Greg Toppo: The promise and perils of AI in schools

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Alex Kotran (aiEDU) (00:01):
Hey everybody, I'm Alex Katran, I'm
the CEO of AIEDU and we're hereat AIEDU Studios DC Edition.
I'm very excited to have GregToppo, a reporter for the 74,
one of the leading publicationscovering the education space,

(00:21):
someone that I've talked toformally and informally over the
past year, two years, and Greghas been both a confidant and
sort of a co-learner as we tryto make sense of this space and
sort of how technology, andspecifically artificial
intelligence, is changingeducation or not, and so I'm

(00:44):
very excited to flip the scriptand put the mic in front of you
and give you a chance to talkabout you know, from the vantage
point of a journalist who'stalked to a really wide range of
folks in our space, what you'relearning, questions you're
asking.
But before we go into sort ofthe day-to-day you know this

(01:05):
sort of this AI moment, this AIzeitgeist I'd love to zoom out a
little bit and hear why don'tyou start with just sort of you
know what you're doing at the 74and then you know how did you
come to cover education?

Greg Toppo (The 74) (01:17):
Sure, sure Thanks for having me and, just
as a quick aside, mostjournalists hate being
interviewed, by the way, so youknow I'm good with it.
I think we're going to have agood time here.
I cover a couple of things atthe 74.
It feels like it keeps changing, but I cover basically the

(01:39):
intersection of AI and education, and I also have sort of a I
don't know if you'd call it likea side hustle at 74, but one of
the things we are veryinterested in is the future of
high school, and so one of thetopics that I spent a lot of
time on is writing aboutvisiting and thinking about and

(02:01):
writing about innovative highschools, and the two haven't
quite merged yet.
I expect we're going to beseeing more and more in the near
future.

I mean there's a lot of reporters
covering AI A lot of good ones.
There's a lot of good ones, anda lot more today than there
were three or four years ago Notthat many covering AI plus
education intersection.
How did you get to be doingspecifically sort of this like
sort of view of the future ofschool and innovation?

Yeah, well, I mean, I was interested in it
for a number of years.
I co -wrote a book in 2021 thatactually looked at AI and
education specifically,interestingly enough, actually
looked at AI and educationspecifically, interestingly
enough, what it was going to doto high school.
We were very interested in justsort of projecting forward what
all these technological marvels, if you will, were going to the

(02:58):
effect they were going to haveon education.
I had a co -author, myco-author and I.
He was an educator, he still isan educator.
He and I wrote kind of.
I thought it was a fun, reallyinteresting book.
We basically did a what if ahigh school principal fell

(03:18):
asleep in 2020 and woke up in2040 and then went and visited
his old school?
What would it look like?
You know, after all thesetechnological changes, we've
been seeing so sort of a Rip VanWinkle story, interspersed with
some reporting on what'shappening actually now and
different models of high schoolso big picture.

(03:39):
I've been thinking about thisfor a number of years.
I've been covering AI you knowAI daily for two years and
change at the 74.
And it really began where Ithink most of us started
covering it, which is November2022, when chat GPTs were burst

(04:01):
on the scene and people startedmessing with it and freaking out
about it and wondering what wasgoing to happen.
Um, and you know, decrying thedeath of the five paragraph
essay, et cetera, et cetera.
Um, that was when we as aninstitution started really
paying attention to it and whenI basically started saying you
know, okay, this is something weneed to cover full time.

Yeah To me cause we were chatting earlier.
You, you know, okay, this issomething we need to cover full
time.
Yeah to me because we werechatting earlier.
You mentioned that you'd evendid you do a story on GPT-2 or
had you just sort of.

In the book .
Actually, one of the thingsthat we looked at was and it was
actually not specificallyrelated to education, it was
just in 2019, I want to say theywere playing with OpenAI, was
playing with the ways you couldessentially make GPT-2 into kind

(04:51):
of an approximation of ajournalist, and actually the
chapter in the book where wetalk about it is where we talk
about, um, what's this going todo to journalism and the?
The example we gave, which, ofcourse, you could imagine even
more clearly now, was that itbasically just they gave it a

(05:12):
couple of words from a, you know, a real existing I can't
remember it was guardian storyor something like that and it
just basically just went to townand wrote the whole story I
remember yeah, and you was lotsof crazy stuff and it was even,
I think, one of the one of thethings they said at Luzon was
imagine a reanimated JFK.

(05:34):
you know, just patently crazystuff.

Yeah, I think GPT-2 was a time when
there was, I think, only acouple hundred people in the
world who had really had thechance to look under the hood of
these language models.
But for me, the zeitgeistmoment that ChatGPT ushered in,
because I've been in the AIspace since 2014.

(05:58):
Ai has been around I mean, youcan actually go quite far back
and some people argue the Turingcomputer.
But machine learning reachedtechnological maturity in
industry in the 2010s, so AI wasnot a new conversation.
There were plenty of peopleportending a lot of the stuff
that is now reality today.
It was sort of lived in therealm of futurism.

(06:20):
I think the challenge forartificial intelligence in part
was it was a concurrenttechnology with this suite of
emerging tech.
You know crypto, you knowblockchain, virtual reality,
much of which you know 3Dprinting the metaverse, all this
NFTs, nfts.

(06:45):
And while there's actuallyarguments that a lot of those
technologies are still nascentand will bloom in later years
and will transform society inways that people have predicted,
they were clearly a bit aheadof their time.
If we're being gratuitous andsome of them are probably just
like patently you know we'reoverhyped and over embellished,
and so I think there was thisfatigue from OK, the futurists
are at it again talking aboutsome crazy future.

(07:06):
I think people wereunderstandably cynical and
incredulous at some of theclaims that were being made.
Chatgpt put everything intotheir hands.
It made it tangible in a waythat it wasn't before.
I think in the abstract, peopleunderstood that there was an
algorithm curating the contentthat they view.
It's very hard to really wrapyour head around.

(07:27):
You know what that means.
You're not really using it.
You don't really understand youknow what the weights are or
even what that means when youhave a tool that's just as and
the user interface was reallypowerful, right.
You just ask a question, youpress enter and you get this
output and these turn, you getthis output, and these outputs

(07:50):
were, you know, astounding, Ithink.
Today we're sort of we are thebar continues to to increase, um
, but I think most people willbe able to tell you like exactly
the first thing that they putinto, whether it was chat, gbt
or gemini or something else.

Where were you and what were you doing at
the time?

yeah, yeah did you have?
You been in education yourwhole, your whole life um, I
actually sure um we were all ineducation, you know, the first
part of everybody's livesgenerally.

Exactly, I spent a long time in education.
When I graduated from college.
I actually got right intoteaching, actually was in a
teacher prep program the Mondaymorning after I graduated from
college, um, and stayed ineducation.

(08:29):
Stayed in, uh, elementary andsecondary secondary education
for about the next eight or nineyears, um, mostly, I have to
say, if I'm being honest,wanting to get the hell out of
it.
As soon as I got in, not quitesure how, I knew I wanted to
write, knew I wanted to dosomething different and about

(08:49):
you know, like I said, eight ornine years in I finally just
said, okay, I need to changewhat I'm doing.
And so started freelancing,doing lots of magazine pieces,
and finally got my first staffjob at a newspaper in New Mexico
at the great great paper stillexists, the Santa Fe New Mexican
, and really just loved it andhad the time of my life there.

(09:15):
Spent three years there and thenwent to Baltimore to work at
the AP.
Spent two years at the AP inBaltimore, two more years at the
AP in Washington.
Was there just in time forthings like the Clinton
impeachment and ended upcovering some of the Monica
Lewinsky stuff.
We actually, as luck would haveit, I was the only person in

(09:37):
the Baltimore Bureau who lived acouple of blocks from the
courthouse where the Linda Trippwiretap case was taking place,
so I ended up covering that Idon't know.
After the AP in Washington,spent 15 years at USA Today and
was the national educationwriter there.

(09:58):
During that time, wrote a bookin 2015 that looked at the
possibilities in learning gamesand really got kind of like a
graduate level education all myown, you know just self almost

(10:20):
entirely like self-taught, ifyou will.
Um, the kind of how that works,how that sector works gamified
learning yeah, yeah, um, andreally had the time of my life,
um.
by then, um, I was reallywanting to do something
different, was kind of sick ofthe you know nclb race to the

(10:40):
top standards, testing on and onand on um, and just wanted to
just explore something new anddifferent and weird, and that
was really.
That was probably the most funI've had in my career.
Um, and I'd love to actuallymaybe at some point talk about
the parallels between what I sawthen so now going on a decade
ago and what I'm seeing now withAI.

Yeah, so you, you found your way out.
You're covering nationalpolitics.
We were just talking to areporter yesterday from one of
the big publications, people ineducation.
I think sometimes there's a lotof stories of folks becoming
cynical.
Education is a tough place forpeople who are well-intentioned,
trying to do work and have animpact, and there's just a lot

(11:27):
of systemic challenges, um, butnational politics is its own uh
cynicism factory.
Um, so you went from one veryyou know uh place where a lot of
jaded people to back toeducation.
Um, but at the 74, you, I meanyour reporting is interesting
because it's a lot more, it'slonger form.
You're doing much more of adeep dive.

