August 23, 2023 • 29 mins
Join us as we continue our summer mini-series where we discuss how to utilize artificial intelligence (AI) as a teaching tool. In this episode of Math Teacher Lounge, we sit down with educator Kristen Moore to discuss how she is using Chat GPT as a tool to revolutionize student learning and streamline lesson planning.
Listen today and don’t forget to grab your MTL study guide to track your learning and make the most of this episode!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Kristen Moore (00:00):
If you're a skeptic, you wanna start by
testing ChatGPT for what itknows and understands.
Dan Meyer (00:05):
Hey folks. Welcome back to Math Teacher Lounge.
It's the summer jam summersession. I'm your host, Dan
Meyer, and I am without myesteemed co-host, Bethany
Lockhart Johnson. She'll beback in the fall, bringing you
all of her insight and commentsand commentary. I've been
really excited this summer tochat a bit with guests who know
a lot more than me aboutartificial intelligence, about
(00:26):
these large language models.
ChatGPT Khanmigo. Do any ofthese keywords resonate for you
folks? I'm hearing lots aboutthem in my world, in education
technology. Lots of huge claimsabout them. I've been in rooms
with people who have made someclaims about how we might not
have teachers in 10 years,might not need 'em, or claims
about how every kid with theircomputer could have a free,
(00:48):
world-class tutor — some claimsthat I have found a bit
outlandish. And , in my lastepisode, I chatted with
Jennifer Carolan, a partner atReach Capital, one of the folks
who invest money in companiesworking in t his space. And she
o ffers one kind of perspectiveon AI and education, and I'm
enormously thrilled to have onthe podcast today, K risten
(01:10):
Moore, t o chat about, at theground level, at the place
where this technology actuallymeets teachers and students, w
hat it can do. K risten Mooreis a high school math teacher.
Let's welcome her to the show.
Hey, K risten, thanks for beinghere.
Kristen Moore (01:21):
Hey, Dan, thanks so much for having me. And I am
a big fan of AI. I don't thinkit's going to replace us, but I
think it's going to be anamazing tool to support us. So
I'm excited to have thisconversation with you.
Dan Meyer (01:32):
Likewise. I've chatted with you for like five
minutes and can already tellthat AI could not do what you
do. So I'm excited to dig inmore to that. Can you give us a
little bit of your backstory,you know, kinda a brief tour
through your, your spin througheducation, teaching? Like, how
did you get into looking — Imean, this technology's been
around for like, I don't know,six months or something. It's
not been a ... but what firstgot you into the tech, and
(01:55):
thinking about how could thisbe useful for me as a high
school math teacher?
Kristen Moore (01:59):
Yeah, absolutely. So I've always kind
of been into more innovativeteaching models and teaching
practices, but when I firstheard about ChatGPT, I really
kind of thought of it as thePhotomath equivalent for
English teachers. Like, it wasgoing to write everybody's
English essay and they weregoing to be grumpy about it. So
I completely wrote it off tobegin with. But then my
colleague kept talking about itat lunch and the interesting
(02:20):
things he was doing with it asa teacher for English. And I
was like, all right , well,let's see what we can do to
bring math teachers into thatfold. And really, I just
started with like, how canmathematics be used to model
polynomials in real life? Andtrying to figure out how I
could make that connection formy students. And that just led
me down the rabbit hole. And Ihave been using AI to co-plan
lessons with me and createtasks with me pretty much ever
(02:42):
since.
Dan Meyer (02:42):
Awesome. So, feeding that back to you a little bit:
Photomath is the tool that usesAI. It's been around for a w
hile to help students ... let'sjust charitably say, learn all
the steps and the answer to alot of different math problems.
Kristen Moore (02:55):
Yeah .
Dan Meyer (02:55):
And so now finally, y ou know, English Language
Arts has their own Photomath todeal with: Type in any essay
prompt and say, "Gimme fiveparagraphs. No. Make it sound
like an eighth grader. No, aseventh grader." And, th en
they're done.
Kristen Moore (03:09):
Yeah .
Dan Meyer (03:09):
So now, we're all even. And so you're in this and
you type in a math topic thathigh school math teachers often
struggle to make relevant andconcrete to students:
Polynomials. You know, I'mthinking the common
applications I see at the startof the chapter are like, "a
roller coaster," which, yes, itcan be modeled by a polynomial,
(03:30):
and no, t he students don't then ca re about polynomials more
after than be fore.
Kristen Moore (03:35):
No.
Dan Meyer (03:35):
First, can you just tell me, when you type in to Ch
atGPT, that prompt, let's talkthe te ch a l i ttle bit. I 'm
curious. I wan na un derstand al i ttle bit better and wh at
you know about it. When youtype in, "Tell me an
application for polynomials inthe real world," what do you
know about what's happeningbehind the scenes in ChatGPT
that's different from, say, ifyou type that into Google?
What's your take on the tech?
Kristen Moore (03:57):
So what I like about ChatGPT is that I can
really filter the answer . Sowhen it first comes up, it'll
come up with, you know, five orsix answers for me that really
probably don't relate a lot tomy students. I saw some things
in there on, like, Béziercurves, and I was like, my
students don't know what thatis. But what I like is that the
more you talk with ChatGPT —and I definitely think of it as
(04:18):
having a conversation withChatGPT — the more you're able
to get it to narrow down intothe things that you want to
see. So, it'll give you a bunchof prompts and you say, "OK,
I'm working with high schoolersin, you know, I'm in suburban
Michigan. And these are thingsthat are interesting to them.
Can we find real applicationsthat are more relevant to
that?" And the more you havethat conversation, the better
(04:40):
responses you're going to getfrom ChatGPT.
Dan Meyer (04:42):
That's super-interesting. And that
does seem like a keydistinction between Google and
ChatGPT. Google will — I guessit'll give the most clicked-on
links across the entire spaceof whatever. Which could
reflect, you know,perspectives, or have an idea
about, you know, contextualrelevance, that doesn't work
for students in suburbanMichigan.
Kristen Moore (05:02):
Right. Yeah.
Google is much more SEO-driven,and ChatGPT, you have the
ability to filter it down tomake it more YOU-driven.
Your-students-driven. And notdriven by who can write the
best long-tail keywords.
Dan Meyer (05:14):
Right, right.
Long-tail keywords, wow. So theconversational element feels
like it's a big difference. Canyou let us know, what were some
of the winners, after youhelped tune up the responses?
Do you recall what ChatGPT toldyou about polynomials? Math
teachers all over the listeningaudience are dying to know
here.
Kristen Moore (05:31):
We are dying to know! You know, one that always
comes up i s cryptography. AndI think that that's really
cool, especially now that we'retalking all about
cryptocurrency. So that'ssomething that I can get my
students to buy in on. We cango a little bit into the
history of it and how i t, youknow, started in mathematics.
And then my students are morebought in, because they wanna
make the next billion dollarson some sort of investment. And
(05:52):
so that connection with themgets us down that rabbit hole a
little bit further. Same thingwith something similar, like
rational functions. Boilingpoints are really good to model
rational functions. And so ifyou can get your students
figuring out a recipe to makemac and cheese or how to make
the optimal fondue, they'relike all of a sudden
super-bought-in. And they'relike, "OK, M s. M oore, I'll do
(06:14):
this math problem for youbecause you're right: I
actually want to know how andwhy I'm going to do this."
Dan Meyer (06:20):
I pride myself on making a pretty mean box of mac
and cheese, and I've alwayssuspected is because of my
mathematical background.
Kristen Moore (06:26):
Absolutely.
Dan Meyer (06:27):
Great. So I'm curious
who need a little morepersuading here to hop into the
process you're describing, thisconversation with an AI agent,
talking to a robot ... whatkind of encouragement can you
offer here? What's their firststep? The next step?
