Subscribe
Share
Share
Embed
The voice of the American Consortium for Equity in Education at ace-ed.org | Host Larry Jacobs facilitates rich discussions with innovative educators, thought leaders, authors and the leaders within the education industry to promote equity, access and opportunity for every student in every school.
I'm David Cicero, and this
is Education Talk Radio.
One thing I've noticed is that
conversations about educational technology
often begin with the technology itself.
We ask whether AI belongs in schools,
whether students should spend
more or less time on screens, or
which products are most effective.
But is that the right place to start?
What if instead we asked what
kind of learning experiences
are we trying to create?
And only then do we ask whether
technology has a meaningful role to play.
Today, we're exploring
that question with Dr.
Chris Dede, senior research fellow
at the Harvard Graduate School of
Education, and one of the world's
foremost researchers on educational
technology and learning design.
Together, we're gonna explore how leaders
can think differently about technology,
not necessarily as the starting
point, but as one of many tools for
creating powerful learning experiences.
Chris, thank you so much
for joining me today.
David, I appreciate your inviting me.
So I, I kinda wanna take
a, a, a step back, Chris.
We-- I feel like we have spent years
asking what technology can do for
schools, and what I wanna ask you today
is, from your perspective, what should
technology make possible for learners?
So I'm, I'm always wary when people
start with technology because then you
have a solution that's looking for a
problem, which is never a good thing.
It's important to realize that technology
itself is not the innovation, but
it is a catalyst for an innovation.
So a lot of reporters will ask
me, um, you know, "Does artificial
intelligence improve learning?"
Or, "Does, you know, um,
VR im-improve learning?"
And my answer is, um, do
books improve learning?
And, and of course, the answer
to that question is, it depends
on the book and how it's used.
And in the same way, um, what we want
to answer is if we know collaboration
improves learning, so does the
technology improve collaboration?
We know that, um, students struggle,
uh, with engagement in school, so is the
technology somehow improving engagement?
Um, we know that formative evaluation
of student work is powerful in
improving that work, so does the
technology aid in formative evaluation?
In other words, when you start with
something that we know based on theory
and evidence improves learning, and then
we customize a technology to increase
that variable, then we can say with some
confidence, although we still would want
to do research and evaluation to verify
this, that the technology is likely
to have a positive impact on learning.
But the mere presence of a
technology in the classroom, not
so much because it's not magic.
I'm, um, a co-PI and the
associate director for research
of one of the National Science
Foundation's, uh, National
Artificial Intelligence Institutes.
Uh, the particular one I'm involved
with is the National AI Institute on
Adult Learning and Online Education.
But there are four other national AI
institutes that are centered in education.
And interestingly enough, one of them
is about collaboration, one is about
motivation and engagement, one is about,
um, different ways of providing, uh,
formative evaluation, one is about,
uh, diagnostics, uh, for students
in terms of identifying difficulties
that they may have in learning.
And then our adult, uh, institute is
about, um, the kind of factors that
we know make it possible for adults to
learn in the context of the workplace.
So all of the national AI institutes
are starting with a theory of learning
and then asking about whether artificial
intelligence can make a difference,
as opposed to starting with artificial
intelligence and then saying, "Let's throw
it at the problem and see what happens."
Right.
And n-now that you're saying this, you
know, once you, once you s- once you see
something done, it, it, it feels a bit
demystified, and it's like, oh yeah, that
makes sense that you would do it this way.
The, what you're saying 100% makes
sense, but I feel like That's not
what a lot of schools are, are doing.
I think in many cases they're
doing what you just mentioned
a moment ago, "Hey, here's this
technology, let's throw it at it."
Or as I am thinking towards, um, the next
question I actually had written, a lot of
these conversations become about outcomes.
"Hey, we have this, um, you know,
let's find something that's known
for increasing these outcomes."
And I'm, I'm not so sure the
evaluation part of that goes as
deeply as what you've mentioned.
Okay, here, here's learning theory.
Let's evaluate this in terms of, um,
these pieces and really figure out if
it is going to elevate what we know, um,
influences learning in a positive way.
Why do you think conversations
end up being about that in many
cases, the outcome specifically,
and never sort of make it to how
the technology is tied to specific
learning outcomes that we know work?
Well, it starts with the vendors, I
think, because the vendors are selling
technology And so they are in a position
really of saying, "Here's a solution.
Whatever your problem is, this
is going to be a solution to it."
And we see, uh, organizations in response
setting goals like, "I want to see every
single teacher using AI," or, "I want
to se- every single teacher to use an
automated grade booker," or whatever the
technology happens to be, and that's not
at all the right way to think about it.