(11:47):
You're not like, I haven't seentoo many process stories from
you, and is that because of yourremit at the 74?
Is that something that you'retrying to be more intentional
about in terms of?
I mean, especially right now,there's so much to cover, you
know, and AI is not a narrowspace anymore.
I mean, how do you think aboutsort of your beat, as it were,

(12:12):
like what is the frame or sortof the aperture that you're
using?

Yeah, that's a great question and I
think, in a way, the DNA of the$74 million is that you know, on
the one hand, we're a nationalpublication, so we don't have to
be writing about the ClevelandSchool Board or the, or you know
what's happening at the, youknow the State Board of
Education in New Hampshire, sowe're we're not worrying about

(12:36):
those little processes unlessthey're interesting to us.
So so if there's somethinghappening on a local level that
we want to dip into, we can dipinto it.
The perfect example to me isthe story I wrote a couple weeks
ago about this case inMassachusetts with the student
who was accused of cheating on ahistory project using AI In a

(13:02):
way, a local story.
It was in a federal court inBoston, but we found it
interesting enough where we justsaid you know what we need to
write about this.

Yeah, I remember you reaching out to me
for my thoughts on that storyand shortly thereafter and it
was around the time that yourpiece had come out, I mean, it
started to really gain nationaltraction.
I don't know if it's actuallythe first, but it certainly
appears to be the first, atleast the first example of this
that has been talked about andcovered nationally.
Do you know, was this actuallythe first time someone has gone

(13:36):
to court challenging anaccusation of cheating because
they used Chad GBT?

Yeah, here's the thing you need to
know, because they use chat gpt.
Yeah, um, here's a, here's thething you need to know.
Um, we will never say first,ever, or, or, or never before,
or, uh, anything like any, anysuperlative like that, unless we
can absolutely prove it.
Because somebody always comesalong and says no, no, no, you
know, I was the first ever and Ifiled a lawsuit in 1988.

(14:04):
And so we're basically amongthe first, or the known first,
so we will always hedge our bets.
That said, as far as we can tellit's a and it's an interesting
case too.

It's an interesting case We'll link to
your reporting on this.
I also wrote a substack thatsort of was a follow-up, you
know, very much inspired by ourconversations um the last, and I
and I and I'll admit I haven'tactually checked in on whether
there have been any updates thelast thing that I have seen was
a preliminary injunction thatwas denied by the judge, and I

(14:51):
remember when we had talkedabout this case, you know, one
of the things that you and Iboth sort of agreed on was we
just don't know enough, like wedon't like there's we have one
side of the story, which is thestudent is, you know, claiming
that they were unfairlypenalized, um, and you know, uh,
for, for cheating, when inreality, uh, from the student's
perspective, um, this was purelyjust a misunderstanding.

(15:12):
And you know a lot, you knowmiscommunication or lack of
sufficient communication fromthe school about, you know what
constituted cheating, and Ithink there was a lot of, um, uh
, reactionary outcry that waslike, ah, this is completely
unfair.
The student, you know,shouldn't be penalized for the
school just not being up to dateor up to speed.

Not having a good policy.

Not having a good policy.
But when I read the preliminaryinjunction the judge actually
quoted specific excerpts fromthe school academic integrity
policy where they say you knowyou're not supposed to use, you
know technology and they ensurethey don't necessarily
specifically name artificialintelligence.

(15:51):
But so there was even even thethe out of date academic
integrity policy, I thinkarguably did cover it.
But then the other other sortof like you know detail that I
think was lost in some of thenational coverage is the school
had actually had a whole classwhere they talked about
appropriate and inappropriateuses of chat GPT.

(16:13):
Right.

It was an English class I think it was an
English class.
And just for the record, thekid had been punished in a
history class.

Yeah, and so so what we, what?
Basically, what the case kindof evolved into was to me very
clearly the student cheated, andthis was basically a semantic
argument about whether you knowa policy in an English class is
transferable to a social studiesclass and you know I'm not a

(16:43):
student of the law and I don'tknow exactly.
You know I'm not a student ofthe law and I don't know exactly
.
You know I don't believe thatthere's been a ruling yet.
In the case, the judge seemspretty aligned with the school
based on his commentary in thedenial of the injunction.
But you know, I think Iactually think that school in

(17:04):
Massachusetts is an exemplar ofwhat most schools should be
doing.
And sure, maybe there, you know, I think there policies school

(17:25):
wide and be really intentionalabout letting kids know.
But I also think that, as astudent, you know there is a
there's, there's these grayareas, even before Chad GPT.
And one of the one of the pointsthat I make when people ask
well, you know, chad GPTshouldn't be cheating because a
student is just, you know, usinga tool.
You know chat GPT shouldn't becheating because the student is
just, you know, using a tool.
You know, one of the questionsI'll ask I'll be curious for

(17:47):
your take on this um is imaginea student's writing an essay and
, um, it's an essay on cold warhistory and their parent is a
professor of cold war history.

Um, no, you're going with this, but
while they're writing the essay,and while they're writing the
essay, their parents are sittingright next to them.

Right, and you know, maybe not literally
writing it, but helping themwrite every single thing like
sort of like you know, providethem with all this feedback, you
know, not just reviewing it andproviding like red lines, like
actually co-writing it with astudent Right.

You know there's a sculpture project in
there.
Dad's a sculptor, yeah, yeah, Imean is that cheating I mean.

I like if you were the teacher in that
situation no, I think it's agreat it's.

It's great, it's a great question and I
think the the to me the kind ofthe short answer to that.
The short response is you know,transparency says you should
indicate that you know I gothelp from my mom, who's a Cold
War historian, or you know, hereare the contributions that that

(18:52):
she, you know that she was ableto offer and I.
You know.
It's interesting, cause I thinka good teacher maybe this goes
back to you know, like basicprinciples, principles like a
good teacher would probably know.
Like you know, alex's mom is acold war historian, so he's
gonna write a killer paper on,you know, the q and missile

(19:13):
crisis.
Um, and I think in in a way,maybe that's beyond the pale for
most teachers.
Maybe they don't need to knowthat much about their students,
but it's probably a good, smartteacher would know that.
It's kind of the same thing.
I think you should be able toexpect from your students a kind

(19:34):
of transparency about theirwork and their sources and the
advantages they've got.

Because the student didn't appropriately
cite their use of ChatGPT andthat was one of the things that
they'd covered in English classwas first of all, if you're
going to use ChatGPT, you haveto.
I forget exactly how theyframed it, but it was like the
spirit was you can't just typein one sentence and press enter.
The goal is that you're and ifyou're going to use it like you

(20:03):
have to cite it in the same waythat you'd cite an encyclopedia
or some other sort of likeonline technology based
information product.
And so I think that there is,but there's generally not.
I think the point with theparent is there's not a right
answer, but there is sort of thegoal of education, the goal of
an assignment is ultimately astudent, you know sort of this,

(20:24):
this like sort of productivestruggle, yeah, um, and I think
and I think this is where I'mvery I'm very biased towards
teachers in the debate aboutwhether chat GPT is cheating,
because the answer is it'scomplicated.
But ultimately I think teachersactually have an instinct and,
to your point, you know,teachers know their students and
like they generally have asense for whether something is,

(20:47):
like you know, kind of wildlyout of scope of what is what
that student was able to do.
Um, and I think the challengeis the parents, students are
looking for sort of a blanketanswer to a question that I
think society and industry hasnot even really quite answered,
because we have not figured out.
I've not seen any company thathas like really honed in on what

(21:12):
does it look like for humansand language models to sort of
co-create together?
And so in this moment wherewe're actually figuring this out
together, I think there needsto be some latitude.

I think you're right, and, and and you
know I mean another way ofthinking about productive
struggle is, you know, effortfulthinking, right?
You know this is what school isabout, right?
This is you.
This is basically practice foryour brain.
You know solving these problems, and I think if if we're
relying on a technology just toget the student to the end of

(21:51):
the assignment, then we've lostthe narrative, and to assume
that kids aren't going to dothat is naive.

Another reaction is well, if students
are cheating, it's simplybecause the assignment isn't,
it's out of date, um, and Ithink there was an assumption
that this was just sort of astandard book report, but again
in the, in the judge's uh denialof the preliminary injunction,
so more was revealed about sortof the nature of the assignment.

(22:24):
Um, my understanding is, uh, thestudent had to choose the topic
, right, they were working ingroups, so a lot of components
that we would actually generallyadvise teachers, like how do
you design uh assignments andwork that is more meaningful to
students?
I mean, there was a lot ofagency provided to students and
it was also multifaceted and,interestingly, the teacher, the

(22:45):
reason the student got such abad score on that assignment was
because, um, the it wasn't eventhe final assignment that was
turned in, it was the, theresearch and the outline.
Um, because I think theteacher's mindset was uh, there
were a few different ways thatthey could present what they
learned, and so, rather thangrading everybody on just sort

(23:06):
of like whether their PowerPointwas good or whether their
report or their write-up wasgood, they were focusing on the
process of gathering theknowledge, and so that's again
something that we've, that'sadvice that we often give
teachers is you have to focusless on the work product and
more on the process.
So could the assignment havebeen honed?

