Latest / IB Unplugged / Inside the IB's AI Principles w/ Clara Hawking

Transcript

Inside the IB's AI Principles w/ Clara Hawking · IB Unplugged

0:00
As an educator could be saying, I mindfully
0:04
and consciously choose not to use technology in my classroom,
0:09
because in my particular subject, I'm developing this set of skills,
0:14
and I believe it's done better in an analog way.
0:23
The IB spent decades defining what it means to educate,
0:25
for international mindedness, for ethical thinking, and for lifelong learning.
0:30
But now for the first time, the IB has published a set of principles to guide
0:35
how artificial intelligence should be used across its entire ecosystem.
0:39
In this episode, we go inside that process.
0:42
Clara Hawking joins us both as a returning guest.
0:45
Her first episode was about policy, practice
0:47
and the promise of AI in education, which is definitely worth a listen,
0:51
if you haven't yet.
0:52
And she's also poised as a rare insider.
0:54
She was one of the external advisors who helped shape
0:57
the IB's five AI design principles from the ground up.
1:01
We dig into what these principles actually mean, the hard conversations
1:04
that shape them, and most importantly, what IB educators need to do now.
1:09
So we're so glad you're here for this episode of IB Unplugged.
1:12
Let's dive in with our guest, Clara Hawking.
1:21
Hi, Clara, thanks so much for joining us again on IB Unplugged.
1:24
Hey, you guys, it's nice to see you.
1:26
And it's really wonderful to be back.
1:28
I look forward to our chat.
1:30
Me too.
1:31
So I want to jump in with some warm up questions.
1:34
And this episode is going to really be focusing on the IB's AI principles
1:38
and I was wondering when you first got the call from the IB
1:42
to help shape these principles, what was your reaction?
1:45
Yeah.
1:46
So on a personal note, I was thrilled.
1:50
Because I love the IB, and I love to be able to support it,
1:54
so getting that call and knowing that my inputs
1:57
were considered valuable meant a lot to me personally.
2:01
But I think, more importantly, I think the timing really mattered
2:06
of these questions.
2:08
Because I think that
2:10
what I certainly see in my work
2:12
is that schools are now starting to ask,
2:18
I would say increasingly
2:19
more complex questions about AI,
2:22
where less in a tool's moment, and there's a kind of,
2:26
suspicion or uncertainty,
2:29
about where AI is heading and how it's impacting schools.
2:33
And, also, educators are really looking for practical guidance, guidance that work
2:38
not just about academic integrity, but in how
2:42
AI is embedding into to their ecosystems.
2:47
And then I think the most exciting thing for me,
2:49
when I got the call, was that this was not an easy conversation.
2:53
And although I'm a computer scientist and do a lot with tech, it was,
2:59
most definitely a conversation
3:01
that wasn't about technology adoption.
3:05
It was really a conversation that was going to be about educational purpose.
3:10
So not what AI can do, but what kind of human beings are
3:16
we trying to develop and specifically trying to develop in the IB.
3:21
So I think those were the first feelings that
3:24
really struck me when I got that call.
3:27
Nice.
3:28
Clara, it really is great to have you back on the show,
3:30
because obviously we had you in season one.
3:32
I have to admit, I went away from that one a little bit
3:35
intimidated and kind of like, oh, we've got to get our stuff together.
3:38
So I've got a a massive amount of respect for you.
3:41
I'm going to interrupt you.
3:42
That is how everyone looked when we walked.
3:44
We saw everyone walk out of Clara's workshop at the global conference.
3:48
You just see everyone's faces,
3:50
just be like, I've got to get to work right now.
3:53
I have so much to do.
3:55
So, you know. So, I'm really glad you're back.
3:58
And I've got so much respect for the work you've done.
4:00
You kind of alluded to it just here with regard
4:02
to how this conversation is kind of shifting.
4:04
And, you know, it sounds like it's mainly not a focus on tech.
4:07
It's a focus on how are we building good humans.
4:10
But I wonder if maybe you could elaborate a little bit more
4:13
from our first conversation we had in season one versus where we are now?
4:15
What feels different in that conversation around AI in IB schools?
4:20
Yeah, I think for me, the biggest thing I see
4:23
is that the conversation now has matured.
4:27
We have moved from this initial 23, 24
4:30
tools moment, and we have to hurry up and develop AI literacy,
4:35
and we need to learn how to prompt,
4:38
and the more experimental work vibe coding early. But,
4:44
I think schools have
4:45
definitely moved from this early experimentation
4:49
to start thinking about implementation.
4:53
And that is something about getting to work.
4:56
It isn't something that's easy to do,
5:00
because AI isn't just specific tools.
5:04
It's really about infrastructure of how AI operates.
5:10
It's about how we make decisions that are now being influenced
5:15
by AI driven technologies without us even noticing.
5:19
It's about world news changing at the speed of a tweet.
5:25
It's about a deep, ingrained
5:28
infrastructure that starts especially for our children.
5:32
You know this the first time they have a device in their hand.
5:36
So I think for schools, they are definitely maturing now
5:39
to going from can use AI to
5:44
how should we use AI.
5:47
And one thing that, as part of that maturation, is
5:51
that leaders are really now asking governance questions,
5:55
partly driven by parents who are entering the conversation.
5:59
In the beginning,
6:00
we heard a lot of parents ask for more Stem, more computer science,
6:04
learning how to use these AI tools, to parents saying
6:08
we want our kids to learn how to be
6:12
thriving in the future, driven by AI.
6:15
But, we have concerns.
6:18
And so leaders are really asking how do we govern the use of this
6:24
AI technology inside of our schools?
6:27
And I think very importantly also, we're seeing regulators
6:31
are now finally entering the conversation in a meaningful way.
