a bit of both. The truth is that this is a multi-billion pound market that grows every year. It is NOT disappearing. There will always be a need for single solutions. I just hope that this does not descend into the mess that is the all-embracing, death-clutch that is ‘Talent management’.