(23:28):
Perhaps, but again, I find thisan example of even when
teachers and schools are doingeverything right.
Kids are crafty.
I guess you should get maybesome points for trying, I think
this was a little bit.

Ham-handed.
Yeah, I guess it was ham-handed, I think this was a little bit,
um, ham handed.

Yeah, I guess it was ham handed, I guess
I guess there's also a concernthat teachers that that and I
think this is actually reality.
I think some teachers areincredulous at the idea that
students would cheat and so,rather than interrogate, this is
actually Victor Lee's researchout of Stanford and Victor's
research, which I want to talkabout.
Victor Lee's research out ofStanford and Victor's research,

(24:06):
which I want to talk about, bythe way, found that incidents of
cheating post-JETGBT have notincreased and the hypothesis is
that students are always goingto use sort of like the latest
tools to cheat, like when I wasin high school, we had our
lunchrooms or these round tables.
And so you know, before our APuh, european history class with
Mrs Janini, um, we kind of had apact where, like, at least one

(24:29):
person would make would have thehomework done and everybody
would sort of pass the the sheetaround and sort of like copy
the answers Um, and you knowthat was sort of an old school,
that was like pre uh, socialmedia and, and you know, now
students have Tik and other Copyand paste.

Your early copy and paste.

So I think part of the argument is that you
know teachers are always goingto.
The worry is that teachers aretaking it personally, that it's
in front to them that studentswould cheat when researchers
like Victor Lee would say youknow teachers need to create
learning experiences thatstudents, you know, sort of
inherently understand andconnect with their own world.
And that are meaningful Right.

I love the word I've seen a number of
instances around this the wordartifacts, because I think
that's a.
To me it's a powerful word.
It's like what are theartifacts you want as a teacher?
Is it just you know?
In Alex's history class, youjust want you know eight

(25:37):
questions answered and ifeverybody gets the right, you
know, if everybody copies andpastes the right answer, then
the work is done.
Is that a satisfactory artifact?
Or do you want a different kindof artifact?
Do you want something whereindividualism is essential?
So I think we're coming upagainst that in terms of what do

(26:03):
teachers feel is like thesatisfactory, um, uh, proof of
learning and uh, I mean in a waylike I, I hate to say this is a
really interesting conversationfor teachers to have and for
parents to have and for studentsto have.
So maybe, like all that, maybethis technology is really like
forced, uh, forced that upon us.

(26:24):
So let's do it, let's have thatconversation.
We certainly weren't havingthis conversation 10, 15 years
ago, that's right.
So in a way.

Progress is upon us.
Yeah and you covered the futureof.
Is it high school, the futureof school?

Future of high school.

Future of high school or the future of
school, Future of high school,Future of high school.
I've been a part of a lot ofconversations with folks and
there's different sort ofmonikers for this.
There's school transformation,reimagining education,
reimagining schools, schools ofthe future and those

(27:04):
conversations have been going onand you can attest to this, for
decades, centuries forcenturies and you know, as a
nonprofit, as the head of anonprofit, I spend more time
than I care to admit, you knowfocusing on fundraising.
And one of the phenomena in theeducation space, in K-12
specifically, has been ashrinking of the number of

(27:28):
philanthropies and the amount ofmoney that philanthropies are
giving education, and in partbecause they're sort of throwing
up their arms, they're justlike we've been trying and we're
just not seeing the results.
it's just too hard and justthere's.
It's very multi-dimensional.
You're trying to, you know,address math scores, but you
know the schools that you'reworking in are dealing with,
like you know, literally ahundred different other sort of

(27:48):
connected issues, um, and Ithink philanthropy has been been
having trouble sort of likepushing that forward.
I think, you know, at a certainpoint people are just
exasperated, not just thefunders, but the stakeholders
and the parents and all thefolks that you need on board for
this.
And I'm curious if you see this.
But from my vantage point, thepowerful thing about artificial

(28:09):
intelligence perhaps even moreso than what it can do in the
classroom and we can talk aboutsort of like what, how AI is
going to transform the classroombut to me the bigger, uh, the
bigger part of this moment is ithas actually reinvigorated
interest in this conversationabout the future of school and
suddenly it feels relevant,tangible and actionable.

(28:33):
And you know, tangible andactionable.
And conferences are selling out,and I think the challenge is
how do we make sure that theconversation stays on this
question of what is the futureof school without moving too

(28:54):
quickly to the answer of?
Well, it's just AI, tutors andhyper-personalized learning and
just throwing products, because,basically, I think the space
has been inundated so quicklywith generative AI products that
and I'm curious for your takeon this is AI products certainly
a component, perhaps, of thisconversation?
How much of the conversationabout the future of school is
truly about the AI tools thatwe're going to use, as opposed
to some of this more maybephilosophical or almost like the

(29:16):
spirit, the conversation aboutwhat is the spirit of what we're
actually?

trying to achieve.
That's a really good question.
So I have many thoughts aboutthis and I think, basically I
think at the very base of thisyou're right there is a lot of
interest in what this is goingto do.
I think people's imaginationsin 2022 were really fired by

(29:40):
this to your point, and nowthey're wondering okay, how is
this going to apply to my kid'sschool?
What could they possibly dowith this?
That will make a difference.
I will say that it's probably asafe bet to say it's not going
to make much of a difference,just in terms of basic

(30:02):
achievement and NAEP scores.
And we're already seeing justlast week we saw the flood of
terrible, disappointing, verybad, no good NAEP scores.

Can you for our audience explain, naep?

Sure National Assessment of
Educational Progress.
It's a test given in variousforms every two years or so.
It basically tracks fourth andeighth graders' reading and math
skills and, as we're seeing now, five years out of the
beginning of the pandemic,they're just not really

(30:36):
recovering in a lot of ways anda couple of places they are
Louisiana, I thought, was areally interesting outlier where
their skills of those kids andI can't remember which group it
was, but really defied sort ofthe expectations.
But mostly, you know, we're insort of a like a very kind of

(30:57):
grim holding pattern and it'sbeen like that since actually
before the pandemic, since like2019, I want to say Certainly
the pandemic didn't help.
So I don't think, on the onehand, it's going to really have
much of an effect, these newtools, these new ways of
thinking on that, because itfeels like very little, if

(31:21):
anything, has any effect on that.
I'm not quite sure systemically, why, other than that, you know
, I mean there are lots ofpeople who would say you know
well, you know we keep, we keepjudging the, the standard for

(31:43):
what it means to be basic andadvanced, and so you know we're
always chasing this, thestandard that keeps
incrementally increasing thisstandard that keeps
incrementally increasing.
But I think to me the reallyinteresting question is is, like

(32:07):
, what do we expect in terms ofthese tools?
And like, how much of that arewe expecting to be student
facing?
I would have thought at thebeginning of all this two years
ago that are we expecting to bestudent facing, right?
I would have thought at thebeginning of all this two years
ago, two plus years ago, that itwas going to be like all
student facing, like we weregoing to just get, you know,
tools that students are going tobe using.
You know, you know 24, seven,and there it's just going to be

(32:30):
just this sort of amazingrevolution of kids just
interacting with the technologyand just doing different things
and different kinds of learning,and it's just going to be.
You know, we can't predict it.
I think to me what I'm seeing,which I didn't predict and
that's just my lack ofimagination is, to me, the most

(32:52):
interesting piece of this is notthe student facing stuff, it's
the teacher facing stuff and theyou know kind of, in some sense

(33:14):
, the drudgery, in some sensefiguring out what students
should be doing at any giventime, and that can be as basic
as, like you know, predictingwho's going to be absent, or you
know, I mean, you're familiarwith all these things.
To me, those are the things Ithink we should really be paying

(33:36):
more attention to.
In a way, it's interestingbecause I talked to someone who
said a couple of weeks ago, whosaid you know, ai is weird
enough right now and it's sounpredictable that I think,
until we, like figure out moreabout it and how it works and
really what it does, I don'tthink we should put it in front

(33:56):
of students at all.
Let the adults mess with it,but students shouldn't have
access to this.

Anyway, I don't know how you feel about
that.
Well, there's this assumptionthat we are the AI education
project, and so oftentimes Iactually will often now start
meetings with.
This is what this meeting isnot.
We are not here to try to putAI in front of all of your
students, or rather, we're nothere to accelerate that process.
There's, I think, a lot ofquestions that need to be

(34:29):
answered in terms of how to dothat appropriately, safely, but
also in a way that complementsrather than replacing education.
I want to talk about ananecdote specifically where
you're replacing teachers, orreplacing, when you say,
complements.
That complements teachers.
Okay.
Yeah, but also that complementsthe learning process.