6:35
The EU was quick there, but we're seeing more
6:37
and more jurisdictions around the world where regulators are coming in,
6:42
either by filing
6:44
lawsuits and courts,
6:48
given their view on how
6:51
AI should be interpreted in real human lives.
6:55
But also on a greater scale,
6:57
national and regional.
7:00
And then finally,
7:01
I think that assessment concerns have become more urgent.
7:07
People are starting to realize that cognitive offloading is real.
7:12
I think a good example was Iceland
7:15
did a pilot on using large language models in all of their schools.
7:21
And students were thriving when it came to formative assessments.
7:24
Grade had never been higher.
7:26
When it was time for the summatives, grades had never been lower.
7:30
And so I think that people are really starting to realize there's
7:33
something going on that needs to be addressed.
7:38
And so
7:38
we've definitely moved from exploring tools to
7:42
wrestling with these consequences of the technology.
7:47
No, I think that makes a lot of sense.
7:48
Yeah. Thanks.
7:50
I want to dive in, because I think that, like you said
7:52
there's a lot of regulators, but there's also,
7:55
there's another side of it of wanting people to get
7:59
schools want advice and want guidance from the IB, as well.
8:03
And I think that that's not a regulator, but it is someone who is a body
8:08
that is providing a path and direction
8:12
for how to think about this in your school.
8:15
So if we could first talk about the process
8:18
of setting up these principles, could you walk us through what happened
8:21
in these working sessions, like what was the process?
8:25
Yeah. How did it go?
8:27
Yeah.
8:28
It was very IB,
8:31
and I just loved it.
8:33
It was really a collaborative process.
8:36
I think the IB really did a phenomenal
8:39
job of bringing together diverse expertise.
8:43
There were a lot of different perspectives.
8:45
So we had educational technology presented.
8:49
We had, sort of educator perspective.
8:52
We had the governance perspective.
8:54
We had the IT.
8:56
And IB oversight perspective and also innovation
9:01
in the education perspectives.
9:05
So having all of these people together,
9:09
I think was very fruitful in this collaborative process.
9:15
Additionally, it was very important from the beginning that this wasn't
9:19
just going to be something developed by this group of experts,
9:25
that this was intended to become
9:29
a global collaborative process,
9:32
and that a survey would be sent out and on the ground feedback would be,
9:38
gathered at the global conferences,
9:41
so that they could be adjusted to meet
9:46
the IB philosophy, but also
9:48
the realities of what is possible on the ground.
9:53
And I think one of the things that became really clear was this idea
9:58
not to create a technology framework, not to park
10:01
this in IT or digital
10:05
systems.
10:06
It was really from the beginning, important for everyone to create
10:11
principles that were deeply grounded in the IB’s educational philosophy.
10:18
So, while were reviewing existing global frameworks
10:22
and taking into account emerging and existing global regulation,
10:27
it was really important that we ended up with something
10:30
that was very recognizable
10:33
IB and I think that
10:37
we managed to do that on a logistic level.
10:40
We've had several meetings and then based on those meetings,
10:44
written feedback was developed that was then pushed out to us.
10:47
So we had some pre-reading to do, some homework to do before each meeting.
10:52
And then during the meetings, everybody had an opportunity to come with input,
10:58
recommendations and adjustments.
11:01
And I think, one of the thing that I really liked
11:05
was that, at the very beginning
11:09
of the first session,
11:12
the question came up
11:14
how do we think of this in the context of our ultimate goal,
11:17
which is to develop this IB learner profile.
11:23
And I would say were very much
11:27
on the same page, actually, from all of the different perspectives.
11:31
So, there was very little disagreement
11:34
on how to approach this, which I thought was,
11:39
sort of evidence of the importance
11:42
of these conversations and doing the work.
11:45
You can see how important it would be to have all those different perspectives
11:48
in the room.
11:49
And actually, I'm really appreciative that, as you've said, it
11:51
didn't just sit in the digital literacy sort of conversation.
11:55
It sat in an entire conversation of how are we developing the IB
11:58
learner profile,
11:59
and of course, then it makes sense that you'd have people kind of,
12:02
you know, reflecting on the OECD stuff and the UNESCO and Brookings Institute
12:06
and European Commission and all these different sort of,
12:09
you know, thoughts and ideas that are contributing to the conversation.
12:12
Yet also, as a school leader, I recognize that if you get too many voices
12:17
in one room, it's like, how do we move this forward at all?
12:20
It just seems like that working group almost becomes a gridlock.
12:24
So I'm kind of interested in like how did you kind of balance that?
12:29
How did you make sure to keep it in the IB distinctive identity?
12:32
But then also like
12:34
what were some of the more contested conversations?
12:36
What were some of the like
12:38
the real sticking points that needed ironing out?
12:41
Yeah, super question.
12:42
So I think that for all of us, we very quickly established that
12:46
the learning profile has to be the anchor for this.
12:50
This is what we're all doing on a global scale,
12:52
regardless of perspectives and priorities,
12:56
whether you're an ad tech vendor or whether you're a parent
12:59
or an educator in the classroom, this is what we're all trying to do.
13:03
We're trying to work together on a global scale to develop the IB learner profile.
13:08
And there are many excellent
13:11
frameworks and amazing ideas.
13:14
So, you know, the OECD framework is phenomenal.
13:18
UNESCO's framework for AI ethics in education.
13:23
The Brookings work has been
13:26
really valuable in the work coming out of the European Commission.
13:29
The research that was done around the regulation there also
13:34
really phenomenal.
13:35
But I think for us, it was helpful that we very
13:41
early on established what is our anchor here.
13:44
Because we want to identify, not frameworks,
13:50
that the IB and the learner profile could fit into,
13:55
but, that they served the IB learner profile.
