Latest / Integrate This! Ed Tech Podcast / AI Built for Teachers, By a Teacher | Thomas Thompson of Eduaide.ai

Transcript

AI Built for Teachers, By a Teacher | Thomas Thompson of Eduaide.ai · 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
0:40
this at tech podcast David and I are really pleased
0:44
to be joined this Today, it's this morning for
0:47
us. That's what I was about to say, but you could
0:49
be listening to this anytime who knows But our
0:52
special guest here today is Thomas Thompson Thomas
0:56
is a former school teacher, but also turned Ed
1:01
tech entrepreneur. He is the CEO and co -founder
1:06
of EduAid AI. And we're really excited to have
1:10
you with us here this morning. Thanks so much.
1:12
I'm excited to be here. Yeah, Thomas, why don't
1:14
you kind of kick things off for us and kind of,
1:18
you know, give our listeners here, you know,
1:21
a little bit of an overview, kind of who you
1:23
are. How did you get into education and then,
1:26
you know, turn that and roll that into, you know,
1:30
founding EduAid? Yeah, certainly. So education
1:34
wasn't The first path I had in mind throughout
1:37
high school, I was quite intent on going for
1:40
pre -law and then going to law school and then
1:42
my 11th grade world history teacher, Mr. Jason
1:45
Clare, convinced me that education in addition
1:49
to being a public good and a foundational human
1:51
right is also one of the more meaningful careers
1:54
one can find themselves in. He also said if you
1:56
ever plan on having kids, you'll be working for
1:58
80 hours as a lawyer You're never gonna see them.
2:01
You might be able to take a nice vacation every
2:03
once in a while But like you're gonna be absent
2:05
for most of their lives and I was like, okay
2:06
This this all is starting to make some sense
2:08
to me. Maybe I'll think about getting into education
2:11
Then I got to for a student project. I got to
2:15
teach a class. I was like, oh this is Pretty
2:17
cool. And then my senior year, the school created
2:20
a whole new elective for me, which was being
2:23
able to shadow Mr. Claire for an entire school
2:25
year, student teach, make his copies, just learn
2:28
at the foot of the master, so to speak. And that
2:31
apprenticeship model was really, I think, what
2:33
gave me the dedication and focus to get into
2:37
it and to do a good job of it and to look at
2:39
it as a craft. And as a craft, there's a bit
2:42
of an art to it. There's a bit of a science to
2:43
it and really understanding what both those parts
2:46
look like. Went to a small state school by where
2:50
I lived studied post -secondary education social
2:53
studies and Started my career as a history teacher.
2:56
I moved out of Western PA. I wanted to go start
2:59
my Start off like on my own strike it up somewhere
3:02
else where I didn't really know anyone That lent
3:04
that put me in the eastern shore of Maryland,
3:07
which is pretty rural area There were cornfields
3:12
literally on all sides of the school building
3:14
where I was teaching like every four corners,
3:17
just cornfields around it, no houses, nothing.
3:20
And I taught there for three years where I met
3:23
the co -founder of EduAid, Thomas Hummel, who
3:26
interestingly enough grew up 15 minutes up the
3:28
street from where I did in Western PA, but we
3:30
never knew each other until we were both teaching
3:33
across the hall. After those three years, I moved
3:36
to another school district closer to home. I
3:39
started my graduate studies at Johns Hopkins,
3:41
which is in Baltimore. I lived in Annapolis,
3:43
so I was driving across the Bay Bridge. It's
3:45
an hour drive to school, an hour drive back,
3:48
then up to school. And it was just a lot. So
3:50
I was like, let me teach closer to home. So I
3:52
started teaching in South Baltimore, did that
3:54
for two years while I finished up my grad work.
3:56
And that's right when we launched EduAid. So
3:59
that's the preface and the life story up to that
4:02
point. Yeah. I think it's really great. David
4:07
and I love first off, bringing social studies
4:09
teachers onto the show, you know, and I think
4:12
once you're a social studies teacher, obviously,
4:14
I think that sticks with you like all time, even
4:18
when you're not in the classroom anymore, because
4:20
it certainly is, you know, a way of looking at
4:22
things. But I also think it's really cool when,
4:25
you know, people can give a shout out to mentors
4:29
that they've had along the way and people that
4:33
have definitely influenced their career paths
4:35
and their life's decisions. And so a shout out
4:39
to your mentor that did that for you. That's
4:42
pretty awesome. And he's on the board of the
4:44
company. So I get to talk to him all the time.
4:47
Yes. to keep that connection going. And I know,
4:51
like, for me, it was Jay Burroughs. And he was
4:54
just, I loved his history classes, just soaked
4:57
up everything that he would bring into the classroom
5:00
and the discussions that we had. And it was just
5:03
really, really cool. You know, I know he's since
5:06
retired, and I've kind of lost track of him since
5:09
then. But, you know, I doubt he's listening,
5:12
but I'd wish him well, too. So EduAid has really
5:18
grown, I think, since the time that you and Thomas
5:22
Hummel got together and kind of fired this thing
5:25
up, that it now is being used by hundreds of
5:29
thousands of educators, even in dozens of different
5:32
countries. What do you think, as far as helping
5:36
to build and grow the company, have been kind
5:39
of your biggest successes and challenges along
5:42
the way? Biggest success is was word of mouth
5:46
right that teachers sharing the tool with other
5:49
teachers I could go out and dump, you know thousands
5:52
of dollars into some robust marketing campaign
5:55
and Target the airports where there's going to
5:58
be conferences and we put up our signs I've seen
6:01
a few of the companies doing that But what's
6:03
really going to convince a teacher at the end
6:05
of the day is another teacher saying hey I've
6:07
been using this it really helped me do my job
6:09
X Y & Z. I think you should give it a try too
6:12
I'm going to take that recommendation much more
6:15
closely than I am, you know some tick -tock I
6:19
saw or something on Instagram reels or some YouTube
6:22
ad or Google ad or whatever We do all of those
6:26
things. Of course, it's the nature of the game.
6:28
But word of mouth is the key lever So to us it's
6:32
build something that is actually speaking teachers
6:35
something that a teacher can actually wrap their
6:37
head around it Folds into their daily practice
6:40
quite nicely, right? We're not trying to get
6:42
them to learn how to do a new system We're not
6:44
trying to force a new teaching method or technique
6:47
on them rather We just want to give the teacher
6:49
a workspace a canvas where they can do their
6:51
work as simply as possible a lot of educational
6:55
technology companies try to I guess shoehorn
6:59
other methods or ideas into the teacher's practice.
7:02
So then the tool ends up feeling like it's one
7:05
more thing I have to do as opposed to something
7:07
that's going to actually assist me in my job.
7:09
So that's one thing we've done well. One thing
7:12
for on the challenge side is certainly understanding
7:17
what every teacher actually does need and doing
7:20
that piece where you try to make it speak to
7:21
a teacher, right? I have my own assumptions about
7:23
what a teacher might find important. It's what
7:26
I found important as a teacher, right? Same with
7:28
my co -founder, but we had to kind of... throw
7:32
off all of those biases and preconceived notions
7:34
and just go out and talk to as many teachers
7:37
as we possibly can, as many building administrators
7:40
as we possibly can, and really as many parents
7:42
as we possibly can to get a sense of how they're
7:45
feeling about their district, what's working,
7:47
what's not. We work with a lot of homeschooling
7:49
cohorts as well, so it's really just staying
7:52
connected and talking to these folks as much
7:54
as possible, getting as much feedback as possible,
7:57
taking what they say quite seriously and making
7:59
changes in the application based on those. Yeah,
8:03
and you know recently you were recognized by
8:05
Forbes as being one of the the top 30 under 30
8:10
in education and you know, I'm certain that has
8:13
had something to do with You know the the efforts
8:17
that you have definitely put behind this Trying
8:20
to make you know, let a teacher first product
8:23
You know that that focuses on those things and
8:27
you know how is that that recognition kind of
8:30
what does it mean to you personally and how is
8:32
it kind of affected how you've approached um
8:35
you know trying to put a good product in front
8:37
of of teachers and you know therefore then by
8:40
extension students yeah i mean there's a lot
8:43
of ways in which you are supposed to run a tech
8:46
company and there's a lot of ways in which edu8
8:48
has not done any of those things, right? You're
8:51
supposed to focus on distribution. You're supposed
8:54
to focus on raising your prices to test whether
8:57
or not you have a moat on your product and, you
8:59
know, dumping a bunch of money into marketing
9:02
and sales and all of these things. But we've
9:04
led with building a tool that is sustainable,
9:07
that can actually do something meaningful for
9:10
teachers and then putting it in their hands and
9:12
then the network effects will take care of themselves
9:14
as they share it around. It's just heartening
9:16
to see. That some other folks noticed what we're
9:19
doing and said hey these guys are doing something
9:21
pretty interesting and it's worth recognition
9:23
So it came as quite a surprise to me as well
9:25
because I thought we're kind of breaking most
9:27
of the typical tech company rules but the fact
9:31
that we got recognized for it is is pretty interesting
9:33
and got quite humbled by the whole experience
9:37
and Finding out the day of that you're actually
9:40
like hey, you're gonna be on this this list congratulations
9:43
like i didn't even realize i was nominated for
9:46
anything this is pretty cool you know it just
9:47
kind of came out of nowhere it was a bit of a
9:48
shock and uh thomas could could you talk a little
9:53
bit just about um the platform itself and how
9:57
i kind of want to touch on a couple things but
10:00
first starting with the what is the platform's
10:03
overall mission what is what is it that um you
10:07
want people to be drawn to that to edge aid for
10:11
more to take away from it or or keep coming back
10:15
to it. What's your overall mission and core purpose
10:18
of the tool? Yeah, certainly. So the. Core mission
10:24
is that. We have a whole bunch of research on
10:27
one hand about how the human mind works, how
10:29
it encodes information, how it retrieves information,
10:32
and those things constrain what does or does
10:35
not work in education. So we have evidence on
10:37
what actually works at the level of the human
10:40
mind in processing, learning, retrieving information.