(34:50):
I think, there is a world whereyou could imagine students.
In fact, if you go to china,they actually I remember this.
There's a picture that hit mereally viscerally uh, it was
students preparing for one ofthe big national exams in china
and they had this giant like Idon't think it was even a
gymnasium, I think it was awarehouse, okay, and it was like
3 000 students in in cubicleswith headphones on getting

(35:13):
personalized learning.
And it just hit me that, evenif the scores are higher, the
idea of just a student sittingdown in front of a computer all
day digesting information,there's something really wrong
with that.
And I see sort of flavors ofthat with this sort of obsession

(35:35):
with AI tutoring.
But it's complicated, right,because actually AI tutoring is
still really exciting andinteresting as an after school
support, as a support forneurodivergent students or, uh,
students in special educationmore broadly.

(35:56):
Um, you're blind If you're anEnglish language learner.
Um, if you're deaf, right, thistechnology is sort of
unequivocally going to addincredible value and, in sort of
like, giving you access tolearning that was otherwise, um,
you know, much harder to access.
So so it's very hard to sayblanket, you know, ai bands, I
think, ai bands, I think forthat reason are um much harder
to access.
It's very hard to say blanketAI bans.
I think AI bans for that reasonare sometimes looked down upon.

(36:18):
To me, if I was asuperintendent, it would be more
like we're going to pump thebrakes.
You described it really well.
We need to understand thistechnology.
We can't just rely on what thevendors are telling us.

Outside of the vendors, I've talked to most
of the smart people who arereally obsessed with this
question and most of thesmartest people will admit that
we don't have the answers yetand most of the vendors, by the
way, in 12 months time, aregoing to be moving on to
something else If they have anycapital left.
Yeah.

Right, right, if they have any capital
left, right, and so I think thebut on the other, the challenges
.
Let's say we're superintendents, we're trying to design this
high school of the future.
Maybe we should do that.
Let's put those hats on.
Let's go through some of thedecisions we have to make.

(37:12):
So decision one we don't haveany new budget.
We have a technology line item.
We've been spending it now onthings like smart boards, uh,
chromebooks for students.
Um, we're keeping an eye on the.
Uh, is it enet?
E the government?

(37:33):
Uh, there's a governmentsubsidy for technology access.
E-rate.
Oh.

E-rate.

Yeah, we're keeping an eye on E-rate
funding.
We're worried that that mightexpire, so we don't have too
many extra dollars to spend, butmaybe some stuff we can move
around.
Okay, and the first decision iswe need to send some sort of a
letter to parents and studentsexplaining whether or not

(37:57):
students are allowed to use chatGPT One step short of writing
an integrity policy, but what'ssort of step one?
Do we blanket ban while wefigure it out, or do we
encourage students to use it butask them to be transparent
about how they use it?

I'll let you go first.
I think you let everybodydownload DeepSeek onto their
Chromebook and call it a day.
This is a really good question,and for our audience.

Deepseek is the language model that was
released by a company in China.
Two weeks ago now.
Technically in December.
There's an earlier version.
Lots of privacy concerns thereyeah.

And it was built for cents on the dollar in
comparison to.

Reportedly, although there's actually some
new reporting out.
We'll come back to that.

Yeah, we can come back to that.
You know, my feeling is I don'tthink you send out a memo to
everybody, to the parents andthe kids.
I think the first, to me, yourfirst step, is you give teachers
some time to play around withit and you really give them the
opportunity to find a comfortlevel with it, find what is the

(39:13):
most likely way they mightemploy it.
Let them figure out what therisks are.
I'm not saying it's likesecurity, because I don't think
most teachers know enough aboutthat.
You can read the 74 million tofind out that, because we cover
it really well.
I, yeah, I think you start withthat.

(39:33):
I think you start with teachersjust playing with it, and I
really do mean playing.
And I want to talk at somepoint about the parallels that
I'm seeing with the reportingthat I did 10, 12, 15 years ago
on games and learning, um,because I see a lot of

(39:57):
similarities, um with that andwith with AI, um.
But I think, I think you do, Ithink you let teachers mess
around with it and see what popsout and maybe, as part of that,
you require them or stronglysuggest that they like get
together, you know, once a week,once a month, and just let each

(40:17):
other know.
You know what the.
You know what the coolestaspect of this is.

OK, I'm in line with that.
So we don't ban.
In fact we just avoid thatwhole all the controversy around
that by just not addressing it,because anyways the kids are
going to access it out of school, they're going to access it on
their phones.
A ban only really makes senseif we feel confident that we can
enforce it.
And then the other thing that aban does is it orients teachers
to this defensive mode where,rather than focusing on learning

(40:49):
the tool, they spend more timefocusing on enforcement.
I think my modification to whatyou're describing is and this is
a plug this is the work thatthe IEDU does.
We have cohorts of teachers andwe facilitate those
professional learningcommunities so that there's some

(41:10):
structure.
I don't think you need to workwith nonprofits like us to to
create those sandboxes, but Ithink it does help to have, you
know, some subject matterexperts who can sort of like
sort of helpfully guide teacherstowards, you know, some of the
best, sort of like some provenways to actually sort of like

(41:30):
like some some scaffolds fortheir exploration.
But the other reason I thinkthat's important is you know,
ultimately, what is likelyhappening in our district is
teachers, and maybe even withthe way we discovered.
You know ChatGPT.
I don't know that we werecovering.
You know, I guess as asuperintendent we're not reading
Wired, or you know MIT TechReview, right, maybe we saw a

(41:54):
headline about ChatGPT, but weprobably before we actually got
the chance to use it.
My guess is, a teacher went tothe lunchroom and was just like,
look at Sally's paper.
It looks like it was literallywritten by a college professor.
And I asked her.
You know, I confronted herabout it and she showed me this
crazy tool called chat GBT.

(42:16):
Like check this out, and butit's with that frame of like
students are using this to cheatand but I, but I think this
idea of, I think as as asuperintendent or a
co-superintendent, my answer tothose teachers who are like,
well, how do we deal with this?
It's like my answer to thoseteachers who are like, well, how
do we deal with this?
It's like you're trying todesign a bank that is impervious

(42:39):
to being robbed.
It really helps to consult witha bank robber who is really
thinking about how to do this.
You have to get into the mindof somebody who is trying to
cheat and the way to do that is,uh, you need to use it yourself

(43:00):
.
There is no sort of cheat code,unintended um, uh, or shortcut
or rule that we can put in.
Um, you're probably going tohave to if not definitely going
to have to adapt some of yourassignments and you're not going
to know how to adapt thoseassignments to be more resistant
to cheating if you don't have asense for how the tool could be
used.

Okay, I like that yeah.
I'll go for that, Find the kidswho are doing the most mischief
with it and maybe even, youknow, have a, have a some sort
of like a regular round table orsomething we're actually doing
that at one of the a couple ofthe upcoming conferences that
we're doing, sessions that we'rewe're having teachers, I think

(43:36):
we're literally calling themlike cheat-a-thons.

We're going to have some like oh you know,
sort of like traditionalassignments and we're like we're
going to spend the next hourliterally cheating on these
assignments and we're going toshow you how, uh, how, you could
do it sort of get into the mindof a of a student.
Have some students get involvedin that?
Um, how are there, you know, asthis wise superintendent who is
, who was around for the time of?
Um, you know, gamified learning.

(43:59):
What are those you mentioned?
You saw some parallels there,like, how, like, can you just
talk about that?

yeah, I mean, you know, just like you're
sort of suggesting here, I meanthe, the, the place where 10,
15 years ago, um games andlearning were the most
interesting and where thingswere getting done that I found

(44:27):
really exciting were placeswhere an individual teacher, um
had discovered something ontheir own, or had discovered
something with a group offriends, or went to a conference
and was like, oh my God, you'vegot to play this game or met
some developer or became, youknow, a couple of them actually
became, like you know, gamedevelopers themselves, just

(44:49):
literally teaching themselveshow to build this stuff
themselves, just literallyteaching themselves, um, how to
build this stuff.
Um, and I, I, I loved theexcitement around that, I, I and
I also loved, um, what kinds ofconversations were happening
because of it?
Because a lot of instances, um,instances when teachers

(45:12):
discovered really good learninggames, especially something like
math, which is really easilyadapted to a game, teachers were
asking totally, totallydifferent kinds of questions
than they had been before then,totally, totally different kinds
of questions than they had beenbefore then.

(45:34):
They were really, in a way,kind of like turning their
curriculum inside out, becausethis one game or this couple of
group of games had justbasically asked them to say to
themselves, like what is algebraLike?
Why do we have algebra and whatthe heck is it?

