Latest / IB Unplugged / Policy, Practice, and the Promise of AI in Education w/ Clara Hawking

Transcript

Policy, Practice, and the Promise of AI in Education w/ Clara Hawking · IB Unplugged

0:00
Welcome to IBDP Unplugged, a Kognity podcast for
0:03
IB educators that tackles your submitted challenges and questions
0:07
by connecting you with experts and peers who share their ideas and stories
0:12
from real classrooms.
0:13
Whether you're a seasoned educator or new to the IB world, each bi weekly episode
0:18
and guest aims to equip you with practical tips and positive inspiration,
0:22
helping you teach with more confidence, creativity, and joy.
0:26
So whether you're walking your dog or commuting to school, hit play and join
0:29
a community of educators ready to learn, laugh, and thrive in your IB journey.
0:35
Hello and welcome to another episode of IB Unplugged.
0:38
I'm John Patton and I'm Courtney Marline, and in this episode
0:42
we got to speak with Clara Hawking, who was talking a lot about AI governance
0:45
and specifically use of AI in schools.
0:48
A hot topic amongst many of our users who submitted questions with regard to
0:51
how do we navigate this new landscape with students and teachers
0:55
having access to AI, and how do we make sure
0:56
that we're using that both ethically, effectively and safe.
1:01
Yeah, and Clara is a fantastic person to have this conversation with.
1:04
She's someone I've had the pleasure of working with for years.
1:07
A little background on her.
1:08
She's the executive director of Kompass Education, specializing in
1:12
AI governance, tech compliance and education.
1:15
Formerly, she was the head of AI for a global school
1:17
group with over 70 schools, and she is a recognized leader
1:21
in AI ethics, policy and responsible innovation.
1:25
With a deep background in computer science, AI ethics, pedagogy,
1:28
and educational leadership,
1:30
Clara advises schools, policymakers, and edtech companies
1:34
on the ethical and effective use of AI.
1:37
She's passionate about technology, learning, and human rights.
1:40
She's a sought after speaker and award winning advocate for safe AI,
1:44
that enhances, not replicates, education.
1:48
Yeah, and for me, in the conversation that was so evident.
1:51
That's what really kind of came out, is this idea
1:52
that, like whatever you choose to use, whatever tools, it's got to come back
1:56
to enhancing the learning, making sure that it's making learning better.
1:59
So not just using AI because it exists and like, oh, we need to do this
2:03
because it's going to be out there for our students later on.
2:06
But more from the standpoint of it can improve education.
2:10
And so really like for educators,
2:12
whether you're, you know, teaching right away
2:13
or if you're on like your policy development team at the school,
2:16
there's a whole lot in this episode with regard to, like, developing policy
2:19
and why that policy is so important for creating
2:23
good learning conditions so that students can actually grow quite a bit.
2:27
And so for me, that was a big takeaway
2:29
that I really appreciated about talking with her.
2:31
Yeah, I agree, I also love this idea where it's not policy versus innovation.
2:37
That's really something that she was trying to have come
2:40
through is like, we want you to innovate,
2:41
we want you to experiment, and we want to celebrate and not diminish the enthusiasm
2:45
of the teachers who want to go ahead and use these AI tools, especially
2:50
if they're really thinking critically about how it can improve education.
2:53
But how do we also make sure that this is safe?
2:56
How do we make sure that we're doing this
2:58
in a way that, above all else, is protecting student safety?
3:02
So I think that that's something that was a really interesting
3:05
part of the conversation for me, especially on the edtech side.
3:08
Also, what kind of questions should you be asking to your edtech vendors?
3:12
What kind of questions are parents going to be asking you about the edtech vendors?
3:16
And really just, yeah, how to build your confidence
3:18
and how to make sure that your school is ready to handle that kind of situation.
3:23
Yeah.
3:23
So get ready for a great episode. Welcome to the show.
3:26
And let's talk with Clara Hawking.
3:34
Hi, Clara.
3:34
Welcome, welcome.
3:36
We're so glad to have you here on this episode of IB Unplugged.
3:39
Thank you so much, Courtney and John.
3:41
It's great to be here.
3:42
I look forward to having a nice chat with you guys,
3:45
about all the exciting things that are going on in the world of AI.
3:48
Yes, us too.
3:50
We've really been looking forward to this episode.
3:52
We have a lot of great questions we want to get to, and in the beginning,
3:55
we'd like to just do some warm up questions before we dive right in.
3:59
So the first question that we like to ask is why education?
4:04
And I think this one's a really interesting one for you, Clara,
4:06
because this was not your first path, your first job that you've done.
4:10
So it's definitely been a shift for you and career.
4:14
Yeah, exactly.
4:18
I started my life backwards,
4:19
I suppose I should say, because I finished
4:23
at third, sorry, eighth grade in third grade.
4:26
And back then there weren't really any alternatives after that.
4:29
So I more or less sat around.
4:31
I spent years to calculate how quickly I could get through high school,
4:35
attending the minimum amount of possible and doing the minimum amount of work.
4:40
Because I had discovered
4:41
this very nice relationship between working hard and money.
4:45
The harder you work, the more you get.
4:47
And I thought that was significantly more interesting than school.
4:50
So I went straight out into the work life, and I did, sort of an old fashioned
4:55
American dream work myself from the mailroom, literally up to
5:01
advanced positions in the corporate America.
5:05
And then after some years, I decided
5:07
I was working too many hours and life was too boring.
5:10
And, I got into the yacht industry managing mega yachts.
5:14
As most teachers start with.
5:15
Yeah, we will talk more about this, but while I was working in that industry,
5:21
I really missed access to learning, which on land,
5:26
you have easy access to libraries, which I was a frequent visitor of.
5:31
So I started buying
5:32
tons and tons of books to take with me wherever I was traveling.
5:36
And, longing for education really started to grow.
5:41
So I decided to start my college education.
5:46
And once I started, I realized how easy it was for me.
5:50
And I just raced through and got this enormous hunger.
5:54
And then I realized that there was
5:55
some kind of a disconnect at university
5:58
level between the high school students that were coming in,
6:02
the abilities that they had and the expectations of the professors.
6:06
And at the time, I was teaching at a university,
6:09
which prompted me to sort of switch
6:11
into secondary education and find out what's happening there.
6:15
Why are they not ready for these more complex
6:19
critical thinking skills and more abstract ideas.
6:23
And I just fell in love with education.