10:43
And then we have classroom practice on the other
10:44
hand. And oftentimes what we know works is not
10:48
always what gets to happen in the classroom.
10:50
And there's many reasons for this. Limited time,
10:52
limited training, limited budget. But more than
10:56
that, it could be money's getting dumped into
10:58
a pet project like inquiry -based learning or
11:01
something like this at the expense of a teacher
11:04
has to do some pre -teaching and actually get
11:06
into the terms and actually do some direct instruction,
11:10
some worked examples and these kinds of things
11:12
in student practice. It's not all just discovery
11:14
-based learning. So all that preface is to say,
11:17
EduAid's main idea is we want to take what works
11:19
in education. and put it in the hands of teachers
11:21
to plan with. So it's a teacher -facing application.
11:24
It does everything a teacher would need in the
11:26
course of planning. So some teachers still have
11:28
to create lesson plans and unit plans and learning
11:31
objectives. We can assist with that. But we also
11:33
know a lot of teachers are past that stage. They
11:35
have lesson plans they've been working with for
11:36
a long time. They don't need to go back and do
11:38
all that. So that's the materials creation side.
11:42
Worksheets, readings, graphical organizers, instructional
11:45
games, review, practice session, spacing practice
11:48
out over time. How do you do that? When do you
11:50
introduce new material to interleave in with
11:53
the old material? What kind of practice is most
11:55
appropriate? Edu8 assists with all of that deep,
11:58
like, instructional design work. So a teacher
12:01
might come on Edu8 and say, you know, I know
12:03
I'm going to be teaching about the War of 1812.
12:07
OK, I'm going to teach specifically about the
12:09
causes of the War of 1812. And I know that I'm
12:12
going to do a jigsaw activity. We'll break students
12:15
up into groups. We'll have four groups. Each
12:17
group will take a cause. We'll do expert groups.
12:20
We'll come back. We'll do a whole group discussion,
12:22
maybe some sort of formative assessment at the
12:24
end of class to see if they got it. Well, to
12:26
create that jigsaw, I need the protocol to figure
12:28
out how I'm going to break these teams up. how
12:30
I'm going to divide the students in my room.
12:32
Then I'm going to need the materials for each
12:34
group. So I'm going to need the readings. I'm
12:35
going to need the questions. I'm going to need
12:37
the note -taking organizers. I'm going to need
12:39
the rules and directions for the students to
12:41
know what to do with those materials. And then
12:43
I'm going to need to be able to time that out.
12:45
So I need to figure out what that actually is
12:47
going to look like from a management perspective
12:48
of my classroom. And then finally, I need to
12:51
figure out whether or not students actually learned
12:52
anything in this whole exercise. So I need some
12:54
sort of formative assessment, maybe an exit ticket,
12:56
an end of class review, something like this.
12:58
You could do all of that on Edu8. You can create
13:00
the protocol. You can create the readings. You
13:02
can create the materials, the questions, and
13:03
the assessments. And we do all of that with generative
13:06
AI. But generative AI itself is not necessarily
13:10
the most useful tool for teachers because you
13:13
have things like hallucinations, biases. It can
13:15
speed up things, but can it do things well? That's
13:17
the big question. So we try to solve that problem
13:20
on EduAid and solve the problem of bringing evidence
13:22
-based instruction into the hands of teachers.
13:24
So that was a lot. I can dive into the actual
13:29
mechanics of the sites. People hear about AI,
13:31
they might think we're talking about like chat
13:32
bots. We're not on EduAid's case, but I don't
13:35
want to ramble on for like a 20 minute block
13:38
and no one got anything out of it, you know?
13:40
No, that's that's okay. And one of the things
13:43
I was gonna ask you actually is You you have
13:46
had to? Iterate and Jade right you you're updating
13:50
it and so one of the more recent things that
13:53
happened is you released the version 3 and So
13:56
some of the things that this kind of what you
13:58
mentioned to just with instructional alignment
14:01
and how The support that's there in edge aid
14:04
can help a teachers and all those different things
14:08
that you mentioned as they're planning for a
14:11
lesson. But can you talk about what's changed
14:13
from maybe, you don't have to maybe go back to
14:16
the first version unless you can, but like the
14:19
most recent update, what changed and what was
14:22
the feedback that you got to get some improvements
14:27
made for this third version that came out? Yeah.
14:30
So we talked to... dozens and dozens of teachers
14:33
who did hundreds of hours of user interviews
14:35
and feedback sessions. We had people clicking
14:36
around, using it, walking us through how they
14:38
plan. And what we've established are a couple
14:41
of things. One, the working with AI in a chat
14:46
format is not that efficient for a teacher. A
14:49
teacher might tell chat GPT or Claude or whatever
14:52
other bot, I need X, Y, and Z. And the bot will
14:55
create it. But it might need some revision. And
14:57
then once it's the information's there, you still
14:59
need to do some structuring, right? You need
15:01
a place for your students to write the name and
15:03
date. And maybe you're going to want some leveled
15:06
readings for this exercise or whatever. Anyway,
15:08
you're going to copy and paste that. So you're
15:09
going to copy it, paste it over to Google Docs.
15:11
Then you do all your formatting stuff. And it's
15:13
like, that's. Worst be saving teachers time.
15:16
I'm doing air quotes if people are just listening
15:18
here So you're not really saving time if I'm
15:21
gonna be going back and forth between two screens.
15:23
I'm copying this I'm pasting this I'm formatting
15:25
it because it was in markdown over here and I
15:27
needed in a different format over here and it's
15:30
all a mess Edu8's whole focus now is instead
15:33
of chatting in a chat based format You ask the
15:37
AI agent to create something it'll create that
15:40
in an actual document It'll look like Microsoft
15:42
Word or Google Docs. You have a toolbar You have
15:45
the full document editor right there in the application,
15:47
but then we added elements to that toolbar So,
15:51
you know on a standard Google Docs toolbar you
15:53
can do headings and font sizes and underlined
15:56
but here you can insert multiple choice questions
15:58
for that particular passage you just looked at
16:01
or you can put a check for understanding as well
16:03
or you could toggle in the name and date section
16:05
of your assignment, just things little and things
16:09
complicated. We try to bring that into the document
16:12
editing experience. So that's part of it. Just
16:14
putting everything in that editor and making
16:16
it feel like an application that teachers used
16:18
before, because every teacher works in Google
16:20
Docs or Microsoft Word. So we wanted to make
16:23
it feel familiar and then add the AI pieces into
16:26
that familiar application in a way that speeds
16:29
up the time. speeds up the process of planning
16:32
and creating materials. The second piece is getting
16:34
insight into what the AI actually does. How do
16:37
I know if this AI generated text is grade level
16:39
appropriate? How do I know if it's actually aligned
16:42
to an eighth grade reading level? What prior
16:44
knowledge the student need to have to access
16:45
this material? I'm not exactly sure of that.
16:48
So we brought in these evaluator tools. We did
16:50
this through partnership with a nonprofit organization
16:52
called the Chan Zuckerberg Initiative. We collaborated
16:55
with them. built out a set of evaluators, and
16:58
now teachers are able to actually get transparent
17:00
insights into the AI generated material. They'll
17:03
do a informative text. They can run the grade
17:06
level appropriateness evaluator. We do a quantitative
17:09
qualitative text analysis. Essentially, we're
17:12
looking at word counts, syllable counts, sentence
17:16
complexity, length of sentences, number of sentences.
17:18
We're looking at the structure of the text itself
17:21
when ideas are presented, how they're prefaced,
17:23
how they're followed up on. Then we look at the
17:25
background knowledge constraints. What did students
17:27
need to do prior to access this? Then we assign
17:30
it two levels. Level one is your grade level
17:33
where you could read this independently. So an
17:35
eighth grader could access this text with no
17:37
assistance. They'd be able to do it. And then
17:40
you have a teachable with scaffolds. uh grade
17:43
level so it's like you could do this with sixth
17:44
to seventh grade students but you would need
17:47
the following five scaffolds it's a one click
17:49
button it will apply those scaffolds to the document
17:52
so it might say you needed a glossary it'll create
17:55
the glossary it might turn those short answer
17:57
questions into multiple choice questions it might
17:59
eliminate some of the distractors from the multiple
18:01
choice questions it will add captions to images
18:04
if you have it and stuff like that so Evaluation
18:07
and transparency on one hand, ease of use on
18:10
the other. We bring these things together and
18:12
created version three of the application. So
18:14
it doesn't sound like we simplified it, but we
18:17
simplified it. I just wanted to give you guys
18:18
some details on the evaluator side so you know
18:21
how it works. And I'm not just talking out the
18:22
side of my neck. Yeah, no. And I think that's
18:24
really important to point out, too, because we
18:27
have a lot of different education, AI education
18:30
tools that have come out in the last few years.
18:32
And there's great things about many of them,
18:37
but what you mentioned there, is that something
18:41
that you feel sets yourself apart as a company,
18:45
where you do provide those evaluation tools,
18:49
because a lot of that, sometimes when we do trainings
18:53
with teachers, you know, you you help them understand
18:56
how those outputs are working, making sure that
18:59
they have ways of checking that output so that
19:01
it's, you know, checking for biases, checking
19:04
for accuracy, all those type of things. But it
19:06
feels very like maybe it could be a bit cumbersome,
19:11
especially when it spits something out at you
19:12
that's really long, and you've got to put way
19:15
more time into it than you may intended to. So
19:19
is so I The first question is, do you have do
19:22
you feel like that sets you apart a bit with
19:24
those evaluation tools? And is there anything
19:26
else, too, that you would add as other features
19:28
that set you apart from other education, AI platforms?