(45:54):
And there's one game inparticular that I wrote about in
my book.
It was Dragon Box.
I don't know if you're familiarwith that, so this is by a
French developer and itbasically asked students to kind
of like almost like inventalgebra from the ground up and

(46:18):
almost like discover, like, whatX is and what X plus one equals
Y is.
So really just kind of like Isaid, just turned everything
inside out.
When you have students doingthat, when you have teachers
getting excited about that, it'sa totally different kind of
conversation than why can't mykids do algebra, why aren't they

(46:39):
interested in algebra.
And I see the possibilities forthat kind of thing with
something like AI.
If students can come at a mathproblem from a totally different
point of view or do differentkinds of problem solving, then I

(47:07):
think we're really in adifferent place.
And you've probably experiencedsome of these things.
I mean you've probably seenteachers, you know, looking at
some of these tools and getting,you know, pretty excited.
I mean the easy way to do it isjust to ask chat GPT to, or the
easy and stupid way to do it isjust to ask chat GPT to.

(47:30):
You know, give me 10 algebraproblems, right, but there are
other ways to, to, to recruitthese tools, to do, I think,
much more fascinating, much morefascinating things.

Yeah, the, the.
So I'm going to sort of drawout from what you're saying that
stands in contrast to you know,presumably there were lots of
companies approaching districtsand saying here is a you know
gamified learning product.
Can I improve your math scores?
You should put it in front ofevery single teacher.

(48:05):
And my guess is that thosedistricts, as is often the case
with new curriculum, theyencountered a lot of pushback
from teachers.
It worked in some classesdidn't work in others and so, as
we.
The challenge is this is likeit'd be really nice if the
answer to this question of like,how does AI going to transform

(48:27):
learning?
If it's just like, well, youjust give teacher, you know,
just have teachers use AI to doX Y Z.
They're going to have more time, they're going to use it to
personalize learning in X Y Zways, and then your goal is just
get that in front of everysingle teacher, train them on
how to use it exactly in thatway and something that's
repeatable, prescriptive.

(48:47):
And for districts like PrinceGeorge's County, where we're
just up the street from here,where we're doing work, they
have 10,000 teachers.
I suspect that the answer to thebest outcomes in terms of how
AI will transform education issimilar to what you're
describing, where you create theconditions for teachers to

(49:11):
hyper-c, hyper, customize andexplore and use it as a, as a
force multiplier with their ownspin.
And I guess the challenge forfor superintendent for schools,
for school leaders, is you know,for PGCPS, they have 10,000
teachers.
You know you start with thesandbox, that learning community

(49:32):
of 50 teachers, but you reallywant this to get to every
classroom and, in fact, if youhave a mind for, you know,
reaching the kids who are atrisk, or most at risk, of
falling behind.
They're very often in some ofthe lowest performing schools
that really struggle to retainthe brightest teachers.
Is there anything whether it'sfrom your time in sort of like

(49:55):
gamified learning, but is likehow can this sort of get into
the bloodstream?
Like what are the conditions,for I guess the way to describe
this would be it's almost likewhat is your understanding of
some of the best ways to buildand enhance, like the teacher
workforce, and upskill them inthe ways that they need to

(50:19):
explore and understand anditerate on the technology in
their own way.

I have a couple of thoughts, by the way,
and I'll just preface this bysaying I think a lot of people
in the games and learning spacewould take exception to your use
of the word gamified.
They would say thatgamification is kind of a lesser
form of what they're doing.
They're actually using games ata higher level than just

(50:50):
getting people to behave acertain, to sort of behavior,
behave a certain way.
They.
They would say thatgamification is a little too low
level.
I don't want to spend much ofour time talking about that, but
I know it's interesting.
You, you use a phrase a secondago get into the bloodstream,
and I guess I would, maybewithout with your permission,

(51:14):
maybe explore that a little bit.
So you're talking about, like avirus getting into the
bloodstream, something beingviral.
I mean, why are things viral?
In the sort of social sense?
They're viral becausesomebody's really excited about
it, or it's new or different orweird, or just, you know,

(51:39):
getting something done in a waythat, um, we've never seen
before, um, and so people getexcited about it.
I'm not sure where you go fromthere, um, but I do see almost
like the necessity of gettingpeople excited, and maybe it's

(52:01):
just one big natural experimentwhere the people who are excited
about something test it out andtry it out and have a certain
amount of success and the vastmajority don't have access to it
and either their results aren'tas good or they're.
Maybe their teachersatisfaction isn't as high, or

(52:24):
um, you know, I, I, I I'm notfamiliar enough with sort of the
ways systems like these work tobe able to say, yeah, this is
all you need to do is just you,you know.

Well, I think even the deepest experts
in teacher professional learningwould struggle to answer this
question right.
There is no quick fix.
It's the challenge, this ideaof like the components of
virality.
It has to catch people'sattention.
In education.
You know there's.

(52:54):
If something works, presumablyyou know that is the type of
thing that teachers are going tobe talking about.
If something doesn't work, orfor whatever reason they're
perceiving that it doesn't work,then they're going to.
They're going to resist andalso they're just in many ways,
especially if a teacher of like10 plus years they're on cruise
control, they've been teachingthis class it really is going to

(53:15):
take something profound forthem to change course.
It's a lot of work.
You have your script, yourlectures down pat, you have the
homework assignments.
It's not trivial, even ifsomething is pretty good.
I do see.

(53:40):
The lever is this idea of ifyou find the right teachers,
they actually can be catalysts.
I think I bought into that.
I'd be curious.
You know we don't have toanswer this now, but I think one
of our wonderings if you comeacross anything like this in
terms of you know analogs thatwe can find where really
successful whether it'scurricula or teaching practices,

(54:02):
project-based learning we'rebig proponents of that.
Most of our curriculum is PBLor inquiry-based learning.
A lot of the schools that wework with have struggled
mightily to get project-basedlearning into the classroom.
And.
I think the answer in part isbecause they just have
project-based learning issometimes perceived to be at
odds with, like, the standardsand all the content knowledge

(54:24):
and the rigor that teachers areexpected, and so their teachers
are getting these conflictingmessages where they're being
told you need to have moreproject-based learning.
And then they're also.
Their salaries, in some cases,are literally determined by how
their students are scoring onthese sort of like very rigid
standards.
Okay, being a superintendent isreally hard, so I think this is
our chance to take off thosehats.

(54:45):
Thank, God.
Okay.

So I don't have to be a superintendent
anymore.
Okay, good, no, not anymore.

Put your journalist hat back on.
There's so much happening rightnow in ai.
What are the stories thatyou're just keeping an eye on?
I'm especially interested ifyou're willing, anything that
you maybe haven't covered yetbut that you're, you know, just
trying to learn more about, thatwe could sort of like dive into
sure, I'm actually going to beum and this is not a
self-promotional plug, but whatthe heck um at by Southwest.

I'm going to be doing a panel with folks
from Curriculum Associates andKhan Academy about AI and
assessment, and that's one ofthe things that I think is
really interesting about AI'spotential to mess with

(55:37):
assessment and improve it andmake it I mean, the word I would
use would be more invisible.
I wrote a story a couple ofmonths ago about curriculum
associates buying this Irishstartup that has developed

(55:59):
speech recognition for kids, andit seems like a minor thing,
but I think it's kind of huge ina way, because you know, it's
one thing for you know aneight-year-old or nine-year-old
to be able to ask Alexa, youknow to.
You know play a song or turn onthe lights or whatever, but

(56:19):
imagine a three or four year oldbeing able to do that.
You know, I actually have apersonal experience with that.
I've got a three and a halfyear old grandson who has
figured out how to very clearlyask Alexa to do stuff, but I
think that's rare, and so theyhave developed essentially a way

(56:39):
to help with AI, help computersunderstand the utterances of
very small children, and this is.
You know, it doesn't take muchimagination to think like, okay,
if I'm a preschool teacher orkindergarten teacher, you know,
this really revolutionizes, um,what I can do with instruction,

(57:02):
because I can sit a an entireclass down and just have them
read into a, you know, a tabletor a microphone or what have you
.
And, um, you know on their ownby the way, and that, you know,
not even in the presence of ateacher, just have them, you
know, open up a passage and readit and have the AI, you know,

(57:24):
assess how well this kid isreading.
In a way, that's one of theexamples of to me I was going
back to something we weretalking about earlier like
that's not really student-facingAI.
In a sense, it's AI for theteacher's sake, and I think

(57:44):
that's a really exciting aspectof this, and I think, you know,
if you want to talk about forcemultipliers, I think that could
be really cool and I'm excitedabout, like, the ways in which
technology can play withassessments in other ways, you

(58:10):
know, maybe make them absolutelyinvisible.
Maybe, you know, take, let'ssay, work a student is doing
already in class and has alreadyswallowed up all the responses
you've just offered over thecourse of a week or a month or a

(58:31):
unit, or what have you?
Getting back to the games piece,this is what games do, right?
Games assess you constantly,all the time, from the you know

(58:57):
minute one to the end of thegame, right?
That's one of the things peoplelove about games.
Right, they're constantly beingassessed, but in a way that is
invisible and I think.
I think with AI we have thepotential to do that in a way
we'd never have before.
Again it's not student-facing,it's totally teacher-facing

(59:19):
right, but it does something wecouldn't do before.
You're skeptical.

