Latest / Integrate This! Ed Tech Podcast / Can AI Truly Redefine Learning? Becky Keene Suggests AI Optimism in Education

Transcript

Can AI Truly Redefine Learning? Becky Keene Suggests AI Optimism in Education · Integrate This! Ed Tech Podcast

0:00
Welcome to Integrate This, your go -to EdTech
0:03
podcast where curiosity meets classroom innovation.
0:07
Join your hosts, David Berner and Jeremy Mikla,
0:09
as we explore the rapidly evolving world of educational
0:12
technology. From game -changing tools to insightful
0:15
conversations with today's top EdTech voices,
0:18
we're here to help educators stay informed, inspired,
0:21
and ready to innovate. So tune in, level up,
0:25
and let's integrate this. Welcome to the Integrate
0:31
This EdTech podcast. Today we're very excited
0:34
to be joined by AI optimism author, Becky Keene.
0:40
Very excited to have you with us today. Becky
0:43
Keene is an educator, author, and speaker focused
0:46
on innovative teaching and learning. She specializes
0:50
in instructional coaching, game -based learning,
0:52
and integrating AI into education to empower
0:55
students as creators. She has developed eSports
0:58
programs for schools and explores immersive learning
1:01
through games. Becky speaks globally on AI and
1:05
education and has spent over 20 years designing
1:07
professional learning experiences for teachers.
1:10
A national board certified teacher, ISTE certified
1:13
educator and certified instructional coach. She
1:17
has spent 15 years teaching, coaching and leading
1:19
programs in public schools. She's the author
1:22
of the book AI Optimism and holds a master's
1:26
in ed in early literacy. So welcome, Becky. Very
1:30
happy to have you here. Join the integrate this
1:32
podcast. Can you give the audience a little bit
1:35
more insight into your background and how the
1:38
last 20 years here have gone for you? And now
1:42
you've got the book release as well. So. Yeah,
1:46
thanks. Thanks. First of all, thanks for having
1:47
me and. Happy to share a little background. Sometimes
1:51
people ask about my like career journey. I've
1:55
spent 25. This is my 25th year in public supporting
1:59
public education, which is nuts to me that it's
2:04
been that long. But I taught elementary school
2:07
and then I taught middle school and I was really
2:10
fortunate to work with people, leaders, school
2:14
leaders, system leaders who allowed me to do
2:18
things I was excited about, who didn't kind of
2:21
keep me in a box or force me to follow some curriculum.
2:25
A lot of the mantra early on in the 2000s was
2:28
if your test scores are good enough, you can
2:33
do whatever you want. And so that was really
2:35
helpful for me. And I got to do a lot of really
2:38
innovative and creative things, which really
2:40
set the stage for me understanding that there's
2:43
a lot possible in education. if we move some
2:47
of those challenges out of the way. And so that
2:50
was really fun. And I got to be a part of a lot
2:52
of different proof of concept programs and pilot
2:54
programs because I said yes. I just kept saying,
2:57
sure, I'll do that with my students. Yeah, I'd
2:59
love to try that out. invited at some point to
3:05
work at a system level in my school district
3:08
south of Seattle, Washington to help support
3:11
and technology district wide and really scale
3:14
the programs that I had been piloting in my own
3:18
classroom with some other teachers as well. And
3:20
so we worked on that for about 10 years and then
3:24
I had the opportunity to join my current. company,
3:28
I2E, and help schools and systems, primarily
3:32
North America, but we work with companies and
3:35
organizations worldwide to help them implement
3:37
technology in new and innovative ways. In the
3:42
business world, we focus on productivity and
3:44
creativity. And then in the education space,
3:46
obviously, we focus on best teaching and innovation,
3:49
which is a ton of fun. So that's a little bit
3:53
about me. And somewhere along the way, I became
3:54
a LinkedIn learning instructor. which is fun
3:58
and I have two books now and a blog and and a
4:02
newsletter and I do all sorts of oh and TikTok
4:04
walks. I do all sorts of little things but it's
4:07
it's been really fun to get out there and get
4:10
connected with educators like you from all over.
4:13
Yeah and that's that's really awesome. I've seen
4:15
a couple of your TikTok walks that you've done
4:18
here in the last couple weeks or so. But it was
4:21
really cool for me to be able to connect with
4:24
you at ISTE this year. I know I came up, you
4:27
were doing a like a tabletop presentation and
4:31
I came by there and it was nice to be able to
4:35
connect with you there. But overall, what would
4:36
you say about an experience like a conference
4:39
like that? You know, being able to not just kind
4:43
of get your message out there about your book,
4:44
but to be able to like find networks of people
4:47
who kind of like are on the same way. of length
4:50
and can kind of like get to get to understand
4:53
kind of what it is that you know it helps make
4:57
Ed technology kind of supportive of of increasing
5:00
student learning you know I mean oh absolutely
5:04
and and I love ISTE it was kind of a dream of
5:07
mine to go when I was working for my school system
5:10
and I finally got up the guts to ask and my supervisor
5:15
said Of course you can go. Why didn't you ask
5:17
before? My teammate and I were like, oh, I guess
5:21
we should have asked earlier. I don't know why
5:23
we thought it was beyond us. So I've been attending
5:27
ISTE for over 10 years straight, and it's an
5:30
amazing event. I love the energy. I love connecting
5:34
with people. It's obviously... one -stop place
5:37
to be in June for EdTech. However, I will say
5:41
that in the past five years, I've spent a lot
5:44
of time in smaller regional EdTech conferences.
5:49
In my role at I2E and also as a featured speaker,
5:52
I get to go out and attend some of these other
5:54
events. And I just want to add that as a word
5:57
of encouragement for all of the listeners that
6:00
going to a smaller regional event, I think is
6:04
often more impactful. because you have more time
6:08
to really connect with people and network and
6:12
there are fewer sessions to choose from which
6:14
actually makes life a little bit easier. I mean
6:18
ISTE is nuts as far as like at any given moment
6:22
there are 15 things you could be doing and that's
6:26
exciting but I do find high value in the regional
6:30
tech conferences both in connection and time
6:33
and just the availability to really dig deeper.