14:01
And of course, the IB already has a strong
14:04
educational framework and it has, I think, been easy for us
14:09
on a global scale, especially when we look
14:13
to the conversation about AI
14:16
and from big tech, for example,
14:19
is to think that we kind of have
14:24
to fit in to that.
14:26
But, really for us, we swapped that conversation to
14:31
how can external frameworks kind of help inform our work,
14:36
but keep that learner profile remaining central,
14:40
and how do we ensure that
14:43
all of these different external frameworks,
14:47
evidence, advice,
14:50
you know, global policies that have been written and so on,
14:54
how does that serve the IB learner profile, rather than replacing it.
14:58
And I think that,
15:00
because went in that direction,
15:02
we actually avoided a lot of disagreement,
15:04
because we all understood
15:07
exactly where we wanted to go with all of these frameworks.
15:11
Nice.
15:12
It sounds like some of the other frameworks, maybe,
15:15
and I don't know entirely, but it kind of sounds like they were more
15:18
let's protect this stuff around education, that sort of safeguard
15:21
to make sure we're doing things right, and that sort of thing where it
15:24
sounds like the AI principles in the IB were kind of built up from
15:29
how are we enhancing the IB learner themselves.
15:31
How are we making sure that they're developing as a learner
15:34
aligned with our IB learner profile,
15:36
and from that. Is that a fair summary of what you're talking about?
15:40
Yeah, very much so.
15:42
I think that what we've seen in many schools,
15:45
I've certainly seen this in many schools, is that that AI has entered
15:50
and then schools have quickly said how do we integrate this into our school?
15:55
You know, we need to use this,
15:57
so let's get started.
16:00
And what has
16:00
been missing is the questions that a lot of school leaders
16:03
and educators are asking now, is it serving our purpose?
16:08
Is it serving our mission, our vision, our values?
16:12
And let's just take a pause for a moment and think about things
16:18
like learna agency, human judgment, authentic thinking,
16:24
and then also balancing our innovation with our other responsibilities,
16:28
as well-being, safeguarding, meeting the curriculum and so on.
16:32
So I think that
16:35
one of the things
16:37
that became really clear for us and where we wanted
16:41
to meet is, not how can
16:46
the curriculum, the IB,
16:48
fit into the technology, but rather how can
16:53
technology be used to serve the IB
16:57
philosophy and mission and vision
17:01
to develop human flourishing and thriving young people that
17:07
go into the world, exhibiting those IB learner profile values.
17:12
So, I think that was absolutely a grounding for this work.
17:17
No, that makes a lot of sense.
17:18
And I think that's a really positive way of going about this,
17:22
so it doesn't feel like an add on but an integral part in the core.
17:27
So, obviously,
17:29
you're now at the point where you're like, right, these feel really good.
17:32
But then you take it to the masses and you're saying that you guys
17:38
are really open to having them
17:39
all kind of torn apart, maybe, for lack of a better word.
17:43
Can you walk me through what it's going to look like,
17:45
for example, if I'm a teacher who has something to say about this,
17:49
at the Leon conference or something like that, what's the process now?
17:54
Yeah, it's a super good question.
17:56
So, once the first five principles
18:00
were identified and shared with the community,
18:04
so, you know, AI for human flourishing
18:08
and as the number one principles,
18:10
but also, inquiry driven
18:13
educator agency, which was really important
18:17
for us to embed there, and then safety and transparency and then also this
18:23
principle of continuously adapting, which is a very IB principle.
18:27
We see that in our curriculums. Every five years we are adapting
18:32
in an evidence led way.
18:35
But feedback has come back
18:38
to the group, to our working group, from surveys that was sent out
18:43
and from on the ground, and really used to make
18:49
real changes, tweaks.
18:53
I think one that was very big and was a big concern
18:59
that came from everyone was around educator agency.
19:04
So, this need for educators to know that they still had,
19:10
some say, in how
19:13
AI is used inside of their
19:17
fields, or their particular subject.
19:21
And this has been a little bit tricky, because
19:26
we have seen educators are seeing
19:29
how AI is used in learning.
19:31
And we have seen in the last couple of years an influx of really
19:37
innovative, educational technology
19:40
that may not actually have aligned with
19:44
developing educator agency and may not have been aligned
19:49
with developing AI for human flourishing for the students and have had a lot of
19:55
cognitive offloading and has been missing some oversight.
19:59
So we really needed to develop a principle around
20:04
supporting educators in
20:07
still having that agency, still being able to
20:12
use AI in the way that is
20:15
best suited for their particular subject.
20:19
Some educators, for instance,
20:22
don't use any AI in their subject at all,
20:26
because they are developing other skills,
20:29
more human centric skills, that AI probably isn't so good yet
20:33
at developing, could be creativity, for example,
20:38
or really complex critical thinking,
20:42
or collaborative skills.
20:45
So some educators say we don't want to use it,
20:48
but we're still being affected by how it's used in learning,
20:52
because when the students leave our classroom, they're using AI at home.
20:56
And so developing that principle was where we got a lot of feedback.
21:02
And I think this is one of the ones
21:04
principles that we will see continue
21:07
to develop as technology develops.
21:10
We are going to see AI agents,
21:12
we're going to see much more advanced AI driven technologies, and we're going
21:17
to see edtech companies embedding
21:21
new technologies into their offerings,
21:25
which will affect the ecosystem, but also big platforms.
21:30
You know, platforms that are used in inside of the school
21:33
could be anything from, you know, Google searches
21:36
to using a large language model
21:40
for a particular purpose inside the organization.
21:43
And so that one I think is where we will very much
21:48
see that we need to return to support educators in this, because they are
21:53
on the truly on the front line and need the most support.