Technology isn't like a vaccine.
With a vaccine, it, it really, once you
have it, um, it just works, and there's
no particular effort involved in your
part on whether it succeeds or fails.
Technology in education
is more like antibiotics.
Antibiotics are very powerful, but
a lot matters on how they're used.
If you just grind them up and
smear them all over your body,
nothing is going to happen.
If you put them on a altar and worship
them, nothing is going to happen.
If you take all of them at once,
something bad will probably happen,
and it only works if, if you take it
in the prescribed manner, which is to
keep a relatively constant level of the
antibiotics at a reasonable amount in your
bloodstream, a- and then it works great.
So the issue is that there are conditions
for success for learning technologies,
just as there are conditions for success
for antibiotics as opposed to vaccines.
And unless we pay attention to,
to the conditions for success,
they're not going to work well.
But the message from the vendors
is more, "Oh, this is, you know,
just do it, and, and good things
will automatically happen."
History proves that that's
not an effective approach.
Right.
Yeah, I, I have worked in educational
technologies for a decent part of my,
my career and, um, many schools I ran
into and worked with, we would- You
know, you would come to a point where
you would realize that the school you
were working with sort of interpreted
what you were selling or implementing
as, quote unquote, "a silver bullet."
Just plug and play.
Like we're going to, um, implement this,
and it's going to solve all our, our
problems or solve whatever the problem
is that we were hoping for it to solve.
And then you get things like,
um, you know, difficulties and
barriers during implementation.
You also get, you also see, um, I don't
know what the right word is, but it, um,
the, the standard they hold it against
of whether or not they'll continue to
use it becomes a very short timeframe,
a year or a year and a half, and they're
ready to move on to, to something else.
And I always wondered why, um,
you know, do you think this is a,
a, a cultural thing in the way we
do schooling in, in this country?
Um, that we, you know, where we're
looking for that silver bullet, and
we're expecting things to happen
quicker than maybe a technology
implementation should succeed.
Are, are you seeing that as well?
That's what I've experienced in, in
my time in educational technology.
I, I like the way that
you're framing that, David.
And I think, again, if I can use
the metaphor to different kinds of
medications, um With antibiotics, you,
you don't take it and then, um, one day
later you say, "Well, this isn't working.
I better try something else."
You understand that it takes days
for the antibiotic at a sustained
level in the bloodstream to be
able to tackle whatever kind of
bacterial infection it's up against.
And, uh, this is similar to
other kinds of medications.
Not every medication, but other
types of medications where
dosage over time is important.
And everything we know about learning
suggests that dosage over time is the
right way to think about it, that you
don't suddenly have a blinding flash of
insight and you understand something.
Rather, you build your knowledge around
what you already know, and, um, some of
what you're told, you then reconstruct
in your mind until you've framed it in
a way that you can understand it and add
it to your body of knowledge, and so on.
So there are no magic shortcuts, but
there are plenty of people who, in
education, uh, year after year, decade
after decade, who are selling magic
beans, and there seems to be an unending
supply of people in education who are
willing to buy those beans and then once
again are disappointed with what happens.
I really like the way you, you,
you framed that, um, as, as well.
And so what I, what I kinda wanna do now
that we've sort of level set, um, and,
and hopefully kind of reframed our think-
our own thinking here about, um, what
technology should do and, and perhaps just
lightly touched on how schools should see
technology, I kinda want to break down
the different roles technology can play.
So I, I, I think we often think
educational technology, and we
see it as one thing or maybe one
classification or category of products.
How would you break it down?
What different roles can technology
play in teaching and learning?
One of the things that I find very
useful is drawing on research about
how people learn, and the National
Academy of Sciences released its first
book on how people learn around the
year 2000, and then they released
an updated version in 2018, um, that
deepened our understanding of culture
and context in terms of how people learn.
And, and so when we categorize
technologies, it seems to me, given what
we've said so far, that what we wanna do
is look at how people learn and then say,
"Is this a technology that affects this
part of learning or that part of learning,
or maybe two or three parts of learning?"
But what I find myself doing with,
uh, many technologies is asking which
are about doing things better and
which are about doing better things.
Because I have a definite,
um Preference for the latter.
So doing things better means that we
take the essentially teaching by telling
and learning by listening model that's
characterized industrial era education
for a couple centuries, and we find
ways to make it somewhat more efficient.
So, um, ninety-five percent of
what I read about AI in education
is about doing things better.
Look, you can take your, your syllabus
and turn it into a PowerPoint.
You can take a reading and turn it into a
podcast, and so on, and so on, and so on.