6:36
with individuals and with the vendors who are
6:39
there. It's just not quite so chaotic. So there's
6:42
a place for both. It's really funny that you
6:45
bring that up because last week David and I both
6:49
hosted regional tech conferences in our different
6:52
parts of Minnesota, where it was just exactly
6:55
like you're describing it, fewer sessions, but
6:57
it's much one. It's much easier for presenters
7:00
because they're dealing with other teachers and
7:03
educators that are like from their area and from
7:07
their schools and deal with similar issues. But
7:10
at the same time, you know, it's it's also like
7:12
you said, the number of choices is just it's
7:15
more manageable and it's easier to find, find
7:18
a track that that you could go down, you know
7:20
what I mean? Yep. Yep. Yeah. So I really enjoy
7:23
those. My regional is, excuse me, called NCCE
7:28
here in the Seattle area. And I've been attending
7:30
that since 2004, I think. And I've presented
7:35
there most of those years. And it's just a great
7:39
event. It's not huge, but you get a lot done,
7:42
so. Yeah. And I know what you're saying about
7:45
ISTE. I think I've literally had FOMO, for another
7:49
session while sitting in a really good session
7:51
at ISTE. So that tends to happen. But yeah, and
7:55
Jeremy and I, like you said, we do a number of
7:58
trainings too. And one of the things that I saw
8:01
you posted recently that you start some of your
8:04
trainings with that I found really interesting.
8:07
was the AI personalities concept that you posted.
8:11
I had seen that on LinkedIn, and I thought it
8:13
was a really great bell ringer, I guess, if you
8:17
want to call it, for those that are attending
8:19
the training to identify. Are you like one of
8:22
these personalities? There's five on there. It's
8:24
like, are you one of these personalities? Or,
8:26
you know, are you a combination of them? And
8:30
it did a nice job of laying out the different
8:32
feelings that people have around AI. Why do you
8:35
think that that was like an important exercise
8:37
to include in your trainings recently? I think
8:42
it's really important that we validate the emotions
8:45
attached to generative AI. And as an AI optimist,
8:50
I don't want to focus on fear or negativity,
8:55
you know, honestly, there's a lot of negative
8:56
around generative AI. And I want to recognize
9:00
the challenges and the issues around privacy
9:03
and safety and bias and the things that are true
9:06
and real. But I also don't want to dwell there.
9:09
And so I've found that giving people an avenue
9:12
to kind of claim their excitement and their concerns
9:16
in a way that was The word I guess is inclusive,
9:22
right? Like, hey, we all have opinions about
9:24
this and they all add value to the conversation
9:26
and it's okay to say like, hey, I really care
9:29
about the human element. There's a flip side
9:33
to all of those AI personalities where we're
9:35
concerned, but I think it frames it in a positive
9:38
light where like this is what you're bringing
9:39
to this conversation. Last week I was working
9:43
with a school in Southern California and one
9:46
of the teachers just kind of had a a moment during
9:49
one of our breakout sessions, and he was kind
9:53
of upset. I mean, I could feel his frustration
9:57
around some of his expectations around AI and
10:00
how those weren't met and how students are misusing
10:02
it at times. And I just stopped the conversation
10:06
and said, thank you for sharing. I think other
10:09
people in the room have similar concerns, and
10:11
it's really important we talk about. These things,
10:14
I said, you sound like a traditionalist and there's
10:17
room for you. You don't have to feel like you're
10:21
not a part of this conversation. And he came
10:23
up to me after the session and thanked me for
10:25
how I handled it because he didn't feel like
10:27
he was an outsider. He was a part of the community
10:30
that we were learning together. So I just think
10:33
the personalities lend themselves to that. type
10:36
of culture in the school system. And of course,
10:41
whenever people feel included and validated,
10:43
that's a good day. Right. And to be honored is
10:47
just going to, you know, to know that their opinion
10:49
matters and that you're honoring their position
10:51
and where they're at currently is just going
10:54
to make it easier, I think, for them in the long
10:56
run to start to maybe come over to that optimist
11:00
side gradually. Right. But, you know, like since
11:05
you brought up the you being an AI optimist here
11:08
and why don't we why don't you spend a minute
11:12
here and tell us you know what was kind of the
11:13
inspiration behind the the writing of AI Optimism?
11:18
Well I have to tell you that the title came last
11:21
and a lot of people are surprised at that yeah
11:24
I didn't like have this vision to write about
11:26
AI Optimism and then write a book. I wrote the
11:29
book and based on what I felt was missing, honestly,
11:36
kind of this hole in the space of education AI,
11:40
where we have some people doing a great job surfacing
11:44
concerns, and we have some people doing a great
11:46
job talking about, you know, tools and what those
11:52
look like in education. And I felt there was
11:56
this gap of not understanding kind of the trajectory
12:01
of AI, if you will. And that's why I use the
12:04
SAMR model in the book because I just, I saw
12:09
so many educators scrambling to quote, use AI.