Well, I'm not skeptical.
I just imagine the perfectiteration of this, where you'd
probably get the best outcomesfrom the perspective of like,
accurately assessing students,would be to embed AI and
obviously the studentsChromebooks, every single
homework assignment.
You know, if they have accessto some kind of a chat bot or a

(59:50):
tutor, that data is all digested.
There's probably like theclassrooms, maybe with video as
well.
And let's just assume, let'stake for granted, that we've
absolutely solved the privacyaspect.
Right, the data is anonymized.
So my skepticism doesn't comefrom a space of like oh, data
privacy.
But in this world, presumablywe wouldn't just use it for an

(01:00:17):
individual class, the students.
You could use the same approachto evaluate the student over
the course of their entire yearand then over the course of
their entire academic career,and perhaps we don't need
college exams or perhaps wedon't need SATs or college
applications colleges.
But in that world it's kind oflike the sorting hat You're told

(01:00:43):
.
Here are the three collegesthat you can choose from.

Greg Toppo (The 74) (01:00:50):
I understand what you're saying.
It feels like you're gettingahead of yourself.

Alex Kotran (aiEDU) (01:00:53):
So I'm absolutely getting ahead of
myself.
I don't know if that's 10 yearsaway, 20 years away.
I do think that it's, it's,it's it feels inevitable and
it's it's sort of like we justhave to make a choice as society
, that that's the directionwe're comfortable going down.
And if not, I suppose and thequestion is simply, how do we

(01:01:15):
build in a sufficient amount ofstudent agency?
And I think we've done similarthings in the space of using AI
to screen resumes, I think, atleast in many places, even
though you, theoretically wedon't use any AI to screen our
resumes, even though we clearlycould, because we just made this

(01:01:36):
decision, so sort of like anormative decision about um not
wanting to surrender thatprocess to to the ai um, I just,
I just wonder if, like, are wesleepwalking into sort of this
world where you know studentsare sort of just?
I mean, there's the, uh, theayn rand book, um, where
everybody's sort of just toldwhat their job is, and I just

(01:01:59):
know your take that's just sofar out, it's yeah, you went
very dark very fast with what Iwas saying.
I guess I was just envisioningand I think a lot of people are
envisioning just a just a moreseamless transition from work
and thinking to um assessmentproduct um, so I guess is there

(01:02:23):
because maybe, then maybe theanswer is um, there's some
amount of this, like because Iassessments are clearly need to
be changed, right, so, like thealternative, the flip side of
this is the current assessmentregime does not work it works in
a certain way, and that's howit works.

Greg Toppo (The 74) (01:02:40):
Some of it's great and some of it's
terrible.

Maybe assessments.
It's not that they do not work,but there's a lot of room for
improvement in assessments.
There seems to be a middleground where we use technology
to improve and enhance andaddress some of the gaps that
exist in the education systemand maybe we just need to be
sort of forward thinking aboutwhere do we need to sort of draw

(01:03:05):
the line.

Greg Toppo (The 74) (01:03:07):
Yeah, no, I would say that I guess maybe
the thing I'm proposing is alittle different.
Maybe I'm not describing itvery well.
My feeling is that the bestassessment is one that you don't
even know just happened that,um, you are assessed on
something because you've beenworking on it, let's say all day

(01:03:28):
or all week, and, uh, at acertain point the system has
enough information about whatyou know and how well you know
it that we're done.
We're done here.
Very much again like a game.
Right, there's that great quoteby Jim G, who's a sort of game

(01:03:48):
scholar.
He says if I get through everylevel of Halo, this is going
back 10, 15 years if I getthrough every level of Halo.
This is going back, you know, 10, 15 years If I get through
every level of Halo.
You don't need to give me, like, a Halo final exam.
Like I got through every level,I'm done, I mastered the game
right.
There's no like final test andI think that's I think, what a

(01:04:10):
lot of people are envisioningNow, whether that's got you know
necessarily, you know Ayn Randimplications to it, because you
know the student isn't choosingto sit for an exam.
That's another reallyinteresting question.
But I guess I'm not quite thereyet.

(01:04:32):
I'm just thinking of, like,what is the process we were able
to just knock out and what canwe replace it with?

Yeah, I think you've convinced me that I
am.
This sort of catastrophizing ismaybe valuable for futurists
Totally, but actually I think Ican also imagine the flip side,
where 10, 15 years from now,people are like there was a time
when you spent all the time inschool and you did all this work

(01:05:04):
and none of that mattered.
All that mattered was just thescore you got on this one test
on one day, and if you happen tobe sick or if you happen to be
having a bad day didn't matter.

Greg Toppo (The 74) (01:05:13):
Or if it was a bad test and oh, by the
way, some of these tests are notvery good yeah, um, and so,
yeah, I, I think, I thinkthere's a lot of, there's a lot
of power.

Then there's all the other piece
about, um, sort of uh,recognition, like sort of
natural language processing, um,one of the things that people
have been saying about.
Well, assessment, assessmentsneed to change.
Homework needs to change.
You know, we need to do moreoral assessments.
Right, and teachers will oftenkind of be a little bit they'll

(01:05:45):
scoff at that because they'relike I just don't have the time.
Yeah, it's, I just like, likehomework is efficient because
people are doing itsimultaneously as opposed to
like once at a time.
Um, so this is perhaps anexample of a problem that AI can
actually, a problem that AIperhaps created, but that it can
also help to solve.
I don't generally think thatabout every single problem AI

(01:06:05):
creates, but I think this isthis might be one of them.
It is a bit challenging.
When have you seen that thishas been?
I think it's one specificperson, but there's probably, I
think probably more than one onsocial media applying for
management consulting interviews.
And he has his setup where hehas chat GBT live transcription
of all the questions that he'sgetting and then it's answering

(01:06:26):
the question and he's reading italoud and literally doing the
interview, acing theseinterviews with the help of his
chat GBT module.
That's great doing theinterview, like acing these
interviews um with the help ofhis sort of like chat GPT module
.
That's great.

Greg Toppo (The 74) (01:06:37):
That's great, but yeah, that's amazing.
Um, yeah, I, you know, um, Iguess I am of the opinion that
um and I and I I'm sure there'sresearch on this, maybe you can
even cite it now but there's alot of wasted time in school.
It now, but there's a lot ofwasted time in school and I
think the product of that isdisengagement on the part of

(01:06:59):
kids and dissatisfaction on thepart of teachers.
So, if there's a way to toclose that gap or not gap, but
there's a way to, to, to, to getrid of some of that wasted time
and you know, and use it forsomething much more interesting
I'm not going to say productivebut, then I think that's a

(01:07:22):
really interesting possibility.

Yeah, teacher retention is
increasingly.
I don't know if it's gettingworse, but it's certainly really
bad right now.
Yeah, right now, Um, you wrotea story a couple of weeks ago
about um, uh, a survey that hadcome out.
It was serving, uh, uh, teacherschools of education.

(01:07:46):
Um, and I forget these specificnumbers, but it was startlingly
few colleges of education haveany kind of a formal curriculum
or approach to talking toteachers about language models
or generative AI, let aloneactually teaching them anything,
and I was curious if anyfollow-ups on that have you.

Greg Toppo (The 74) (01:08:09):
No, I haven't but.
But.
But to me that you know, on theone hand I would say it's not a
surprise Because I think a lotof teachers' colleges, a lot of
teacher education programs arejust, you know, they're kind of

(01:08:30):
trapped in the past.
They're kind of trapped in thepast.
I don't think they've got muchincentive to do something new
and different and innovative,and some of them probably with
good reason, because they'rebeing asked to turn out a
certain kind of product.
Stepping back a little bit fromthat, I would say I'm not sure

(01:08:53):
if it's the job of a teacher'scollege to teach a, you know, a
23 year old about artificialintelligence.
I mean, we're all learning aboutit on our in our spare time and
you know, and you know overbreakfast, and you know like

(01:09:14):
it's to me it's it's kind of ain a way it's sort of like a
responsibility that we all have,um, just to you know, just to
like.
We follow politics and theweather and you know climate and
culture and music and all therest.
Um feels like this is justanother thing we should all be
kind of learning about at a verybasic level.

(01:09:36):
My kids are interested in it.
They're by no means liketechnically savvy, they just
think it's an important thing tolearn about.
So I don't know, I don't wantto say I give teachers, colleges
a pass, I just think it's justsomething it's in the air.

(01:09:57):
So it seems reasonable toexpect that young people just
figure this stuff out at acertain basic level, not, you
know, saying they don't need tolearn all the ins and outs, but
they certainly need to know howit's used and what it is and the
difference between the models,et cetera, et cetera, et cetera.

And this is we recently, over the past like
two years, we started shiftingfrom talking about AI literacy
to AI readiness, and thatinsight, I think, is sort of
part of the reason behind that,sort of part of the the reason

(01:10:39):
behind that, and is this ideathat, you know, when I close my
eyes and try to look into thefuture, um, yeah, I see a world
where ai is just, it's soubiquitous and seamless and
invisible.
Um, it just seems it'll feelvery like quaint that we would
have to teach people how to useit.
In the same way that my guessis that teacher colleges, um,
you know, 20 years ago, didn'thave a class called like the
internet, right?

