Tuesday, April 30, 2013

Can Computers Replace Teachers?

Can/Should computers replace teachers? It is clear to me that computers are not capable at this time of eliminating the role of teachers. But it is also evident that the number of students one teacher ought to be able to teach should be increased by use of technology for the mass presentation part of lessons. That should free up teachers for the parts that computers aren't yet able to handle, like open ended Q&A time, evaluating the quality of essays, projects and talks, and perhaps providing individual tutoring. The human aspect of teaching, which goes beyond the narrow bounds of "lessons" and covers things like motivating, comforting and counseling are additional weak spots in anticipating what a computer can do at this time (but do see the video linked to my blog post "Reading Faces"). It isn't so clear to me how a computer would cope with an unmotivated unruly student.

Not surprisingly, teachers are upset at the notion that fewer teachers would be needed in a school equipped with sufficient technology and that increasing the student/teacher ratio could result in big budget savings. But it is great fun to watch a good fight. To that end, I share with you today this 21-minute talk "Can Computers Replace Teachers?" by Katherine Mangu-Ward, and in rebuttal, see this blog post by a teacher, Joe Bower: http://www.joebower.org/2013/04/can-computers-replace-teachers.html. He provides a link to another article that is severely critical of the KIPP schools: "Poor Teaching for Poor Children...". I briefly talked about KIPP schools in my blog post about the book "How Children Succeed".

You may have noticed in my blog article that there were already differences between the approach of a KIPP school in an impoverished neighborhood vs. the approach taken by the more upscale Riverdale Country School. I suppose the moral of the story is that different kids have different needs and the "system" (whether that means a computer system, a school system or a teacher) needs to be able to sense and adapt to those needs. Perhaps this argues in favor of establishing separate tracks for different academic abilities and perhaps even for different social backgrounds. Now there's a controversial area! If the tracks are separate, how to assure that they are, in some sense, equal to each other? How "equal" ought the tracks to be?

I suspect the actual savings provided by use of technology is significantly smaller than the vendors project it to be. The "Total Cost of Ownership" over the long term has to factor in the cost of maintaining the network and systems and replacing obsolete equipment and keeping the software up to date. Prefabricated, canned, lessons need to be refreshed frequently to keep up with changes in the world. Arguably, math is a subject that doesn't need much change from year to year, but social studies clearly needs ties to current events to stay at all relevant. Biology is an example of a field with rapid change from year to year. Unchanging lessons in such a subject will be terribly far from state of the art in just a few years if significant effort isn't budgeted to keep the material current.

Perhaps in the future the pipeline for educating teachers will have different tracks for teachers interested in working directly with the students vs. folks interested in developing the class materials (the software). Of course the 2 sides need to stay in touch with each other, but I can see the need for fairly distinct skill sets.

As I've noted in past blog posts on the topic of "education", I'm a software engineer, not a professional teacher. I'm particularly interested in hearing other points of view on this topic of technology as a possible way to displace teachers. Feel free to share opinions in the comments section here or share a link to a blog post of your own if you'd like to have a less confined space to write in. e.g. Given a "canned" self-paced course like Udacity CS101 or some other MOOC course, what role, if any, is there for a locally available human teacher?