<<back to publications<< ........    ...........>> to Teach0logy.xyz>>

A comment to “No Child Left Behind Is Gone, But Will It Be Back?”, by Melissa Tooley, published in “The Atlantic” on 12/24/2015 (http://www.theatlantic.com/education/archive/2015/12/no-child-left-behind-is-gone-but-will-it-be-back/421329/)

Thinking about becoming a STEM teacher? Think again!

Valentin Voroshilov, teachology@teacgology.xyz

Thinking about becoming a STEM teacher? Think again! Soon you may be at risk of being replaced by a robot.

Current progress in artificial intelligence (AI) is fascinating. Various devices can see and recognize us, understand our questions and tell us the answers, guide us toward our destination, advise us on our health, even buy and sell stocks without human intervention.

IBM’s Watson computer has beaten the world chess champion and won Jeopardy. Recently in Japan AI Todai Robot passed a college entrance exam with 80 % chance of being admitted into more than 400 universities. In the US AI GeoS system can solve SAT geometry problems at least as good as an average student.

So, it is not totally unreasonable to expect that soon computers will learn how to solve standard high school math and physics problems after reading them or listening to them. And then it would not take too much time to combine this AI with an advanced computerized tutoring system, and here it is – a robot teaching math or physics.

We talk about math, or physics, and maybe other STEM courses, because those subjects are highly structured and have a very well known reasoning logic, which can be described in terms of specific models and algorithms (similarly to the game of chess).

There are only two reasons for delaying the upcoming “invasion” of robots teaching STEM courses.

Firstly, teaching is more than just managing human knowledge; it also involves managing students’ behavior during a lesson (the simplest model of teaching is: teaching = tutoring + class management). So, we could imagine a class where a robot (or robots) works hand in hand with a human assistant helping managing a class (and who does not require a deep knowledge of a subject). That would require however a major change in education policy.

Secondly, even one-on-one teaching involves a lot more of interpersonal communication than can be provided by even the best computerized tutoring system.

All such systems are based on the logic dictated by a subject: e.g. “if you need to learn H first you need to learn G, and before that F, etc”. Those systems do not take into account the fact that different people learn differently. In fact, currently there is no computerized tutoring system which could recognize learning style of a student and tune up accordingly the tutoring process.

An experienced tutor knows very well that depending on a student the same topic might need a very different presentation and explanation.

There are vast amount of studies of how people learn which describe a general process of developing or acquiring new (for a person) knowledge. Those studies often include the notion of “multiple intelligences”, or “student mind modalities”, or “learning particularities/styles”, or “individual traits of students”, or similar, acknowledging the existence of differences in learning processes of individuals.

However, despite the fact that we know rather well how people learn – in general, no one can describe the best teaching strategy which should be employed to teach a given subject to a given student.

By “a given student” we mean a student of a certain gender, age, race, social background, economical stratum, cultural traditions, behavioral traits, psychological and cognitive modalities, scholarly experience, and, of course, learning stiles. Clearly, if we use those parameters to classify ALL current students, the classification will have a large, but finite (!) number of classification classes (groups, “cells”, elements).

Accepting the fact of the existence of such classification leads to a natural question: is teaching approach effectively used to teach a subject to students of a given class (from the classification described in the previous paragraph) also effective for teaching the same subject to students from another class?

Based on everything we know about learning the answer to this question should be: “No”.

The next question, naturally following the first one, asks: what is the most effective teaching approach/technique/method/tactic/strategy which should be used to teach a given subject to students from a given classification class?

The answer to this question currently does not exist.

Even worse, currently there is no research group which would be looking for the answer to this question.

Every developed science began from a classification: astronomy – from classifying stars and planets, biology – from classifying species, chemistry – from a periodic table. In this sense, there is no yet science of education. There are of course numerous scientific activities – like alchemy had before it became chemistry. Although study of education yet remains in a pre-science phase – like alchemy was before it became chemistry.

However, presently there are no reasons for not developing the classification of types of learners and for not designing the best teaching strategies corresponded to each class of that classification. Except one – it would be a highly financially consuming project involving a large team of experts – a “Manhattan Project of education”, so to speak.

Such project would need designing and building from scratch a large school and supportive facilities, where hundreds of kids would be learning all possible subjects under the guidance of the best teachers closely collaborating with the best educational scientists (that’s right! – if there is no yet a science, it does not mean there are no scientists; they are the people who are building the new science via various scientific activities).

Seventy-five years ago only the government could finance such an ambitious project. But seventy years after the end of the “Manhattan Project” there are now private citizens who could provide the comparable funding (and creating a “Space X project of education”, so to speak).

Why neither the government nor large philanthropists want to invest in creating the science of education?

I do not know.

Maybe scientists from different schools of thoughts on learning and teaching cannot find a common ground on what learning is and how it should be facilitated and guided.

Maybe this field seems not as “sexy” or commercially viable as space tourism, or a self-driving car, or a robot that disobeys a human.

The co-creator of the Breakthrough Prize Yuri Milner, for example, pledged to invest one hundred million dollars into the search for extraterrestrial life. It did not occurred to him, though, that many teachers look at students they teach almost the same way they would look at aliens from “a galaxy far far away”.

So, if you are thinking about becoming a STEM teacher - think again. Soon you may be at risk of being replaced by a robot.

However, if the current trends in science of education will continue, that risk will not be significant enough to take it into account.

Is this a good thing or a bad thing?

No one knows.

The best we can do these days is to open a discussion.


What should be taught and how - has been known for DECADES. The real question is why despite new laws we have every 10 – 15 years in 10 years we need a new law again to fix what had to be fixed by the previous law? Possible answers are there:

1. “How is The Third Program of the USSR Communist Party related to education reform in the USA?” http://www.teachology.xyz/3pc.htm

2. “Why have hundreds of millions of dollars been spent on developing the common core math standards if content-wise they are not much different from the ones they replace?http://teachology.xyz/3r.htm

3. “Education reform needs a new paradigm.” http://www.teachology.xyz/np.htm

<<back to publications<< .......        ............>> to Teach0logy.xyz>>