Greg Toppo (The 74) (01:10:59):
Um, how do you use your?
How do you use your iPhone?
How do you use your?
I know, I think that's exactlyright.

Mobile is even a better example, because
that was, you know, the earlyinternet was actually kind of
how to use your phone.
Um, and and teachers, collegesactually have a lot of important
work to do, which is how do youbuild those durable skills,
which is when we talk about AIreadiness.
Sure, ai literacy is a part ofit.

(01:11:25):
We're assuming that that'sactually going to be.
Maybe one of the easiest partsis like teaching people about
what AI is, and how to use it.
Um, the more important pieces oflike building agency, critical
thinking and those sort of softand hard skills that you know.
I think most teachers, whetherthey realize it or not, they are
.
They like they have theknowledge and skills to get kids

(01:11:45):
ready for the future.
I just don't know that you know, 21st century skills has kind
of been it's been talked, wasit's been.
It's been talked about so muchthat I think it sort of lost its
its punch.
I, I my my big sort of likescheme is this is our chance to
sort of reinvigorate 21stcentury skills and sort of
almost like reframe it, as, asyou know, ai is really just

(01:12:06):
another way of talking about a21st century skills.
Yeah.

Greg Toppo (The 74) (01:12:09):
I mean, I think in a way to me I don't
know what, as I think about it,like how I use AI in like my
daily life.
You know I'm by no means like atechnophile.
You know I'm interested in itand I like to kind of think
about it and explore it.
But you know, in terms of likebeing a journalist like to me
the most I mean, once you dig inand start looking at these

(01:12:33):
different models, you knowwhat's caught.
Or or chat, GPT, or any of theother ones.
Google, the Google one,fascinating, you know that came
out a couple of months ago.
You know that allows you tocreate these crazy little 10
minute podcasts notebook.
LM.
We had a lot of fun with that,not only in my house but at work
.
We were just we were playingwith feeding our news stories

(01:12:57):
into it and just seeing what itcreated.
So once you get past that, thesort of basic literacy with that
, and knowing what these thingsdo and what they don't do and
how one is different than theother, then to me the big
challenge and I'd be interestedin hearing your take on this is

(01:13:22):
keeping them at arm's length.
I think that's really maybe themost important skill of all, and
what I mean by that is like Iwould never use one of these
models, certainly to write astory, okay, that's like
verboten, okay, and I wouldnever feel comfortable doing

(01:13:45):
that, right, um, I wouldn't evenfeel comfortable like taking
like a report.
Let's say you know somegovernment agency you know sends
out a PDF of a report.
Um, like a report.
Let's say you know somegovernment agency you know sends
out a PDF of a report.
I wouldn't even feelcomfortable asking Claude to
like summarize this reportbecause, like I, you know, I,
who the hell knows what Claude'sprocess is in doing this in a

(01:14:09):
way, like you want to, youalmost want to do, you know,
like a kind of like a test of it.
You know, you want to takesomething that you know really
well inside and out and askthese models to summarize it and
then just see, you know whatit's, what it's going to tell

(01:14:31):
you.
So I, I really feel like, um,there's, there's almost no way
in which, like, relying on thesemodels for something with any
stakes um, is a good idea for me.
Yeah, no, I others may othersmay differ.

(01:14:52):
Um, I would, but I feel like,at this point, february 2025,
this may change in March orApril but I feel like the place
of something like AI is verylow-stakes territory.
I'm curious about something.

(01:15:14):
I want to see something.
Territory, right, I just wantto.
I'm curious about something, Iwant to see something.
I want to get the, you know,the AI's take on something, but
I'm not going to rely on it inany way, shape or form, and I
don't know if that where thatcomports with your conception of
this stuff, but I, you knowit's just, it's almost like a as
a journalist, I think stuff,but I, you know, it's just, it's

(01:15:39):
almost like a.
As a journalist, I think, um, Ihave a kind of an like an
allergy to relying on this stufffor for something that really
matters well it does.

The challenge is that it's uh, um,
it's really good in some, insome cases, Um, and you are a
journalist with integrity whohas respect for the craft, but I
think I don't know the studythat it was.
I think this is very hard toactually put a put a number on
Um, but there's a large portionof the content now on the

(01:16:12):
internet as AI generated and ifyou go and read any sort of like
the, just like sort ofclickbaity, SEO optimized blog
posts like a lot of that is AIgenerated.
I know ad agencies are using AIto write things like ad copy.
Like if you're doingprogrammatic ads, you now can
not just do an ABC test, you cando ABCD, efg, all the way, like

(01:16:33):
300, 10,000 hyper personalizedum.
But then there's like certainvery specific uh uh applications
where you know quality is notnecessarily the measure, there
isn't necessarily like sort ofan integrity aspect and you're
not necessarily even asconcerned about accuracy, yeah,
um.
So setting those aside, um, Iworry more about what you just

(01:16:57):
described.
So, if you're reading a reportand you're going to write about
it, the way to really be able towrite intelligently about
something and add value beyondjust all the noise because we're
in a world now where they'rejust like everybody has
something to say abouteverything and it's mostly
service, service level toactually add value, you have to
really like deeply understandthe thing that you're writing or

(01:17:18):
talking about, and I just donot think there is any
substitute.
I was talking to one of myboard members.
She's she's a corporate lawyer.
Um, she has her own, uh,consultancy now, um, and so she
advises.
You know really big, you knowblue chip clients and she gets
tapped to speak at these.
They're like, they're esoteric,um, you know legal conferences.
So she was like, I think, doinga conference on, um, you know

(01:17:42):
financial services, uh, you knowpolicy in in brussels and like,
oh, I've got this like keynotetomorrow.
It's presentation.
I have to make all these slides.
I have to like down, digestthis like 70 page um regulation
or something I was like well,perfect use case for ai.
Why don't you use ai tosummarize it?
I hadn't even gone to the placewhere I was still sort of

(01:18:04):
taking for granted that peoplethat ai was good at
summarization.
I have since found that itactually hallucinates maybe 20
of the time whenever I ask it tosummarize.
And so, setting that aside thefact that that just incorrect,
you should never use it tosummarize if you're trying to
like accurately reflect what'sin a report.
But let's just assume that,even if it was a hundred percent
accurate, her point was my likeprocess for preparing to go and

(01:18:25):
speak and the reason I'm tappedto speak is not because I have
I'm reading off of a like aseries of bullet points or
talking points that I'vememorized.
I actually am going up and likespeaking as an authority figure
and the only way that I'm goingto be able to like speak
authoritatively andintelligently about this topic,
like the process of actuallyreading those 70 pages and like

(01:18:48):
writing out everything that Iwant to say, like that's how I
get to a place where I'm readyto go and speak off the cuff,
and my experience with writinghas been absolutely like that's
how I get to a place where I'mready to go and speak off the
cuff, and my experience withwriting has been absolutely.
I just wrote something on deepseek on sub stack and I used it.
Actually, I will admit I usechat duty quite a lot and the
way I used it was I downloaded,I watched a lot of.

(01:19:09):
There's a lot of good YouTuberswho had sort of like done like
much deeper, longer, like a lotof the coverage was like you
know, a thousand words.
It's like hard to get into the,into the weed, so I wouldn't
watch like hour long YouTubevideos, got the transcripts,
plug that all in the chat GBThad to sort of synthesize it and
then created my outline andthen sort of like had chat GBT

(01:19:30):
sort of expand out in particularlike sort of like the technical
explanations of how DeepSeq waspotentially how DeepSeq was
trained.
But CJ and I worked on thistogether and I think what we
found is that it would get yousomething that's maybe like a C+
, maybe a B-, but to getsomething that's actually worth

(01:19:53):
publishing we basically rewrotealmost every word.
Now it was helpful to have alittle bit of a scaffold, but at
the end of the day, I'm stillunclear about whether the juice
is worth the squeeze, because Icouldn't have written that
without watching all thosevideos myself, and I think that

(01:20:13):
when you talk to folks,especially people who are really
deep in the AI space, this issomething that's really common.
Most people are not using theselanguage models.
The exception, I think, iscoding or data science.
I think there's someapplications where it's just
actually legitimatelygame-changing, but I think for
the humanities, I don't knowabout law.

(01:20:37):
I've only I only know the storyfrom the lawyer that got
disbarred or penalized forsubmitting a I don't know if it
was an injunction, some documentthat was like wildly
hallucinated and um right Making, making up cases or what have

(01:20:59):
you yeah?
Um, so my guess is, most big lawfirms are not allowing their
lawyers to.

Greg Toppo (The 74) (01:21:06):
I'm hoping that they're not what else are
you covering?
Uh, and if not covering whatelse?

are you just like tracking?