6:25
So, I have stayed in some form within the education form
6:31
since then, and I know that this is the calling for me.
6:36
I have since then studied, I have become a computer science,
6:41
specializing in IT architecture, first becoming an IT architect
6:45
and then specializing in AI, where I quickly realized that
6:49
my passion area was AI governance, AI ethics, and how
6:56
this technologies are impacting our social communities.
7:01
And, none are more impacted than our children.
7:05
And, so I'm just in love with education.
7:07
I think it's wonderful. And that's why.
7:10
Yeah.
7:11
Well, and I think that idea like,
7:12
AI governance and its effects on children
7:14
is like exactly what we want to get into here.
7:16
But it also any educator right now is kind of thinking like
7:19
what's going on, like what's going to happen in these coming years.
7:22
So another warm up question we like to ask all of our guests is, you know,
7:25
what's the landscape of education going to be looking like in the next ten years?
7:29
Yeah, that's a really good question.
7:31
Because it is
7:33
I think frustrating in many ways for educators to try and listen
7:37
to the narratives that are taking place from big tech,
7:42
big tech books that are coming out to what's actually
7:45
happening on the ground in the classroom, and then also knowing that innovation
7:50
always sort of races ahead and regulation kind of follows behind.
7:54
And then somewhere ten years after that is educational reform.
7:58
So teachers are definitely feeling a bit lost in this.
8:02
One thing I will say is that teachers should not expect to be replaced by AI.
8:07
Not anytime soon, certainly not this decade.
8:12
And probably not the next.
8:14
But they should be expecting that these technologies are going to start
8:18
augmenting their workplace.
8:20
Some of these technologies will be used in their daily work,
8:24
just like they're using Google Search, or
8:27
certain kinds of educational technology with very simple forms of AI.
8:32
More advanced forms of AI will make an appearance
8:36
in their workspace on a more standardized, regular level.
8:40
I think also, the role of the teacher will shift.
8:44
It has slowly shifted from where the teacher was more of
8:47
the sort of conveyor of knowledge, the person in the front room
8:51
who was telling students some issues and concepts around knowledge,
8:57
but students today can much quicker and prefer to find that knowledge themselves.
9:02
And so the teacher really needs to be the person who helps to
9:05
guide them through this knowledge to develop their
9:09
AI literacy skills, critical thinking skills, and also the decision making.
9:13
How to make decision making in the world of enormous stimulation.
9:18
And then I think finally we will see that for our secondary students,
9:22
we are going to see more personalized trends
9:26
in two parts, both because the students are wanting to more
9:30
individualize their learning experience, but also because technology
9:34
makes it possible for students to have a more personalized experience.
9:39
And so teachers need to find their way in that.
9:42
And the teaching methods and strategies for how to do that well
9:46
are only just starting to be thought about now.
9:49
They haven't really been developed,
9:51
but I think we will see this in the coming decade.
9:54
Yeah, definitely.
9:55
And I think what you're talking about there with regard
9:56
to like the fundamental skills that we're really trying
9:58
to develop, that sort of like literacy and decision making skills,
10:01
and literacy as it expands into AI. I mean, it seems like,
10:05
you know, this is becoming essentially a fundamental skill
10:07
that we need to integrate more fully into our classrooms.
10:10
As as educators that are practicing,
10:12
we've got to be looking at how we can work with this,
10:14
be more like the guide on the side and that sort of thing.
10:17
So, you know, in your opinion, what are kind of the fundamental skills.
10:20
When we talk about AI in education, what are like those fundamental skills
10:24
that the teachers and students definitely need to develop?
10:27
Yeah, I think one of the most fundamental skills
10:30
is what's referred to by some as AI literacy.
10:33
But I think the concept of AI literacy means a little bit
10:37
of different things depending on who you ask.
10:39
For some, AI literacy means being able to use these systems
10:43
in a way that improves educational goals.
10:47
But in AI governance, AI literacy means also understanding both
10:51
what the AI can and what it cannot do, and then the real world impacts that it has.
10:57
So being able to critically evaluate the AI's outputs,
11:02
anything from spotting bias, questioning information,
11:07
the ability to identify misinformation,
11:11
hallucinations that might occur, because a large language model
11:16
is making some kind of error, and understanding
11:20
how to manage those situations when they happen.
11:23
And I think also, safe ethical use of AI.
11:28
So that includes not just understanding
11:31
whether you have said yes, you can use my data,
11:34
but really understanding what it is that you're saying yes to,
11:38
informed consent when it comes to privacy,
11:41
informed consent when it comes to transparency,
11:46
in terms of the teachers use of AI, of the students work.
11:51
And then I think academic integrity, very important.
11:54
How do you navigate that space, with all of these new technologies.
11:59
And how do you use the AI effectively to truly learn more
12:03
and not just skip through assignments.
12:06
And then finally,
12:08
we could call it responsible
12:09
digital citizenship, understanding
12:13
how to be a good citizen in the digital space,
12:16
which is a very challenging task.
12:19
But skills can be learned around digital citizenship as well.
12:25
So we've got a big job ahead of us, is what you're telling me right now.
12:28
We do, we do.
12:29
You've been going into schools and working with schools on this, right?
12:33
That's been a fundamental part of your consulting now.
12:36
Correct? Absolutely.
12:37
I've been to, just in the last year, been to more than a 100 schools.
12:42
So I travel a lot,
12:46
and I like it there.
12:48
Although there is many opportunities for online,
12:52
and digital training and learning, being in the space,
12:56
your foot on the ground, and you have the students,
12:59
you have the parents, you have the teachers
13:01
and the senior leadership sort of right in your face saying, these are my needs.
13:05
And how do we do this?
13:07
And often I experience in that space that, the needs in the schools
13:12
is a different conversation that what you hearing from
13:16
big tech, when they are predicting the future.
13:20
And, so there is a disconnect there
13:24
and I find it very important to be out there and really support
13:29
our schools in figuring out
13:30
how do we make this shift, what is it that we should focus on
13:35
right now, in order for us to turn this enormous
13:38
container ship in the right direction.
13:41
And it's a big challenge for everyone.
13:43
And there's both excitement and anxiety about it,
13:46
in different schools.
13:48
Something that I think I've noticed, Kognity, as we were starting
13:52
to definitely be getting a lot more questions,
13:54
people being like, well, what are you doing with AI?