19:32
Yeah. I mean, you got directly to it doing an
19:35
evaluation over the text yourself. Anyone, really?
19:38
I'm not just saying just teachers, everybody
19:40
who's looking at an AI output. Unfortunately,
19:43
it's all based on vibes. It's like, yeah, this
19:45
looks right. It sounds right. I think that's
19:47
correct. But now you're taking something that
19:50
may have a bias or hallucination in it. And the
19:52
way that we're judging it is with our minds thinking
19:55
very quickly. And when human minds think quickly,
19:57
we have these things called heuristics, little
19:59
shortcuts that we take to make judgments. And
20:01
those introduce all kinds of problems, too. So
20:04
that check is not the most reliable. And you
20:07
don't blame anyone for that. It's just human
20:09
nature. We try to expend as little energy as
20:13
possible. That's kind of by design. Yeah, put
20:18
the evaluation tool in their hands so they can
20:20
actually get a reliable look into the material
20:22
instead of having to rely on, I'm going to read
20:25
this line by line or I'm going to just kind of
20:28
graze over it and take a look. Neither one of
20:31
those is the most efficient. So it certainly
20:33
sets us apart. I hope it doesn't set us apart
20:35
for much longer. Like I hope everyone else does
20:37
this as well. Teachers deserve high quality instructional
20:40
materials regardless of where they get it from.
20:42
They get it from us. Of course, I'm happy. I'm
20:44
a business owner. But if they're getting it from
20:46
somewhere else, I hope it's of quality because
20:48
I'm a teacher as well. And I want teachers not
20:50
to be using slop in the classroom. I want them
20:53
to use high quality instructional materials.
20:55
That's the thing I'm most passionate about. Other
20:58
pieces that set us apart on EduAid is just to
21:01
that piece about high quality instructional materials.
21:03
So we have a knowledge graph on the back end
21:05
of our system that we use to design every single
21:08
one of our prompts on EduAid. So the way this
21:10
knowledge graph works. is there's about 160 entries
21:15
on the knowledge graph. And each one of those
21:18
entries is a research paper. You have a selection
21:21
criteria for the research paper. So it's like,
21:23
are you using a controlled study? Or what are
21:26
the confounding variables? Can I actually trust
21:28
the evidence that this paper's putting forth?
21:30
So once we verify and select the papers, we tag
21:33
it. What area of instruction does this paper
21:36
relate to? Does it relate to direct instruction?
21:38
Does it relate to student practice? Does it relate
21:40
to assessments? Does it relate to classroom management
21:43
and culture? What does this paper relate to?
21:46
Then once we tag the paper, we say, OK, based
21:49
on that, which tools on EduAid does it relate
21:51
to? Based on that tag. So it relates to all of
21:53
our assessment tools. Great. Then we take that
21:55
paper. We say, what are the instructional implications?
21:58
What is it actually saying about teaching and
21:59
learning? Oh, it's saying that students can use
22:03
highlighting as a study technique, but only if
22:05
they turn those highlights in the flashcards
22:07
because they need to review it over time. They
22:09
need to practice it over time. Well, then we'll
22:11
create a rubric based on that paper and we'll
22:14
grade our outputs. It's like, well, does this
22:16
lesson plan include time for student deliberate
22:19
practice? Does it explain how a teacher should
22:21
do that practice session? Does it explain what
22:25
prior knowledge a student would need to access
22:27
it? And it's like, if it doesn't pass that rubric,
22:29
then we go back through and we revise the output.
22:32
And then we do that for all of our tools. So
22:33
every single one of our tools is essentially
22:36
as close as we can get with AI today. A tool
22:41
to provide the evidence -based instructional
22:43
material backbone for the class. Then we provide
22:45
the transparency tools on top of that. And the
22:48
idea is hopefully we're providing teachers with
22:50
a pretty close approximation of a high quality
22:53
instructional material that will work for their
22:55
grade level. work for their student population,
22:58
work for the reading levels they have in class.
23:00
They can differentiate it with tools on the site
23:02
to make it easy to use. So it should fit every
23:04
diverse population in their classroom. So then
23:08
you kind of are covering most of the bases on
23:10
the teacher facing side. And I think that's what
23:12
sets EduAid apart. A lot of tools might do high
23:16
quality materials. A lot of tools might, well,
23:19
not many tools at all right now are doing evaluation,
23:21
but they might do that as well. It's how we bring
23:24
these things together that really changes the
23:26
game. It's subtle stuff. I mean, because at the
23:27
end of the day, everybody's using ChatGPT or
23:29
Claude at the backbone of a generation, right?
23:33
Brisk, Magic School, School AI, Diffit, Eduade,
23:37
Curipod, all of us, we're all reliant on ChatGPT,
23:40
Claude, Gemini, or the Llama models. We're all
23:45
using the foundation models. No one's a team
23:47
of AI scientists making education -specific AI
23:50
models. If someone's telling you that, they're
23:52
blowing smoke, and you should be skeptical as
23:54
a as a tech procurer for your school, for sure.
23:57
Really, at the end of the day, for me, it's like
23:59
strip away all the technology, just strip it
24:01
all the way. Don't even think about the technology
24:03
for a second. Look at that tool and say, you
24:05
know, how what problem is this actually solving
24:08
for me in the classroom? What's its theory of
24:11
learning? Like, how does it actually relate to
24:13
how students learn? And if it can't? cross the
24:15
bars on those two things, regardless of how flashy
24:17
the tech is, it's going to fail in your classroom.
24:19
Because one, it's not going to support how students
24:21
actually learn. And two, it's not solving a real
24:24
problem for you in the classroom. So sorry, I
24:26
went off on a tangent there. But I think that's
24:28
probably something with AI, it's all like this
24:31
smoke and mirrors stuff. Like none of the companies
24:32
want you to know that they're just an API. I
24:34
don't want to say just a chat wrapper, because
24:36
there's ways in which you could prompt an AI
24:37
and get terrible responses. There's ways where
24:39
you can open it up, allow teachers to create
24:41
hundreds of different resources. And who knows
24:44
what the hallucination rate will be because you're
24:46
stretching it beyond the AI's knowledge set,
24:48
you're stretching it beyond what the AI can reasonably
24:50
do under constraints. And it's like when tools
24:52
do that, then you're just muddying the water
24:54
for everyone. else. Okay, so Thomas since you're
24:58
we're kind of on that that. discussion here of
25:01
you know being tool first or being pedagogy first
25:03
and I think you described it perfectly here when
25:05
you say you can really tell if you have something
25:08
that's good and high quality if you strip away
25:11
any of the tech that's involved and focus on
25:14
you know the pedagogy part you know is it solving
25:17
a problem that I have in the classroom is it
25:19
something that's going to increase student learning
25:22
or is it something that's going to improve instruction
25:25
you know like Um, why is it so important? Do
25:28
you think let's just go on that tangent then
25:30
a little bit more about why is the pedagogy first
25:33
so important? Yeah, I mean that lesson was beaten
25:36
in my head by my graduate school advisor, uh,
25:39
dr. Jim diamond He has a great educational technology
25:42
program and and that's really a lesson they beat
25:44
in your head, which is Strip away the technology
25:47
first ask the first principal's questions What
25:50
problems it's solving and what kind of pedagogy
25:52
pedagogy is it putting forth? What kind of theory
25:55
of mind? And you have to be able to answer those
25:57
questions. With some tools, it's not entirely
25:59
clear what either one of those answers would
26:02
be. Sometimes it feels like solutionism, where
26:05
you have a piece of technology, but it's not
26:07
really solving a problem. It's just kind of cool,
26:10
maybe, more frustrating if you have to learn
26:12
how to use it. And then the other piece is No
26:15
matter what tool you use in the classroom, they
26:17
all have a theory of how learning works. It's
26:20
either explicit, like their creators actually
26:23
think about this question and design a very specific
26:26
theory of teaching and learning into the tool.
26:28
It's embedded. Or they don't think about it and
26:32
they create the tool, but one comes forth anyway,
26:34
because it's interacting with human minds and
26:36
it's interacting with the teaching and learning
26:38
process. So it will have an effect. Let me give
26:40
you some examples, right? Look at something like
26:43
Kahoot! Great thing, I used it in the classroom
26:47
all the time. Most teachers have at one point
26:48
or another. Strip away all the technology. Ultimately,
26:52
what is Kahoot all about? It's just like a trivia
26:56
review thing, dude. Did you remember some certain
26:59
fact from the lesson you had? Certainly. So it's
27:02
a kind of formative assessment. Teacher can write
27:05
a question, get some answers. You do it live
27:08
with the class. I could see roughly where my
27:11
whole class falls. what percentage of my students
27:13
don't know this thing versus what percentage
27:15
do. Okay, then like what's the theory of learning
27:18
there? It's like what is that tool prioritizing?
27:22
It's not like constructivist learning. It's not
27:24
students going out and doing discovery -based
27:26
learning. It's much more of a recall -based thing
27:29
or a retrieval -based thing if the questions
27:31
are well structured. This is fine. But if a teacher
27:35
becomes reliant on this and every day is a cahoot,
27:37
we're going to do a cahoot for the warm up, I'm
27:39
going to check in with you with a cahoot, we'll
27:41
do a cahoot at the end of class. It's like that
27:43
whole thing's going to wear off because it's
27:45
only one part of what a teacher actually does.