12:13
And that's the wrong approach. And that has always
12:16
been the wrong approach to me as an instructional
12:18
coach. We shouldn't be focused on what tool am
12:21
I using or how much am I using technology. It's
12:24
really about meeting student needs and developing
12:28
learning experiences that transform and empower
12:31
learning and so I felt like people were just
12:35
you know getting kind of distracted by the the
12:38
sparkly glitter of AI and all these all these
12:42
little apps popping up and I was seeing social
12:44
media posts like I want to use AI with my students
12:47
where should I start and and those are heartbreaking
12:49
to me because it's the wrong lens. It's tool
12:54
centric, it's use centric, you know, instead
12:57
of thinking about what I want to achieve with
12:59
my students. And so as I started having that
13:02
roll around in my head, I felt like the SAMR
13:04
model was a really good tool to help people understand
13:07
maybe where they had started with using AI and
13:10
then what was possible to really redefine. And
13:14
then after I started looking at all these different
13:16
use cases, I started thinking about how they
13:21
fit into these categories. And that's where the
13:23
AI Optimism Framework started. So this whole
13:25
thing kind of was designed in a backwards way,
13:29
right? Like backwards design. Just give them
13:32
a shout out right now. Because I had the ideas
13:35
and then I thought, okay, these fit into six
13:37
categories of how we use, like the things we
13:40
have to do in education every day. We have to
13:42
design and create and support and analyze and
13:45
evaluate and manage. And those are the things
13:46
we're doing. And so how can I help educators
13:49
categorize and say, OK, if I want to evaluate
13:53
student use and student growth in my classroom,
13:57
what are some tools I can use and how much do
14:00
I want to be involved? Is it just a substitution
14:03
level or am I going to really modify what I've
14:06
been doing or redefine something completely unique
14:09
and different? And so the And then I had to write
14:12
the beginning of the book, which is like, OK,
14:14
if you're going to do all those things, I need
14:15
you to understand first what I believe about
14:19
privacy and praxis and prompting. And so really,
14:23
the whole thing came about backwards. And the
14:25
very last thing was the title, which was a conversation
14:29
I had with my editor when he said, OK, but you
14:32
really need a title because we need to work on
14:35
a cover and. We have like the book was already
14:37
through developmental and kind of far along the
14:40
process when I finally landed on what I wanted
14:42
to call it. And that's simply because the middle
14:47
of the book to me was what I wanted to say. And
14:50
then it took me a while to figure out how to
14:52
put that into one succinct to two or three words.
14:57
And the SAMR model has been around for a while.
15:01
And that was developed by Dr. Ruben Pantadora,
15:08
who did the forward for the book as well. And
15:12
so what was that like? I got two questions, I
15:16
guess. What was it like talking with him about?
15:19
this process or, you know, what conversations
15:22
did you have with him being that you were going
15:24
to use the SAMR model? Ultimately, why did you
15:27
feel like that would be the most complementary
15:29
model to what you were trying to convey in the
15:32
book? Well, I actually started field testing
15:37
SAMR within AI a year over a year before the
15:41
book published because I'd had it kind of rumbling
15:44
around in my head. So. Some of the things I did
15:47
is I started presenting with the SAMR model infused
15:52
in my AI workshops. So we just talked about NCC,
15:56
for example, February, 2024. So a year and a
16:00
half ago now, I did a couple of workshops for
16:02
NCC as just a presenter at the conference and
16:05
I infused SAMR into the AI workshop I was leading.
16:11
rave reviews like people are coming up to me
16:13
saying this makes so much sense. I'm so thankful
16:16
I it makes so much sense now to you know think
16:19
about my purpose first and then go back and look
16:21
at how I'm using the tools and just got amazing
16:25
feedback. I did the same thing at idea con last
16:28
year. I was there as a featured speaker and I
16:29
just I just put it in and the feedback I was
16:33
getting was that it was really helpful. Educators
16:35
were like this is beautiful. And then Matt Miller's
16:39
team at Ditch that textbook invited me to write
16:41
a guest blog. And I said, hey, I'd like to do
16:45
Sam or AI. Like let's see how that lands and
16:47
put that out really broadly because he has massive
16:50
reach. And so I did that guest blog and that
16:53
went out and he had people later come up to him
16:55
and say, this could be a book. Like Becky's blog
16:59
makes so much sense. Have you thought about doing
17:01
a book? And Matt and I had had a little side
17:03
conversation and he was like, I think you're
17:05
going to get something for that. because he was
17:09
kind of in the loop on the whole book idea earlier.
17:12
And so I got really good feedback from educators
17:16
that that was the right path to be going down.
17:20
There's a lot of different models we can use
17:22
that all add value. But I think for what I wanted
17:25
to communicate, it really was a perfect fit.
17:29
Yeah, and I think the SAMR model is great just
17:31
in terms of like the depth in which you are integrating
17:34
technology and has been for a long time. But
17:36
what I really like is how you've applied AI to
17:39
that and be particularly because it's new. And
17:42
a lot of people are kind of like, how do we use
17:44
this? And to what degree? And, you know, with
17:47
our students, what does that look like? So and
17:49
I'm sorry, you had a two part question and I
17:52
can't let the conversation move on without giving
17:54
credit to Dr. Quintedora. He He was so great.
18:00
So I had a mutual colleague connect us and he
18:06
didn't reply right away and that's totally okay.
18:08
Like I don't reply instantly to everything that
18:11
hits my inbox either. Sometimes we're drowning.
18:14
So I gave it a little time and then I sent him
18:16
another little ping and said, hey, you know,
18:18
just checking in, is this something you'd be
18:20
interested in? And he got right back to me. He
18:23
was like, yes, I thank you for the second email,
18:26
which is a great reminder that like it's OK to
18:28
send a second email sometimes. He's like, thank
18:31
you so much for getting in touch. Yes, I would
18:33
love to be a part of this. You know, let me know
18:35
what you need. And so I just knowing that life
18:39
is busy for everyone and I'm sure he has a lot
18:42
on his plate. I didn't I didn't really want to
18:44
hold my breath on like it would have been OK
18:47
with me if it came to the time he was ready to
18:49
write a forward and he was. swamped with something
18:51
else in life, like that would have been OK. My
18:54
publisher was super excited. They're like, this
18:56
is amazing. He agreed to write it. And then he
18:59
did. We just were in communication from then
19:02
on. As far as the timing, I would give him updates
19:05
on, OK, I think I'm going to have a really solid
19:08
manuscript to you by this time and so on. And
19:10
he was so sweet. He wrote me back after he read
19:14
it the first time and was like, he said, I have
19:16
seven pages of notes. it's can I have a little
19:19
extra time to write the forward like he was so
19:22
complimentary and positive about how I reflected
19:25
his work in an AI sense and that was just really
19:29
exciting for me as well to have him be a part
19:32
of the project and lend his voice to something
19:34
that I think really makes sense. And definitely
19:39
it does and I I'll admit here that I didn't really,
19:44
when I first started even presenting on AI, really
19:47
consider the SAMR model as something that should
19:49
be integrated through the training of teachers
19:54
in the different levels and where they're at
19:56
and how it can even change depending on the type
19:59
of activity and stuff you have. But I remember
20:01
presenting on the SAMR model when I was still
20:03
in the classroom and doing some of the the part
20:06
-time technology integration stuff. When we were
20:10
first going one -to -one and trying to get students
20:13
and teachers to use their devices more purposefully
20:16
and not just as that substitution tool to, hey,
20:21
don't just write your notes on your notebook
20:23
paper anymore. Let's type them on our iPad or
20:27
like some of those things. I think one of the
20:30
things that I really appreciated too from the
20:32
book is how you make that ease of entry into
20:37
AI use through the SAMR model, I think much more
20:41
accessible to teachers. But I'm also kind of
20:44
wondering, do you think there's like, what sort
20:46
of barriers do you think teachers face when it
20:49
comes to adopting AI into their practice, even
20:52
at like some of those basic substitution levels?