13:57
And it's interesting
13:58
because we're really at that phase of learning
14:00
which questions were even supposed to ask,
14:02
yet alone which ones to answer, like schools aren't really sure
14:05
what they're supposed to know or what they're supposed to be worried about,
14:11
or be concerned about, but they're just like, AI - let's talk about it.
14:16
And so, I mean, when you're going into all of these schools,
14:19
what's the Swedish expression like the red thread, like,
14:22
what are you seeing kind of consistently coming up?
14:25
I think two things that consistently come up are the
14:28
10% on each end: 10% of staff that are incredibly excited and they want
14:34
to experiment and they want to go all in and they are ready for,
14:39
you know, cyborg teachers, robots, they're all in.
14:42
And then we have 10% on the other end that are sincerely afraid.
14:48
They are worried.
14:49
They're concerned, and not just
14:52
concerned for their ability to learn all of these new things,
14:56
but really concerned for whether it's even a good idea.
14:59
And so there is also a push against like,
15:03
can we still be teachers and not use tech?
15:06
And so we have these hard conversations,
15:08
and they go to the bone because,
15:11
everybody has to find their own way in how they want to use
15:15
these technologies.
15:16
But in the big red thread 80% in the middle
15:20
are some, very sort of, common things.
15:24
One of them is about work, like the role of the teacher.
15:28
What skills do I need to learn in order
15:32
to still do a good job for my students and still have a good work life balance?
15:37
And how do I fit that in to a schedule that's already incredibly packed?
15:42
I am so busy.
15:43
How am I going to learn this?
15:46
Do I need to be a computer scientist?
15:49
And then another big question is really about
15:53
what makes sense in the classroom in terms of the use of technology.
15:57
If we bring this in, does it make education better?
16:01
And I think that's a critical question that we are missing
16:05
sometimes, when I'm just with the tech conferences and the techie
16:10
people, that very fundamental question is often missing.
16:15
But teachers are always going into that.
16:17
You don't bring a piece of technology into the classroom
16:20
just to entertain the students and have fun.
16:23
You bring it in because you believe that that piece of technology
16:26
can substantially create a better learning
16:29
experience for those students, and because you know that it's safe.
16:33
And so what's coming up now with generative AI that hasn't come up
16:38
much in the last 20 years, with different types
16:41
of educational technology has been: wait, is this safe?
16:47
Is it okay for me to make an account for a student?
16:51
Because I don't really understand how this is different.
16:54
So they are starting to ask those questions.
16:56
So, as you say, Kognity is getting those questions.
17:00
And I believe that we will see this more and more
17:03
in the years to come, that the communication,
17:06
the dialog between edtech vendors and schools will need to increase.
17:12
It will need to be more effective, informative and transparent.
17:17
Because, certainly in the EU, we can maybe talk about that a little bit.
17:21
But certainly in the EU, the EU act requires that
17:26
the users in the schools are understanding
17:29
the technology that they are using, and they are in fact AI literate.
17:34
It's a mandate.
17:36
And, so I think those questions where before you could take
17:39
a piece of technology like Kognity, or some other piece of technology,
17:43
and just use it and feel safe in doing that, teachers are starting to wonder,
17:49
is it safe?
17:49
We just need to know.
17:51
And, I think that that is a trend of the times.
17:54
Indeed, yeah.
17:56
This episode's KogTip
17:58
comes from Anu Kokko in Birkerød Gymnasium.
18:02
She says, I like the AI guidance, and will give it to
18:04
my students to read and to be able to check that they have read it.
18:07
It is a good starting point for the IA process.
18:10
In addition, I love that there are good inspirational embedded
18:13
YouTube videos that help me in finding inspiration for the topic
18:16
to investigate further with my students.
18:21
I mean, there are some fundamental things like when it comes to
18:23
using different platforms where obviously they have like age limits
18:26
and that sort of stuff.
18:27
There is their specific policies
18:29
for how students, or people, can interact with that.
18:32
And I noticed, you know, as a practicing educator,
18:34
we see this with our students as well, where
18:36
suddenly they're using some form of tech where already you're
18:39
just in the back of your head.
18:39
You're like, well, you're you're not old enough actually to be using that tech.
18:43
So it lends itself to the policy development and the like,
18:47
implementation of that policy, the practices that come with it.
18:50
So I guess, you know, you just mentioned you're seeing over 100 schools a year.
18:54
Are there any schools where you see like they're really doing it quite well,
18:56
like this is kind of
18:57
got it going on there, got there I governance well,
18:59
their policy into practice is working well.
19:01
And you see it in acting with the students and the teachers quite well.
19:06
That's a super question,
19:07
because we're always looking for good cases.
19:09
And I think, in many schools, it is a work in progress.
19:15
But I have seen a set of schools in Morocco, it was a set of French
19:21
and English bilingual,
19:23
and one of them was trilingual, private schools, in Morocco.
19:27
And, they took a decision almost a decade ago
19:30
to put in their senior leadership position at digital
19:36
coordinator for the teachers.
19:38
So this person's role in the senior leadership team is to work with
19:44
the teachers on identifying technology that could be useful,
19:48
and then how to use it well, for their particular subject.
19:52
And so they have many different types of technology.
19:55
And, they go through a series of activities
19:58
before deciding on a technology, including filming lessons,
20:02
measuring the outcomes, doing little pilots,
20:05
and then they scale it out, and it's a safe
20:09
feedback form for the teachers and the learning coordinator.
20:13
And, that person’s sole job
20:16
is to work on methodologies and technology.
20:19
And that has been amazing to see. When I've gone
20:22
to these schools, I'm just blown away with how smartly the technology is used
20:27
and also how considered. There's lots of analog learning where the kids
20:33
are sitting and doing other things without any technology at all.
20:36
They do a lot of stations where they rotate students
20:40
in and out, of different learning environments, including technology.
20:48
They’re very clear on their understanding of why that technology is there.
20:51
And I think that's a good case for us,
20:54
because the parents that are coming up now,
20:57
our youngest millennium parents and our oldest Gen Z parents,
21:03
they have grown up with technology as opposed to the Gen Xers.
21:07
And so they have learned the hard way what works and doesn't work.
21:11
And they will be asking much more critical questions, including,
21:15
why are you using this technology with my kid?
21:18
And so I think having someone there that really works with teachers,
21:21
developing methodologies, is in a fantastic case.
21:25
It's not in the budget for everyone.
21:28
But that's an example that comes to mind.