27:48
It's not all just assessment recall in a game
27:50
-based environment. You have to do more than
27:52
just that one thing. But if a teacher grows very
27:56
fond of a tool, comfortable with a tool, it has
27:58
student engagement or a student's motivated to
28:01
get engaged with it and at the expense of, say,
28:05
exploring other options, then it can become kind
28:07
of stale, right? I might just be doing this for
28:09
the wrong reasons. Because all of my students
28:12
are at least are in their seats and not running
28:13
around the classroom like hooligans while we're
28:16
doing this, right? It's actually a controlled
28:18
environment. I knew far too many teachers who
28:20
that was the thing. Like that's how they decided
28:22
which ed tech tool to use. I'm going to do a
28:24
book it because look, it keeps my kids at the
28:27
laptops. And at least they're not on their phones
28:30
or saying something rude to someone else or,
28:34
you know, distracting other kids or something
28:36
like this. It's like, well, those are some of
28:38
the wrong, the wrong reasons, right? Like it
28:41
seems more, it's occupying their time rather
28:43
than fully engaging them. Exactly. So if a tool
28:46
is really good at occupying the time and has
28:49
no solid theory of learning or no solid pedagogy
28:51
behind it, but can occupy, occupy time, occupy
28:54
the students attention, get them engaged. It's
28:58
kind of hollow. It might actually hurt learning
29:01
gains over time, right? It's just a distraction
29:03
that you use to fill the class, fill time in
29:06
your classroom. I worry that if you're not paying
29:10
attention to the pedagogy of the tool, if you're
29:11
not paying attention to how it's going to affect
29:13
teaching and learning. Some teachers, I'm gonna
29:16
say especially new teachers. I mean, I've been
29:18
in that boat trying to figure out classroom management
29:20
my first couple of years teaching. Maybe those
29:23
veteran teachers who have lost. the spark. I
29:27
mean, I've seen that happen, too. You get burned
29:28
out. You've been in the career for 17 years.
29:32
Everything's changing around you. But at the
29:34
same time, nothing's really changing, too, right?
29:36
I get that feeling. And then there's like, I
29:40
need kind of coast type of thing. Then you can
29:43
see how some of these tools that are just kind
29:46
of hollow, engaging and only engaging, seep in
29:50
to classrooms. You see a lot more game based
29:53
engagement in classrooms today than you did.
29:56
10, 15 years ago, 20 years ago. I forget how
29:59
fast time flies. I'm like, oh, 10 years ago,
30:02
that's already 2016, geez. Okay, so yeah, we
30:04
got to go back 20 years to 2006. I'm not sure
30:09
I 100 % answered your question. I threw out some
30:12
examples and kind of pointed to that as a trend.
30:14
So if I was to like summarize what all of this
30:16
means. You have to design with the pedagogy in
30:19
mind first, because the tool is just that. It's
30:22
a tool. It's something you're going to use to
30:24
assist in the act of teaching and learning. And
30:26
if the tool does not actually understand teaching
30:28
and learning, then it's not really going to assist
30:30
with teaching and learning. It's something you're
30:31
kind of putting on top of that process. Maybe
30:35
to get student engagement. maybe for motivation.
30:37
Maybe it does one thing really well, which is
30:39
assessment. You have to understand why that tool
30:42
is going to work in that particular moment before
30:44
you use it, right? If you don't know why you're
30:46
using the tool, you're just using it because
30:47
the district has a subscription and everyone
30:50
else uses it. you're going to run astray. So
30:52
you need to have a firm understanding of, oh,
30:54
I'm going to use this for this formative assessment
30:57
because I need to know whether or not students
30:58
understand this concept because I'm going to
31:00
do something in class tomorrow that's going to
31:03
build on that concept. And if they don't know
31:04
it, I got to go back and reteach it because they're
31:06
not going to be able to do tomorrow's lesson
31:08
unless they've mastered it. It's like, well,
31:10
now I know why I'm doing the formative assessment
31:12
here. And I know Kahoot might be the good tool
31:15
to use for it because it's engaging, but it also
31:17
does the assessment piece. asking those questions,
31:21
you might run run aground. So it kind of leads
31:25
me to this idea here, like, and I thought some
31:28
of this was kind of like Barry, but I keep seeing
31:30
articles about it popping back up in my feeds,
31:34
and I'm like, LinkedIn and stuff like that. But
31:37
it's, it's this idea, AI is going to take all
31:40
of our jobs away. And I thought that was kind
31:42
of like, No, no, that's not really. And then,
31:45
you know, it's going to be like, those people
31:46
that don't have jobs will know how to use AI
31:48
as part of their job. And I'm reading some articles
31:53
then recently like Spotify is laying off 17 %
31:59
of their staff because according to one article
32:02
I read that their software engineers haven't
32:06
done any real coding since like December or October,
32:10
whatever that month was. I don't have it in front
32:12
of me anymore, but I'm hearing stories like that.
32:15
And it makes me think here again that there are
32:18
these avenues, these career paths out there that
32:21
AI can replace more easily than others. But then
32:25
I come back to this idea in education here. And
32:28
I think because of the things that I'm hearing
32:31
you say, that it is still, I think, essential
32:35
that a student has a high quality teacher and
32:39
that high quality teacher is going to be the
32:42
thing that's going to make their learning the
32:44
greatest, that high quality teacher is most essential
32:47
for student learning. In fact, that's one of
32:49
the maxims when I go out and I present to groups
32:53
of teachers that I try to stress. But, you know,
32:57
like, so along with that, though, I think teachers
33:00
also need to understand AI, and we've been talking
33:04
about that. And I really appreciate your focus
33:07
on the evaluation piece of that. recognizing
33:11
whether you have high quality and using the quantitative
33:15
and qualitative measures as part of that to identify
33:19
those things that are high quality teaching practices
33:22
that I think can benefit both new teachers and
33:25
teachers that have been in that field for quite
33:27
some time. But I think I'm kind of leaning now
33:30
towards the question here. How does or how can
33:34
teachers responsibly use AI without giving up
33:41
their expertise and their human centered approach
33:44
to learning, building those relationships with
33:48
their students. That's a loaded question. First
33:51
things first, if you're scrolling around online
33:55
and you see anybody say that AI is going to take
33:58
a bunch of jobs in the field of education, that
34:01
tells me two things. One, they don't really understand
34:03
AI and they don't really understand education,
34:05
which sounds like a flippant answer, but once
34:08
you start to dive into the technology and the
34:11
practices of a teacher, it's not so flippant.
34:13
It just sounds like, oh, he's just stating a
34:15
fact of reality, which is you look at a lot of
34:18
different practices that a teacher engages into
34:20
the AI. It's still well beyond the reach of an
34:23
AI system. So a large language model at base
34:27
is running a relatively shallow computation over
34:30
a vast amount of data, which essentially takes
34:32
your input, tokenizes it into words or word parts,
34:35
and then makes a predictive analysis of which
34:39
words or word parts will follow that input. And
34:43
it's quite effective. It looks like thinking,
34:45
right? It does a great job at this computation.
34:47
But the problem you start to run into is let's
34:49
go early elementary, early ed. It can't do phonics
34:53
instruction. It can't do phonetic spellings of
34:56
text. It struggles to craft reasonable decodables.
34:59
It can create verbose text for college students
35:03
and high school students. But can it accurately
35:06
convey information in very simple text for, say,
35:09
a first grader or a second grader? It cannot
35:10
do that very well. Then you move into the tutoring
35:13
piece. Even if an AI was encoded to Ask great
35:18
guiding questions, ask great leading questions,
35:21
have wonderful work examples on deck, be able
35:23
to actually jump in when a student makes the
35:26
same mistake multiple times. Currently, it can't
35:28
do that very well, but let's say it could. Even
35:30
then, I'm not sure an LLM will make such a great
35:34
tutor that it will place teachers in the classroom.
35:38
I just don't see that happening. One, you have
35:40
a lack of engagement and motivation to really
35:42
dive in deep with that AI. If I'm a student,
35:45
I'm going to put IDK and wait for swaps to get
35:48
the answer. Or I might just do the base minimal
35:52
amount of work. But I'm not really engaging deeply.
35:54
When I'm questioning a student, I'll ask a question
35:57
of the student. I'll get the answer. And then
35:59
I'll dive in and ask a follow -up and ask a follow
36:01
-up. And so as my class discussions would stretch
36:04
out pretty long, because I'm asking so many follow
36:06
-up, elaborative questions of the student who
36:08
answered, because I'm trying to probe deep where
36:10
that answer is coming from. What connections
36:12
are they making? How are they thinking? And I
36:14
really do not see an artificial intelligence,
36:17
an LLM tutor doing that in the near to medium
36:21
term. That's over the next 10 years. And even
36:24
if they could, you're still going to have a human
36:26
in the room who is going to engage with that
36:28
student and debrief them after that one -on -one
36:31
experience with that AI tutor. It just seems
36:33
to be the direction is going for me. So how does
36:36
a teacher actually use an AI system? Again, to
36:41
not undermine their autonomy, to actually assist
36:44
them in class. A couple of parts for that. One
36:48
would be, again, strip away all the technology
36:52
and ask what problem am I trying to solve? So
36:54
as a teacher, I'm using a technology because
36:56
I have a problem I want to solve. What is it?
36:58
Once you understand what you're actually trying
36:59
to do, engage with the AI to that problem, right?
37:03
Don't like veer off and try to reinvent the wheel
37:05
in the classroom. So it might be, my problem
37:09
is... I am teaching an ELA course. We are looking
37:13
at a text of some sort, let's say a novel or
37:17
something, but I have students who are not at
37:19
the level to be able to access the text of that
37:21
novel. So I have a gap. I need to solve for this
37:24
gap. So I need scaffolds. That might be where
37:27
it's useful to engage with an AI because currently
37:30
a teacher trying to solve that problem, it is
37:32
a long drawn out process. I might ask my colleagues.
