[Xmca-l] Re: Reaping what you sew

Glassman, Michael glassman.13@osu.edu
Tue Jun 6 09:16:08 PDT 2017


Mike and interested others,

What is interesting is that the idea of using algorithms and gamification processes to learn is hitting some really big and hard walls.  It will be interesting to see if the Silicon Valley types respond to them or if their approach is mostly just thoughtless, ideological belief in the virtues of a certain and limited perspective on technologies and their applications in education (which of course also includes a little bit for the business models of the progenitors).  A lot of this is coming from machine learning, where computer scientists are finding out that learning is - you know - really complex.  For instance the two most powerful machine learning experiments Alphago and Libratus to not just rely and search and retrieve methods (which have dominated so much of cognitive science) but realize they have to use positive reinforcement incorporated into their rule systems (is it okay if I don't use the word algorithms?).  It is basic operant learning taken from your favorite tattered copy of Walden Two.  It is the only way to really move forward.  My guess is that soon they are going to find that learning is more complicated than that.  I was really amused by the story of the mother using Mathbox because her son was doing the math exercises to improve his avatar and not to actually learn math.

There are also the deep flaws in the basic associationist learning methods used in the big data models, related to the ones discussed in the article about figuring out and responding what students need at the moment.  One of the things people (hopefully) are finding out is that when you have big data that includes all the flaws and biases that exist in that data.  You might be able to code your way out of this but the coders would have to know a hell of a lot about socio-cultural theories.

It will be interesting to see if anybody owns up to, let alone addresses these limitations.

Michael

-----Original Message-----
From: xmca-l-bounces@mailman.ucsd.edu [mailto:xmca-l-bounces@mailman.ucsd.edu] On Behalf Of mike cole
Sent: Tuesday, June 06, 2017 11:37 AM
To: eXtended Mind, Culture, Activity <xmca-l@mailman.ucsd.edu>
Subject: [Xmca-l] Reaping what you sew

XMCA-o-philes-

Below is a story about the infusion of high technology pedagogical regimes into American schools. What struck me was both the extent to which the underlying pedagogy in some of these effort (rhetoric about student-driven educational
regimes) and the technological structure (individualization through use of learning algorithms) resonate with views one might find on xmca and/or are derived from research on the use of learning theory ideas that can be traced by the mathematical learning theory ideas popular when i was a graduate student.

Anyway, a sign of the times. For your amusement if you are so inclined.

mike

PS- Putting this url in my browser worked for me but its pretty long. Good luck.


https://www.nytimes.com/2017/06/06/technology/tech-billionaires-education-zuckerberg-facebook-hastings.html?hp&action=click&pgtype=Homepage&clickSource=story-heading&module=second-column-region&region=top-news&WT.nav=top-news



More information about the xmca-l mailing list