20:56
Oh, so many, so many challenges. I think if I
21:01
could categorize them into one, Topic it would
21:04
be lack of trust and that has several different
21:08
pieces to it. One is not trusting the tool to
21:12
keep their data safe or their students safe.
21:15
Another one is not trusting that it's not going
21:18
to be used to train the model or use their data
21:23
without their permission long term. So short
21:25
term and long term. Then there's the lack of
21:28
trust that it will return great results that
21:30
are accurate and reliable. I hear a lot from
21:34
educators how they've tried using a tool, you
21:38
know, whatever it is, it doesn't pick an LLM.
21:41
And they've been disappointed with the results.
21:43
And so they've chosen to walk away from all of
21:47
AI because of that experience. And we know that
21:50
in education specifically, having been a trainer
21:53
in education for 20 years, if educators don't
21:57
see valid, reliable, trustworthy, responsible
22:00
use on the first They're not coming back to that
22:04
tool. Like that's pretty universal. And so it's
22:07
been really difficult because generative AI is
22:10
still new in the realm of the universe, right?
22:14
We're two and a half, three years in. And the
22:18
tools are changing at lightning speed. And so
22:22
when you've seen something work or not work last
22:25
week, it might be completely different this week.
22:28
And I've been talking to educators about I'm
22:31
just calling it AI grit and reminding them that
22:35
25 years ago they had to learn how to search
22:37
Google, right? We had to learn how to type in
22:40
queries and like all the hits were wrong and
22:43
it didn't work the first few times. And we've
22:45
been teaching students that ever since in digital
22:47
literacy lessons. And for some reason, I think
22:51
we've gotten really comfortable with Alexa and
22:54
Siri and the tools that we've had for a long
22:57
time. And so this tool, when it didn't respond
23:01
the way we expected it to, to throw it out, feels
23:05
a little bit easy, you know, like easy quitter.
23:10
And we don't, you know, not everyone has time
23:12
to learn and learn prompting. And I've heard
23:14
that argument, but I've also just kind of come
23:16
back and said, hey, you spend the time to learn
23:20
what's important to you. When you got an air
23:22
fryer, You know you you got out all these recipes
23:25
and you figured out how to use it and you made
23:27
sure your Trader Joe's orange chicken turned
23:28
out amazing like you put the time in so If you
23:33
want to use these tools like you're gonna have
23:35
to put some time in and that's why my third core
23:37
principle is prompting because the reality is
23:40
it doesn't always do what we think it's going
23:42
to do and Throwing it out because it didn't return
23:45
the experience. We hope is the wrong approach
23:49
And it was definitely in the classroom teaching
23:53
kids how to do Google searches. I'm that old,
23:55
too. Yeah. Oh, same. I remember even like I and
23:59
I when I would first kind of talk to teachers
24:01
to try to remind them or to maybe build their
24:05
trust or to tell them that they feel supported,
24:07
that if we want to go way back, you know, and
24:10
talk about even like when I was in school, the
24:13
librarian in the school library, because we didn't
24:16
even call the media centers then was showing
24:18
us how to use the card catalog. and how we would
24:21
find journal and periodical articles in the Reader's
24:25
Guide to Periodical Literature. And those have
24:28
their own strategies that you need to employ
24:31
to get the best results possible for the information
24:34
that you're looking for. And I really do feel
24:36
like with prompting and to get the best results
24:40
that you're looking for, there are skill sets
24:42
there that you need to learn also. And so I'm
24:45
right with you with what you're talking about
24:48
when it comes to that. Appreciate that. Yeah,
24:51
and I like I like the AI grit term too because
24:54
like I feel like as much as I've used AI and
24:58
these different programs like I feel like failure
25:01
is Almost essential to the whole process because
25:04
it's like I can't you know, every single time
25:07
I go to it I basically think like it's gonna
25:09
take several iterations of me talking to this
25:12
and Explaining what it is. I want to actually
25:14
get what I have in my mind, you know what it
25:17
is that I want and And just to expand on that
25:21
briefly, different AI tools respond different
25:24
ways. And I've heard lots of people say, you
25:27
know, you can think of them as different friends.
25:29
We don't want to overhumanize AI, but it is a
25:32
nice lens to be thinking about. You know, I have
25:34
one friend I might ask for for advice on making
25:38
a meal for my family because she cooks really
25:41
well. I might have another friend that I ask
25:43
for advice on a fitness plan or another friend
25:46
that I'm like, hey, what's a good vacation spot?