Kristen Moore (06:42):
Yeah. So I think if you're interested in using
ChatGPT or any form of AI likethis to help you with writing
lessons or creating differentprojects and products for your
classroom, I think if you're askeptic, you wanna start by
testing ChatGPT for what itknows and understands. So, I
like to go in there with aspecific standard in mind and
ask ChatGPT what it thinkswould be a good learning target
(07:05):
or what it thinks the rightsuccess criteria are. And as I
go through and I have thatconversation with it, I now
know that it's filtered to theexact same lens that I wanna
look at it through. So if Iwant a procedural lesson on,
you know, GSTR-8, we're workingwith similar right triangles
here, I can ask it that. And ifit's giving me the right kind
(07:26):
of learning targets, I knowthat it's going to give the
right kind of results that Iwant. And it's not just going
to give me something off thecuff or that I could have
randomly found in a Googlesearch on TPT, or, you know,
Google.
Dan Meyer (07:38):
Yeah. Great. OK.
That makes sense. So again, theconversation is so interesting
to me. Like, I resist the ideathat every student will have a
more capable teacher throughChatGPT, because I just think
the difference between what youdo, what you can do with a look
and a raised eyebrow, vastlyexceeds what ChatGPT can do.
(07:58):
And you're there in the room.
But I do love this idea thatteachers ... many teachers
don't have a teacher assistant.
Like, my first assistant, I got'cause I had taught this kid
for a while and she came in andstapled tests for me or
whatever. But what you'redescribing feels like someone
who's not just a student whohas graduated out of your
class, but rather someone who'sa fairly capable adult. Who has
(08:20):
read the entire internet,basically. And can be reasoned
with.
Kristen Moore (08:24):
Yeah.
Dan Meyer (08:25):
Which to me feels pretty exciting. I'm getting
excited thinking about it rightnow. Is that a useful way to
think about this? Or are thereother ways to think about this
for you?
Kristen Moore (08:33):
Yeah, I think so. I think it's like having a
thought partner. Which i sweird, because it's not having
its own actual thoughts. Butit's bouncing ideas off of you,
or y ou're bouncing ideas offof it. And between the two of
you, it sparks a lot of reallyinteresting ideas that you may
not have had, but you're stillthe teacher, you're still the
expert of your content and ofyour classroom. So you're going
(08:56):
to be able to say, "OK, I likethis idea, but let's tweak it
this way," so that it's bettergoing to meet the needs of the
students in your classroom. Sowhile AI is an amazing tool,
you've really gotta make surethat you're focusing in on your
expertise as well, and saying,"How can I use this to make
something better?" And not,"How can I just use this to
make something."
Dan Meyer (09:16):
Right. This "human i n t he loop" idea is one that I
see tossed around a lot, bypeople who are thoughtful about
AI and education and elsewhere.
"Keep yourself in the loop,"they say. Let me ask you:
You've mentioned, l ike at thestandard level, like "What's
the standard about?" or at thecontext level, like, "What's an
a pplication for this context?"You've also, it sounds like,
(09:38):
done some work with ChatGPT atthe project level, at much
larger levels. Multi-dayexperiences for students. Can
you tell us how in this dialogue wi th ChatGPT you wer e
ab le to pull out of th at amulti-day project for students?
Kristen Moore (09:56):
Yeah, absolutely. So, project-based
learning is a practice that isnear and dear to my heart. It's
something I've been doing forover a decade now in my math
teaching. And so my ability tocreate better projects and more
current projects and projects alittle bit quicker using this
AI was really important to me.
(10:16):
So when you go ahead and youfind some of those real-world
applications and you're able toconnect them to your students,
or you're able to say, "I knowthat my students would be
interested in that," then youcan go down and have it go
through like the entire PBLprocess, or if you don't wanna
go full-on PBL, you can help itconstruct a project with you.
So, for example, I've done aproject in the past all about
(10:38):
how art and math unite usacross cultures. And I've
always found that fascinating.
And I've done this project offand on for the last 10 years.
And so I've got a pretty goodbasis for it. But I wanted to
see what ChatGPT could do tohelp me take that to the next
level. And what I loved aboutthat connection was that
(10:59):
ChatGPT was able to not onlyidentify cultures that I've
never heard of before, getreally specific with that, but
it was able to tell me where Icould go and find these pieces
of art — whether it was localto me and my students at a
university or at an art museum,or if I could do a virtual
field trip, or where I couldfind it online. So ChatGPT was
(11:19):
able to make that experiencefor me even more present and
authentic to the situation thatwe're in now. And that's where
I think it can be reallypowerful — in that kind of
search engine/co-creator role.
Dan Meyer (11:32):
That's really interesting. Yeah. So , you
have your own cultural lensbackground, your own knowledge
and competencies, and if youwere to create the art project
on your own, it might reflectthat. I understand there are a
lot of concerns about like,"Who is ChatGPT listening to
and reading, and does thatreflect a diverse group of
people and cultures?" But itdoes sound like you were able
(11:53):
to tell this teachingassistant, "OK, here's our
premise, go to the library,whatever library you have
inside your brain, and go lookup other cultures for me." That
sounds pretty exciting.
Kristen Moore (12:03):
Yeah, it is really exciting that you can
use it, like you said, as thatteacher assistant, and it does
some of the heavy lifting foryou. We know that teachers
already have so much on theirplate, right? So if you can do
a little bit of that heavylifting, and sort through that
Google search faster than Ican, I'm all about that life.
Dan Meyer (12:20):
Definitely. Can you back up just a little bit? And
so the first prompt is ... youcame to it with art in math.
Walk us a bit through theconversation you had with
ChatGPT. Roughly speaking,where did that go next? You
have a rubric on that lessonplan; is that something you
came up with from PBLframeworks or t hat did ChatGPT
offer you a rubric?
Kristen Moore (12:42):
So you can ask ChatGPT to just design a
generic rubric for you. You cantell it the number of points
that you want it to assign. Youcan tell it what kind of
categories you want. So itreally depends on how
well-versed you already are increating your own rubrics, or
if you have a set standard thatyou have to use with your
school or your classroom. Ifyou're doing a mastery-based
(13:02):
approach, really you can haveChatGPT customize it however
you want. You just have to bereally good at inputting the
information that you're lookingat it to give back to you . The
better input you give, thebetter output you're going to
get back. And then you can takeit and you tweak it and you
know, you make it in the rightformat for you, because it
might come up with a littletable or it might come up with
a bunch of bullet points, butit's not going to be the same
(13:22):
format that you have to have.
But, like you said , when Istarted the project, I went in
and I asked about geometry andart and where they saw it
across cultures. And then Iasked, "OK, what kind of
transformations are mostvisible in those types of art?"
And as it gave me some morebackground information, then I
could ask it to help me form anessential question that's going
to drive the learning for thatwhole project. And from there,
(13:45):
I can chunk it down personallyinto what I know the best
milestones are going to be. Mystudents are going to need to
know what transformations are,first. What a translation is is
easier than what a reflectionis. Or what a rotation is. And
I can build on that. And as Iidentify those learning goals
and targets that I want mystudents to hit, I can ask
ChatGPT to help me with anumber of things, whether it's
(14:06):
generating extra ideas on ahands-on activity that I could
include in that piece, or ifit's asking it to help me
create a performance task or amini-rubric for that
understanding that builds ustowards that final product.
Dan Meyer (14:18):
You mentioned a couple parts that are really
exciting to me, especiallyextension problems. And making
context more accessible. Like,I think that I've seen, "Take
this article and rewrite it inlanguage that a sixth grader, a
fifth grader, a fourth gradercould understand," to me is
where I got those, thosegoosebumps, of like, "Wow, this
does feel like a kind ofsorcery." And the same for, a
(14:40):
student finishes my projectearly or needs more challenge;
what are some ways I couldchallenge that student? Just as
a first draft, obviously.