21:30
I think that's great because I think there's also, I personally,
21:34
I see such a difference in teachers who feel confident in using the tool.
21:39
And I'd imagine if they have that person, like you said, that feedback loop,
21:42
just somebody to bounce the ideas back, like, let me make sure I understand this.
21:45
And that way
21:48
it's very scary to go up to a
21:50
group of teenagers and start using a tool that you're not fully comfortable with,
21:54
and it can derail you so quickly as a teacher if you start
21:59
hesitating and being like, oh, actually, I don't know what I'm doing
22:02
with this thing.
22:03
And so you're, you're going to be more resistant
22:06
to trying something new if you're not confident in using it.
22:10
And so, I mean, I've done a lot of teacher training.
22:12
And that's always one of the biggest things that I want
22:14
to make sure that we achieve is this confidence level.
22:16
But I guess it's not just confidence in utility
22:19
and using the tool, but also in why you're using it,
22:22
and what are the impacts of the students, like what is going on.
22:27
It's really taking it to a much deeper level of not just,
22:31
I know how to use this program, but I know why I'm using it.
22:34
The potential outcomes,
22:36
the ethics component, like you were saying, like,
22:39
now, these kind of dialogs also need to be happening with your classes.
22:43
And I guess
22:43
that's probably part of the reason that there's
22:45
almost like a delay, as you were commenting earlier, Clara,
22:48
like, we kind of innovate really quickly and then we like figure out
22:51
how to enact that.
22:52
And then behind
22:53
that comes the policy and the behind that comes like practice of it.
22:56
Right?
22:56
It's like the development comes a little bit behind what's going on there.
22:59
But I mean, obviously as you're talking about AI governance,
23:02
you know, in the back of my head, I work as part of a senior leadership
23:04
team at the school.
23:05
I'm constantly thinking, okay, how does this inform our policies?
23:08
And the obvious one for most schools is, you know, academic integrity,
23:11
but it extends beyond that.
23:13
And like the actual policy itself, like an AI policy,
23:16
how do we actually do things at our school itself?
23:18
I wonder if you could comment with regard to
23:20
AI governance in schools and education,
23:22
young people like, why are those policies so important and how like, what are ways
23:26
that teachers can kind of
23:27
easily get involved in that in a way that doesn't feel intimidating,
23:30
but they can actually like, move forward with it?
23:32
Yeah.
23:33
I love policies, but I love life policies.
23:37
I love them as live documents.
23:40
I just heard every teacher be like, oh, I love policies, too.
23:45
I do, I think, I love clear instructions,
23:48
on, you know, what we can and cannot do.
23:53
And I love them when they are live policies, when they're hands
23:57
in the pudding, that we are constantly working with those policies.
24:01
I go sometimes on school websites and look up some of their policies.
24:05
And they were last signed in 2013, you know,
24:08
and I think okay, so, you know, it has no meaning.
24:11
And, when it comes to AI, there's so much,
24:15
it's such a difficult terrain to navigate
24:19
for any person, that,
24:22
it is not realistic to expect them to be able
24:25
to take all of their own decisions on what they should do
24:29
in x, y, z situation. And their policies
24:33
are really a tool to help them feel safe.
24:36
in making decisions
24:38
around the use of technology in the classroom. And that,
24:42
it is also a tool for them to build trust
24:46
with their most important stakeholders,
24:48
the children, the students, the families, both
24:52
in primary, secondary education, but really extending into universities.
24:57
We won't have a relationship or even very many people
25:01
enrolled in our schools, if we don't have a trust relationship.
25:04
Once that trust breaks down, everything else falls.
25:08
Parents certainly need to know that
25:10
when they send your child to you, they're going into a safe space.
25:14
And policies around AI can
25:17
really help both the communication,
25:20
on safety,
25:22
but also on
25:25
everybody understanding exactly why it is that we are doing the things we're doing.
25:31
Moreover, of course, for schools,
25:34
policies that are helping schools,
25:37
align themselves with law, data protection law,
25:41
privacy laws, transparency laws and laws coming up in AI,
25:46
not just in the European Union, but we're seeing AI law,
25:52
many of them risk based laws,
25:55
developing in many regions of the world.
25:58
Policy also, I think,
26:00
and this is one of my favorite about policies,
26:03
they really help vet vendors.
26:06
There's a lot of very sparkly, very fancy,
26:10
technology in the educational marketplace.
26:15
I was just at an ad tech conference, and I think there were 170
26:19
new startups, and all of them had AI inside.
26:23
And I thought,
26:25
good policy will help schools figure that out as well.
26:29
Because it will help schools really figure out
26:31
how to ask the right questions from their vendors
26:35
and make sure that they are using school approved accounts
26:38
and not their personal accounts, which is a lot,
26:40
there's a lot of risk and liability around establishing personal accounts.
26:45
And then, policies also help with training.
26:48
They help schools identify exactly what needs are, in terms
26:53
of training students and the staff in being responsible in their AI use.
26:57
So I think, a policy is
27:01
one of the most fundamental
27:04
concepts to get in place well,
27:07
as the school begins to ready themselves for building
27:10
an AI governance framework.
27:11
So much of what you're talking about seems to be like
27:13
it's effectively safeguarding, right?
27:15
We're really looking after young people.
27:17
We're looking after the people in our schools.
27:19
So of course we're mission driven and we're trying to abide by the laws
27:22
of where we might be located.
27:23
But it seems like so much of this is actually just like protecting
27:27
young people as they engage in this wider world.
27:31
And again, like become citizens of a digital space,
27:34
not only of the physical space.
27:36
And so I guess in that respect, like, to what extent do you see this as like
27:39
a standalone kind of policy development, or is this more of a
27:42
this is basically safeguarding or is there more than that?
27:45
Yeah. Good question.
27:47
You'll be amazed that how many vendors I have spoken to where I ask them,
27:52
tell me how safety is built into your design.
27:55
Tell me about,
27:58
you know,
27:59
imagine I'm a parent and tell me about the safety.
28:02
Go into details.
28:03
And they struggle with this question, because a lot of edtech is not built
28:10
starting with safety and design.
28:13
And so, yes, essentially a good AI policy is safeguarding,
28:19
because that must be the number one reason why we go to work every day,
28:24
is that the people that are put in our charge are safe.
28:28
And then the second that they are thriving
28:30
in that safe space and third, that they are learning.