But the real gold in any new technology
is doing better things, doing something
that wasn't possible before because
the technology now makes it possible,
and when that something is, is
related directly to how people learn.
So let me give you a couple examples.
Uh, the National AI Institute
that I'm part of on adult
learning and online education has
developed seven different tools.
One of those tools is called Vera,
and it's a way of taking a videotape,
an instructional videotape, and
making it interactive so that you
can pause the videotape at any
point, and you can ask questions.
And the questions not only can be about
what is in the videotape, the questions
can be about the larger context of
knowledge in which the videotape was made.
Now, there are many people who
learn better from video than they
learn from books or, or lectures or
other standard forms of instruction.
We see that in the popularity of the
many explanatory videos on YouTube.
But it ta-- it's going a step further.
It's doing a better thing when you ask a
question of, of an instructional video,
and it actually can answer in a way
that addresses the specific issues that
you have with learning the material.
There's another tool called Sammy.
Sammy makes it possible in a large
online class for students to identify
potential students in the class who
might be good learning partners for them.
It's, uh, like an in-in-intellectual
dating service, if you will, where you
get some possible matches, and then you
have the option of following up on some
of the possible matches, reaching out
to them, maybe doing a little learning
together and seeing how well that works.
Now, we know that if you have effective
groups of learning partners, that
online learning becomes much more
powerful, both in terms of engagement
and in terms of the learning.
But it's just not possible for an
instructor in a large online course to
somehow facilitate the accurate formation
of these kinds of learning groups.
And so here again, you have AI making
it possible to do things better.
So I'm suggesting two things.
I'm suggesting that the logical
categorization for any technology is
to look at all the ways that we make
learning more powerful and tag it
with which of those ways it can help.
But I'm also suggesting that if it's
simply about making sort of weak
models of instruction somewhat more
efficient, as we see with teaching by
telling and learning by listening, it's
going to be less valuable than if it
does something new and powerful, like
collaborative learning or personalized,
uh, response to questions, uh, that
wasn't possible before the technology.
One of the takeaways I, I would
love listeners to walk away with
is just how to think about-- Right.
So the, our audience is building
leaders, superintendents, um,
district leaders, teachers.
If we shouldn't make a decision
about technology because of the
technology itself, and we shouldn't
make a decision about technology
just simply because, "Hey, I see
negative outcomes in student scores.
This vendor says that
they'll take care of that.
Boom.
Let's buy."
We shouldn't do that.
One of the ways in which we-- I've learned
from you to think about a role technology
can play in teaching and learning is
technology can help us not just make
what we're already doing better, but it
can show us how we can do better things.
And so I kind of have that as
a, as a checkbox, maybe as I'm
evaluating technologies to solve
whatever, whatever problem I'm
hoping to solve in my district.
How do we, as education
leaders, look at our curriculum?
How should we be thinking about the
relationship between the learning
experiences that we want students
to have and the role technology
may play in creating them?
Can you help us just think about
how we walk back from that?
If we're not going to make a
decision based on the tech, based
on the outcome, how-- at what point
does the technology come up in the
conversation, and now I can decide
if technology has a meaningful role?
What should that look like?
So, um, I think that the place
to start is to see learning
technologies as sociotechnical.
Too often the companies just sell
something that's technical, and so you'll
say, uh, if you wanna use AI, you need
to do prompt engineering, and you need
to understand vibe coding, both of which
are technical, and then you're set to go.
But technologies are also always
sociotechnical in education, and they are
implemented by human beings in contexts
that are determined by human beings.
And one of the reasons why so many
educational technologies that have
come out of research, including
technologies that I've helped to
develop, have increased our knowledge,
but they haven't gone into widespread
use because it's very difficult to
design something for the sociotechnical
environment of today's classrooms.
Let me give a specific example.
One of the other national AI institutes,
um, is about helping teachers to form
groups in which students can construct
knowledge, so that instead of giving a
lecture in the front of the room that
transmits knowledge to students, you
have, say, ten groups of three students
who are all working in, as individual
groups to construct knowledge together.
We know from theory and evidence that
that's powerful, but we also know that
it's very difficult for a teacher to
orchestrate, that if you're standing
in the middle of ten groups, each one
in a different place, it's hard to know
which group to go to next and how to
help them if they're getting stuck.
Yeah, and the professional
development that you need isn't
simply in how to use the technology.
It's in how to effectively use groups
of students working together to
get, develop their critical thinking
and their ability to, uh, go beyond
routine problem-solving, and so on.
So, um, I think that the challenge
is to recognize first that it isn't
just investing in the technology, it's
investing in professional development that
creates the conditions for success for the
technology, as we talked about earlier.