You're bringing your expertiseto it , to whatever ChatGPT
throws at you. But as a firstdraft, that sounds really
exciting. I would love to ...
so just declaring my biaseshere, I taught for six years,
and I left teaching in part — Ihaven't left education, but I l
(15:03):
eft teaching — in part becausecreating curriculum was a huge
part of what I was doing everyyear. And that became extremely
exhausting for me. And whatyou're describing sounds like a
way to make this lessexhausting. But I did go into
curriculum development becausemy belief is that every teacher
should not be reinventing acurriculum, or every district
should not be creating theirown core curriculum. There's
(15:25):
advantages to the coherence tohaving people who are like, "We
have 20, 30 people working onevery lesson in our
curriculum."
Kristen Moore (15:31):
Oh my gosh.
Dan Meyer (15:32):
Teachers don't have the time. They're working for
multiple days, a n army ofpeople, on these lessons. And
there's no way one teacher,even with ChatGPT, could create
that kind of thing. So what I'dlove to do is venture outside
of what we've talked about sofar, curriculum development.
And there are lots of peoplewho need this, because they
don't have a core curriculumthat works for them, right? But
(15:56):
I found myself wondering, whereelse could Kristen and I start
to brainstorm, justwhiteboard-sketch ideas, for
where we could put ChatGPT towork for us? And I went to
Deborah Loewenberg Ball, aprofessor in Michigan , U M
ichigan, University ofMichigan, who has researched
teacher practice for greateramount of time and a t more
(16:16):
depth than really just anybodyelse. And she has a list of 20
or so high-leverage practicesfor teachers. She describes
them — her group,TeachingWorks, describes them —
as used constantly and criticalto helping students learn
important content. That makesthem a high-leverage practice.
I'm looking through them; I'mlooking through them; and I'm
like, "Oh! There's Kristen!Right there! In t he one that
(16:39):
is listed as "designing singlelessons and sequences of
lessons." One of thehigh-leverage pra ctices.
That's what you've been talkingabout. So there's that.
Kristen Moore (16:49):
Absolutely.
Dan Meyer (16:50):
And there's others where I feel like we would
probably, definitely not askChatGPT to get involved. I'm
gonna read a couple of them. Iencourage you to check me to
check me if you think like,"Well, hold on now."
Kristen Moore (16:59):
Oh, I will .
Dan Meyer (17:00):
Because you've used this. You've used this tech in
ways that I definitely havenot. So, a couple of them ...
there's five where I was like,"Nah." So one was leading a
discussion. Building respectfulrelationship. Communicating
with families. Learning aboutstudents. Checking,student
understanding.
Kristen Moore (17:20):
I'm g onna say we can use ChatGPT for all of
those.
Dan Meyer (17:25):
Whoa! No! OK, awesome. I'm gonna learn
something here. Gimme anexample. What do you think?
Kristen Moore (17:29):
OK. So what was that first one? Leading a
discussion?
Dan Meyer (17:32):
Yeah, yeah, yeah.
Kristen Moore (17:33):
Right. So that's actually one of my favorite
things to do with ChatGPT, isto ask it to help me scaffold
and understand what studentsare going to think. And you can
ask ChatGPT to act as a ninthgrade student. What are
questions you would have aboutthis? And so it kind of gives
you an insight into what thestudents are thinking. And then
you can use it to help youcreate questions to guide that
(17:56):
dialogue. I will frequently goin and say, like, "Give me 20
questions about this topic,"and then I can select and
sequence those to kind of pluginto my classroom and lead that
conversation. Or get mystudents to raise those
questions by themselves.
Dan Meyer (18:11):
OK, hey, I love this. I'm learning a ton here.
So you're asking the chatinterface to play the role of
someone opposite you — astudent, in a discussion — and
getting questions that mightkeep that discussion moving.
Kristen Moore (18:27):
Exactly. And then you've got that as your
backup script so that when youknow you're in the heat of the
lesson, and you see the ideasthat are coming out from your
students and you know whatorder you wanna sequence them
in, you can feed in those extraquestions to keep that dialogue
going with your students andmake sure that you're reaching
your mathematical goals.
Dan Meyer (18:44):
This opens up a whole new category for me, of
teacher development. I don'tthink that while you're in the
discussion itself — and again,correct me — I'm not gonna
pause a discussion, wherestudents are in their seats and
we're talking together. I don'tk now if I'm gonna pause that
and go consult ChatGPT.
Kristen Moore (19:01):
Yeah .
Dan Meyer (19:01):
But what you're describing feels like the
preparation for the lesson. Therehearsal.
Kristen Moore (19:04):
Yeah.
Dan Meyer (19:05):
And maybe even there's an angle here for
teacher education and staffdevelopment. Is that a category
that we should think abouthere? Teacher development?
Kristen Moore (19:15):
Yeah. So you know, the five practices for
orchestrating productive mathdiscussions has that whole set,
where you're thinking about theanticipation of what your
students are going to bethinking. And that's where I
think you can start going toChatGPT and figuring out, what
are some common misconceptionsstudents might have? What are
some questions students mightraise in this? And how can I
(19:36):
lead a discussion to connectthem from this to that or to
clear up a misconception? So inthat process of preparing for
the lesson, that's when I wouldgo ahead and figure out how to
use those questions to guidethe conversation. Definitely
not in the middle of thelesson.
Dan Meyer (19:51):
I love that so much.
Actually, before this episode,I went to ChatGPT and I asked
it, "What are some common waysthat students would get the
question wrong?" Graph t heinequality, w hy, less than,
whatever? Yeah. Like a linearinequality, where you create a
line and you shade on one sideof it, make it dotted or solid.
Anyway. It was reallyinteresting. It was like,
students might shade above theline instead of below the line.
(20:12):
They might make the line dottedinstead of solid. They might
shade the wrong side of theline. So ChatGPT gave me the
same one twice, but whatever.
It was cool, you know? It's allgood between friends, and
you're free, you know. So I'mnot gonna knock it too much. As
a new teacher, you know?
Kristen Moore (20:26):
Oh, absolutely.
And those are things you knowas a teacher. But when you have
to plan for two, three, fourpreps, or you're teaching 150
kids, it's helpful to haveChatGPT there to feed you those
ideas and go, "Oh yeah, that'sright. They might do it this
way. Absolutely. How can Ipre-plan for that?" So it takes
some of that cognitive load offof you, and then you're going
(20:49):
to go work your magic in yourclassroom. Because it gave you
that idea and you're like,"You're right. Let's do that.
That's how I'm gonna handlethis."
Dan Meyer (20:54):
There's a lot of curricula that that will not
let you know the common wronganswers. Like, they tell you
the right answer, for sure. Inthe answer key. But that
question with the common wronganswer, it feels like it's like
folk knowledge, that you almosthave to develop, you know, like
through having taught it onceor twice. And being surprised
that students do what they do,that they have these ways of
(21:14):
applying old knowledge to newscenarios that don't quite
generalize. And so for teachersthat are teaching a course for
the first year or a new grade ,and you know, how intimidating
and terrifying that can be, itdoes seem like this m ight
offer practice, a practiceround, for instance, of a new
lesson.
Kristen Moore (21:32):
Absolutely.
Dan Meyer (21:33):
You really rocked my world with leading a
discussion, which I had justwritten off as, you know,
"ChatGPT can't, you know, mopmy floors. It can't do
everything for me. It's fine.