21:58
So yeah, I would say that's the one we got the most feedback on.
22:01
The rest really resonated with the community.
22:05
I think that a lot of people really needed to hear that the IB is doing this
22:11
and that AI is now being thought of as, you know, how do we
22:19
use AI or shape the AI use in schools,
22:22
I should say, to serve the learners that the IB is trying to develop.
22:28
Yeah.
22:29
I love that you said it's tricky, because
22:32
I can absolutely agree with that in that,
22:36
you know, you've got educators on one hand being like, we just need some policies
22:39
and some actual principles that we can follow.
22:41
Just tell us what to do and we'll do it right.
22:42
And then you get others who are like, don't get into my classroom.
22:45
Let me do it the way I want to.
22:46
So it's a very messy space I can imagine.
22:49
And so as you develop these principles, you kind of send them out there
22:52
for, you know, stress testing.
22:53
I guess, it's this idea of like, okay, well, here's
22:56
the guiding principles when it comes to AI use and let's get into that.
23:00
You've alluded to that already, but let's get into this
23:02
kind of second segment here of our show and really get into
23:05
what are these five principles and what do they mean for our schools?
23:10
The first principle you've already alluded to, it's
23:12
caring and balanced and AI for human flourishing.
23:15
And I guess the idea
23:16
there is it's going to put the emotional, the social, cognitive development first.
23:20
Talk to us about why that framing was chosen
23:23
and what does that look like in practice for a classroom teacher.
23:26
Yeah.
23:30
Speaking of AI for human flourishing
23:32
is really connecting with the IB mission and vision.
23:36
This is the reason why the whole curriculum
23:39
exists, that humans come first.
23:42
And what we have heard quite a lot is that AI
23:46
somehow has been the goal, we have to bring AI into to our schools,
23:50
we have to teach students about AI.
23:53
You know, you'll be replaced, not by AI, but a professor who knows how to use it.
23:58
Just panicked messages, huh?
24:00
Yeah.
24:00
And I think this is really returning to hang on for a minute.
24:04
Who are we? Why do we exist?
24:07
What is our mission and vision for the IB on a global scale,
24:12
and also encouraging schools to connect with this, not just to the IB
24:19
curriculum, but to other things that they might be part of.
24:22
They might be green flag schools that are
24:26
meeting certain
24:28
ecological goals.
24:31
They might be
24:34
UN sustainable
24:36
goals schools and so on.
24:38
But this idea is that what is the IB existing for?
24:42
It’s existing because humans come first.
24:45
So human development is the goal, not AI.
24:49
And education, I think, is certainly the way that I look at it,
24:54
has always been about developing people.
24:58
I know we speak a lot
24:59
about the factory model, and somehow we landed in classrooms
25:03
where many students sort of sit in a row and learn by rote learning.
25:08
And then they come out with a set of skills.
25:12
But the truth is that, you know,
25:15
the reason why we started educating was about developing people.
25:19
When we look at some of our first official academy,
25:22
Plato set up his his academy, the goal there wasn't
25:26
to drive a specific set of you have to learn these very things.
25:32
You know, you have to learn this much math,
25:34
you have to learn this much Greek, and so on.
25:38
It was really about developing a human being,
25:41
that had lots of different characteristics.
25:46
And this principle, it really
25:49
then begins with the idea that technology
25:53
should serve human flourishing, that we shouldn't
25:57
try to fit into the technology, but the technology should help support
26:01
those principles, those ideals, and that that, you know, our emotional
26:06
our social, our cognitive development, our ethical development,
26:11
must really remain central when we use the tool.
26:16
So I think for that principle, absolutely,
26:19
was this idea that AI
26:23
is a tool.
26:24
It's also now infrastructure.
26:26
But human flourishing has to remain the objective and technology
26:31
has to fit into that objective, not the other way around.
26:35
Yeah.
26:35
Yeah, I love that because I think one of the, you know, there's
26:38
a lot of different articles out there on like, why education exists.
26:42
And I think we can always acknowledge that it's partly for the economy.
26:45
We want to educate people to be contributors to the economy.
26:47
And I guess, you know, when AI sort of pops in, there's that immediate alarm
26:51
bell of like, well, okay, we got to be upskilled in this,
26:52
got to be ready for this so you can go and have a job in it later.
26:55
But it's really, for me, heartening
26:58
to hear that this is one of the main purposes, because
27:01
education for the economy is only one factor of what we're doing in education.
27:05
It's not just about preparing people for a job market.
27:07
It's about preparing young people to be positive citizens
27:10
and good contributors to communities, local and global.
27:13
And so I love that the framing, just as you've outlined here, first
27:17
and foremost, is let's create some good humans
27:19
who have some good emotional, social, cognitive skills,
27:22
ready to contribute positively to their local and global communities.
27:26
Plato's Academy
27:27
was found in ancient Greece around 387 BCE.
27:31
It's often considered
27:32
one of the earliest institutions of higher learning in the Western world.
27:36
Clara is reminding us that education is historically
27:39
not only been concerned with transmitting knowledge.
27:42
Researchers such as Gert Biesta or Gary Becker,
27:45
Mortimer Adler, John Dewey, Martha Nussbaum, Parker Palmer,
27:49
and earlier sociologists like Emile Durkheim and Pierre Bourdieu,
27:53
all present ideas which synthesize around five main
27:56
purposes of education.
27:58
In no particular order,
28:00
these include one,
28:02
economic purpose, preparing learners for work and contribution.
28:06
Two, the academic purpose,
28:08
which of course is about developing knowledge and intellectual capability.
28:11
Three, the social purpose, so that we can participate in cultures and communities.