25:48
So you know looking at not just using one a lot
25:52
of us are we have this core desire to find like
25:55
the one ring to rule them all and we think there's
25:57
one tool is going to do everything and with generative
25:59
AI I just I don't see it I see different tools
26:03
fitting different needs and that takes time to
26:05
learn as well so it can feel so overwhelming
26:08
for educators who are already tapped out in you
26:11
know the amount of things they're supposed to
26:13
do and manage and learn so that was another hope
26:16
of my book is that they could use it as a guidebook
26:19
and just flip right to a strategy they were interested
26:22
in, see a couple of tool recommendations or prompt
26:25
recommendations and go from there. Instead of
26:27
having to feel like they're starting with this
26:29
empty slate, at least they're getting something
26:31
to start with. Yeah, and it's been interesting
26:35
too, like we've had a lot of conversations with
26:37
our like our special education department on
26:40
the various things that they can do to save time
26:43
and work with students on. And one of the things
26:47
that you bring up in the book is personalized
26:49
learning. And so you look at that as one of the
26:54
boons of AI, right? The personalized learning,
26:58
how can that be something that can really be
27:02
utilized by any teacher, in your opinion? Well,
27:07
I so I think that term is being overused a little
27:11
bit at the moment Like everybody's like and then
27:14
AI can personalize learning and without a an
27:17
actual application of what that looks like I
27:20
think it can be really difficult to understand
27:22
what that even means I've seen it work a variety
27:27
of different ways. So even at a substitution
27:29
level, I just Talked about this with some math
27:32
teachers last week. I said hey give give like
27:37
a chat GBT or a copilot or whatever your kids
27:40
have or a Gemini, give it what you're trying
27:43
to learn in math and then their interests. And
27:47
it created these story problems that were maybe
27:49
more relevant for the students. Now, that's just
27:51
a strict substitution level. And one of the teachers
27:54
said to me, I already have a book full of story
27:57
problems. Like, why would I do this? And I said,
28:00
hey, I think you missed my point. And that's
28:03
good feedback for me as the presenter. Like,
28:04
I think you maybe missed something. And I said
28:07
the whole point here is that students, which
28:10
is what I said, and he like missed it, but is
28:13
that students would be writing the story problems
28:17
based on them being interviewed by the AI and
28:20
coming up with story problems that they're actually
28:22
interested in. And we were talking about graphing
28:24
linear inequalities. And so with graphing linear
28:29
inequality, two of them specifically, now there's
28:32
a decision point. And so they were actually writing
28:36
Story problems that caused them to question their
28:39
own value system and think about what they would
28:42
choose based on a scenario. Like there was so
28:44
much more here than a story problem. And it was
28:46
something that resonated with those individual
28:48
students. So in my sample lesson, the students
28:53
were then telling a story around, okay, if I
28:56
have two choices in how to spend my afternoon,
28:58
either doing homework or. a hobby with friends
29:02
but I only have a finite amount of time and I
29:04
have to get my homework done like where's that
29:06
intersection that's a that's two linear inequalities
29:09
so I'm like isn't that more interesting for your
29:14
students to like talk through and think about
29:16
when it actually connects to their own life than
29:19
a story problem out of a book and so with all
29:24
that in mind I think personalized learning has
29:26
this huge range right it can be as simple as
29:28
just literally personalizing the content that
29:31
you're trying to generate, but it can be as powerful
29:33
as a full personalized learning pathway where
29:36
students are being taken along a journey of growth
29:39
from their own starting point through meeting
29:45
standard at a certain level. And there's specialized
29:47
apps that do that that are really good. That's
29:50
what they're designed for. So personalized learning
29:52
has kind of gotten this overkill at the moment.
29:55
But the reality is that there are a wide variety
29:58
of tools and options that fit in that umbrella
30:01
modality. And AI can definitely, I think, or
30:05
generative AI, assist you in finding ways to
30:09
or to help a teacher connect the content to the
30:12
interests of their students. In fact, we just
30:15
had an interview the other day with an instructional
30:18
technology specialist from out east in Massachusetts.
30:22
Her name is Liz Boese. And she has a project
30:26
that she works with elementary age teachers in
30:30
her buildings that she helps to support called
30:33
the Perfect Book Project. And really, it is a
30:40
reading lesson, a reading project, and not a
30:43
writing or an AI project, right? But it's to
30:46
have students work with AI. to help create books
30:51
and stories with plots and characters and subjects
30:55
that they're interested in. And then they help
30:59
to design outlines for these chapters of the
31:03
book. The students work to add their own ideas
31:07
in with ones suggested by AI. And in the end,
31:11
they produce a manuscript that they turn around
31:15
and they go to like staples and they have it
31:17
published. So they get a physical copy. of the
31:19
book as well. Um, of those books in their o So
31:24
now you're building a of these different things
31:28
personalized, right? And ma they're using those
31:33
to hel better readers, um, which i how you're,
31:39
you're able concepts and put that all same thing
31:43
could apply li like you're talking about. So
31:51
Becky, one of the things that you brought up
31:53
was prompting. And I know when I do trainings
31:55
and stuff, I have a prompting strategy that I
31:58
like to encourage people to follow. And I know
32:02
there's several acronyms, because we're teachers
32:05
and educators. We have an acronym for everything.
32:08
But I'm kind of curious, is there a frame that
32:11
you kind of think of when you are talking with
32:15
educators about how to become better prompters?
32:18
Or is it maybe even, like I'll give you maybe
32:21
even part of my answer to this too. I'm starting
32:24
to become a little more of a convert to the conversational
32:27
method of just having that dialogue developed
32:30
too. But I don't know if that's the best either.
32:33
But is there kind of a format that you tend to
32:36
follow? There is. I do a whole session called
32:40
The Art of Prompting sometimes when invited.
32:44
And we dig into just a myriad of ways you can
32:47
prompt powerfully. In the book, I use Saba Kinway's
32:51
model Spark because she's just a wonderful advocate
32:56
for using AI. And she kindly said, yes, I could
33:00
include that in the book as one of the models.
33:01
And I wanted to promote her as a woman in AI.
33:04
But I do a lot of co -pilot training in my day
33:09
job. There is a model that Microsoft uses that's
33:14
just goal context expectations and. Oh my gosh,
33:21
why? Why am I playing on the fourth thing? This
33:23
is weird. I do it all the time. Source there
33:27
we go. Thank you. Goal source expectations in
33:30
context and so that kind of just. Is the simplest
33:34
to me and it doesn't require anything like fancy
33:38
or interesting. It's not an acronym. which is
33:41
probably why I blanked on source, but it really
33:44
makes sense to me. Like most of us prompt AI
33:47
with a goal initially, but adding context and
33:51
expectations is something that we often don't
33:54
and AI will over deliver at a frustration level
33:58
because it doesn't know what you want, right?