And then it's about respecting the fact
that for human beings to change how
they're doing something takes time.
And typically, if you change what
you're doing, things get worse before
they get better, because you're going
from something that you're deeply
familiar with doing to something where
you're figuring out how to do it well.
And even though it may, in the long
run, be much more powerful, that
doesn't mean that from day one you're
going to be seeing all of that power.
So design-based implementation research,
DBIR, is about understanding that
between the design, design-based, and
the research, studying what's happening,
there needs to be a sophisticated
implementation, and that implementation
is sociotechnical and then focuses
on professional development as well
as just developing technical skills.
But what I'm hearing from you, in some
way, what I, what I kind of drew from
that when you were talking was, you
know, a lot of times, um, the length
of the implementation, when, when
schools buy a technology from a vendor,
that's part of the sales process, is
talking about how long implementation,
how long until we're ready, right?
How long until we're using this, what
we would say, with fidelity, right?
And kind of what I took from you is
that's not a great way to base your
decision, just because implementation of
X solution is shorter than Y solution,
that's not a good judgment call on the--
how the technology will impact learning.
And I also, and you correct me if I'm
wrong, what I'm also taking away is
some of the most meaningful technology
is, is going to take some time, right?
And so we should expect,
um, you said it's theâ¦
I'm, I'm forgetting the exact,
the exact phrase you used, but
that technology is socioâ¦
What did you say?
Sociotechnical.
Is sociotechnical,
right?
And it sounds like the more a potential
technology has the potential to help us
do better things, I'm wondering if that
technology is even more sociotechnical.
That implementation might be a little more
involved, and thus, if we're looking to
implement meaningful technologies that
are really going to impact learning and
help us do better things, our idea of how
long, how long a technology should take to
be well implemented might need to change.
Am I onto something here, Chris?
Am I, am I thinking about
this in the right way?
Absolutely, David.
And in, in fact, um, I've been struck
as I read, you know, articles about
AI and surveys that are done of
educators and students, how students
in particular, but some educators
as well, will say, "I really like AI
because it takes this complex thing
and gives a clear and concise answer."
Well, the one thing I can say
about a clear and concise answer
to something complicated is, um,
you're guaranteed it's wrong, right?
You're, you're guaranteed that
something complicated cannot be
captured in a clear, concise answer.
So it's fool's gold in a way that,
um, people are selling with AI when
they say, "Oh, you know, it's so smart
that it can simplify things to the
point that I can understand them."
No, it's simplifying things to the
point that you don't understand
them, but you think you do.
And I think this is true for
educational innovation as well,
that there is no magic bullet.
There is no route that, uh, very quickly
leads to some kind of transformation.
Transformation is certainly possible,
but it is a process that requires a lot
of investment, a lot of time, clarity
of purpose, the willingness to not be
deceived by early outcomes that may
not be as positive as one would hope.
And in many other sectors of
society, people understand that.
I mean, in medicine, for example,
people do not expect miracles.
And in fact, if a drug or a medical
intervention promises miracles,
they're rightly very skeptical about
what the side effects might be or the
issues associated with it might be.
But education seems mired in this
kind of silver bullet mentality,
and unfortunately, a lot of the
ways that technology is sold and,
and, um, taught, um, reinforce
that rather than undercutting it.
So what I, what I kind of want to do
is quickly, I wrote a few things down
while you were talking, and I just
want to not fully summarize it, um, but
what some of the things we were talking
about, um, were how schools should
think about technology, implementing
it into the curriculum, and the sorts
of things that they should look for.
And we're not giving a list of things
you should, you know, you know, 20
things that each technology should do.
I think some of the major takeaways
were it should be based on
teaching and learning research,
what we know works, right?
It's not the technology itself necessarily
that's working, but we know collaboration
works, engagement works, right?
We should be looking through, um,
we should be starting with maybe the
problem is, you know, testing outcomes,
and we should be then investigating
what parts of teaching and learning
are breaking down in order, um, you
know, that, that may have caused this,
looking at lesson planning, looking
throughout the curriculum and saying,
"Okay, you know, we-- collaboration
could be a problem, engagement could
be a problem," starting to identify
those underlying problems, then looking
at technologies But through that
lens, we have a collaboration problem.
Which technologies are best at that?
We have an engagement problem.
Which technologies are best at that?
So that, I feel like that
was one big takeaway for me.
And then the second one was, in that
process, as you, you know, as a school
may approach vendors eventually after
we sort of know the underlying issue,
and we, we s- issue, and we know what
we're looking for, we need to sort
of, you know, this may force a rethink
of the length of the implementation.