28:16
Four, the personal purpose, to help develop identity, agency and flourishing.
28:21
And five, the civic purpose to contribute ethically to society.
28:25
What Clara is essentially framing is that AI should serve all five purposes,
28:30
but especially the personal and civic ones.
28:33
These are often overlooked when education becomes
28:35
too focused on qualification and economic outcomes.
28:39
For the second principle, it says AI should deepen
28:42
learning with inquiry, critical thinking and creativity at the center.
28:47
I have a lot of thoughts on this, but I would love to hear your thoughts on
28:49
how do you actually test whether an AI tool is doing that,
28:53
versus maybe quietly replacing it?
28:55
Yeah, a fantastic question, and I think this is a hot topic
28:58
around the world because, we're hearing schools, we're hearing,
29:02
universities, we're hearing,
29:05
other tertiary educations coming out
29:07
and saying these students are cognitively offloading,
29:11
our staff is cognitively offloading,
29:14
and we can see it, even if we don't have all of the
29:17
empirical evidence yet
29:19
to support that, because we're still just a few years in.
29:23
We are seeing something that isn't jiving.
29:26
And I think that when we talk about is AI deepening learning,
29:31
it's really a question about who's doing the thinking.
29:35
And to me, I think that
29:37
we might, on a global scale, kind of have discovered ChatGPT 3
29:42
when it came out in in the end of ‘22, early ’23.
29:46
But, students who are already using, not all students, but many students
29:51
were already using other tools
29:54
to take over their thinking,
29:57
there was a relatively thriving economy
30:00
of students hiring ghostwriters, for example,
30:03
other humans around the world, who could do their writing work for them.
30:09
Students were grabbing for pieces of text,
30:13
and, you know, putting it into their own writing as their own.
30:18
And of course, there was ChatGPT 2, that was already available,
30:22
for students to use and some students certainly were.
30:27
I know, myself worked with some students
30:30
already in 2019 on looking at ChatGPT 2,
30:35
and identifying who's doing the thinking.
30:39
You know, and is this enough to support
30:43
the learning outcome for the IB learner profile in the subject.
30:47
But I think one of the things we went back to in conversation
30:53
and that I also
30:56
felt was important, both from what
30:59
I've heard from schools,
31:02
but also, what we are starting to hear in the news is that,
31:07
you know, is the learner thinking more deeply?
31:11
And sometimes that requires struggle.
31:14
We want our students to build up resilience
31:18
by struggling with difficult contexts, difficult concepts.
31:23
Also,
31:24
another thing is their curiosity increasing.
31:28
I think also are the students asking better questions.
31:32
If we look at the PYP, curriculum and the MYP curriculum,
31:37
and the DP curriculum, some of those questions that
31:40
are being asked around developing ATL skills and later embedding TOK
31:46
into their abstract thinking is are they learning to ask better questions.
31:51
So we're hunting not always for the right answer,
31:56
but for their ability to ask those discerning questions.
32:00
Is creativity being strengthened through their use of technology
32:05
or is AI doing this
32:08
cognitive work for them?
32:10
So I think for me, certainly a useful test would be
32:13
whether the learner is thinking more or less than before
32:19
they use the technology.
32:21
And I think we can't lose sight of that, because that is so critical
32:26
for developing that IB learner profile is
32:29
that we're really developing our human through
32:33
experiences, cognitive struggles and building resilience over time.
32:38
Nice.
32:39
Courtney, can I turn it on you for just a moment?
32:41
You said, I've got lots of thoughts on this area, too.
32:44
What kind of comes to mind
32:45
when you think about, like, AI deepening learning
32:47
and maybe some of the risks associated with that?
32:49
I think that AI is the flashy,
32:54
maybe alarm bell version of a lot of things
32:56
that have been happening for a long time with all kinds of edtech,
33:01
and even in just the publishing space, I think edtech and publishing houses,
33:06
which I realize I'm part of this world, but I think a lot of them
33:10
historically and still presently try to say
33:14
we’re your core resource, think of us like the spine
33:16
of your classroom, or think of us as the heart of everything,
33:21
and everything branches in, whether it's an LMS or an ad tech tool
33:25
or even a textbook.
33:26
Like, we are the center.
33:27
And it's such a bad way of thinking about your classroom, where it should be,
33:34
that the students are the center of the classroom, the teachers learning
33:39
design should be the spine, and then the tools should be the ways
33:45
that the teacher can achieve the goals that they're trying to do.
33:49
And I think by, if you look at these principles,
33:53
if you look at all of them together, I think it's a great way of evaluating
33:58
any of the tools that you're using in your classroom.
34:02
I think making sure you're using all of them for the right reasons
34:06
is something that I always come back to, and I think in a lot of different episodes
34:10
we've come back to is, why are you using anything that you're doing?
34:14
What's the intentionality behind it?
34:16
And do you still have
34:19
control
34:21
of what you're trying to achieve in your classroom,
34:24
or are you getting swept away by what
34:28
either other companies are telling you
34:31
you're supposed to be doing, or other teachers in your school?
34:35
You know, maybe there's one teacher who is an edtech enthusiast,
34:39
and they're doing incredibly advanced things, but that's great for them,
34:45
in their classroom, because they have a different mindset and learning
34:50
objectives and understanding of the way that they want their curriculum to go.
34:54
But what I love about the teacher, that educator agency, it's
34:59
saying you still need to be held responsible
35:02
for the human flourishing part, and there are tools
35:06
that will help you with human flourishing.
35:10
There are tools that will help you deepen learning.
35:13
There are ways of setting up your classroom with different degrees
35:17
of technology use, that are all really positive.
35:21
If you keep that agency front and center
35:25
with the intentionality of am I deepening learning by using what I'm doing?