37:35
which maybe you should still do that, right?
37:36
How would you solve this problem? They're a great
37:38
resource. Maybe if you're lucky and you teach
37:41
at a huge school district that does these co
37:43
-plannings, you might have a co -planning with
37:44
everyone in your department. You're talking to
37:46
all the other English teachers. Maybe that's
37:48
useful. Maybe not. My experience co -teachings,
37:51
they varied from very useful to not so useful,
37:54
depending on the group I was working with, but
37:57
not naming names here or schools for that matter.
38:00
Then you might go online, right? And at that
38:04
point, that's where I think we start to see the
38:06
real benefits of AI, because a lot of teachers
38:08
are going online and they're pulling lesson ideas
38:10
from Pinterest or lesson ideas from something
38:12
they saw on a Facebook community group, or they're
38:16
going on teachers pay teachers or they're just
38:20
kind of beg, borrow and steal as a teacher. That's
38:23
kind of what you do. You try to find us reading.
38:26
Yeah, I mean, that's Harry Wong, First Days of
38:30
School, that's one book they assign in class.
38:32
He says teachers steal everything, which you
38:34
do. You see a great lesson online somewhere,
38:36
like I'm going to do that. Or the National Archives,
38:40
if you're a history teacher, has a great lesson
38:41
on this historical document. So you pulled that.
38:43
And AI kind of can take all of that research
38:46
time and compress it. You could run a deep research
38:49
agent on, you know, I need to do source analysis
38:52
of this document. How might I go about it? It
38:54
might pull all of those resources for you. So
38:56
it's a great research assistant. It's a great
38:58
instructional design assistant when you need
39:01
to do all that differentiation, right? I have
39:03
varied reading levels in class, but I need all
39:05
students to be able to access it. So you go about
39:08
doing all the personalization, the differentiation
39:11
of those learning materials. That to me are the
39:13
two most compelling areas where a teacher might
39:16
use AI today. Assist with finding the materials,
39:19
assist with creating the materials, and assist
39:21
with refining and differentiating them. Outside
39:24
of that, I don't really see it. I don't see the
39:27
we're going to put students in front of tutors.
39:29
Maybe there's something to be said for analyzing
39:34
assessment results and suggesting personalized
39:37
pathways of instruction or different interventions
39:45
you might make based on those results. But I
39:48
don't know, I think a lot of people overblow
39:49
the AI hype. And it will ultimately be to the
39:52
detriment of those companies that are overblowing
39:54
it, because everyone's going to get disillusioned
39:55
when they realize that all those promises weren't
39:57
really there. It's not saving them the 20 hours
40:00
they were promised. It's not improving their
40:03
students' test scores by X, Y and Z amount that
40:05
they promised. So then they start to back off.
40:07
But if you say, we're going to do some really
40:09
simple things, and we're going to do it right,
40:11
and we're going to be there for you, That's much
40:13
more compelling. Part of what you said, I think
40:17
I harken back to and I think this was from Ethan
40:19
Malik, but go to AI when it can be better than
40:22
the best human available to you at that time.
40:25
Exactly. That seems to me like exactly what you're
40:28
saying. I mean, a lot of times a teacher runs
40:32
out of the best humans, right? We've asked everybody
40:35
in the building, and they're still stumped. So
40:37
you're like, well, where am I going to go now?
40:38
And then you're doing research online. And that's
40:40
where the planning time gets burned, because
40:42
you're trying to find, well, what's the right
40:44
source for this lesson? For me, it was always,
40:49
I like to start a class with a political cartoon,
40:51
a map, or an image related to the content that
40:53
we'll be doing in class. And we'll do some sort
40:55
of source analysis there. So during World War
40:57
II, it might have been a map of the different
40:59
battles of the Pacific. We look at the tactics.
41:01
We look at, what do you think was strategically
41:03
important for this side or this side? What resources
41:06
do you think were there, et cetera? So you do
41:08
that great source analysis. Finding those sources
41:10
was always. a pain and a half, because one, is
41:14
this actually conveying the message that's going
41:16
to relate to the learning objective for class?
41:19
Is it actually going to elicit student responses?
41:22
Is it over their head? Do they have the context
41:24
for it? With an AI assistant, I might be able
41:27
to, like on EduEd, you could do image generation,
41:30
but we also included image search. So now you
41:32
can actually search and find that image, inject
41:34
it right into your source analysis assignment,
41:37
give our AI assistant that image, and say, I
41:39
need some deep source analysis questions to get
41:43
to X, Y, and Z theme or concept. And then you
41:46
can build off of that. So it's like taking what
41:50
teachers already do and then only adding AI where
41:53
it's like, this is where I got stumped when I
41:55
was doing it, right? So I might have it. I might
41:57
have my first two questions, but I need my next
41:59
two questions. And I'm not exactly sure what
42:00
to ask. That's when an AI might be most effective.
42:05
So we've covered a bit of the teacher side of
42:08
things. And so I'm wondering too on the Student
42:11
side, you know, that's that's the. I think what
42:15
I've noticed is a lot of schools aren't there
42:17
yet because they haven't even really fully covered
42:19
the teacher side, right? But if you're maybe
42:23
just in schools that you've talked to or. In
42:27
working with Edu aid, where do you see the students
42:30
fitting into all of this discussion? Like when
42:32
should they start using it? What what would be
42:37
the parameters? that they should use AI? Is it
42:42
always going to be under direct teacher supervision
42:46
in the classroom? Because we know of all these
42:50
different issues that go on, some of which we
42:52
can't even really fully mitigate ourselves. So
42:56
then what's the best way to start talking with
42:59
students about it and to open the door to them
43:03
in a guided way? you know, because in the I guess
43:07
the other piece of this is that we know that
43:09
they're using it on their own anyway. Right.
43:12
It's not like they're just, you know, sitting
43:14
at home and not not using it. Right. Pretending
43:18
it doesn't exist. So what what would be your
43:22
opinion on student use and how we should go about
43:25
that? Yeah, I mean, this that's the tough question
43:28
of the day, because, as you said, students, your
43:32
students might not be using AI. But the kids
43:35
are definitely using AI. And if you think about
43:37
that, it's like, OK, they're using it at home,
43:39
but they're not using it in school. So any time
43:41
there's schoolwork that will be going home, you
43:44
can definitely see that there's going to be some
43:47
use of that AI that they're using at home on
43:49
that schoolwork. So maybe that means homework
43:51
is dead. Can you really trust students once they're
43:54
outside of the school building to not use the
43:57
tool that we most definitely would be using if,
44:01
let's say, your building admin comes to you and
44:03
says, hey, We're doing some curriculum mapping
44:06
and we need unit plans and we need all this paperwork
44:08
or something. And you're like, I know this is
44:10
gonna go in a filing cabinet. No one's really
44:12
gonna read it. What are you gonna do? You're
44:14
probably gonna use some form of an AI to help
44:17
assist creating those papers. So that preface
44:21
out of the way. Students might not be using it.
44:24
The kids definitely are. When is it appropriate
44:27
for the kids to use it? I mean, my honest answer
44:29
is I'm really not sure yet. because there's papers
44:33
coming out saying, you know, there's an accumulation
44:36
of cognitive debt when using an AI assistant
44:38
for, say, essay writing tasks. There's papers
44:41
coming out that are showing the negative effects
44:44
of AI usage on developing minds when they offload.
44:50
So I'm worried about cognitive offloading. How
44:53
much of kind of the basic work and mechanics
44:57
of, say, writing or of thinking about something
45:00
just kind of go out. the window. Some other pieces
45:05
that kind of build off of this. I see a troubling
45:09
trend where a lot of specifically curriculum
45:13
writers, but I've also seen it among teachers
45:16
and definitely among a lot of researchers on
45:18
teaching and learning, which is, you know, kind
45:20
of discounting facts, recall, declarative knowledge,
45:25
like actually understanding concepts. We want
45:27
students to think critically. And anytime I hear
45:29
that, it just like It just like nails on a chalkboard
45:33
to me because it's like, okay, what is critical
45:35
thinking then? Can you tell me what critical
45:37
thinking is? And no one ever says that it's taking
45:39
the accumulation of knowledge that you have,
45:42
applying that to solve problems, and then doing
45:44
those problems effectively. It's applying knowledge
45:48
to novel situations effectively. Ultimately,
45:51
that's what critical thinking is. I take my knowledge,
45:54
I apply it to a situation, I solve that problem
45:56
where it might not be apparent that I would use
45:58
that knowledge to solve that problem. and I do
46:01
it efficiently and effectively. If you undermine
46:03
a student's declarative knowledge about any subject,
46:07
they're not going to be able to think critically
46:08
about it. How is anyone going to think critically
46:10
about government policy in the U .S. if you don't
46:12
have a basic understanding of our constitution,
46:15
of our laws, of our government structure? There's
46:17
basic facts underneath every discipline that
46:20
you need to know. And if teachers are using,
46:23
or if rather students are using, AIs to... essentially
46:27
offload that part, all that recall, all of that
46:30
knowledge stuff. They're not going to have the
46:32
substrate for the higher order thinking. They're
46:34
not going to have the necessary knowledge and
46:36
skills. So I do worry about introducing AI too
46:39
early, but I also understand that regardless
46:42
of when we as teachers or schools introduce AI,
46:45
students will use it on those devices that they
46:47
already have. So you have to balance the two.
