I have been reading Machine, Platform, Crowd, a book written by Erik Brynjolfsson and Andrew McAfee. It is not a book written for educators and the process of education receives little attention. I would describe it as a book about how technologies result in innovation and about the role AI will play in our futures and how AI will combine with human capabilities and expertise.
I must admit some of the insights from the book surprised me. I often protect my own self-interests by imagining that technology will advance human capabilities by taking care of the routine and the data-intensive and allow humans to take advantage of expertise and intuition. The authors immediately dispute the notion that human expertise is more productive than carefully constructed rule-based AI systems. Humans are biased and prone to stick to personal insights even when proven wrong. It is not the rule-based AI systems however that the authors see dominating the future. It is the type of AI that finds relationships in massive amounts of data resulting in effective predictive models they see presently offering the greatest opportunities.
I don’t necessarily recommend that educators read this book because of the discussion of the major approaches to AI, but because of the insights offered regarding technology and innovation. Allow me to offer a perspective without spoiling (or attempting to remember) all of the big ideas from Brynjolfsson and McAfee’s book.
Technology and Innovation
History does offer some opportunities for seeing present circumstances through analogy. The book begins with a description of the transition from steam power to electrical power in driving major manufacturing and in sorting out which major industrial powers survived and which did not. Huge companies and fortunes were developed based on being the first to master the use of steam power. Being first allowed the accumulation of wealth needed to continue to lead in the application of steam-based infrastructure. To be effective, a single power source (the steam engine) was used to run coordinated equipment through complex systems of pullies and belts. Plants with multiple levels were constructed to take advantage of this single power source. When electricity and the electric motor was introduced, some tried to just use it as a power source for the same approach. This fit well with the existing infrastructure and provided a small increment in efficiency. Of course, this is not the approach that ultimately ended being most effective and flexible. Modern assembly lines on a single level were more suited to multiple electric motors and required a different building design. Multiple motors also allowed greater flexibility in that an entire process did not have to be reworked should there be a better way to execute one of the components of that process. Hanging on to the one power source model and the expensive infrastructure that supported this model ended up resulting in the ruin of many once powerful companies who were too big to change or at least too afraid to innovate.
Educators might take different lessons from this description. Mine is likely different from most. I imagine the student as the factory. My take on learning – a form of constructivism rooted in the science of human cognition – sees the productivity of learning resulting from the individual learner (hence the learner is the factory). What the learner needs to be most productive is to be able to arrange the use of tools to optimize the use of his or her learning factory. Depending on the goals of the factory and the needs of the goal specific learning processes, one size does not fit all. Many educators might see the inflexible factory as the system of education, the school, or even the classroom. My recommendation – begin by attempting to understand the work of the learner and not the work of the environment surrounding the learner.
My point – we have yet to make use of technology to respond to the needs of the individual learner. Technology should be thought of as a way to individualize the learner experiences. Noting this opportunity is not new. I think my take on individualization prioritizes different things that many education pundits promote. In considering my personal recommendations for individualization, I have decided that I see the education as meeting the goals of both learners and society. I am not a proponent of learn whatever you want whenever you want to know it. The self-indulgent model fails to recognize the importance of some common goals to society.
Individual differences technology can presently address:
- Starting point
- Aptitude differences in speed to learn
- Individual differences in content interests
The classroom may fail to support the individual learner in these ways. I often argue the first two factors which interact in practice as demonstrating the need for mastery learning (an old term, but why change the term when the problems remain the same). These are lockstep systems (group-based instruction) which operate with limited regard for what learners already know and for differences in their speed of learning. More than this, when the optimal speed of learning is ignored many learners are either held back in what is possible or develop greater starting point problems (lack of background knowledge). For the second group of students who are constantly left behind, learning becomes so frustrating because of the lack of background knowledge or skills that declines in motivation kick in creating all kinds of further problems. To use the presently popular notion of mindsets, how can you convince someone to approach learning with a growth mindset when their personal history lacks any evidence that such an approach is realistic?
Content interests mean that learners are allowed to pursue different things. There are shreds of this logic in allowing students to select course options, to pursue different majors, 20% or passion projects, or to select a different book to meet a reading practice expectation. I do think it important to consider when the goal is to encourage the development of different declarative knowledge or procedural skills and when the goal is to encourage the development of specific declarative knowledge and procedural skills in the different ways. These are very different issues. We may want all learners to be effective critical thinkers and problem solvers or to understand the basics of our system of government. We may also want some students to be able to explore computer programming and others to explore instrumental or vocal music. These are different categories of goals, both are important, and both can benefit from the benefits technology offers learners.
I attempt to acquaint students in the grad courses I still teach with the notion of blended approaches to education because the concept allows most of these ideas to be integrated. Technology is what allows the transition toward a blended approach much in the say the electric motor allowed a transition away from the less flexible steam powered approach to manufacturing.