35:30
Or am I scared that I'm falling behind?
35:33
Therefore, my students will be falling behind
35:36
and everybody else uses an AI tutor.
35:39
So I guess I should use an AI tutor.
35:43
You know, it's such a different way of looking at everything.
35:46
And I just think that these principles are a really great framework for evaluating
35:50
the way that you are making decisions in your classroom,
35:54
from the print textbook all the way to the most advanced technologies.
36:00
What I appreciate about what you both just mentioned
36:02
is this idea that you've got to be looking at developing the learner,
36:06
and this tool supports the development of the learner can,
36:10
but the goal should be to deepen learning, nonetheless.
36:13
And that of course, then reflects right back to what that principle is all about.
36:17
And then, Cortney, I mean,
36:18
you just sort of led into that idea of like that educator agency.
36:22
And I know that's the third principle.
36:24
And it's to be
36:25
the idea is like guided by capability, grounded in responsibility.
36:30
I wonder, Clara, could you tell us a little bit more about
36:34
what's the risk if we don't get that right?
36:35
Yeah, I think, certainly for me, and I know
36:39
I've spoken a great deal about this, but in the classroom,
36:43
the teachers really must remain the decision makers.
36:45
It cannot be taken over by technology.
36:48
And I think Courtney is landing exactly
36:51
in the right place with what she's saying.
36:54
There's been a push of getting all of this technology into the classroom.
36:57
And a lot of educators have felt like, oh, okay, I need to do this.
37:03
But I think I've sometimes spoken of something
37:06
that I call an analog educational specialist.
37:10
It doesn't really exist yet.
37:12
That sounds like you're describing me.
37:14
I'm an analog educational specialist.
37:18
Yeah.
37:18
Fantastic.
37:20
And I think this, for me, lands with there are going to be
37:23
teachers who, in order to do their work well,
37:28
in order to reach the learning objectives, in order to develop
37:32
the student profile, that they're part of the curriculum entails.
37:38
They need to stay analog, because they might be teaching skills
37:42
that simply cannot be replaced by technology.
37:47
But a lot of educators have said, we feel that we need to do this.
37:52
We feel that we have to do this.
37:55
And there's a lot of pressure
37:57
in the global messaging that we're getting,
38:00
but also from different sides inside of their ecosystem.
38:04
But I think using AI well,
38:07
as an educator, could be saying
38:10
I mindfully and consciously choose
38:14
not to use technology in my classroom,
38:17
because, in my particular subject, I'm developing this set of skills,
38:22
and I believe it's done better in an analog way.
38:26
So, supporting professional
38:30
judgment is a great thing for AI.
38:35
So I don't know, Courtney,
38:37
if you could perhaps speak a little bit to this, but I think it's
38:40
such a good example of the feedback function that Kognity
38:45
has in its programing, ‘cause I got very excited about it, because
38:51
the feedback from a teacher cannot be pushed out,
38:55
unless the teacher has sort of actually looked at the feedback,
39:00
and either approved, edited, or rejected.
39:03
This idea that the human,
39:05
the teacher, should always, always be between
39:10
the student and technology output.
39:13
And I think that that is a world where teachers
39:18
don't just become operators of AI systems, but where their human expertise,
39:24
their professional expertise, remain essential.
39:28
And also that human oversight really matters.
39:33
And interestingly, we see that
39:36
very much in emerging and existing regulation.
39:40
There are requirements for human oversight.
39:43
So, I think this idea of teachers,
39:46
as professional decision makers, who use AI
39:50
as one of many options that they have, I think is absolutely essential.
39:56
No, I definitely agree on that.
39:57
I think with Kognity’s feedback and kind of the way
40:01
that we've been developing AI, it's been down to kind of our
40:04
learning principles and our way of working, is that we amplify a teacher's ability
40:10
to teach for our students, to have amplified learning.
40:13
And when you remove a teacher out of it, and kind of
40:17
don't have the human in the loop, I feel like that's such a cliche
40:20
saying now, but it really is a core value for us
40:24
to make sure that, again, when companies are, edtech
40:29
products are trying to do things without the teacher, it's like,
40:33
make sure you understand why and make sure that that's
40:36
actually going to lead to quality learning.
40:41
And it's not about, oh, you need to micromanage as a teacher.
40:45
It's making sure that the feedback that students get
40:49
is what it should be, and really making sure that these tools
40:53
are helping you teach, not just taking up space in your classroom.
41:03
So, I know that the principals
41:05
also introduced this idea of red lines, these non-negotiable boundaries.
41:09
And there's an example here.
41:10
Just like what we've been talking about, how like high stakes decisions
41:13
or feedback about grading and progression should never be fully automated.
41:17
It should still remain with the teacher.
41:20
Clara, are there other red lines that you think are
41:23
equally important, but didn't get included in that final draft?
41:27
Yeah.
41:27
So, this is where I'm going to climb up on my soapbox,
41:32
because I come very much from AI governance lens.
41:36
So, I think a lot about not just what AI can do,
41:40
but what we are allowed to do in the context
41:43
of regulation, but also aligned with our
41:47
ethical approaches, our values, and then ultimately our mission.
41:52
And, one of the tricky parts,
41:54
of course, was for us to be able to say, is there a red line
41:59
for the International Baccalaureate, something that we can competently
42:03
and confidently say, this is where we draw the line on a global scale.
42:09
And that is really tricky, because the issue of equity very much came up.
42:14
You know, what is potentially a functional red line,
42:18
I don't know, in the Netherlands might not be a functional
42:22
red line in Ghana, or Brazil.