34:00
If you're prompting vaguely, it's like just throwing
34:02
things back at you. And so I do tend to just
34:05
kind of stop And I close my eyes a lot of times,
34:10
honestly, and I think, okay, if I was telling
34:12
a human this direction, like someone who worked
34:16
for me, let's pretend I have a real assistant,
34:18
a human assistant, what would I say? And often
34:23
I find it's a whole lot more details than I was
34:25
gonna tell AI, and that feels backwards. Like
34:28
why would I tell AI less? They don't know less,
34:31
right? And so I feel like I have to, at some
34:35
point, remember that AI is not, it doesn't know
34:41
everything I think it's going to know. It can't
34:43
read my mind. I have to give it more details
34:45
than I think. And when I do that, I get much
34:49
better results. And so one tip I'll give that
34:52
I really appreciate as well that my colleague
34:55
Nathan Wells says a lot is that he'll at the
34:58
end of his prompt say, and is there ask me any
35:00
questions you need, you know, to know about this
35:04
before you answer me or Is there anything I should
35:07
have added to this prompt? And the AI will tell
35:09
you. Claude is actually my favorite model for
35:14
that method because I think it does a great job
35:16
of asking questions and prompting the user on
35:20
what else it needs to do a good job answering
35:22
your question. It will do that even without you
35:25
prompting it. It'll just be like, actually, before
35:27
I answer you, can you help me with these questions?
35:30
And I love that because it does feel more like
35:32
a thought partner that way. So I think the biggest
35:36
advice I can give on prompting is to offer more,
35:40
not less. And you know, I think along those lines,
35:43
I think teachers kind of naturally are good explainers
35:47
in providing background and context to things
35:49
because we do that every day when we introduce
35:52
an assignment or an activity or a lesson, we
35:54
kind of have to fit it within the frame of connecting
35:57
it to the students prior knowledge and let them
35:59
know what we're doing so that we know where we're
36:02
going. And I find keeping that in my head, too,
36:06
as I'm starting to prompt really helps get better
36:09
results. But I really am going to take your suggestion
36:12
of asking the question, what else am I missing
36:14
that should be included? Because I often overlook
36:17
that. And because it's a thought partner and
36:21
highly creative, AI will come up with things
36:24
that we wouldn't have. And so when I'm designing
36:29
new workshops or new strategies, I'll sometimes
36:32
take, and I'm not exaggerating here, an hour.
36:36
of dialoguing with AI like I'm thinking it's
36:40
asking me questions and I'm like oh okay yeah
36:42
how would I approach that what do I believe about
36:44
that um and we're having this conversation back
36:48
and forth and at the end of that hour I have
36:52
really refined my ideas it's incredibly helpful.
36:57
Have you have you had any conversations with
36:59
teachers about like um you know I I think with
37:02
our roles I know that Jeremy and I particularly
37:05
spend like compared to teachers or we spend a
37:07
bunch of time in front of our computer with these
37:10
systems so like like chat gbt or maybe if it's
37:13
like magic school and you know we we have time
37:16
to kind of really dig in and do tutorials on
37:18
them and things like that have you had any conversations
37:20
with teachers that kind of streamline their process
37:24
a little bit because as we know like when they're
37:26
getting back into like back to school week it's
37:29
crazy They've got all sorts of trainings that
37:33
are required of them. And so to sink as much
37:36
time as we do into AI, it's really impossible
37:40
for them. Have you had any good conversations
37:42
with teachers about like, how do they start planning
37:45
and organizing for their classes and provided
37:50
the time constraints that they have? Yeah, the
37:54
one advice that I often give when they say like,
37:57
I just don't feel like I have time to learn it.
37:59
As I say, don't worry about learning it as a
38:04
goal. We would say the same thing about Minecraft
38:08
or Canva. I'm not just going to sit down and
38:10
do a bunch of tutorials. I always say, find a
38:14
purpose. When you have a problem to solve is
38:18
my big recommendation. Let's say that I had this
38:24
conversation with educators just late last week
38:27
where I said, What's something that you wish
38:31
was different about your classroom or like you're
38:34
trying to come up with a poster or a better way
38:36
to explain something, like what's a problem that
38:38
you want solved? And then can you ask AI for
38:41
help with that? And I tell them, be very specific,
38:45
like give lots of details on what you want. You
38:48
know, do that good prompting, but they don't
38:50
have to like memorize some prompt framework to
38:52
just give it lots of, you know, lots of description.
38:56
And so giving those use cases as far as having
39:00
a problem to solve, I think is really helpful.
39:02
When I see teachers get frustrated, it's when
39:04
they feel like they're supposed to just go use
39:07
it and they don't even know what for. You know,
39:10
they're like, OK, well, I don't need it to write
39:12
an email for me. Like I'm good. Thanks. I can
39:15
answer my email. So what am I using it for? And
39:19
focusing on that, what is a problem you want
39:21
help solving is really helpful. Yeah, now let's
39:28
kind of get into the book here a little bit because
39:30
I feel we've had great conversations, but I also
39:32
want to bring up some of the great points in
39:37
the content that's in the book. You know of course
39:39
you've mentioned and discussed how it's based
39:42
on the SAMR model or you apply that model to
39:45
generative AI use within education, but you know
39:49
what are some of your kind of things that stick
39:51
out in your mind about you know picking activity
39:54
if you wish or a topic and maybe kind of give
39:58
us an idea of how a teacher might approach it
40:00
from the different levels within that SAMR model.
40:04
Oh, goodness. Is that too vague? I can be more
40:07
specific if you want. Yeah, that's super. There's
40:11
like a hundred -year -old strategy in the book.
40:13
A teacher would like, you know, a special ed
40:16
teacher is, you know, needing some assistance,
40:19
maybe scaffolding, you know, some directions
40:23
within their lessons and stuff like that. Like,
40:26
how could they do that using AI on some of those
40:29
different levels? Well, OK, that's a little narrower.
40:34
So what's interesting about what you just said
40:36
is that we start at the substitution level without
40:40
even using generative AI. And this is a good
40:44
time to call that out, is that AI is this big
40:48
umbrella. And I'm optimistic about all of it.