I'm not expecting it to." ButI'm convinced t hat it has more
applicability there than Ioriginally did. The other four,
just in case you wanna knockone more off? Building
respectful relationships,communicating with families,
(21:53):
learning about students,checking student understanding.
I'd be curious if you seeanything in there that's worth
talking about, from yourperspective.
Kristen Moore (22:02):
You know, I think that in general,
communication is something thatwe've seen a lot of. People are
using ChatGPT to help themcraft newsletters, and weekly
emails home, and how to write aletter of recommendation. I
know as a high school teacher,I get inundated with letters o
f rec. And using ChatGPT forsomething like that is really
(22:23):
helpful. I like that you can doit, with ChatGPT for
communicating home withfamilies, because you don't
always know the reading levelof everyone that you're talking
to. And so you can ask it tomake it in really friendly
terms. You can ask it to changethe tone so that when you read
it, it's coming off andpresenting the same way that
you are intending it to. Youknow, i t can be really
(22:45):
difficult to have aconversation through text or to
communicate effectively throughtext. So being able to run it
through ChatGPT a couple oftimes and really wordsmith it
out can be helpful.
Dan Meyer (22:54):
There we go. OK. Let me offer ... there's a few
areas where I thought, maybethere's something here. And so,
I imagine given what you didwith the areas I thought that
there was nothing there, you'llsee one of these four I imagine
will catch your eye in a bigway. And I'd love to just
tackle that one and see whatyou do with it. These, again,
are the high-leverage practicesas defined by Deborah
Loewenberg Ball's groupTeachingWorks, out of
(23:17):
University of Michigan. Sothese four include setting up
and managing small group work,attending to patterns of
student thinking — which, youknow what, I think we just, we
just tackled that one, in termsof like, "how do students come
to understand these differentideas?" So we got that one, I
think. So I'm gonna check thatone off my list as nailed by
(23:39):
Kristen Moore. And then theother two are implementing
norms and routines fordiscourse, selecting and
designing assessments. Assumingyou don't have a resource that
offers you assessments alignedto your curriculum. Like, I
kinda wondered if you had anyexperience — you've talked
about problems, standards,projects, assessments, another
(23:59):
huge area. What you got on anyof those?
Kristen Moore (24:01):
Oh my gosh.
Well, I think it can definitelybe really helpful in designing
those assessments, right? Ifwe're already sure that we can
use ChatGPT to give us a goodquality learning target and
success criteria, we know thatit's giving us what we want and
we can use it to identifystudent misconceptions. We can
also use it to help us kind ofscaffold what the problem
(24:23):
process is going to look like.
And, you know, people doassessments in all different
ways. If you're doing, forexample, mastery-based grading
and you want them to be, youknow, at the Level Three,
that's your target level, youcan use ChatGPT there to input
what your problem is and whatyour context is and say, "How
could we make this more of.
..." and then fill in what yourLevel Two criteria is. Can you
(24:43):
give me five different examplesfor this? And it can come up
with all these different waysthat you can then take and
tweak and suddenly you can makeyourself two, three, four
different versions of anassessment in about the same
amount of time that it would'vetaken you to craft just one.
Dan Meyer (24:58):
Yeah. That's really inspiring. I gotta ask
know, as well as anybody, thestate of teacher morale coming
out o f the pandemic, andvirtual teaching.
Kristen Moore (25:08):
Oh yeah.
Dan Meyer (25:08):
You know, there's endless surveys just
demonstrating that teachermorale is n ot in a great spot
right now. If teachers gotinvolved with this, it's
obviously helpful. Is ithelpful, do you think, in a way
that could improve teachermorale to a point where some
teachers might decide to stickaround longer than they would
otherwise? Like how, howeffective do y ou think this
(25:29):
might be? I'm just curious ifyou have any kind of sense of
... how excited are you whenyou do this kind of work? Is
this something that teachersshould get into, so as to
improve morale?
Kristen Moore (25:40):
So I think that ChatGPT has really helped me to
streamline a lot of theprocesses that I already use.
It helps me to get things donea lot quicker and more
efficiently. And I think themore that you can get done in a
quality manner, in an effectiveamount of time, the happier
you're going to be with yourwork life. You're able to have
(26:02):
more of that work-life balance.
So I d on't k now. I d on't know.
Dan Meyer (26:09):
It's hard to quantify, but it's good to hear
your sense of what it does, notjust for ... there are teachers
who will hear this and say,"Yes, I can spend more time
than I already do now usingChatGPT to make even better
lessons for students," and thatkind of teacher is one of my
favorite kinds of teachers andnot one that I'm super-worried
about, as far as stickingaround. I'm really worried
about the ones who are on theedge of burnout, and who need
(26:31):
to gain back time , forinstance. And I guess I just
want to name my own hope thatthis could give some hours
back, like you're describing.
Help teachers recover some timefor binge-watching and whatnot.
And keep them around. Doing thestuff that they can uniquely
do, the things that ChatGPTcannot and maybe will not ever
(26:53):
touch. So thank you so much,Kristen Moore, for hanging with
us and really altering myperspective on this tool,
broadening my horizons, and Ihope that of our listeners as
well. Me, personally, I wouldlove to keep tuned into your
learning journey with thesetechnologies and understand
what you're learning, in realti me, i f possible. Do you
have any kind of media you canshare, or can keep track of
your learning?
Kristen Moore (27:14):
Yeah, absolutely. Everyone can
connect with me on all thesocials. I'm at MooreThanJustX
Z — "Moore" with two Os —because I believe that math is
about so much more than justsolving for X. And you can also
catch me on my own podcast thatjust launched this last month,
The Modern Math Teacherpodcast. So I'll be sharing all
(27:34):
the things that are happeningin my classroom, and how AI is
transforming it in real ti me.
Dan Meyer (27:38):
Awesome. As the year kicks off. Thank you so much.
And we'll put all of that inthe show notes for folks to
tune into. Thank you again.
Kristen Moore (27:45):
Awesome. Thank you so much, Dan. This has been
great.
Dan Meyer (27:49):
Thank you so much folks for listening to my
conversation with KristenMoore, high school math teacher
and STEM instructional coach inMichigan. Check out the show
notes for a link to her recentpiece in Edutopia, titled
"Using ChatGPT in Math LessonPlanning." If you're looking
for more of what you heard herein this show, we'd love to hear
from you with your thoughts onthis episode, this series on AI
(28:11):
and ChatGPT more broadly. Maybeyou also have some tips of your
own to share. So hit us up. Letus know in our Facebook
discussion group, Math TeacherLounge Community, or on
Twitter, at MTLShow. You canfind more info on all of
Amplify's shows at our podcasthub. You can go to
amplify.com/hub. And next timeon Math Teacher Lounge, get
(28:32):
this, we are launching Season6. That's right. We're gonna be
focusing on the topic offluency. We've had some past
episodes on fluency that reallypacked it out, resonated with
listeners, and now we're goingto do a full season-long deep
dive . I am so excited tolearn. My co-host, Bethany
Lockhart Johnson, she hasactually changed my own work
(28:54):
with , the kids who live in myhouse with me on fluency. She
has excellent perspectives,especially at K–5. I'll be
thinking about the secondarylevel. We're gonna talk to
leading researchers on whatfluency is, why it matters ,
how it can help students andalso hurt students sometimes,
and how we go about it. Andthen we're also gonna delve
more into best practices fordeveloping fluency. It's going
(29:17):
to be a hoot. The best way toensure you don't miss out on
any of that is to subscribe toMath Teacher Lounge, wherever
you get fine podcast products.
Also, please do us a favor andrate and review us. We love to
read your reviews. That'swhat's ahead on Season 6 of
Math Teacher Lounge. Can't waitto see there in the Lounge.
Thanks so much for listening.