42:26
So, really being able to identify what is something that absolutely
42:33
is a red line on a global scale,
42:36
and then, what I thought
42:40
was also important,
42:43
was not able
42:45
to come together, and I think,
42:48
for obvious reasons,
42:50
red lines is something that we will see develop over time.
42:54
So, this idea that the key messages, certain decisions,
42:58
should always remain human, is something that we can all agree on.
43:04
We don't want fully automated,
43:06
high stakes decision, we don't want an AI to completely
43:11
take over grading and progression decisions.
43:14
We want that human in there to have human eyes on these things.
43:22
Systems that undermine learner agency
43:25
also must remain human systems, that compromise child safety
43:30
is something that we can all agree on a global scale,
43:35
that we cannot compromise.
43:37
And then I think, for me,
43:40
one of the things that I really wanted in there
43:43
in a more stated way and I think we'll come with time,
43:48
is this opacity that we're seeing in the systems.
43:51
Because the truth is that there is this
43:55
odd dissonance between wanting,
43:58
for certain decisions, to remain uniquely human,
44:02
but AI functioning as an infrastructure,
44:05
that has guided decisions all along the way
44:09
in many different ways from teacher education to leadership
44:13
skill development and our day to day
44:16
family life at home.
44:19
And many of those decisions
44:21
are really made in a very opaque environment.
44:25
So, we don't always have systems now,
44:28
that can be meaningfully understood.
44:31
That will become increasingly difficult as we move forward.
44:36
I think it's probably very difficult to articulate that
44:39
right now, but I think with time,
44:42
that that will come in
44:44
as its own kind of a red line that
44:48
if the systems can't meaningfully be understood,
44:52
in the way that
44:54
decisions can remain human, then that will be its own red line.
44:58
But I think it will come with time and it will be supported by
45:03
regulation, and it
45:04
will be supported by court rulings, that we are now
45:08
seeing courts come out and rule on some of these things.
45:12
So, now that we've gone through all of these principles,
45:15
I want to zoom out a little bit and wonder about you've gone
45:19
into a lot of different schools, and what is the difference
45:23
that you can clearly see between a school that has an AI policy,
45:27
which might end up being a bit of a superficial, we made a document,
45:30
or a school that has genuine AI governance culture?
45:36
Are you laughing?
45:37
We made a document? I love that, yeah.
45:39
Because you're right, sometimes it does feel that way.
45:41
That's what it is sometimes.
45:43
I wrote a policy.
45:44
We have one. Yeah.
45:47
That was funny.
45:47
Sorry. Go on.
45:48
I love it, and I think we actually all are on a global level.
45:53
Everyone that works in education kind of gets what that means,
45:57
because we have all of these policies, and then they just become
46:01
part of the policy stack, that we can show people live if they ask.
46:05
And that doesn't quite meet the demands of
46:11
AI technology driving into the school ecosystem.
46:16
So, I think if we put it in context of something that everybody's
46:19
familiar with, if we asked instead, what's the difference between
46:23
a wellbeing policy and a wellbeing governance culture,
46:28
we understand that this is less about a policy
46:32
telling people what to do and more about governing
46:37
how decisions are made, and then how we follow up with that.
46:42
So a policy is just a,
46:46
sort of silent, or dead document,
46:49
unless we actively work with it all the time.
46:53
So, governance of AI, just like governance of well-being,
46:58
involves accountability.
46:59
We have identified who's accountable for all of the different steps
47:04
in the process of wellbeing or AI.
47:09
We also know that governance involves oversight.
47:13
So, who is looking after this?
47:16
Who is sending out surveys to understand what's happening in the ecosystem?
47:21
This could be how AI is being used.
47:24
Have any staff member set up individual accounts
47:27
where they're uploading information
47:29
they shouldn't be using, is AI being used by students
47:32
in a way that harms other students,
47:35
and who is reporting back on that, on
47:39
a semiannual, or annual level.
47:43
AI governance, as opposed to AI policy,
47:46
also involves ongoing review.
47:50
You know, one of the things that
47:52
we have been used to is something
47:55
around giving consent to use.
47:58
But if you've given consent to use ChatGPT 4,
48:03
but now the model has updated and now it's ChatGPT 5.7,
48:09
you have given consent to two different models.
48:11
We have to go back in now and review
48:15
what's the drift on that model,
48:17
what kind of biases does it have,
48:19
what kind of risk pop up
48:22
with these new models, and so on.
48:25
So, this idea of ongoing review is different than just having a policy.
48:31
It's really about building that culture.
48:33
And then I would say, also, a successful school is
48:38
just like well-being, a successful school
48:41
is where well-being is part of the institutional culture.
48:45
AI governance is exactly the same.
48:47
So, governance of this technology becomes part of the institutional culture.
48:53
We're consistently working on it.
48:55
We're consistently reviewing, we're consistently trying to improve.
49:00
And if some bad things happen, we have a way for addressing it
49:04
and redressing it.
49:06
So, I think for me, the big one is that policy is a document,
49:10
kind of like Courtney says,
49:11
but governance of this technology is a practice.
49:16
Nice.
49:16
It's making me think of if we compared it, you compared it to well-being.
49:21
But I wonder if you even take it to the step of, like, if you just write a policy,
49:25
is it basically like a dress code policy in your school?
49:29
It's like we have it.
49:31
Nobody's looking at it actively, except when something's gone
49:34
really wrong and you need proof that it's bad.
49:37
And I feel like that's kind of the same problem.
49:41
If it's just a policy that's used for deterring
49:46
bad behavior, compared to maybe the way that your school works
49:49
with their values, where it's like, how do I shape this in a positive way,
49:53
how does this help me make decisions?
49:55
All of that.
49:56
I think that can make a huge difference.
49:58
Absolutely.