46:50
So this is why I see schools going like... I
46:55
don't wanna say backward, but essentially taking
46:58
on a role. I think about the more like churches
47:02
than I do about like most other things. Like
47:05
a church is a very, it's a tradition, it's a
47:08
structure, it's based on a spiritual side of
47:12
human nature that maybe isn't addressed by the
47:15
modern world as well. So in this way, it's kind
47:18
of ageless. I wonder if education will take a
47:21
similar path, but for the brain instead of spirituality,
47:23
it's like we're going to be working on paper
47:25
and pencils, and we're going to be looking out
47:27
of actual books, and we're going to be doing
47:28
the arcane form of learning because that's the
47:32
most closely related to some cognitive function
47:35
in the human brain. Is reading and writing really
47:38
an analog to thinking? Well, if it is, we shouldn't
47:41
be typing on keyboards with AI assistance then.
47:44
I see schools talking about blue books coming
47:45
back and things like this. I'm not going to be
47:48
one of those Luddites who say don't use technology
47:50
at all, but I'm just not sure the tech is there
47:52
where it would be beneficial for a student to
47:54
be using an AI for. I can't think of any area
47:59
where it would be useful for a student to use
48:01
an AI. I'm sure Sal Khan and a number of other
48:07
founders would tell me I'm wrong and say there's
48:09
plenty of uses for students to have an AI. They'll
48:12
use personalized learning, which every tech company
48:14
has been using since like, 1996 because it's
48:17
easy to hide behind as well. It doesn't work
48:20
for everybody because you know the student Had
48:23
an extraordinary case and you know, they fell
48:26
outside the parameters It's like well a teacher
48:28
would adapt to these things or attempt to adapt
48:30
to these things better than Technology can so
48:33
that's my pessimistic answer. I didn't really
48:35
give you a great one I didn't say when you should
48:38
you I don't I'm not sure I answered the question
48:39
I kind of evaded it with a lot of words Does
48:43
it become part of the school's responsibility
48:45
to teach AI literacy safely, ethically, effectively,
48:50
you know, like we would, you know, because it's
48:52
not happening in other places, right? Same thing
48:55
with like drug and alcohol, you know, why did
48:58
it become part of the schools to teach kids about
49:02
the dangers of drugs and alcohol? Let me push
49:05
back on that though. I mean, have rates of drug
49:08
and alcohol addiction actually fallen from that
49:11
intervention? No, probably not. No, I mean, opioid
49:14
epidemic is still raging in the United States.
49:16
But would anything happen with AI usage too?
49:19
I mean, do we we have some online literacy courses?
49:22
I see that schools about like social media and
49:25
bullying and things like that. But rates of teenage
49:28
suicide onset by depression related to some sort
49:31
of online interaction is still Unconsciously
49:35
high. I mean, I'm not sure if you make it like
49:38
this pull -out thing someone comes into your
49:40
class and talks about the dangers of drug addiction
49:43
or whatever is Necessarily the most effective
49:46
way right if you're going to do AI literacy if
49:49
you're gonna do awareness around any social harm
49:51
It needs to be embedded into the curriculum into
49:54
the different subjects right into the daily work
49:57
of the students So you might talk about? specific
50:01
examples of technological innovation in your
50:04
history class and how that changed the way people
50:05
think and then relate that to today. In an English
50:08
class, you might be talking about the production
50:10
of a text, you might relate that to today. So
50:12
a lot of this, unfortunately, like everything
50:15
in education, falls onto the shoulders of the
50:17
teachers to get creative and think about how
50:19
this thing in society today relates to the subject
50:23
that you're teaching. And this is why teachers
50:25
are always like a... Thought about in political
50:29
terms because they have to interpret a lot of
50:31
what's happening in society For their students
50:33
and relate it to the subject that the students
50:35
are learning in that class. It's a difficult
50:38
job But I'm not sure it's it's going to be solved
50:41
by this is our AI literacy week at school It's
50:44
like you're gonna have to do more than that It
50:45
needs to be something deeply embedded and I'm
50:48
not sure anyone's in a position to do AI literacy
50:51
very well yet I see a lot of the training programs
50:53
out there What they say is, you know Double -check,
50:58
there's bias. Don't use it for everything. OK.
51:03
Thanks. I'm still going to use it, though. It's
51:06
embedded into my Snapchat. It's embedded into
51:09
every single application I use. It's embedded
51:12
into new devices now. It's on their iPhone. So
51:15
OK, we can tell students to proofread and double
51:18
-check and look for bias. Can they actually identify
51:21
bias? And even if they can, are they going to
51:23
expend the cognitive effort it takes to look
51:26
for bias when the whole point of using the tool
51:28
is to cognitively offload? I don't have great
51:31
answers for you on that. I think I'm raising
51:34
more problems than I am putting forward answers.
51:38
But they're worth considering. I mean, I just
51:41
hope these things don't get swept under the rug
51:43
in the name of we're just doing our AI literacy
51:45
part. It's like, that's not enough. You got to
51:47
do more. Right, and I think one of the things
51:49
too I just wanted to go back to is you mentioned
51:53
background knowledge about subjects, and it seems
51:57
like part of the issue that we have, particularly
52:00
when it comes to what you mentioned with student
52:03
use, is that if you don't have any background
52:06
knowledge, the cognitive offloading is just going
52:08
to automatically happen because of the way that
52:11
AI spits out answers. and gives you information
52:15
without considering because it's not going to
52:17
consider for you, whether you have the background
52:19
knowledge, right? You have to have had that established.
52:21
And that's probably one of the key distinctions
52:24
between when we when we have teachers use it
52:27
for certain things. We do it with the understanding
52:29
that they've they've gone to college. We assume
52:32
that they've got the pedagogical side of things
52:34
figured out. And so then their their background
52:36
knowledge of that and the content area helps
52:39
inform how they might use AI. When we talk about
52:42
students using it, that might not be there at
52:46
all for whatever given. That's why they're in
52:48
school. They gotta get that background knowledge,
52:49
exactly. Yeah, yeah. So that just stuck out to
52:52
me when you had mentioned that, and I think that's
52:54
really important when we talk about student use,
52:56
because we have to understand that that background
53:00
knowledge might not be there, and they might
53:02
just be getting answers automatically. Not that
53:04
teachers can't as well on things, but. But the
53:08
point you're honing in on is that's the hopeful
53:10
piece of it. Like you've identified the correct
53:13
cognitive function, which is students have to
53:16
retrieve information from their memory. That's
53:18
retrieval practice, right? Not just recalling
53:20
something that's in their short term memory,
53:22
but actually going back to long term memory storage,
53:25
retrieving a piece of information and applying
53:26
it. So that's retrieval practice. Then you want
53:29
to do that over time, which is distributed practice.
53:31
If you just do something in one mass during one
53:34
week and never do it again, that's why students
53:37
struggle with statistics. You do it at the end
53:39
of the year for a couple of weeks in math class,
53:42
and then it's done. In my mind, statistics is
53:44
one of the most important branches of math that
53:46
students should be studying for everyday life,
53:48
but that's another tangent for another time.
53:50
The third piece is interleaving. I've got to
53:52
bring in new pieces of information that I'm picking
53:55
up and relate it to the old pieces. That is where
53:58
an AI could be most effective for a student.
54:00
And how do you orchestrate retrieval practice
54:03
over time and interleave all those practice questions,
54:05
right? If the whole point is we want deliberate
54:08
practice, we want students to get as many reps
54:10
as possible on this information. And AI being
54:13
able to orchestrate that practice over time for
54:16
the teachers. So where it's kind of running in
54:17
the background, the students are doing their
54:19
practice tests and doing the reviews, but it's
54:21
not like a tutoring session. It's really just.
54:24
a bunch of different practice tests and assessments
54:27
over time. Maybe it's creating flashcards based
54:29
on student notes or things like this. That's
54:32
where it is really interesting to me for student
54:34
use. I'm not sure any tool is there yet. I'm
54:37
not sure what that actually looks like, how you
54:40
kind of mass the content to create the practice
54:43
questions and to actually structure it. Like
54:45
you would probably have to. There's some significant
54:48
amount of fine tuning, really some very specific
54:51
instructions and parameters for the AI to work
54:53
in. I'm not sure it's on a standard device like
54:56
a laptop or a phone. It would have to be some
54:58
other kind of thing. I don't even know what that
55:01
looks like either. But there are some hopeful
55:04
points. It's like we know how learning works.
55:06
So if we can create an AI system that engages
55:10
with those processes well and helps orchestrate
55:13
those processes so students can engage with them.
55:15
through time effectively, then, then you're onto
55:19
something. But not sure it's entirely there yet.
55:22
You could also do the same thing for teachers.
55:24
If you can embed professional development into
55:27
the tools that they use, such that, you know,
55:30
you go to use an assessment on EduAid and you
55:32
get explained why it might actually work or where
55:35
it might break down or what to look for, like
55:37
you actually help a teacher along with giving
55:39
them the tips and tricks, so to speak, as they
55:41
go through the process, then it might be effective
55:43
as well. There's a lot of great research going
55:46
on on this. There's some great companies and
55:48
a couple of great organizations based in some
55:50
schools of education that are actually looking
55:52
at creating an AI assistant for, like, teacher
55:55
professional development, where as you're working
55:57
on the application itself, you're learning about
56:00
not just the application, but the practice that
56:02
the applications apply to. Those are two interesting
56:05
areas. So that might be the sliver of hope as
56:07
we... we're nearing the the end and I've just
56:10
been just constantly saying no this is awful
56:12
there's no great answers there's there's the
56:14
few glimmers of hope I would put out yeah so
56:18
is you know if you were in charge of the AI world
56:22
right what's the what's the future path look
56:27
like you know an AI changes so fast so I don't
56:29
want to say like 10 years down the road I just
56:31
want to say like two to three years from now
56:34
If you're in charge of AI systems for the entire
56:37
universe, what's the ideal path we would all
56:41
be going down? Oh, gosh. I wouldn't even know
56:46
where to start. The cynical answer would be distribution,
56:51
distribution, distribution. I need my AI system
56:53
to be in front of everyone, so I need to create
56:55
this application that has business uses, enterprise
56:59
uses, as well as consumer uses. I'm not sure
57:02
it's going to just be a standalone chat bot because
57:04
everyone else is producing chat bots. Now, if
57:06
I said I want to direct humanity towards best
57:09
possible outcome for humanity and not my stockholders
57:12
as Lord Commander of my AI company. That's a
57:19
much tougher question to answer. I mean, clearly,
57:23
I'm not I'm not qualified for the job. You see
57:27
it being down down a path towards equity. Do
57:29
you see it being a path? towards, you know, human
57:34
centered approaches to things and not letting,
57:37
you know, the Terminator robots dominate all
57:40
of our lives, you know, those sort of things.