Um, because we sort of established a
few moments ago that, you know, a lot
of the most meaningful technology is,
is likely to, to take a while, right?
It's not a vaccine, right?
So we need to think about change
management and changing the, our thought
process around what a technology imple-
implementation looks like if we're really
searching to solve that core problem.
Um, so Chris, I, I, I feel like we've,
we've spent the, the last good while
thinking about different way leaders
can think about technology in general,
what it can do for, um, learning.
And instead of asking for what
tech to buy, we've sort of been
asking what kinds of learning
experiences we're trying to create.
And so once leaders begin thinking
that way, seems like an entirely
new set of possibilities open up.
And so since we've sort of reframed
the role of technology around learning
experiences, what kind of learning
experiences are becoming possible
today that weren't practical before?
Well, something that I spend a
lot of time thinking about is
immersive authentic simulations,
and an example would be negotiation.
So negotiation is a skill that's sort
of universally useful in life, um, both
in, uh, occupational situations like
negotiating a raise from your boss, in
economic situations like, uh, working
with your landlord or working with a used
car dealer, in family situations where
different members of the family want
different things, and so on and so on.
And the ways that we teach negotiation
historically, uh, while, while, you
know, people have worked very hard on
this, ultimately have failed because
there wasn't a good assessment mechanism.
I mean, you can go to the Harvard
Business School, you can take five
courses on negotiation, you can take
straight As in the courses, you can
get a nice letter on Harvard stationery
from the negotiation professor saying
how well you did in the courses, but
none of that says anything about whether
you can actually negotiate or not.
It's like taking five courses on tennis
and being able to recite all the rules
and, and indicate where you should hit
the ball in a particular situation,
but not necessarily ever stepping on
a tennis court to see what you can do.
And what the technology now makes
possible, a combination of AI and, uh,
different kinds of XR, different kinds
of immersive interfaces, is the ability
to construct a chatbot that behaves like
a used car dealer or like a landlord or
like, uh, your boss, and then to practice,
to rehearse, uh, negotiation skills that
you are learning in a course that are
valuable when they, you have the chance
to actually practice with them so that
you're more effective in real life.
And I think what that underscores is
not only that the technology can do a
better thing by providing the opportunity
for this kind of rehearsal, but that
the assessment system then is directed
to a performance outcome as opposed
to a psychometric testing outcome.
And that, um, is another crucial
part of the difference that leads to
doing better things with technology.
I kind of have two questions, I guess.
I kind of want to ask, how do you
see this being applied to education?
And I'm also thinking that you might have
a, the inside scoop as to where this is
headed in terms of education applications.
Well, something that I spend a lot of
time thinking about is not just AI,
but IA, intelligence augmentation.
And that term refers to what happens
when a person works with an AI
agent, and each does what that
part of the partnership does well.
So the AI is good at something called
calculative prediction, which is
essentially absorbing huge amounts of data
and making forecasts based on that data,
much more than any human being can absorb.
But the human being does judgment,
which is applied wisdom, that we
understand many more things about the
world and context and human behavior,
all the socio part of sociotechnical,
than any AI system does, or really
than any AI system can at this point.
And so our part of the partnership is
taking that calculative prediction,
that reckoning, and using it to
inform our judgment and making better
decisions through applied wisdom.
So that then provides a kind of a metric
for where our educational system should
be going, that we don't want to spend
huge amounts of time teaching calculative
prediction, which is what we're doing
now with psychometric tests, because
then we're just preparing people to
lose to AI in terms of their skill set.
What we need to focus on, and, and
many people are saying this now, is
what AI doesn't do well, the critical
thinking, the cultural understanding,
uh, what, what people find motivating,
and to, and to then form these
partnerships that are IA based.
Several thousand years ago, w-
the Greek goddess of wisdom, uh,
Athena, um, is always portrayed as a
maiden with an owl on her shoulder.
And that animal sidekick is somehow
important because when the Romans
conquered the Greeks, they changed
the name of the goddess to Minerva,
but it still is a maiden goddess
with an owl on her shoulder Now, in
modern society, AI perhaps is that
owl, that kind of a sidekick that we
have that aids with our being wise.
But understanding what the AI can do,
and in particular that we're not sitting
on the shoulder of the AI, it's sitting
on our shoulder, uh, is important
to understanding where we're going.
Thank you, Chris, so much.
Wonderful insight and great talk.
Thank you so much for
coming on the show today.
Thanks for joining us on Education Talk
Radio, a part of the B Podcast Network.
If there's a topic you'd like us to
tackle or have a guest idea, send me
a message at dciceroedutalk@gmail.com.
Thanks for listening, and
we'll see you next time.