28:33
But if we can't keep them safe, then we're missing the mark.
28:37
And I certainly think, very highly, of safety.
28:41
I start and end my day with it, but I think also very highly of innovation.
28:46
And I actually believe that safeguarding when it comes to technology
28:51
will lead us to better innovation, because there's plenty of technology
28:55
out there that should have never been in schools to begin with.
29:00
And this will allow for schools to make informed decisions
29:04
and get some of that technology out of the schools as well.
29:07
Yeah.
29:08
I mean, obviously people were using edtech solutions
29:12
before AI came out, and there were a lot of really important questions
29:16
that we would be asked about the technology that we're providing,
29:19
things like that, in the safeguarding space.
29:21
But is there anything that you think has become,
29:25
kind of increasingly important to be asking, when you are evaluating
29:30
these AI tools in the school, that maybe didn't come up as much beforehand?
29:34
Yeah, I think there are issues that we see in a different way.
29:38
So, of course,
29:41
there have
29:41
been privacy laws in the US, there's the copyright regulations.
29:46
In Europe you have the GDPR,
29:49
and vendors mostly have been able to deal with that very easily.
29:54
What was different in that regulation that's coming, is that
29:58
people have to make informed decisions, on the use.
30:02
And we are seeing that important
30:07
concerns are already taken place.
30:09
So, for example, we see a lot of
30:12
edtech companies that have embedded
30:14
some kind of a large language model wrapper into their tool.
30:18
So they have a little chat bot, or something like that,
30:21
that is driven by a large language model.
30:24
But at the end of the day, we really don't have a good understanding of
30:29
how the data in that chat bot is exactly used.
30:33
We don't exactly know what servers that information is going to.
30:38
And we keep in mind here
30:39
when we're talking to schools, primary, secondary, we're talking children.
30:43
So these are children's unique data, that are being sent off to
30:48
we don't know where.
30:50
There's also a lot of unnecessary data collection.
30:53
And so we need to know that the vendor is not collecting data
30:58
that is anything other than what's absolutely necessary
31:02
for the technology to work well,
31:05
but that there's not anything unnecessary data coming up.
31:08
So that could be something like,
31:11
an edtech that is
31:13
tracking the mouse clicks of a student,
31:16
to track how engaged they are in x, y, z task.
31:21
That seems probably at first like a good idea,
31:23
but it seems like an unnecessary data collection.
31:26
And it's also potentially a breach of certain regulations.
31:30
So we need to be better at sort of asking those types of questions,
31:35
because it is students information that is being sent out.
31:39
And then, of course, we also know that a lot of the
31:44
generative AI has some errors and some biases.
31:49
It has a preference for the male prototype.
31:53
It has more data,
31:55
that is produced by young males, and it has algorithms
32:00
that have trained the AI that's primarily,
32:03
trained by male employees.
32:07
We are missing data from two thirds of the global population.
32:13
There's a billion people that still don't even have electricity,
32:16
much less are hooked up to the internet producing data.
32:20
We know that there are risks around language.
32:24
For example, the American large language models process in American English,
32:29
not in British English or any other language.
32:32
So even if I asked a question in, say, French,
32:36
the model would then translate the French into American English,
32:41
process the information, produce a response, translate it back to French,
32:45
and push it out to me.
32:46
So I'm getting this sort of American
32:50
biases in my responses, which is why students
32:54
that are using a large language model in French often feel
32:57
that they are disconnected from the system.
33:01
And these are real risks, because this comes down to an equitable issue.
33:06
Are the students getting equitable access in their education?
33:09
And we know that with some forms of AI, they are not.
33:12
So we do have to address that as real risks.
33:15
Yeah. Well, you know, if you think about like the IB
33:17
mission itself and ways in which the IB is really trying to push things out
33:20
through the diploma program, we want to develop
33:22
people who have intercultural understanding and respect.
33:25
And when you see that widening of the gap, it's kind of like the,
33:29
a preference for for certain models of education and almost
33:32
like removing of indigenous styles of education, for example.
33:36
And so, you know,
33:37
as you were talking there, I think quite a bit about
33:40
theory of knowledge, I was thinking quite a bit
33:42
about like developing critical thinking skills.
33:44
And really essentially just questioning, is this worth time or not?
33:49
Like, how do I know if I can actually trust this?
33:51
Where's the value in this particular platform?
33:54
And I guess, does it lend itself to school approved
33:58
AI tools and AI accounts?
34:00
Does that kind of help the process or is it like exacerbate it even more
34:04
because you're getting like this narrowed window view
34:06
of what you're looking at.
34:08
Yeah, I think the importance about school approved accounts
34:12
is that schools can, maybe as part of the AI policy,
34:16
set up a sandbox for the 10% of the super innovative one experiment,
34:22
you know,
34:23
teachers, where they can, in a safe space, test out some of these tools,
34:27
before they use them for the whole school.
34:31
And, it will probably always be the case that in a school environment, students
34:37
will not have access to the super newest model that have just come out.
34:42
And nor should they, because we need to make sure
34:45
that the technology that we bring in has an educational value that is safe.
34:51
And, we can't figure that out on a whim.
34:54
So of course, there's lots to be excited about, but it's really important
34:59
that the school goes through the process of identifying a vendor
35:05
that meets the standards of the school and the value of the school.
35:09
I think the IB, of course, is a great example.
35:11
As I said, I have a soft spot for IB, but I actually believe that
35:15
the IB learner profile
35:17
is a really good profile to develop in a time, where you live
35:21
with very advanced technologies that are coming in very quickly.
35:25
I mean, we're talking generative AI now, but in ten years we'll be talking
35:30
quantum computing and it's a whole different conversation.
35:35
But still the IB learner profile really lends itself well,
35:41
because it is about developing holistic thinking,
35:44
risk taking, abstract thinking, and engaging with other viewpoints,
35:48
whether they come from a machine or another human being or another country.
35:53
And I think that there is a lot of strength
35:55
in that for the IB.
35:58
The challenge there then is to ensure that when technology is
36:03
used, that it's used in a way that aligns itself with the values,
36:07
and identifying when it may not, and then how to deal with that.
36:13
Yeah.
36:13
Have you seen any really powerful
36:15
or really great ways that schools have been teaching
36:19
whether it's the ethics component and the academic integrity
36:22
component with AI, is there anything that you feel like
36:25
has really resonated with students where they're like, okay, I get this.