57:45
The problem with any new technology is that everybody
57:49
who's living through that technological revolution
57:52
will make a prediction about what it will do.
57:54
And those are almost assuredly always wrong.
57:57
Here's why I'm saying this. It's going to open
58:01
up things in ways that we can't even imagine
58:02
simply by existing, right? The way that this
58:07
system now takes, say code and enables people
58:12
to program in natural language can open up avenues
58:17
for creativity that I don't think wherever before
58:20
possible. So what we might see is now a teacher
58:24
might not have to buy educational technology.
58:27
They could log into their computer. They can
58:29
log into Codex and they can spin up an EdTech
58:32
application for personal use in their classroom
58:34
without needing to pay any other third party
58:37
money to do it. That would be the ultimate best
58:40
path forward is goodbye software companies, goodbye
58:44
software as a service, goodbye troublesome subscriptions
58:46
where we don't own anything and pay everybody
58:49
monthly rents to use their online device, right?
58:53
Now you take back some form of ownership because
58:55
you're able to create these things yourself.
58:58
That's why you see a lot of software engineers
59:00
perhaps getting laid off because really the killer
59:02
application of a large language model might be
59:04
its ability to code. If that's the case, students
59:08
should probably still study coding because you
59:10
got to be able to troubleshoot and things go
59:12
wrong and you need experience. I do worry about
59:15
the state of entry level programmers then because
59:19
you need an entry level job to build up the skills
59:21
to be able to do all of the debugging. If you're
59:24
only hiring senior engineers to debug the AI,
59:27
then you kind of have a pipeline breakdown there.
59:29
So that's an issue. But the for the sake of a
59:32
teacher, I think the proposition in, say, two
59:35
years of being able to spin up an application
59:38
for your class that is going to live locally
59:41
on your device is pretty appealing. Or for schools
59:44
to take all that data they're sitting in from
59:46
all those tests, from all those assessments that
59:48
you get the year after you take the assessment.
59:50
So you don't really have a great benchmark for
59:52
the students in your classes. I still don't understand
59:55
the testing cycle in that way, but being able
59:57
to, say, create some simple data dashboards to
59:59
be able to understand and make sense of that
1:00:01
data and be able to actually plan on top of it
1:00:04
is also powerful. I mean, these AI coding assistants
1:00:07
are great at spinning up dashboards. You give
1:00:09
it access to your folder with the relevant data.
1:00:12
You say you want a dashboard to display it in
1:00:14
X, Y, and Z ways. It can spin that up pretty
1:00:16
quickly. Those are the most interesting use cases
1:00:18
for administrators and for teachers in the short
1:00:21
term. From there, I'm not sure how things change.
1:00:24
It could blow up a good portion of the world
1:00:26
economy. It could show that the markets are giving
1:00:30
very mixed signals. You're seeing some negative
1:00:34
market actions by people saying, OK, AI is not
1:00:38
really delivering on what it said. And that's
1:00:41
a problem. So that's where the bubble comes from.
1:00:43
People talk about the AI bubble. And on the other
1:00:45
hand, you're seeing some market activity that
1:00:46
suggests people are scared because of the AI's
1:00:49
effect on the job market. It's affecting hiring
1:00:53
practices and it's affecting jobs, but it's not
1:00:56
having an effect So the bubbles going to it's
1:00:58
like we have two conflicting ideas if it's having
1:01:01
such a direct impact on the market on hiring
1:01:04
practices on talent Then why are we? Underestimating
1:01:08
it on the one side of the market saying it's
1:01:09
not delivering on promises and on the other side.
1:01:11
We're saying it's over delivering We're not hiring
1:01:13
we're having to lay people off and look at all
1:01:15
these new cost savings and it's very disruptive
1:01:17
So that's why I'm not comfortable answering your
1:01:21
question entirely other than saying You ever
1:01:25
see that Monty Python skit? It was like the the
1:01:27
boring prophet and he's like someone will have
1:01:29
an idea that they are unhappy tomorrow and it's
1:01:32
like No one cares about the boring prophet, but
1:01:36
things will change they might change for good
1:01:38
and they will also change Poorly depends on who
1:01:42
you are and what your job is that day I really
1:01:46
like your approach. I really like your approach
1:01:48
to that question. And I think that's a pretty
1:01:51
insightful answers. And I think we just do need
1:01:54
to let it play out and see where it's going to
1:01:55
go. That's contrary to what you hear. Everyone
1:01:58
says, oh, you have to move fast or you're going
1:02:00
to be in the path. No, time will keep flowing.
1:02:02
You're not going to lose out. You're not going
1:02:04
to miss anything. I mean, if someone would have
1:02:07
told you 20 years ago now, hey, maybe hold off.
1:02:12
on making that Facebook account for your 13 -year
1:02:14
-old. No one's ever going to say that was terrible
1:02:17
advice these days. They're going to say, oh,
1:02:18
that would have been quite prudent, pretty smart,
1:02:21
right? I hated that my parents didn't let me
1:02:23
have a cell phone until I was a senior in high
1:02:25
school. I felt like a loser. I'm like, oh my
1:02:27
gosh, a girl I was talking to had to text my
1:02:30
dad once. Like, that was embarrassing, right?
1:02:32
Awful. But in hindsight, now I'm like, you know
1:02:35
what? I can actually set my phone in another
1:02:37
room without having a nervous breakdown. Maybe
1:02:39
it was a good thing. So I'm not sure. pausing
1:02:42
and not using a technology is going to lead you
1:02:45
in the past, right? They keep saying these things
1:02:47
are so intuitive, anyone can pick them up anyway.
1:02:49
So I don't know what the rush is. Yeah, I think
1:02:54
that's a good kind of segue to lead us into kind
1:02:57
of our last little segment that we have here,
1:02:59
which is a lightning round. David and I will
1:03:04
just kind of alternate asking some rapid fire
1:03:07
questions. You know you can respond with one
1:03:10
word or you know a paragraph or sentence, whatever
1:03:12
you think it needs. I need the limits. I tend
1:03:15
to ramble so. No, never noticed that. Yeah, so
1:03:22
we'll have a few and then we'll get into a this
1:03:25
or that kind of thing and then wrap up here.
1:03:27
So David, you want to kick us off with the first
1:03:29
one? Yeah, so what do you miss most about being
1:03:32
in the classroom? Actually being able to connect
1:03:37
with humans. on a personal level and not through
1:03:40
my computer. Like actually going out and working
1:03:45
with students, being able to pick up on it's
1:03:49
much more than just the words coming out of our
1:03:50
mouth. It's facial expressions. It's the it's
1:03:52
your body language. It's like a student comes
1:03:54
in slumping and clearly maybe something was going
1:03:57
on and then you have a great lesson and they
1:03:58
come out of the class energetic and they pick
1:04:00
up on something. It's like, OK. You get a bit
1:04:02
of energy from that, too, because you saw that
1:04:04
kid's day change like right in front of your
1:04:06
face or someone's going through some sort of
1:04:09
interpersonal problem with other students. And
1:04:11
you work that out. You do conversation and kind
1:04:14
of reconcile that situation. You get some energy
1:04:16
out of that, too, because it's like, OK, I helped
1:04:17
this this kid figure out a personal situation
1:04:20
that they didn't have the social skills to navigate
1:04:22
before. But now they're figuring out how to work
1:04:24
their way through it. I mean. Teaching is very
1:04:27
rewarding. This is rewarding because in theory
1:04:29
I'm helping hundreds of thousands of teachers,
1:04:31
but when you get to that scale I don't actually
1:04:34
see those teachers I don't work with those teachers
1:04:36
one -on -one like I used to and I miss a lot
1:04:38
of that um Human one -on -one interaction. I'm
1:04:41
just like stuck in the cave over here Rambling
1:04:45
my computer screen Yeah, and on the opposite
1:04:49
side, what are you grateful for not having to
1:04:52
do anymore as a classroom teacher I mean As a
1:04:56
classroom teacher, I was always thinking like
1:05:00
big about the system and about the state of public
1:05:02
education and research on teaching and learning,
1:05:06
but you never really have a lot of time to do
1:05:08
that. That's like sitting there, you know, scarfing
1:05:11
down lunch, just kind of daydreaming about there
1:05:13
could be a better way. And it's like, yeah, maybe
1:05:16
pal, but what are you going to do about it? Now
1:05:18
I actually get to think about those questions
1:05:20
seriously. I get to think about. the state of
1:05:24
education and about technology and what interesting
1:05:27
things happen when these things collide and actually
1:05:29
have a hand in perhaps shaping folks interactions
1:05:33
with a new technology. Hopefully if I say things
1:05:36
loud enough and maybe intimidate some of my competitors,
1:05:39
change the state of this educational technology
1:05:41
for the better, right? If they're like, oh, he's
1:05:43
doing evaluations and that's getting some views,
1:05:46
we better do evaluations ourselves. It's like,
1:05:48
well then, you know. the tide rises all boats
1:05:51
or something like that, right? That's the idea.
