Day: 2018-01-06

AI in education


I’ve been focused on these two questions from my friend Peter McAsh in his recent post on the ECOO site.

“How will AI impact education?
How will education adapt to teach students who will be part of a world with AI?”

You can’t pick up any reading or news report these days without learning about some new take on Artificial Intelligence, Machine Learning, Digital Assistants, etc.  It’s natural to be concerned about education as Peter points out because these things are forming the world students and teachers will live in.

I mean, fast forward, to the perfection of self-driving cars.  Does that put a whole generation of driving instructors out of business?  If you are involved with computers and technology, you know that the fear of computers replacing teachers has been around seemingly forever.  And, for every fear story about computers replacing teachers, there’s a comeback interview or report indicating that it will never happen.

In the near future, students will have to deal with a world with imperfect AI as AI learns and grows.  Of that, there is no doubt.  I wonder how many classrooms will have Alexa or Google device sitting on the teacher’s desk when things resume next week.  (or earlier for my US friends).

And yet, they’re not about to change the classroom significantly.  Remember the old adage about don’t ask questions that Google can answer.  Your typical assistant falls into that category.  Watching Canadian television over the break, I wish that I had a dollar for every time an Alexa commercial was aired.  Still, this isn’t going to make or break things in the classroom.

I love playing with the current thoughts about AI/Machine Learning – https://experiments.withgoogle.com/ai but I want more than drawing recognition or being beat by a computer playing Rock, Paper, Scissors.

Screenshot 2018-01-05 at 08.39.31

Now, I’m not naive enough to believe that this is it.  It’s just the tip of the iceberg when it comes to Machine Learning.  I remember being excited at one of the CSTA conference outings to the University of California-Irvine where we had a presentation about Machine Learning.  I got an overview before the presentation before it started and almost forgot to introduce the professor who was going to talk about it and then almost forgot to thank him afterwards.  My mind wanders when I hear and think about the possibilities.

I think K-12 education is ripe for a good implementation of AI.  Now, before you skip the rest of the post to hastily reply that I’ve lost it, please read on.

We’ve seen the growth of how computers in the classroom are going to change everything.  Cross Country Canada was the beginning of a promise.  Then, we got excited because we could edit images with tools like Photoshop or Gimp.  Then, we got excited because we could run an add-on to Google Sheets to automatically mark multiple choice tests.  Then, school districts got really excited because we could do report cards on computer or we could send email to parents.  And the list goes on.  But, like a broken elevator, they don’t take us anywhere near the top floor.

This computer science teacher would argue that programming is the one are that hit the educational scene with the promise of delivering results in the classroom and continues to deliver.  Sure, the tools have changed, but the promise and the results are there and even better today.  You know what makes it work?  The presence and guidance of an informed educator or a peer working along side the programmer.

Despite this, we’re still not there.  I can’t help but think about the promise and the potential of AI.  My crystal ball doesn’t see any implementation that doesn’t involve an educator alongside a student.  I definitely think of AI as being more than just a bookkeeping tool though.  I imagine a world when AI is there and can inform you immediately when a student “gets it” and offers suggestions for moving on.  Or, when a student has challenges and needs remediation.  We think we have it now but imagine a world where that classroom teacher has access to everything and an intelligent resource that immediately offers the right suggestion and implementation.  How often are these moments missed and then only caught later when doing some sort of assessment?  I know it’s supposed to be happening now with ongoing assessment and that has varying results.

And then there’s the concept on online resources.  Who doesn’t collect links with content specific to a particular topic of study.  Most connected educators do so.  Often, they’ll have a nice collection of bookmarks, a blog or wiki to accumulate these sort of things  But, are they the best available?  Imagine AI that is constantly at work, making sure that links are still active, and is constantly looking for better resources than what you currently have.  Wouldn’t it be wonderful to know that you’re always working with resources that are best of breed.

Unless it’s a well-kept secret, I suspect we’re years and years away from it.  Right now, we’re fascinated with chess playing computers, or better GPS, or self-driving cars that do silly things.

Let your mind free of the shackles that we have now with our thinking about educational technology.  Where could the right AI take us?

Advertisements

OTR Links 01/06/2018


Posted from Diigo. The rest of my favorite links are here.