If you're a skeptic, you wanna start by
testing ChatGPT for what itknows and understands.
Dan Meyer (00:05):
Hey folks. Welcome back to Math Teacher Lounge.
It's the summer jam summersession. I'm your host, Dan
Meyer, and I am without myesteemed co-host, Bethany
Lockhart Johnson. She'll beback in the fall, bringing you
all of her insight and commentsand commentary. I've been
really excited this summer tochat a bit with guests who know
a lot more than me aboutartificial intelligence, about
(00:26):
these large language models.
ChatGPT Khanmigo. Do any ofthese keywords resonate for you
folks? I'm hearing lots aboutthem in my world, in education
technology. Lots of huge claimsabout them. I've been in rooms
with people who have made someclaims about how we might not
have teachers in 10 years,might not need 'em, or claims
about how every kid with theircomputer could have a free,
(00:48):
world-class tutor — some claimsthat I have found a bit
outlandish. And , in my lastepisode, I chatted with
Jennifer Carolan, a partner atReach Capital, one of the folks
who invest money in companiesworking in t his space. And she
o ffers one kind of perspectiveon AI and education, and I'm
enormously thrilled to have onthe podcast today, K risten
(01:10):
Moore, t o chat about, at theground level, at the place
where this technology actuallymeets teachers and students, w
hat it can do. K risten Mooreis a high school math teacher.
Let's welcome her to the show.
Hey, K risten, thanks for beinghere.
Kristen Moore (01:21):
Hey, Dan, thanks so much for having me. And I am
a big fan of AI. I don't thinkit's going to replace us, but I
think it's going to be anamazing tool to support us. So
I'm excited to have thisconversation with you.
Dan Meyer (01:32):
Likewise. I've chatted with you for like five
minutes and can already tellthat AI could not do what you
do. So I'm excited to dig inmore to that. Can you give us a
little bit of your backstory,you know, kinda a brief tour
through your, your spin througheducation, teaching? Like, how
did you get into looking — Imean, this technology's been
around for like, I don't know,six months or something. It's
not been a ... but what firstgot you into the tech, and
(01:55):
thinking about how could thisbe useful for me as a high
school math teacher?
Kristen Moore (01:59):
Yeah, absolutely. So I've always kind
of been into more innovativeteaching models and teaching
practices, but when I firstheard about ChatGPT, I really
kind of thought of it as thePhotomath equivalent for
English teachers. Like, it wasgoing to write everybody's
English essay and they weregoing to be grumpy about it. So
I completely wrote it off tobegin with. But then my
colleague kept talking about itat lunch and the interesting
(02:20):
things he was doing with it asa teacher for English. And I
was like, all right , well,let's see what we can do to
bring math teachers into thatfold. And really, I just
started with like, how canmathematics be used to model
polynomials in real life? Andtrying to figure out how I
could make that connection formy students. And that just led
me down the rabbit hole. And Ihave been using AI to co-plan
lessons with me and createtasks with me pretty much ever
(02:42):
since.
Dan Meyer (02:42):
Awesome. So, feeding that back to you a little bit:
Photomath is the tool that usesAI. It's been around for a w
hile to help students ... let'sjust charitably say, learn all
the steps and the answer to alot of different math problems.
Kristen Moore (02:55):
Yeah .
Dan Meyer (02:55):
And so now finally, y ou know, English Language
Arts has their own Photomath todeal with: Type in any essay
prompt and say, "Gimme fiveparagraphs. No. Make it sound
like an eighth grader. No, aseventh grader." And, th en
they're done.
Kristen Moore (03:09):
Yeah .
Dan Meyer (03:09):
So now, we're all even. And so you're in this and
you type in a math topic thathigh school math teachers often
struggle to make relevant andconcrete to students:
Polynomials. You know, I'mthinking the common
applications I see at the startof the chapter are like, "a
roller coaster," which, yes, itcan be modeled by a polynomial,
(03:30):
and no, t he students don't then ca re about polynomials more
after than be fore.
Kristen Moore (03:35):
No.
Dan Meyer (03:35):
First, can you just tell me, when you type in to Ch
atGPT, that prompt, let's talkthe te ch a l i ttle bit. I 'm
curious. I wan na un derstand al i ttle bit better and wh at
you know about it. When youtype in, "Tell me an
application for polynomials inthe real world," what do you
know about what's happeningbehind the scenes in ChatGPT
that's different from, say, ifyou type that into Google?
What's your take on the tech?
Kristen Moore (03:57):
So what I like about ChatGPT is that I can
really filter the answer . Sowhen it first comes up, it'll
come up with, you know, five orsix answers for me that really
probably don't relate a lot tomy students. I saw some things
in there on, like, Béziercurves, and I was like, my
students don't know what thatis. But what I like is that the
more you talk with ChatGPT —and I definitely think of it as
(04:18):
having a conversation withChatGPT — the more you're able
to get it to narrow down intothe things that you want to
see. So, it'll give you a bunchof prompts and you say, "OK,
I'm working with high schoolersin, you know, I'm in suburban
Michigan. And these are thingsthat are interesting to them.
Can we find real applicationsthat are more relevant to
that?" And the more you havethat conversation, the better
(04:40):
responses you're going to getfrom ChatGPT.
Dan Meyer (04:42):
That's super-interesting. And that
does seem like a keydistinction between Google and
ChatGPT. Google will — I guessit'll give the most clicked-on
links across the entire spaceof whatever. Which could
reflect, you know,perspectives, or have an idea
about, you know, contextualrelevance, that doesn't work
for students in suburbanMichigan.
Kristen Moore (05:02):
Right. Yeah.
Google is much more SEO-driven,and ChatGPT, you have the
ability to filter it down tomake it more YOU-driven.
Your-students-driven. And notdriven by who can write the
best long-tail keywords.
Dan Meyer (05:14):
Right, right.
Long-tail keywords, wow. So theconversational element feels
like it's a big difference. Canyou let us know, what were some
of the winners, after youhelped tune up the responses?
Do you recall what ChatGPT toldyou about polynomials? Math
teachers all over the listeningaudience are dying to know
here.
Kristen Moore (05:31):
We are dying to know! You know, one that always
comes up i s cryptography. AndI think that that's really
cool, especially now that we'retalking all about
cryptocurrency. So that'ssomething that I can get my
students to buy in on. We cango a little bit into the
history of it and how i t, youknow, started in mathematics.
And then my students are morebought in, because they wanna
make the next billion dollarson some sort of investment. And
(05:52):
so that connection with themgets us down that rabbit hole a
little bit further. Same thingwith something similar, like
rational functions. Boilingpoints are really good to model
rational functions. And so ifyou can get your students
figuring out a recipe to makemac and cheese or how to make
the optimal fondue, they'relike all of a sudden
super-bought-in. And they'relike, "OK, M s. M oore, I'll do
(06:14):
this math problem for youbecause you're right: I
actually want to know how andwhy I'm going to do this."
Dan Meyer (06:20):
I pride myself on making a pretty mean box of mac
and cheese, and I've alwayssuspected is because of my
mathematical background.
Kristen Moore (06:26):
Absolutely.
Dan Meyer (06:27):
Great. So I'm curious
who need a little morepersuading here to hop into the
process you're describing, thisconversation with an AI agent,
talking to a robot ... whatkind of encouragement can you
offer here? What's their firststep? The next step?