40:52
Generative AI is a subset. of artificial intelligence
40:55
that we've been laser focused on for a long time
40:57
now because of all the controversy and all the
41:00
possibilities. So for a special education teacher
41:04
that's looking to scaffold or support would be
41:06
the category of the AI optimism framework. We
41:10
start with just substitution, like can we be
41:13
giving audio directions? Can we be giving graphic
41:16
directions? Can we be, do we need to allow for
41:19
translated directions? for an assignment or for
41:22
a task, and all of that is done with what we
41:25
would call more traditional AI resources, right?
41:30
Dictation, translation, read aloud, that's all
41:34
just kind of normal, normal AI that's been around
41:36
for a really long time. But then we could get
41:39
into visualizing those directions. We now have
41:42
tools like Notebook LM and Co -Pilot Notebooks
41:45
that will generate a word online. generate that
41:49
audio overview kind of podcast style of a set
41:52
of directions or content information. Google
41:57
LM just notebook LM for Google just announced
41:59
video as a possibility. And so giving kids that
42:03
multimodal input for looking to just kind of
42:06
raise some understanding and going to a more
42:10
transformative level. is where we can have students
42:13
actually engaging. So speaking of personalized
42:16
learning, scaffolding the directions to an assignment
42:19
in a way where they also are asking questions
42:22
and not just absorbing information. And I think
42:25
that's a big shift for me in the framework. As
42:29
far as the S and A levels of SAMR being very
42:32
focused on pushing kids content, right? Giving
42:35
them things where the M and R levels are more
42:38
about kids. being engaged in a part of the generative
42:43
AI, a part of the content. So maybe we have students
42:46
copy and paste the directions for a task into
42:51
napkin .ai and that creates a whole little infographic
42:55
for them on the steps involved that's very visual.
43:00
And because they initiated that, they weren't
43:02
just handed a graphic. They're then able to go
43:05
in and interact and and even evaluate that graphic
43:09
and be like, is this correct? Are these actually
43:12
the steps? And then they can add to that graphic.
43:14
They could add in, OK, this is what I'm going
43:16
to do at this step in the process, which takes
43:19
us into supporting executive function and self
43:21
-regulation and all these great skills. At another
43:24
level, we could be setting up a school AI sidekick
43:28
space that gives students an opportunity to ask
43:31
the chat bot questions. And that's a really safe
43:36
place for many kids. I know the word safe can
43:38
be weird, but the reality is it's low risk, maybe
43:44
is a better term. It's low risk to ask a chat
43:47
bot questions about how to complete an assignment.
43:50
Way lower risk than asking a teacher and for
43:53
sure lower risk than asking a classmate and feeling
43:56
like you're stupid. which is a whole other set
43:59
of SEL we could talk about. So giving kids access
44:02
to those tools as supports that they get to then
44:06
choose when they access and when they use them
44:09
based on their own needs is incredibly exciting.
44:12
And it gives them that AI literacy of making
44:15
a decision of when to use AI appropriately. So
44:18
we're doing way more. than just scaffolding an
44:21
assignment now, we're also empowering them to
44:23
say, when does AI limit my learning? And when
44:25
does it amplify my thinking? And that's, I think,
44:30
where things get really exciting. And could we
44:33
dive into that a little bit too? There's a ton
44:37
of stuff with regard to students where a lot
44:41
of conversation's happening around there are
44:44
massive benefits to it. There's also some concerns
44:48
about the cognitive offloading side where we're
44:54
having the machines basically perform the task.
44:59
And then there's conversations around what are
45:01
the best ways to go about using these. And I
45:03
always say, too, in our trainings, it's not just
45:06
students. It's us as professionals, too. There's
45:08
that danger of offloading your own thinking into
45:12
these so that you're not really going through
45:15
the creative process or the struggle and things
45:17
like that. Have you had conversations around
45:22
that piece of cognitive offloading that is a
45:26
danger? And then also with that, do you have
45:31
any concerns about the social aspect too? A lot
45:36
of articles lately have been saying things about
45:38
students having relationships with chat bots
45:41
and you know the fear of that replacing the social
45:46
aspect or like emotional growth. So basically
45:49
the kind of the two things I'm wondering your
45:51
thoughts on are the dangers of the intelligence
45:55
piece, the cognitive offloading, and then the
45:57
emotional piece of, I guess you could say social
46:00
offloading, like you're not engaging with real
46:04
humans, you know, face to face, or even like
46:08
in this version where we're talking virtually,
46:12
obviously. Do you have any concerns there with
46:13
those two things? A hundred percent. Any, and
46:18
I have lots of TikTok walks where I talk about
46:21
outsourcing. and outsourcing humanity and outsourcing
46:25
our creativity, like those are things that are
46:28
kind of scary. Now, I will say the MIT study
46:31
with chat GPT, I feel like that was headlined
46:34
to be clickbait. It was a really unfortunate
46:39
framing of that entire study, in my opinion,
46:42
because they didn't actually, they didn't study
46:47
people who used AI. appropriately. Like the study
46:52
shows yeah kids generated essays with AI and
46:57
then copied and pasted it into a document and
47:00
hit submit and didn't know what they said. Like
47:02
oh yeah okay like who is funding these studies?
47:05
Sometimes I want to be like I would like a study
47:07
with a no -brainer ending. So yeah obviously
47:12
that didn't require any cognitive lift. That's
47:16
not how we should be using AI. So It's really
47:20
interesting as we think about my central question,
47:24
does this use of AI limit my learning? That I
47:27
think does help students evaluate, what am I
47:30
supposed to actually be doing here? And this
47:33
gets us into like a whole trajectory rabbit trail
47:36
conversation on what are we actually assessing?
47:40
Because I'll give you a really simplistic example.
47:44
But when my son was a first grader, He came home
47:47
and told me about a timed math assessment where
47:50
he had to write numbers into a grid, write like
47:54
a hundred grid. And then there was a backside
47:57
and he was telling me how he didn't get to finish
47:59
because the timer went off. But he told me, he's
48:03
like, but if I would have finished, I would have
48:05
gotten to 245. And I was like, well, how do you
48:08
know that you didn't finish? And he goes, oh,
48:10
because The timer went off, we had to put our
48:12
pencils down, but then I just counted the rows
48:14
remaining, and I know I would have gotten to
48:15
245. So as a first grader, he knew the math.
48:21
And he goes, but I didn't get to show that I
48:23
knew that because the time ran out. And I said,
48:25
oh. So it wasn't assessing your ability to count.