1:05:54
So even if the company itself long term isn't
1:05:58
like a wild success, if I can in some way influence
1:06:02
the development of educational technology writ
1:06:04
large and help all teachers, then I did some
1:06:07
good with my day. How about if you were to give
1:06:09
a Ted talk and it can't be about education or
1:06:12
technology, what would the topic of that Ted
1:06:15
talk be? I mean, I think I already have. I've
1:06:17
been giving a TED talk all morning. Sorry about
1:06:20
that. That's education and technology. We got
1:06:22
to go outside those places. Look, I'm looking
1:06:24
at my bookshelf right now, like seeing what I
1:06:26
read about a lot lately. I mean, outside of education
1:06:31
and technology, I'm not sure I'm much of anything.
1:06:35
If I were to guess at looking over your shoulder,
1:06:38
it would be like plants. Oh, maybe horticulture,
1:06:43
perhaps. Although I'm kind of terrible at it
1:06:46
right now. I got two plants fighting for their
1:06:48
lives right now. I would do a dangerously close
1:06:55
to education. I'll give you that. But the importance
1:07:00
of classical thought. um how the modern world
1:07:04
has kind of gotten a little ugly in my opinion
1:07:06
like everything's all about efficiency but nothing
1:07:08
looks pretty or nice anymore right that kind
1:07:11
of bugs me out quite a bit it's like not everything
1:07:14
has to be minimalist design minimalism is not
1:07:17
always cool sometimes it's boring giver i'm looking
1:07:20
for a house right now if you go on zillow like
1:07:22
the number of like really cool old looking houses
1:07:24
and go inside the building and they redid it
1:07:27
so everything's just white and black and there's
1:07:29
no color in the house whatsoever and nothing
1:07:31
is exciting. I kind of get we could do better
1:07:34
as a modern world. You don't have to sacrifice
1:07:36
beauty for efficiency. You can have the two things
1:07:38
live together. That'd be my Ted talk for sure.
1:07:41
Yeah. Now I'm passionate about it. Now I kind
1:07:43
of want to go do a Ted talk on architecture in
1:07:46
the modern world. Maybe it'd be a snooze fest.
1:07:50
All right. What piece of old school tech still
1:07:55
deserves respect? I mean, no one really thinks
1:07:58
about it as technology, but paper, pencil. This
1:08:02
is the most enduring pieces of technology we've
1:08:04
had. I mean, we still use them in classrooms.
1:08:08
We still, I have four monitors on my desk right
1:08:11
now, but I still have paper and pencil notes.
1:08:13
I mean, we don't think about it, but we should
1:08:15
because paper's ultimately like one of the most
1:08:19
ingenious inventions that we've ever had. Yeah,
1:08:22
it's kind of funny. I just thought about this
1:08:23
but like yeah I was on the phone the other day
1:08:26
and they had to give me a number or something
1:08:28
and They'd never say you got a computer nearby.
1:08:31
It's always you got a pen. Yeah exactly Exactly
1:08:35
you think of that All right, so we're gonna give
1:08:38
you just some rapid -fire and it's this or that
1:08:41
you got to pick one or the other. Okay? Structured
1:08:44
lesson plan or open -ended project structured
1:08:47
lesson plan Road trip put playlist or true Crime
1:08:51
podcast. Playlist. It would have been podcasts
1:08:56
like 10 years ago, but I'm kind of off the true
1:08:58
crime now. Sure. AI for ideation or AI for feedback?
1:09:03
Feedback. Always start with the human idea. Would
1:09:06
you rather possess super strength or super intelligence?
1:09:11
Super intelligence. I mean, then I could just
1:09:13
build an awesome Iron Man suit and I have super
1:09:16
strength, you know. Sure. Rubric -driven output
1:09:20
or creative output? Depends on the context. I
1:09:26
didn't pick either one, but it really does depend.
1:09:28
Both work. We're easy going. We let you have
1:09:31
that. Would you rather cook at home or order
1:09:34
takeout? Cook at home. David just added this
1:09:39
one. He snuck this one in. Steelers or Eagles?
1:09:42
Oh, Steelers. I'm from Pittsburgh. And really
1:09:46
I'm a Dallas Cowboys fan. My uncle is an Eagles
1:09:48
fan, but I can't give him the satisfaction. So
1:09:50
Steelers 100%. OK, what about Aaron Rodgers?
1:09:55
Because I I live in Minnesota, but I'm from Wisconsin,
1:09:58
so I'm OK. I mean, so that's where that comes
1:10:01
from. You're like one of those guys who like
1:10:03
McCarthy. You're in Rogers. I love those guys.
1:10:06
But when they play for another team, it must
1:10:08
be that you don't want to leave in Cleveland.
1:10:09
It's like you went and played for Miami. Get
1:10:11
out of here. Yeah. And what about being at conferences?
1:10:16
Is it more important to have those hallway conversations
1:10:19
or the keynote sessions? I mean, definitely it's
1:10:22
the one -on -one interactions you have with people.
1:10:24
I mean, that's how we met. It wasn't a keynote.
1:10:26
It was chatting casually at a conference. Those
1:10:29
are much more rewarding than seeing whatever
1:10:32
big name they got to come speak. They can usually
1:10:35
be good. I don't know. Some keynotes. You can
1:10:40
always tell I mean if you're sitting in the conference
1:10:41
look around if everyone's on their phone You
1:10:43
know the quality of that keynote. I Shouldn't
1:10:45
give a keynote right because I'd go on for three
1:10:47
hours. We'd all be sitting there Like is this
1:10:49
guy gonna stop? I mean, he's really going But
1:10:53
I try to speak with passion. So, you know, they'll
1:10:55
pick up their heads from their phones. They'll
1:10:57
have to listen That's the idea. But anyway, yeah,
1:10:59
sorry side conversation No, and this is a good
1:11:04
follow -up question because if you were to give
1:11:06
a keynote presentation tomorrow What would be
1:11:09
your walk -up song that you would have to play
1:11:11
before you go on stage? I Have I actually have
1:11:16
to give a keynote in March. I'm sorry in April
1:11:19
early April and I've been thinking about this
1:11:21
very question It's got to be something It can't
1:11:26
be boring, right? So I used to I also box and
1:11:29
I always I always think about the walk -out song
1:11:30
What is it gonna be because it's got to be something
1:11:32
great I tried once to What's the name of that
1:11:40
song? It's a remix the the beat is from the Godzilla
1:11:44
theme and then and then and then And then there's
1:11:50
like a beat my wife said I wasn't muscular enough
1:11:52
for that song. So 21st century boy by What's
1:11:58
the name of that band? I forget the name of the
1:12:00
band, but it's a great song Everyone should go
1:12:02
listen to it right now and be like, oh, yeah,
1:12:04
that's one hell of a walk -up song I'll look
1:12:07
it up T Rex that's the name of the band They
1:12:12
did We might have to look that up and play that
1:12:16
as an outro or something here What's your guys's
1:12:19
intro for this podcast that was going pretty
1:12:21
hard I like like that four on the floor techno
1:12:23
be like I was nodding my head I was playing every
1:12:26
breast you're still remembering that from an
1:12:28
hour ago No, you know, Thomas, speaking of that,
1:12:33
you know, we're just really appreciative to be
1:12:35
able to have you on the show. And we had a lot
1:12:37
of great conversation and really want to thank
1:12:39
you for sharing your time with us. But we want
1:12:43
to give you a moment here, too. You know, I know
1:12:45
you mentioned maybe have a keynote address that's
1:12:47
coming up or is there anything you want to share,
1:12:50
anything you want to promote in our last little
1:12:53
bit that we have here or or even, you know, like,
1:12:57
how could people reach out to edge aid if they're
1:12:59
interested in learning more? Yeah, I mean, if
1:13:01
I was going to promote anything, it would probably
1:13:02
be, go check out Edu8. That would be, maybe it'd
1:13:06
be worth your time if you're listening and you're
1:13:07
looking for something to do. Nice way to spend
1:13:09
an afternoon. But yeah, seriously, you can go
1:13:12
to Edu8 .ai. All the contact information's right
1:13:15
there. There's a Contact Us button. Those emails
1:13:17
go out to our team, so I get access to those.
1:13:20
You can reach us right there at contact .edu8
1:13:23
.ai. All the social medias as well, we're on
1:13:26
them. We have a nice Facebook community with
1:13:28
like 2000 teachers right now. Nice close Nick
1:13:31
group, um, share lesson plans. And we have like
1:13:34
contests where you can create something on EduAid
1:13:37
and share it. And we do giveaways and gift cards
1:13:40
and all kinds of fun stuff over there. So yeah,
1:13:42
I mean, check out EduAid AI. That's kind of the
1:13:45
home base. You'll see our blog, all of our socials
1:13:47
from there. You can get ahold of us as well.
1:13:49
And, um, I might see some of you on the conference
1:13:51
circuit. I don't do many of those. I try to batch
1:13:53
them together. So like April is my conference
1:13:56
month. I'll be away from my house for like 20
1:13:58
days just traveling around. So I'll see you.
1:14:02
I'll see some of the folks there. I hope. Perfect.
1:14:05
Well, thanks again. We really appreciate having
1:14:07
you. Thank you so much. Thanks Thomas. Thanks
1:14:15
for listening to Integrate This. Want to take
1:14:18
your learning further? Head over to our YouTube
1:14:20
channel for quick tutorials, deep dives, and
1:14:24
more EdTech tips you can use right away. Don't
1:14:27
forget to subscribe and catch new podcast episodes
1:14:30
every Thursday on Apple, Spotify, or wherever
1:14:34
you get your podcasts. Until next time, keep
1:14:37
exploring, keep innovating, and keep integrating
1:14:41
this.