Kristen Moore (06:42):
Yeah. So I think if you're interested in using
ChatGPT or any form of AI likethis to help you with writing
lessons or creating differentprojects and products for your
classroom, I think if you're askeptic, you wanna start by
testing ChatGPT for what itknows and understands. So, I
like to go in there with aspecific standard in mind and
ask ChatGPT what it thinkswould be a good learning target
(07:05):
or what it thinks the rightsuccess criteria are. And as I
go through and I have thatconversation with it, I now
know that it's filtered to theexact same lens that I wanna
look at it through. So if Iwant a procedural lesson on,
you know, GSTR-8, we're workingwith similar right triangles
here, I can ask it that. And ifit's giving me the right kind
(07:26):
of learning targets, I knowthat it's going to give the
right kind of results that Iwant. And it's not just going
to give me something off thecuff or that I could have
randomly found in a Googlesearch on TPT, or, you know,
Google.
Dan Meyer (07:38):
Yeah. Great. OK.
That makes sense. So again, theconversation is so interesting
to me. Like, I resist the ideathat every student will have a
more capable teacher throughChatGPT, because I just think
the difference between what youdo, what you can do with a look
and a raised eyebrow, vastlyexceeds what ChatGPT can do.
(07:58):
And you're there in the room.
But I do love this idea thatteachers ... many teachers
don't have a teacher assistant.
Like, my first assistant, I got'cause I had taught this kid
for a while and she came in andstapled tests for me or
whatever. But what you'redescribing feels like someone
who's not just a student whohas graduated out of your
class, but rather someone who'sa fairly capable adult. Who has
(08:20):
read the entire internet,basically. And can be reasoned
with.
Kristen Moore (08:24):
Yeah.
Dan Meyer (08:25):
Which to me feels pretty exciting. I'm getting
excited thinking about it rightnow. Is that a useful way to
think about this? Or are thereother ways to think about this
for you?
Kristen Moore (08:33):
Yeah, I think so. I think it's like having a
thought partner. Which i sweird, because it's not having
its own actual thoughts. Butit's bouncing ideas off of you,
or y ou're bouncing ideas offof it. And between the two of
you, it sparks a lot of reallyinteresting ideas that you may
not have had, but you're stillthe teacher, you're still the
expert of your content and ofyour classroom. So you're going
(08:56):
to be able to say, "OK, I likethis idea, but let's tweak it
this way," so that it's bettergoing to meet the needs of the
students in your classroom. Sowhile AI is an amazing tool,
you've really gotta make surethat you're focusing in on your
expertise as well, and saying,"How can I use this to make
something better?" And not,"How can I just use this to
make something."
Dan Meyer (09:16):
Right. This "human i n t he loop" idea is one that I
see tossed around a lot, bypeople who are thoughtful about
AI and education and elsewhere.
"Keep yourself in the loop,"they say. Let me ask you:
You've mentioned, l ike at thestandard level, like "What's
the standard about?" or at thecontext level, like, "What's an
a pplication for this context?"You've also, it sounds like,
(09:38):
done some work with ChatGPT atthe project level, at much
larger levels. Multi-dayexperiences for students. Can
you tell us how in this dialogue wi th ChatGPT you wer e
ab le to pull out of th at amulti-day project for students?
Kristen Moore (09:56):
Yeah, absolutely. So, project-based
learning is a practice that isnear and dear to my heart. It's
something I've been doing forover a decade now in my math
teaching. And so my ability tocreate better projects and more
current projects and projects alittle bit quicker using this
AI was really important to me.
(10:16):
So when you go ahead and youfind some of those real-world
applications and you're able toconnect them to your students,
or you're able to say, "I knowthat my students would be
interested in that," then youcan go down and have it go
through like the entire PBLprocess, or if you don't wanna
go full-on PBL, you can help itconstruct a project with you.
So, for example, I've done aproject in the past all about
(10:38):
how art and math unite usacross cultures. And I've
always found that fascinating.
And I've done this project offand on for the last 10 years.
And so I've got a pretty goodbasis for it. But I wanted to
see what ChatGPT could do tohelp me take that to the next
level. And what I loved aboutthat connection was that
(10:59):
ChatGPT was able to not onlyidentify cultures that I've
never heard of before, getreally specific with that, but
it was able to tell me where Icould go and find these pieces
of art — whether it was localto me and my students at a
university or at an art museum,or if I could do a virtual
field trip, or where I couldfind it online. So ChatGPT was
(11:19):
able to make that experiencefor me even more present and
authentic to the situation thatwe're in now. And that's where
I think it can be reallypowerful — in that kind of
search engine/co-creator role.
Dan Meyer (11:32):
That's really interesting. Yeah. So , you
have your own cultural lensbackground, your own knowledge
and competencies, and if youwere to create the art project
on your own, it might reflectthat. I understand there are a
lot of concerns about like,"Who is ChatGPT listening to
and reading, and does thatreflect a diverse group of
people and cultures?" But itdoes sound like you were able
(11:53):
to tell this teachingassistant, "OK, here's our
premise, go to the library,whatever library you have
inside your brain, and go lookup other cultures for me." That
sounds pretty exciting.
Kristen Moore (12:03):
Yeah, it is really exciting that you can
use it, like you said, as thatteacher assistant, and it does
some of the heavy lifting foryou. We know that teachers
already have so much on theirplate, right? So if you can do
a little bit of that heavylifting, and sort through that
Google search faster than Ican, I'm all about that life.
Dan Meyer (12:20):
Definitely. Can you back up just a little bit? And
so the first prompt is ... youcame to it with art in math.
Walk us a bit through theconversation you had with
ChatGPT. Roughly speaking,where did that go next? You
have a rubric on that lessonplan; is that something you
came up with from PBLframeworks or t hat did ChatGPT
offer you a rubric?
Kristen Moore (12:42):
So you can ask ChatGPT to just design a
generic rubric for you. You cantell it the number of points
that you want it to assign. Youcan tell it what kind of
categories you want. So itreally depends on how
well-versed you already are increating your own rubrics, or
if you have a set standard thatyou have to use with your
school or your classroom. Ifyou're doing a mastery-based
(13:02):
approach, really you can haveChatGPT customize it however
you want. You just have to bereally good at inputting the
information that you're lookingat it to give back to you . The
better input you give, thebetter output you're going to
get back. And then you can takeit and you tweak it and you
know, you make it in the rightformat for you, because it
might come up with a littletable or it might come up with
a bunch of bullet points, butit's not going to be the same
(13:22):
format that you have to have.
But, like you said , when Istarted the project, I went in
and I asked about geometry andart and where they saw it
across cultures. And then Iasked, "OK, what kind of
transformations are mostvisible in those types of art?"
And as it gave me some morebackground information, then I
could ask it to help me form anessential question that's going
to drive the learning for thatwhole project. And from there,
(13:45):
I can chunk it down personallyinto what I know the best
milestones are going to be. Mystudents are going to need to
know what transformations are,first. What a translation is is
easier than what a reflectionis. Or what a rotation is. And
I can build on that. And as Iidentify those learning goals
and targets that I want mystudents to hit, I can ask
ChatGPT to help me with anumber of things, whether it's
(14:06):
generating extra ideas on ahands-on activity that I could
include in that piece, or ifit's asking it to help me
create a performance task or amini-rubric for that
understanding that builds ustowards that final product.
Dan Meyer (14:18):
You mentioned a couple parts that are really
exciting to me, especiallyextension problems. And making
context more accessible. Like,I think that I've seen, "Take
this article and rewrite it inlanguage that a sixth grader, a
fifth grader, a fourth gradercould understand," to me is
where I got those, thosegoosebumps, of like, "Wow, this
does feel like a kind ofsorcery." And the same for, a
(14:40):
student finishes my projectearly or needs more challenge;
what are some ways I couldchallenge that student? Just as
a first draft, obviously.
