Risky predictions about the future of e-learning
The Guardian recently republished an article from 1964 about a new venture called the University of the Air. When it was launched, this University became better known as the Open University, and since then has helped millions of people to follow courses without going to traditional universities or college.
Looking back to events like this also tempts one to look forward. So let's speculate about how online learning will develop in the coming years and consider some of the problems the industry still has to solve.
Well aware of the observation that "making predictions is hard; especially about the future", here are my forecasts:
The e-learning of the future
Online learning began as a data dump. Vast quantities of written material were digitised and published with no thought about how people learn or the differing requirements of online. We were still in the age of the book.
Current trends suggest that online learning will be more adaptive. As we learn, software will learn about us and how we respond to content, questions and challenges. It will also tailor difficulty to our levels. A student of climate science, for example, might quickly learn about atmospheric feedback but struggle with the idea of tipping points. Software will be able to recognise this individual's progress and accelerate a course or slow it down to suit.
The e-learning of the future will be more user-friendly. User interface designers have already swept away the idea that e-learning needs to copy books or present a series of YouTube lecture videos. Instead, content is already becoming more fragmented and interactive. It uses imagery, sound and video to reinforce points, and it can be dropped and returned to much more easily. The ideas of people like Brett Victor will reach the mainstream.
Content will be much better at providing feedback to learner behaviour – something teachers remain best qualified to do at the moment. Take a look at this answer to a problem of proportion:
The learner has looked at the change in height and noticed that it increases by 6. Therefore, she thinks, you should add 6 to the width too. Most e-learning tests simply say this is incorrect. If the learner had written 26 or 106, it also says it's incorrect. An experienced teacher, however, will note that the learner is on the right track even if the wrong answer came out. Human feedback is therefore able to gently guide the learner back to the correct path and eventually elicit the correct answer. In future, software will be able to provide more intelligent feedback and guide learners to the right answer. There's no question that teachers will still be with us in ten and even a hundred years, but they will have the opportunity to work with the robots rather than be replaced by them.
What about economic models? Here we get onto shaky ground and predictions get less certain. The development of the MOOC seems to point to vast amounts of free learning in future. And the fact that there are free YouTube tutorials on everything from nuclear fission to flower-arranging suggests that in future learning will be peer-to-peer and ordinary people will be teaching colleagues, friends and complete strangers without charging.
But it seems more likely that current models will continue largely unchanged. The state will provide some education where it thinks best, ordinary people will also teach online, and there will still be a large fragmented private sector market for training.
The risks of predictions
"Predictions are preposterous", Jackie Mason once said. In the field of online learning, rapid technological change makes this comment particularly relevant. Who knows whether neuroscience will suddenly discover new learning models or whether the whole business of learning will be abandoned by governments?
Whatever the case, human beings will still need to learn. The only question is how this will happen.