Filed Under B2B, Big Data, Blog, Education, eLearning, Financial services, healthcare, Industry Analysis, internet, Pearson, Publishing, Search, semantic web, STM, Uncategorized, Workflow | Leave a Comment
“Well”, said my questionner, “all this stuff about workflow is just fine, but what we really need to know now is where this goes next. So lets just imagine that we integrate all the data into all the systems, and everyone on every screen has all they need to be more productive, make better and more cost-effective decisions and to be wholly compliant with all relevant regulation and best practice. What happens next?” There is, I have come to know, a certain class of manager (possibly deprived of breast feeding at too young an age), whose cruel sport is seeking to confound itinerant consultants with all this “next” guff. Hopefully, I had been thinking, workflow and the semantic web would see me out. But you have to answer the question, so I looked at the ceiling in what I hoped was an image of wisdom and muttered something about intelligent systems. And I ignored him when he challenged me to show him one.
But it was a bit worrying. I guess the answer is that far more of the tasks over which we slave will become subject to machine to machine communication and an increasing range of Artificial Intelligence (AI) applications. I remember when AI was seen as the “always there, never delivers” technology (just like GIS in the 1980-1990s). Yet GIS in the broadest sense gave us spatial location and even SatNav. AI is yet to hit the point of large scale integration with what we do with content and data in solutions, systems and services. And I was worrying about how little I knew in this area when I bumped into Sir Isaac Newton’s dog, Diamond.
I bumped into him in a March 2012 paper published by Dr Glenda Eoyang, Executive Director of the Human Systems Dynamics Institute (www.hsdinstitute.org). She is an expert in adaptive systems, and having pointed out how comparatively easy it is to learn and assess learning of something like Sir Isaac’s equations, she then turned to his dog. Here was a learning problem of a different type, as your experiences of the dog altered your learning perspectives and your learning had to become “adaptive”:
“On the other hand, when you learn Newton’s dog, the expectation is that you develop the ability to 1) recognize Diamond; 2) interact with him; and 3) get better at recognizing and interacting with him over time. The assumption is that teachers, too, recognize and interact with Diamond, and that their performance continually improves through a process of life-long learning. The purpose of learning the dog is adaptation. The measure of success is adaptive capacity. The best pedagogy is adaptation, and adaptive action is also the best way to assess performance. Learning Newton’s dog is about engaging with an ever-changing environment in ways that are creative, courageous, sustainable, and sensitive. When the goal of education is to prepare the younger generation for a complex and emergent and unpredictable future, we must teach them Newton’s dog.”
Quite so. And here, clever questionner, is part of your answer. The next leap forward looks like it will be in education, as we begin to tackle personalised learning en masse. And that takes me back to Knewton Technologies (www.knewton.com), whose deal with Pearson last November will lead to some 10 million users of MyLab/Mastering (mostly US college students) having coursework which adapts to their pace and their needs. So they get a test, and the machine does them some diagnostics and sets them off on a new route (or rote)? No, the system is continuously adaptive, it tracks everything they read and write, and it continually adjusts and resets, not just for strengths and weaknesses, but for each individual’s unique learning style. No less a learner than Bill Gates has praised the two first year college maths courses launched last year at the Arizona State University, and Pearson and Founders Fund led the new investment round which raised $33 m in the fall of 2011 for Jose Ferreira’s brilliant start-up.
It is unlikely of course that Knewton will be the only answer. One man’s proprietory algorthym is another man’s challenge, as we have learnt many times in the last 22 years. But what interests me at the moment is that our first approaches to mass customization and personlization in depth should be in education. In education we know for sure that even if we try to teach a class of 30 from the front, a certain percentage will be bored because progress is too slow, and a larger group will be bored because they lost the thread. For a handful the teacher will hit the mark. So we continually, and rightly, stress that learners should learn for themselves and collaboratively with other learners, and the teachers’ role is to moderate, not drive, the process.
And yet, customization does not yet get a look in elsewhere. We try to build workflow as if every company was the same just because it has the same end objectives in terms of revenues or margins. Yet cultures and the ability to improve performance are idiosyncratic and often unique. But we want to sell a “productized” service, not a solution per client, and it may be that the technology of the classroom is in fact taking us towards supplying individual needs without re-assembling the coding. Maybe, indeed, adaptability becomes the new service offering, down the road that takes us ever closer to AI – penetrated services.
Thats All, Folks! There now follows a short intermission before I return in October.