You're bringing your expertiseto it , to whatever ChatGPT
throws at you. But as a firstdraft, that sounds really
exciting. I would love to ...
so just declaring my biaseshere, I taught for six years,
and I left teaching in part — Ihaven't left education, but I l
(15:03):
eft teaching — in part becausecreating curriculum was a huge
part of what I was doing everyyear. And that became extremely
exhausting for me. And whatyou're describing sounds like a
way to make this lessexhausting. But I did go into
curriculum development becausemy belief is that every teacher
should not be reinventing acurriculum, or every district
should not be creating theirown core curriculum. There's
(15:25):
advantages to the coherence tohaving people who are like, "We
have 20, 30 people working onevery lesson in our
curriculum."
Kristen Moore (15:31):
Oh my gosh.
Dan Meyer (15:32):
Teachers don't have the time. They're working for
multiple days, a n army ofpeople, on these lessons. And
there's no way one teacher,even with ChatGPT, could create
that kind of thing. So what I'dlove to do is venture outside
of what we've talked about sofar, curriculum development.
And there are lots of peoplewho need this, because they
don't have a core curriculumthat works for them, right? But
(15:56):
I found myself wondering, whereelse could Kristen and I start
to brainstorm, justwhiteboard-sketch ideas, for
where we could put ChatGPT towork for us? And I went to
Deborah Loewenberg Ball, aprofessor in Michigan , U M
ichigan, University ofMichigan, who has researched
teacher practice for greateramount of time and a t more
(16:16):
depth than really just anybodyelse. And she has a list of 20
or so high-leverage practicesfor teachers. She describes
them — her group,TeachingWorks, describes them —
as used constantly and criticalto helping students learn
important content. That makesthem a high-leverage practice.
I'm looking through them; I'mlooking through them; and I'm
like, "Oh! There's Kristen!Right there! In t he one that
(16:39):
is listed as "designing singlelessons and sequences of
lessons." One of thehigh-leverage pra ctices.
That's what you've been talkingabout. So there's that.
Kristen Moore (16:49):
Absolutely.
Dan Meyer (16:50):
And there's others where I feel like we would
probably, definitely not askChatGPT to get involved. I'm
gonna read a couple of them. Iencourage you to check me to
check me if you think like,"Well, hold on now."
Kristen Moore (16:59):
Oh, I will .
Dan Meyer (17:00):
Because you've used this. You've used this tech in
ways that I definitely havenot. So, a couple of them ...
there's five where I was like,"Nah." So one was leading a
discussion. Building respectfulrelationship. Communicating
with families. Learning aboutstudents. Checking,student
understanding.
Kristen Moore (17:20):
I'm g onna say we can use ChatGPT for all of
those.
Dan Meyer (17:25):
Whoa! No! OK, awesome. I'm gonna learn
something here. Gimme anexample. What do you think?
Kristen Moore (17:29):
OK. So what was that first one? Leading a
discussion?
Dan Meyer (17:32):
Yeah, yeah, yeah.
Kristen Moore (17:33):
Right. So that's actually one of my favorite
things to do with ChatGPT, isto ask it to help me scaffold
and understand what studentsare going to think. And you can
ask ChatGPT to act as a ninthgrade student. What are
questions you would have aboutthis? And so it kind of gives
you an insight into what thestudents are thinking. And then
you can use it to help youcreate questions to guide that
(17:56):
dialogue. I will frequently goin and say, like, "Give me 20
questions about this topic,"and then I can select and
sequence those to kind of pluginto my classroom and lead that
conversation. Or get mystudents to raise those
questions by themselves.
Dan Meyer (18:11):
OK, hey, I love this. I'm learning a ton here.
So you're asking the chatinterface to play the role of
someone opposite you — astudent, in a discussion — and
getting questions that mightkeep that discussion moving.
Kristen Moore (18:27):
Exactly. And then you've got that as your
backup script so that when youknow you're in the heat of the
lesson, and you see the ideasthat are coming out from your
students and you know whatorder you wanna sequence them
in, you can feed in those extraquestions to keep that dialogue
going with your students andmake sure that you're reaching
your mathematical goals.
Dan Meyer (18:44):
This opens up a whole new category for me, of
teacher development. I don'tthink that while you're in the
discussion itself — and again,correct me — I'm not gonna
pause a discussion, wherestudents are in their seats and
we're talking together. I don'tk now if I'm gonna pause that
and go consult ChatGPT.
Kristen Moore (19:01):
Yeah .
Dan Meyer (19:01):
But what you're describing feels like the
preparation for the lesson. Therehearsal.
Kristen Moore (19:04):
Yeah.
Dan Meyer (19:05):
And maybe even there's an angle here for
teacher education and staffdevelopment. Is that a category
that we should think abouthere? Teacher development?
Kristen Moore (19:15):
Yeah. So you know, the five practices for
orchestrating productive mathdiscussions has that whole set,
where you're thinking about theanticipation of what your
students are going to bethinking. And that's where I
think you can start going toChatGPT and figuring out, what
are some common misconceptionsstudents might have? What are
some questions students mightraise in this? And how can I
(19:36):
lead a discussion to connectthem from this to that or to
clear up a misconception? So inthat process of preparing for
the lesson, that's when I wouldgo ahead and figure out how to
use those questions to guidethe conversation. Definitely
not in the middle of thelesson.
Dan Meyer (19:51):
I love that so much.
Actually, before this episode,I went to ChatGPT and I asked
it, "What are some common waysthat students would get the
question wrong?" Graph t heinequality, w hy, less than,
whatever? Yeah. Like a linearinequality, where you create a
line and you shade on one sideof it, make it dotted or solid.
Anyway. It was reallyinteresting. It was like,
students might shade above theline instead of below the line.
(20:12):
They might make the line dottedinstead of solid. They might
shade the wrong side of theline. So ChatGPT gave me the
same one twice, but whatever.
It was cool, you know? It's allgood between friends, and
you're free, you know. So I'mnot gonna knock it too much. As
a new teacher, you know?
Kristen Moore (20:26):
Oh, absolutely.
And those are things you knowas a teacher. But when you have
to plan for two, three, fourpreps, or you're teaching 150
kids, it's helpful to haveChatGPT there to feed you those
ideas and go, "Oh yeah, that'sright. They might do it this
way. Absolutely. How can Ipre-plan for that?" So it takes
some of that cognitive load offof you, and then you're going
(20:49):
to go work your magic in yourclassroom. Because it gave you
that idea and you're like,"You're right. Let's do that.
That's how I'm gonna handlethis."
Dan Meyer (20:54):
There's a lot of curricula that that will not
let you know the common wronganswers. Like, they tell you
the right answer, for sure. Inthe answer key. But that
question with the common wronganswer, it feels like it's like
folk knowledge, that you almosthave to develop, you know, like
through having taught it onceor twice. And being surprised
that students do what they do,that they have these ways of
(21:14):
applying old knowledge to newscenarios that don't quite
generalize. And so for teachersthat are teaching a course for
the first year or a new grade ,and you know, how intimidating
and terrifying that can be, itdoes seem like this m ight
offer practice, a practiceround, for instance, of a new
lesson.
Kristen Moore (21:32):
Absolutely.
Dan Meyer (21:33):
You really rocked my world with leading a
discussion, which I had justwritten off as, you know,
"ChatGPT can't, you know, mopmy floors. It can't do
everything for me. It's fine.
I'm not expecting it to." ButI'm convinced t hat it has more
applicability there than Ioriginally did. The other four,
just in case you wanna knockone more off? Building
respectful relationships,communicating with families,
(21:53):
learning about students,checking student understanding.
I'd be curious if you seeanything in there that's worth
talking about, from yourperspective.