48:29
It was assessing your ability to write quickly.
48:34
That's what she's scoring. She's scoring your
48:36
ability to write quickly. And not to be critical
48:40
of his first grade teacher who was wonderful,
48:42
but I think a lot of times in education we forget
48:44
what we're actually assessing. And sometimes
48:47
we give these assignments and we want to assess.
48:50
Like the student writing and we forget that we're
48:54
also assessing. along the way, a lot of other
48:59
things that we maybe don't need to be. Like maybe
49:01
we're assessing student creativity or we're assessing
49:04
their spelling or we're assessing their ability
49:06
to stay on task for significant periods of time,
49:09
or we're assessing their ability to take a very
49:11
large project and break it into chunks and then
49:13
self -manage all the normal steps that go into
49:15
a project. Like that's what we're actually assessing.
49:18
And so removing some of those barriers for kids.
49:22
does allow us to get to what we actually want
49:25
to assess, which is can they produce high quality
49:30
work in many cases? And if we don't want them
49:34
to have any supports to produce high quality
49:36
work, then that is now part of the assessment,
49:39
is can you do this 100 % independently? Or did
49:41
you get help? And who did you get help from?
49:44
Or what did you get help from? Because we've
49:47
had these support mechanisms in place for a long
49:49
time. The difference with generative AI that
49:52
I hear about a lot is that it over helps, right?
49:56
It goes beyond the line that a human would or
49:59
a tutor or someone at home or some other tools
50:03
that we've had in the past. And so I believe
50:06
we have to allow students to help to understand
50:09
that line because I don't think we're helping
50:12
them for their future. If all we do is tell them
50:15
like, no, you're not allowed to use AI on this.
50:17
or yes you are, and we don't help them see why.
50:22
Like, what am I assessing and what am I trying
50:23
to do? Does that make sense? Yeah, absolutely.
50:27
It's just like a much bigger conversation than
50:29
like... Yeah, and I'm sure all the listeners
50:31
aren't going to see me keep nodding my head on
50:34
the screen, but like through all of that part
50:37
that you just talked about, that's exactly what
50:39
I was doing. I know we're coming up here to the
50:42
end of our time that we have for today, but so
50:45
can we just kind of go back to, you know, obviously
50:48
I know that the title didn't come till the end,
50:51
but still, you know, why should educators be
50:55
AI optimistic? Because most educators, I'm just
51:01
going to be a blanket statement here. Most educators
51:04
went into education because they believe they
51:06
could make a difference. And they love that feeling
51:09
of like imparting knowledge or helping others,
51:12
helping students. If we look at personality assessments
51:15
of educators, because we're big on that as a
51:18
company at I2E. And anyway, they're like hands
51:22
down relators. They're connectors. They're people
51:26
who want to invest in other humans. And so I
51:30
think our only option is to be excited about
51:35
the possibilities of where we can let this go
51:38
to support our learners. And ultimately AI Optimism
51:43
is about acknowledging the challenges, but focusing
51:48
on the potential. And that's an important caveat
51:52
where we're gonna nod our heads like you were.
51:56
Like yes, there are some issues here, but I'm
51:58
gonna choose to look at what's possible for the
52:02
betterment of my students. and what this means
52:05
to support them in a world that will inevitably
52:07
require them to use AI. And that's what AI Optimism
52:11
is all about. Great. Well, we really appreciate
52:15
you joining us for the podcast. And we would
52:20
really encourage everybody to remain AI optimistic.
52:24
Please go get the book from Amazon or wherever
52:28
you can find the book. It's a good read. You'll
52:32
learn a lot of great things from it. And check
52:34
our show notes. We'll have links to the book
52:37
in the show notes. Becky, how can people get
52:41
a hold of you if they'd like to? The socials,
52:44
are there any upcoming events and things that
52:46
you want to plug to let folks know? Well, a great
52:50
way to get in touch with me is my website, beckykean
52:53
.com. I'm the only Becky Keene, like not in the
52:57
world, but if you just like Google Becky Keene,
53:00
I will come up or ask Chad GPT, how do I get
53:03
in touch with Becky? I will come up. So that's
53:06
easy. And that links to all my social sites.
53:09
People can follow me there. It's also links to
53:11
both books. And there's book resources as well,
53:14
like a book study guide for people who are looking
53:16
to do it as a book club group. I do virtual author
53:19
talks for those groups. So lots of resources
53:21
there. Becky Keene dot com. Um, and then coming
53:25
up, let's see. I mean, I don't know when this
53:27
is going out, but I, I, there's a couple of,
53:32
uh, conferences that I'll be at as featured speaker
53:35
in the coming school year. And I'm not able to
53:39
announce that quite yet. Um, because I need to
53:43
wait for them to announce me first. Um, but I'll
53:46
be around, uh, yeah, especially during conference
53:51
season. So that'll be, that's exciting. Okay,
53:54
well, if you can, we'll for sure. Keep an eye
53:55
on my socials. Yeah, check the socials. We'll,
53:58
for sure, if we can, we'll put them in the show
54:00
notes if we, if we're allowed to, but. Perfect,
54:03
yeah. If not, after the fact. Once it's out,
54:05
I can share more, yeah, but it'll be fun to see
54:08
educators around the U .S. in the coming year.
54:11
Good. Well, thanks again. We appreciate it. Thanks
54:14
so much for having me. Yeah, I appreciate it.
54:17
All right, bye guys. Have a great day. Thanks
54:20
for listening to Integrate This. Want to take
54:23
your learning further? Head over to our YouTube
54:25
channel for quick tutorials, deep dives, and
54:29
more ed tech tips you can use right away. Don't
54:32
forget to subscribe and catch new podcast episodes
54:35
every Thursday on Apple, Spotify, or wherever
54:39
you get your podcasts. Until next time, keep
54:42
exploring, keep innovating, and keep integrating
54:45
this.