Greg Toppo (The 74) (01:21:11):
Yeah, we're just like tracking.
Yeah, we're, I mean, you know,as you'd expect we're.
You know we're covering theadministration and you know
seeing what's happening therewith you know their work, um, uh
, with, uh, executive orders andand whatnot, the fate of the
department of education.
Um, we're, uh, really stillinterested in the future of high

(01:21:33):
school.
I'm always on the hunt forsuggestions for the next school
to visit.
Just a couple weeks ago was upin Albany, visited a really,
really fascinating school calledTech Valley High School.
One of my favorite aspects ofit is that when you walk through

(01:21:55):
Tech Valley High School you seealmost no tech at all.
It turns out that this schoolwas built, was originated, at a
time when people were very, verybullish on the idea of like
incorporating schools into sortof like tech hubs.
And so there's this area inAlbany called Tech Valley and

(01:22:20):
the school sort of got got hisname from that.
But it's very it's, it's almost, you know, a humanities
centered high school that justis settled with the name Tech
Valley.
A really cool high school,really interesting, you know,
individualization, lots of lotsof project-based learning, lots

(01:22:40):
and lots of um real focus on umstudents ability to get up and
just like speak about topics.
Uh, really interesting place,and I think it's it's a model
that I that I'm really seeingmore and more, which is schools
that are focusing on givingtheir students the ability to be

(01:23:05):
experts, if you will.

If there's someone listening, that is the
head of a school and they wantto reach out to you.

Greg Toppo (The 74) (01:23:13):
We'll share your contact information or
your email or whatever, I getthe 74 million on ORG.

What?
What advice would you give to abuilding principal who is
trying to pitch their schoolLike what are the types of
things you're looking for?
I'm sure you get a lot.
I mean there's we actuallydon't get a lot of pitches.

Greg Toppo (The 7 (01:23:33):
Interestingly enough, we've been writing
these stories for a while nowand I do not get emails from
people saying come visit peopleto know if they're really doing
something innovative.

Yeah, I suppose more powerful for you to
hear about a school notdirectly from someone who's sort
of like promoting it.

Greg Toppo (The 74) (01:24:08):
No, I think that's right.

Maybe just some parting advice for folks.
I mean, I think one of thecommon threads of this
conversation is, you know, youclearly have this curiosity.
It's been part of your entirecareer and it's it hasn't always
been ai.
It was really sort of much morebroad about sort of like the
future of education, um, but youare now really, I think, one of
like a very small group of ofjournalists who are, true, I

(01:24:33):
don't know if you call yourselfan expert, but but as close as
you can get to someone who justdeeply understands the
intersection of these spaces,what advice would you give to
our audience in terms of folkswho are early on in the learning
journey?
They get it.
Yeah, we need to learn more.
Maybe they want to use ChatGPTand experiment, but are there

(01:24:53):
any specific outlets that you'refollowing?
Maybe they want to use chat,gpt and experiment, but are
there any like specific likeplace the outlets that you're
following?
You know, I don't know, is itlike Reddit, like, like, how can
they sort of just like stay,keep their finger on the pulse?

Greg Toppo (The 74) (01:25:02):
Hmm, I mean , my problem is, I read
everything?
So what are your?

like what's the best, what's the what's?

Greg Toppo (The 74) (01:25:11):
the I mean what actually?
Um.
Interestingly enough, um youknow, one of the podcasts that I
like is the times podcast aboutum tech the hard fork.
Okay.

Um love those guys.

Greg Toppo (The 74) (01:25:22):
Kevin Bruce Um, I I think his coverage
specifically is really um,really really well done.
Um, and he's covered a lot ofthe education slash, safety
aspects of it.
Um, you know I uh, I cover, orexcuse me, I, I, I, I read and
follow.
You know the the skeptics, ifyou will Um, you know the Ben

(01:25:47):
Riley.
He's got a um uh, cognitiveresonance.
Who you?
Um I'll I will actually um ifyou don't mind you um I'll I
will actually um if you don'tmind, my um flogging you guys as
incredible podcasts that youdid with the disagreement, which
I'm hoping, by the time thisruns, it will be public.
Um, you guys uh and I even saidit at the time, I think, on

(01:26:11):
Twitter, x or wherever the heckit is Um that the conversation
you guys had back in the fallwas like the best conversation
I've heard a very high level,with lots and lots of respect
for one another, but also wereable to very carefully pry the

(01:26:52):
threads from where you disagreedand not scream at each other.
I really enjoyed that, so Iwould recommend that highly.
Um, I follow um Dan Myers blog.
Um, he's got a very specificfocus on math Um, and I think
that's a really important umthing to think about when we're

(01:27:16):
talking about this world.
You know I've always been a bigfan of Audrey Waters, who never
met a piece of technology sheliked, and it's interesting
because, like somebody likeAudrey, for people who may not

(01:27:43):
know very much about her, butyou know she's been thinking and
writing about this stuff foryears and years and I think one
of the the sort of the bigthemes she has really been
focused on for as long as I'veknown her, is this idea that
Silicon Valley is not yourfriend, that there's a lot of

(01:28:08):
dark thinking coming out of thisplace and I uh, just speaking
from, uh you know some of thestuff we're seeing with our new
administration.
I mean I, without getting likepolitical about it, I mean I
think you can see theself-serving aspect of a lot of

(01:28:31):
what these guys are doing, and Idon't think that's something we
saw very clearly until now.

I think it's surprising they got to pass
for as long as they did,because I don't think people
would look at car companies, oilcompanies, electricity
companies, all these megalithcapitalist institutions.
Their incentive structures arehighly oriented towards one,

(01:29:00):
only one thing.
Yeah.
And they will push in whateverdirection that they can right.
So I think, but I guess you're,I guess I think it's
interesting that you're namingsome skeptics, because I
actually think that is.
We have a lot to learn from theskeptics, because there's
plenty of optimists and there'sa lot of legitimate optimists

(01:29:22):
who are super smart, but thenthere's also a lot of optimists
who are shilling right, Likethey are self-interested,
whether it's they're either atcompanies or invested in those
companies invested in the futureof AI being successful and I
think the kind of sift throughall because I think the origin
of my question, what I want toleave
with the audience is like yes,there's a lot of noise, you

(01:29:42):
don't have to read every singlething.
Not everybody is a journalistcovering AI in education but I
think it is important to findsome smart people that you trust
.
And when you're trying tofigure out, is this like, legit?
Like, is this legit, is there athere there?
You don't want to be relying onthe.

(01:30:03):
You don't necessarily be on thebandwagon while you're
evaluating that.
And especially for decisionmakers who are education systems
.
I think they're dealing withthis a lot right now.
A lot of AI conferences are 95%of the people who are there and
the sponsors are allaggressively pro-AI and AI

(01:30:24):
products.
In reality, I think it's muchmore of.
I don't know if it's 50-50exactly, but it's definitely not
95% good and 5%.
We just have to figure it out.
It's much more complex um and sothe information ecosystem is
not.
I don't.
I don't think it's verybalanced, I think it's oriented
very heavily and then you have,like the, and a lot of the

(01:30:45):
skeptics or a lot of thepessimists are not useful
because they're like theexistential risk folks who are
worried about, you know thesingularity.
I find that a very interestingphilosophical discussion.
I just don't think there'sanything actionable that
certainly an education leadercan do.
Besides, pay attention and keepyour eye out for AGI.

Greg Toppo (The 74) (01:31:05):
I mean to be fair.
I mean a lot of skeptics, Ithink, have a different view of
education than maybe you or Imight have.
I mean they see it as adifferent kind of enterprise and
some wouldn't even be willingto.
You know entertain, you knowfunctions of a teacher, you know

(01:31:29):
being, you know, augmented incertain ways by technology.
So you know, I want to be clearon that.
I guess where I come into it isI'm really interested to read

(01:31:50):
the and to talk to the.
You know the makers, thevendors, the, the people who run
a lot of these outfits, just tosee what they think they're
doing, like what, you know wherethey, what their place is and
all this.
And some of them have reallyinteresting ways of thinking
about this stuff, um, and reallyinteresting ways of thinking

(01:32:10):
like.
You know, here's our niche orhere's all we want to do.

But I guess you can't plug yourself but I
will.
I think, while it's noteverybody can necessarily talk
to, you know the founders ofcompanies like this.
I think you know reporting likeyours.
You're not blindly excited, butyou're intensely curious.
But you're intensely curiousand I think you're willing to.
I think his orientation towardsthe future of high school is an
interesting place to come atthis from as a journalist,

(01:32:43):
because your incentive is forschools to transform for the
better, and so there's a desirefor things to be.
I can sense a desire for usrealizing the potential that AI
has, and yet I think you've beenin this long enough that you
know that there's generally moreto it than some of the first

(01:33:06):
claims, I suppose, that are made.

Greg Toppo (The 74) (01:33:08):
Yeah, no, I think that's right.
I think that's right.

All right, greg.
Thank you so much for coming by.
It's been fun.

All Episodes

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

CrimeLess: Hillbilly Heist

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Greg Toppo: The promise and perils of AI in schools

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Stuff You Should Know

Dateline NBC

CrimeLess: Hillbilly Heist

All Episodes

Greg Toppo: The promise and perils of AI in schools

Stuff You Should Know