49:59
I think that that's also why,
50:03
even an educators that are choosing not to use technology
50:07
in their classroom, can be a very active part of shaping
50:13
an AI governance culture in the school system.
50:18
And I think, for many schools, AI policies,
50:21
in the beginning, was about academic integrity.
50:25
What we're allowed to do and what we're not allowed to do.
50:28
But good AI policy isn't just a free standing document.
50:32
It's thinking AI into all policies aligned with the school's
50:37
value and the school's mission and with the curriculum.
50:41
So, it is thinking AI
50:44
as an infrastructure into everything from hiring policies
50:49
to employee contracts to
50:52
well-being documents to
50:56
your field trip
50:58
policies and so on.
51:00
So, that it becomes a part of that consistent thinking about
51:04
how do we do that in our school in a conscious, deliberate way, rather
51:09
than letting technology happen to us in all of these different aspects.
51:15
Governance and policy are often confused.
51:18
A policy tells people what the rules are.
51:21
Governance determines who makes decisions, who is accountable,
51:24
how risks are monitored, and how practices are reviewed,
51:27
as technology changes.
51:29
UNESCO, the OECD
51:31
and the World Economic Forum all emphasize that AI is developing so quickly
51:35
that governance systems must be adaptive and continuously reviewed.
51:39
In other words, schools may need less focus on writing the perfect policy
51:44
and more focus on building
51:45
cultures and structures that can respond to change responsibly.
51:49
So, for this episode, this is kind of back to school, fresh start,
51:53
we have these schools that have seen the AI principles now.
51:57
What's a good practical first step for them at this kind of time of year
52:01
and place that they're in?
52:03
So, I think the most important part is to put together an AI working group,
52:09
not for the purpose of establishing a pilot, but a working group
52:14
that can work towards, first and foremost, mapping the existing AI use.
52:20
We have an enormous amount of ghost use of AI.
52:24
People are using it in a variety of different ways.
52:28
There's both legal and also well-being questions that arise from that.
52:33
But it's also important for the school to kind of map what have people learned.
52:37
What are they not doing? What are they doing?
52:39
What is that subculture that has developed
52:42
that we may or may not want to
52:47
alter, or modify.
52:49
And I would say, for a lot of schools, often
52:52
AI working groups have become something like
52:56
IT, representative from leadership, maybe a high performing,
53:01
tech loving, student and some of the early adopters of technology.
53:05
But a good working group should have representation from
53:10
lots of different aspects.
53:13
So, definitely someone
53:15
from the well-being and safeguarding team.
53:19
Definitely
53:21
students, that are not necessarily
53:25
tech savvy and may not even be high performing
53:29
students, but are still using this technology.
53:33
I also encourage
53:36
parent representative to the extent that is possible.
53:39
I think it's important for the parents to be part of this
53:43
working group, to have these transparent conversations
53:46
and connect how technology is being used at home with
53:50
how it's been used at school, and then mapping that existing AI use.
53:54
So, literally send a survey out in the beginning of the year, anonymous,
53:58
so that everybody can be frank about how they're using AI.
54:03
And then I would say the working group should then identify one high risk area.
54:07
And that will be different for every school, depending
54:10
on where they're located in the world and what their priorities are.
54:16
And then start the staff
54:18
conversations around that one high risk area.
54:22
So, that really gets the minds going,
54:25
and it also allows the school to tap into the massive talent
54:29
that they have from professional educators
54:32
in their ecosystem.
54:34
So, this idea that, you know, we're going
54:37
to focus on governance now before we focus more on tools.
54:41
And I would say you can't do that unless you first
54:44
have mapped what it is that you're trying to govern.
54:48
So, that would be step number one: working group,
54:50
map AI use, and identify one high risk area
54:54
and then work with that throughout the organization.
54:57
Clara, I so appreciate your passion for this and your your expertise in this area.
55:01
And, you know, when I think of
55:02
AI governance and understanding policies and the impact of this stuff
55:05
and the meaning behind it, you're one of the first people
55:07
that comes to top of my mind, of just like this person knows what's going on.
55:10
And so I'm so glad that you've been involved in the process
55:12
of designing these five IA principles for the IB.
55:16
And I have to be honest,
55:18
the first time we spoke about AI, I kind of went away, I'm going to say
55:21
motivated, not intimidated, but there was probably a bit of intimidation
55:25
there, like,
55:25
oh, I got to get stuff done.
55:27
But now I'm feeling really quite encouraged.
55:29
And I feel like there's some good moves that are being made
55:32
that are going to really help educators out there,
55:33
and I hope that people listening now are feeling that way, too, that,
55:37
you know, there's good thinking going into this and there's ways
55:40
that we're trying to support good learning,
55:42
again, with the human at the center of that.
55:44
And so, Clara, I'm so glad you're involved in it.
55:46
And again, I just so appreciate you taking the time to chat with us.
55:48
Once again. You're an excellent, excellent, person.
55:51
Really, really smart.
55:51
We enjoy getting to spend time with you. So thanks for joining the show again.
55:54
Thank you so much.
55:55
And thank you so much to all the listeners and joining us
56:00
on this journey on thinking about how we support human flourishing,
56:04
and how technology can support what is important to us
56:08
and not the other way around.
56:09
So, it's just thrilling to be part of this conversation.
56:12
And I really appreciate you having me.
56:14
IB Unplugged is a Kognity podcast, hosted by
56:17
John Patton and Courtney Marline and produced by Katilina Gaffney.
56:20
Be sure to subscribe or follow and share with your fellow IB educators.
56:24
If you'd like to submit a topic, request or question, please do so
56:27
on kognity.com/ibunplugged.
56:30
Let's keep the conversation going
56:32
and support each other in teaching with confidence, creativity, and joy.
56:36
See you next time.