36:29
Not like, yeah, yeah, okay, I'm just going to use ChatGPT anyway, kind of thing?
36:33
Yeah.
36:34
So I think, you know, why do students use ChatGPT for their schoolwork?
36:40
Well, some use it because they think the schoolwork is a waste of time
36:43
or because they think it's boring, or because they,
36:47
you know, prioritize their time poorly and they need to speed up the process.
36:53
But, I have seen teachers that are using it very well.
36:58
What they have done very well is to mark out with clear instructions,
37:03
particular intersections where they want for the student to use generative AI.
37:08
So it could be something like, initial idea generation work with a generative AI.
37:14
Here's what I want you to do at that step.
37:17
These are the instructions.
37:18
And then, the student turn in
37:20
their prompting, a history as well to kind of show how they interacted.
37:25
That allows for the teacher to give the student feedback on
37:28
how they could interact more effectively,
37:31
with the AI as a sort of a technical skill,
37:33
but it also allows for them to see where the student
37:37
underuse their own critical thinking and overuse the technology.
37:43
And then in different stages of these assignments, they
37:46
then say, here you can again go in and use.
37:50
Also, I've seen teachers very
37:51
effectively use generative AI for feedback on drafts.
37:55
AI, I know we hear in the news, that it will replace grading,
37:58
but it's not very good at grading.
38:03
Certainly not the large language models.
38:05
But it's very good at giving feedback on drafts.
38:08
And teachers have used that very effectively,
38:11
which allows for the students really to go in
38:13
and address their papers much more efficiently, than they were
38:17
if the teacher had rushed through a draft feedback
38:21
in too little time to get it done.
38:25
So I think that there are some wonderful cases appearing.
38:29
There's also some instructional frameworks that are starting to come out, and some
38:34
empirical research that is coming out about what works, what doesn't work.
38:39
There's a super interesting study that came out of Nigeria
38:42
where students did two years of learning in six weeks using
38:47
advanced AI, to take them through the learning process.
38:52
And these were students who had no
38:54
computational experience at all, and had enormous learning gains.
38:59
So I think it can be used well, but it can also be used poorly.
39:05
And, of course, that's the rub.
39:09
Becoming a better IBDP educator is a daily practice informed through reading,
39:14
research, personal experience, and episodes from IB Unplugged.
39:18
Like the one you are listening to right now.
39:20
But sometimes the show can go a bit too quickly, or we can simply forget
39:23
some of the key points once we are out of the car
39:25
or back from our walk and in the office, preparing for our next lesson.
39:29
That's why it's easy to download the episode summary
39:32
at kognity.com/ibunplugged
39:34
Simply navigate to today's episode and click Free Summary
39:37
to receive a condensed summary of today's episode, including links to other sources
39:41
and other exciting bits of information you can use immediately
39:44
in your daily practice.
39:45
It's free and it helps.
39:47
Now back to the show.
39:50
You mentioned it kind of earlier.
39:51
It's like you want your ten percenters to like go and try it and innovate it
39:56
and get comfy with it, but you don't necessarily
39:58
want them to just immediately start using it, right?
40:00
Because they need to know that this is a safe and ethical thing, you know?
40:03
You spoke about, providing feedback on drafts and whatnot.
40:06
And that's definitely been a conversation
40:08
at my school where I work in terms of like,
40:10
yeah, this can be an effective way
40:12
to give some positive feedback for students.
40:14
They can learn very quickly.
40:15
But, you know, we have a digital learning coordinator
40:17
who's like, be sure you use this AI and not this one,
40:20
because this one protects data and any sort of
40:23
like information of school or student information, whereas this one doesn't.
40:26
And I think
40:27
if you've got your ten percenter who's kind of running with it,
40:30
who doesn't do that homework, then you may actually
40:32
be causing some breaches of, you know, policy,
40:34
as we were talking about before, or even, you know, creating more
40:37
of a digital footprint for somebody than what they initially sort of signed on for.
40:40
So there's that importance, I think, of
40:42
really wrestling with it for a bit and almost like holding yourself back,
40:46
even though you're like, oh, this is a great tool.
40:48
Let me just kind of make sure that it's good in our context.
40:52
And then the other one, I just want to comment on what you mentioned before.
40:55
Like, we do like a traffic light system.
40:57
We pick that one up.
40:58
So it's kind of the red, yellow, green.
40:59
And like green is like that full use of AI.
41:01
That's cool. Use that to generate some ideas.
41:03
And so that way you can, red means like don't use it.
41:06
So it kind of easily sets up each assignment.
41:09
You just like all right.
41:10
This is a green light sort of AI usage.
41:12
This assignment's a red light.
41:14
And then trying to pull out that critical thinking.
41:16
But what I found really interesting, especially with the IB recently on their
41:20
on their policies on academic integrity and the use of AI is not only do
41:24
students have to sight that they use AI,
41:25
but they have to sight their prompt, like what you mentioned before.
41:28
And that kind of brings out that critical thinking a little bit more.
41:31
So when we think of it from an educator perspective,
41:34
how much of this is like teaching, not just teaching
41:37
use of AI, but teaching, like how to ask great questions.
41:41
And ultimately that's just kind of coming back to good teaching.
41:43
Yeah. Yeah, exactly.
41:44
I have said sometimes, over the last few years, that
41:49
English teachers are a valuable resource right now because they know language.
41:55
And, any language teacher really, but
41:59
they understand how to weave sentences together, how they understand syntax.
42:04
They understand all the nuances of expression and the better you can do that,
42:10
the better you can communicate with a large language model.
42:13
And, so those who have really good results
42:18
with large language models are the ones who already have those skills.
42:23
And so learning to express yourself
42:26
is part of learning AI literacy.
42:30
And, which is why
42:32
we won't see languages disappear out of our classrooms,
42:36
because they are really needed more than ever.
42:39
And there are actual people who have majored in language degrees
42:44
that are making a significant amount of money
42:47
working in the tech industry because they understand language.
42:52
And, I think that is an interesting dynamic, because
42:57
when you hand in your prompts, it's a valuable resource
43:00
for the teacher, actually, to see how that student communicates
43:04
with the machine and then guide them to do it more effectively.
43:07
I might add to that bit, of course, that,
43:12
there are also some advantages for students,
43:15
maybe, who have struggled with communication before,
43:19
because we are getting more and more of a conversational language
43:23
with the large language models, and it actually allows the pathway
43:27
for students who may have struggled, let's say students who are in a IB
43:32
English language pathway or IB Spanish pathway who don't speak that language
43:36
as their first language, and are sometimes struggling with
43:40
getting their points across or being part of the conversation.