Kristen Moore (22:02):
You know, I think that in general,
communication is something thatwe've seen a lot of. People are
using ChatGPT to help themcraft newsletters, and weekly
emails home, and how to write aletter of recommendation. I
know as a high school teacher,I get inundated with letters o
f rec. And using ChatGPT forsomething like that is really
(22:23):
helpful. I like that you can doit, with ChatGPT for
communicating home withfamilies, because you don't
always know the reading levelof everyone that you're talking
to. And so you can ask it tomake it in really friendly
terms. You can ask it to changethe tone so that when you read
it, it's coming off andpresenting the same way that
you are intending it to. Youknow, i t can be really
(22:45):
difficult to have aconversation through text or to
communicate effectively throughtext. So being able to run it
through ChatGPT a couple oftimes and really wordsmith it
out can be helpful.
Dan Meyer (22:54):
There we go. OK. Let me offer ... there's a few
areas where I thought, maybethere's something here. And so,
I imagine given what you didwith the areas I thought that
there was nothing there, you'llsee one of these four I imagine
will catch your eye in a bigway. And I'd love to just
tackle that one and see whatyou do with it. These, again,
are the high-leverage practicesas defined by Deborah
Loewenberg Ball's groupTeachingWorks, out of
(23:17):
University of Michigan. Sothese four include setting up
and managing small group work,attending to patterns of
student thinking — which, youknow what, I think we just, we
just tackled that one, in termsof like, "how do students come
to understand these differentideas?" So we got that one, I
think. So I'm gonna check thatone off my list as nailed by
(23:39):
Kristen Moore. And then theother two are implementing
norms and routines fordiscourse, selecting and
designing assessments. Assumingyou don't have a resource that
offers you assessments alignedto your curriculum. Like, I
kinda wondered if you had anyexperience — you've talked
about problems, standards,projects, assessments, another
(23:59):
huge area. What you got on anyof those?
Kristen Moore (24:01):
Oh my gosh.
Well, I think it can definitelybe really helpful in designing
those assessments, right? Ifwe're already sure that we can
use ChatGPT to give us a goodquality learning target and
success criteria, we know thatit's giving us what we want and
we can use it to identifystudent misconceptions. We can
also use it to help us kind ofscaffold what the problem
(24:23):
process is going to look like.
And, you know, people doassessments in all different
ways. If you're doing, forexample, mastery-based grading
and you want them to be, youknow, at the Level Three,
that's your target level, youcan use ChatGPT there to input
what your problem is and whatyour context is and say, "How
could we make this more of.
..." and then fill in what yourLevel Two criteria is. Can you
(24:43):
give me five different examplesfor this? And it can come up
with all these different waysthat you can then take and
tweak and suddenly you can makeyourself two, three, four
different versions of anassessment in about the same
amount of time that it would'vetaken you to craft just one.
Dan Meyer (24:58):
Yeah. That's really inspiring. I gotta ask
know, as well as anybody, thestate of teacher morale coming
out o f the pandemic, andvirtual teaching.
Kristen Moore (25:08):
Oh yeah.
Dan Meyer (25:08):
You know, there's endless surveys just
demonstrating that teachermorale is n ot in a great spot
right now. If teachers gotinvolved with this, it's
obviously helpful. Is ithelpful, do you think, in a way
that could improve teachermorale to a point where some
teachers might decide to stickaround longer than they would
otherwise? Like how, howeffective do y ou think this
(25:29):
might be? I'm just curious ifyou have any kind of sense of
... how excited are you whenyou do this kind of work? Is
this something that teachersshould get into, so as to
improve morale?
Kristen Moore (25:40):
So I think that ChatGPT has really helped me to
streamline a lot of theprocesses that I already use.
It helps me to get things donea lot quicker and more
efficiently. And I think themore that you can get done in a
quality manner, in an effectiveamount of time, the happier
you're going to be with yourwork life. You're able to have
(26:02):
more of that work-life balance.
So I d on't k now. I d on't know.
Dan Meyer (26:09):
It's hard to quantify, but it's good to hear
your sense of what it does, notjust for ... there are teachers
who will hear this and say,"Yes, I can spend more time
than I already do now usingChatGPT to make even better
lessons for students," and thatkind of teacher is one of my
favorite kinds of teachers andnot one that I'm super-worried
about, as far as stickingaround. I'm really worried
about the ones who are on theedge of burnout, and who need
(26:31):
to gain back time , forinstance. And I guess I just
want to name my own hope thatthis could give some hours
back, like you're describing.
Help teachers recover some timefor binge-watching and whatnot.
And keep them around. Doing thestuff that they can uniquely
do, the things that ChatGPTcannot and maybe will not ever
(26:53):
touch. So thank you so much,Kristen Moore, for hanging with
us and really altering myperspective on this tool,
broadening my horizons, and Ihope that of our listeners as
well. Me, personally, I wouldlove to keep tuned into your
learning journey with thesetechnologies and understand
what you're learning, in realti me, i f possible. Do you
have any kind of media you canshare, or can keep track of
your learning?
Kristen Moore (27:14):
Yeah, absolutely. Everyone can
connect with me on all thesocials. I'm at MooreThanJustX
Z — "Moore" with two Os —because I believe that math is
about so much more than justsolving for X. And you can also
catch me on my own podcast thatjust launched this last month,
The Modern Math Teacherpodcast. So I'll be sharing all
(27:34):
the things that are happeningin my classroom, and how AI is
transforming it in real ti me.
Dan Meyer (27:38):
Awesome. As the year kicks off. Thank you so much.
And we'll put all of that inthe show notes for folks to
tune into. Thank you again.
Kristen Moore (27:45):
Awesome. Thank you so much, Dan. This has been
great.
Dan Meyer (27:49):
Thank you so much folks for listening to my
conversation with KristenMoore, high school math teacher
and STEM instructional coach inMichigan. Check out the show
notes for a link to her recentpiece in Edutopia, titled
"Using ChatGPT in Math LessonPlanning." If you're looking
for more of what you heard herein this show, we'd love to hear
from you with your thoughts onthis episode, this series on AI
(28:11):
and ChatGPT more broadly. Maybeyou also have some tips of your
own to share. So hit us up. Letus know in our Facebook
discussion group, Math TeacherLounge Community, or on
Twitter, at MTLShow. You canfind more info on all of
Amplify's shows at our podcasthub. You can go to
amplify.com/hub. And next timeon Math Teacher Lounge, get
(28:32):
this, we are launching Season6. That's right. We're gonna be
focusing on the topic offluency. We've had some past
episodes on fluency that reallypacked it out, resonated with
listeners, and now we're goingto do a full season-long deep
dive . I am so excited tolearn. My co-host, Bethany
Lockhart Johnson, she hasactually changed my own work
(28:54):
with , the kids who live in myhouse with me on fluency. She
has excellent perspectives,especially at K–5. I'll be
thinking about the secondarylevel. We're gonna talk to
leading researchers on whatfluency is, why it matters ,
how it can help students andalso hurt students sometimes,
and how we go about it. Andthen we're also gonna delve
more into best practices fordeveloping fluency. It's going
(29:17):
to be a hoot. The best way toensure you don't miss out on
any of that is to subscribe toMath Teacher Lounge, wherever
you get fine podcast products.
Also, please do us a favor andrate and review us. We love to
read your reviews. That'swhat's ahead on Season 6 of
Math Teacher Lounge. Can't waitto see there in the Lounge.
Thanks so much for listening.
Popular Podcasts
Las Culturistas with Matt Rogers and Bowen Yang
Ding dong! Join your culture consultants, Matt Rogers and Bowen Yang, on an unforgettable journey into the beating heart of CULTURE. Alongside sizzling special guests, they GET INTO the hottest pop-culture moments of the day and the formative cultural experiences that turned them into Culturistas. Produced by the Big Money Players Network and iHeartRadio.
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com