43:44
And because we have more of a conversational large language models,
43:48
they actually now have a better opportunity
43:51
to be more of an equitable part of their school community,
43:54
which I also think is is a very important part for us
43:58
to consider in this context as well.
44:00
Well, a really positive use of AI as well.
44:03
I mean, the number of students I've come across who,
44:05
you know, their language skills don't allow them to fully participate
44:09
in the conversation or participate in the dialog.
44:11
And therefore, you know, there's that fear of like,
44:14
I am smart, I know what I'm doing, I know I can contribute to this conversation.
44:17
I just can't do it in this particular language.
44:19
And I think it's those students that we see very often kind of going to,
44:24
a large language learning model
44:25
or some sort of translator that sort of helps them get that across.
44:29
And yet at the same time as educators, then we've got like our antenna
44:32
that's like, oh, you use this tool that you're not supposed to.
44:35
And so it almost becomes like a double edged sword for a student who's
44:38
trying to grow up in that.
44:39
And they're going like, I want to contribute to the conversation.
44:41
I just don't have all those skills yet.
44:43
So I use this thing.
44:44
But now you're telling me I can't use that thing?
44:47
So I think it creates a real tension there with,
44:49
you know, you want to develop these skills,
44:51
you want students to be part of the conversation,
44:53
you want to, like, bring out their best.
44:55
But also how do we do that in an ethical and safe way?
44:59
And so I guess
45:01
this kind of leading to that question, like,
45:03
how do you navigate those ethical lines of AI use, technology skills acquisition
45:08
in realistic, specifically international school settings,
45:10
where you've got like students from all over the place
45:12
and they're not all just one
45:13
nationality learning the IB curriculum, but for example,
45:17
they come from many different backgrounds, not just their families.
45:19
The way have come used to education in general, the way they perceive grades,
45:23
the pressure that might come
45:24
from performance and therefore they may use these tools in a different way.
45:29
How do we navigate that ethical and that sort of skill development?
45:32
Yeah, I think one of the things
45:35
we've seen, for example, from the IB is that IB was very quick to come out
45:39
with instructions for how to use AI ethically in IB work.
45:46
What many then immediately
45:48
said, okay, this sounds great, but how do we do that in practice?
45:51
And the truth is..
45:53
I was one of them.
45:56
I think I might have asked that question myself
45:59
at a conference a couple of years ago.
46:02
But I think one of the things for us to remember is that we don't have
46:06
all the answers yet because we haven't lived with this technology before.
46:11
So if we sort of dial
46:12
the clock back to 21 years ago when Google first came in the classroom,
46:17
many of the conversations were a bit the same.
46:19
How do we use this?
46:20
Will it replace teachers? Will we still need a library?
46:23
How do we do all these things?
46:25
And over time, teachers found really good use for having something
46:30
like a search engine ready in the classroom where they could go in
46:33
and ask questions immediately
46:36
to which they may not have had the answers.
46:40
Large language models is a bit the same way.
46:42
Things like spellcheck and Grammarly very like initial signs of AI that now
46:47
when we talk about AI in the conversation of a large language
46:50
learning model, it's like, oh, that's a completely different thing.
46:52
But you also have to say, well,
46:53
no, we've been using a lot of these tools for a while.
46:56
Spell spellcheck, even Clippy back from Microsoft, that was still AI, right?
47:00
So there's still elements of, you know, we've got used to this in the past.
47:04
It's just a matter of like being adaptable
47:06
and thinking about the skills we're trying to train along the way.
47:08
It is and I think the significant difference here
47:12
is that the other types of use of the AI have been relatively inside of a frame.
47:17
We were able to sort of say, okay, the AI do x, y, z, and then that's what it does.
47:23
But these new large language models are more unpredictable, you know,
47:27
and a lot of it is a black box.
47:29
We don't really know exactly what's happening on some levels.
47:35
And that I think is the big difference that now requires us for really
47:40
for all of us to make informed decisions when we are using these technologies.
47:45
And for students needing really clear guidance on when is it
47:50
okay for me to do it and when is it not okay and when it's okay
47:55
how do I do those things?
47:58
And I think also importantly for teachers
48:01
to say, now, I don't want you to use AI and here's why.
48:06
Now you have to learn something human to human, a truly analog
48:10
human skill that AI cannot replicate.
48:14
And, that also requires
48:16
AI literacy to be able to actually make that decision
48:20
and validates the need for training, I think.
48:24
Yeah.
48:24
Where do you think that a school,
48:26
if maybe they can't get somebody to come in as a consultant
48:29
or they can't have somebody in senior leadership that can be leading the way.
48:32
If you're in a school where you're
48:34
maybe a bit more on your own, but you know that this is something that's important
48:38
and that you want to focus, where would you start for those teachers?
48:41
Like what could they do on their own?
48:43
I would definitely start with coming together as a staff and
48:48
given input on AI policy.
48:51
And I think, for those
48:54
sessions, I would also invite parent groups and student groups in,
48:59
and start with a sort of major brainstorm to identify
49:02
the unique needs of each group of people in the school.
49:06
And very importantly, I think also marketing admissions
49:10
officers, the front office, they really have
49:14
a very different experience of how AI is used, really important
49:18
for to get schools to right away get those perspectives in,
49:23
and then, for the policy team to sit down and really develop
49:29
a meaningful policy and get that communicated out.
49:32
Aligning with efforts to identify how AI is used,
49:37
because you can't really
49:38
say much about how you want to use AI if you don't know how it's already used.
49:42
And I think many schools will be surprised, about how many educators
49:47
and students have private accounts with different types of AI.
49:53
And that
49:54
might actually seem like a good idea at first, but in the long run,
49:59
maybe could be done more effectively by the school,
50:05
making sure that it's a decision that's taken together.
50:10
And then there are some free resources.
50:13
We need more, because schools have no time and they have no money.
50:18
These are always the two things.
50:22
So, we need resources.
50:23
We need accessible resources for training.
50:27
I know there are some of the big tech that have created resources.
50:31
So a lot of schools, for example, have
50:34
Microsoft for education, or they have Google for education.
50:37
And both of those companies, for example, have
50:42
free resources online, whether you have a company
50:45
account with them or not, you can go in and take the training there.
50:49
There are other vendors that offer free resources.
50:53
There are lots of webinars as well online.
50:57
But I think that we will see more and more national governments
51:01
not only finalizing their guidelines,
51:04
but also creating resources for schools.
51:09
In the relevant regions and countries.
51:12
Saudi Arabia comes to mind.
51:14
They just finished and published an amazing instruction
51:19
for the use of AI in education for Saudi Arabian schools.
51:23
And they will then follow up
51:25
with training for their school systems as well.
51:29
And we will see this more and more.
51:31
And, although it feels like a very urgent need,
51:35
these things do take time.
51:37
You know, it does take a little time to develop the resources.
51:40
But, there are free resources online,
51:44
and then of course, in some of the online courses
51:49
that you can take as well, where there are training resources.
51:53
But time is complex.
51:56
And many teachers really need to know practicalities.
52:00
What happens in x situation?
52:02
What do I do?
52:03
And so many prefer the face to face workshops and
52:08
environments, and the need for that is enormous.
52:12
Absolutely enormous.
52:14
So, I mean, as you were speaking there, I was kind of making me think about, like,
52:17
the almost
52:18
a luxury of kind of having that top down governance, you know, like as a school,
52:22
if you can lean on your countries or your, you know, in the EU, for example,
52:26
you mentioned it earlier at the top of the show here, like the
52:29
the EU AI act and sort of now like having that there kind of offers
52:33
a little bit of extra governance where the school doesn't feel like
52:35
we have to just be navigating this water on our own.
52:38
And it was good to hear you say that. Okay. That's already started.
52:40
That kind of thing is being replicated in other places.
52:42
You mentioned Saudi Arabia.
52:44
And so I guess I'm wondering, like for those places that don't have
52:48
that sort of national governance yet, and that time might be coming soon,
52:52
is the best thing that that school can do.
52:54
Just kind of looking for these free resources as you're talking about?
52:56
Or is it like developing on policies, or is it like what do you suppose
53:00
the schools should be doing if they don't have that top down governance?
53:04
So I think there is a document that's formed to be
53:07
a kind of a global guidance from Unesco.
53:10
Unesco has an enormous resource library, including recommendations
53:15
for the use of AI and education, which has a lot of really practical
53:21
steps, and advice, guidance and that schools can take.
53:25
But I think it's very important for every school to really have an AI policy.
53:30
You know, schools have different curriculum.
53:33
They have different values, they have different nationalities and languages.
53:37
It's really important for the school
53:38
to start with saying, what are our values, what's our mission and vision?
53:42
And then how can we use AI in a way that we help
53:46
grow that and sustain that for our community?
53:50
And it is work.
53:53
And it does require some deep digging.
53:56
But I think it's work that maybe should have been done, really,
54:00
10 or 15 years ago, when the big push
54:03
for digitalization happened and everybody was buying laptops, and
54:08
nobody was telling teachers how to use these laptops to make education better.
54:13
Which is why today you often go in classrooms.
54:15
And what you see is 20 students in an office landscape.
54:19
And you're not really sure if they're learning or if they're just
54:22
surfing around on the internet.
54:24
So this is an enormous opportunity for school to say, well,
54:28
why are we using technology?
54:30
What is it we want to use it for, in order for us to meet our mission
54:35
and our vision for our school,
54:36
and in order for us to keep the students safe and make education better.
54:40
And those are important questions to ask, on a regular basis anyway.
54:46
So I think it's an amazing opportunity.
54:48
I think Europe, EU is lucky in the sense that we are now compelled to do it.
54:53
We're mandated to do it under law, so the work will be done.
54:58
But I also think it's an amazing opportunity for us
55:01
to really ensure that we can say with certainty
55:05
that when we're using technology, the kids are safe using it,
55:09
and they are learning from it, and we know why we are using it.
55:12
And, I think that other regions are not so
55:16
fortunate where when it comes to those things.
55:19
Great. I really, really appreciate this conversation.
55:22
I think it's been super interesting
55:25
to kind of get into some of the nitty gritty parts.
55:28
I, of course, from the edtech background space,
55:32
I love the idea of having these conversations, because I think that
55:37
from the vendor side, it should be a conversation
55:41
that you're comfortable having,
55:43
because you shouldn't be building anything that isn't bringing value.
55:47
You should be able to stand behind.
55:49
This isn't just a feature we released because it's high tech and new and shiny.
55:53
It's like, what's the pedagogical reasoning behind building this?
55:57
And what's the gain in the classroom?
56:00
And are you, of course, always keeping your students in a safe space?
56:03
So, you know, I think it's a fantastic conversation that we will be continuing
56:08
to have more and more this year, in the next decade, as you said.
56:12
And yeah, I'm just really grateful that we were able
56:15
to have this conversation with you, Clara.
56:17
Yeah, Clara, you're brilliant.
56:18
This was tons of fun. I really appreciate you taking the time.
56:20
This is a lot of fun.
56:21
Yeah, well, thank you to both of you.
56:23
I had a blast, and I love talking to edtech vendors, because
56:28
the whole
56:30
marketplace for edtech vendors is really shifting,
56:33
because these questions are coming up for schools in a different way.
56:37
And I think it's just super exciting for us to have more meaningful
56:40
conversations about, you know, we all want the same thing,
56:46
at the end of the day, hopefully, and, the vendors who don't, will
56:50
find their way out to schools anyway.
56:53
And, I think it's just, of course, fantastic,
56:57
just to see you, Courtney, because I don't see you enough.
57:01
And it was so nice meeting you, John.
57:03
And really, I appreciate both of you,
57:05
taking the time to chat with me as well.
57:07
I enjoyed it immensely.
57:08
Ah, wonderful. Thank you. Clara.
57:09
Great stuff.
57:10
The pleasure is ours. So, yeah. Thank you so much.
57:13
And, yeah, hope to
57:15
talk to you again soon about this conversation and keep it going.
57:19
IB Unplugged is a Kognity podcast hosted by John
57:21
John Patton and Courtney Marline and produced by Katilina Gaffney.
57:25
Be sure to subscribe or follow and share with your fellow IB educators.
57:29
If you'd like to submit a topic, request, or a question, please do
57:32
so on kognity.com/ibunplugged.
57:35
Let's keep the conversation going
57:37
and support each other in teaching with confidence, creativity, and joy.
57:40
See you next time.