Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome back to the deep dive. Today, we're diving into
something really fundamental, maybe the most frustrating challenge in education today.
It's this impossible equation. How do you give that personalized
one on one attention that really helps kids thrive? You know,
when you've got these massive classes, budgets stretch to the limit,
(00:21):
and teachers who are just completely overloaded.
Speaker 2 (00:24):
Yeah, that tension, that's right where our source material for
MESA Doubt Solutions kicks off.
Speaker 1 (00:28):
Yeah.
Speaker 2 (00:28):
Their whole argument is that the way we do things now,
you know, one teacher, thirty kids, everyone learning at the
same speed, it's basically obsoletely strong work. Well, they argue
it was built for the industrial age, like a factory line.
It just doesn't fit the information age we're actually living in.
Can't handle the complexity.
Speaker 1 (00:45):
And you really feel the human impact of that old
system and the examples they give. They talk about Sarah,
a seventh three teacher. It's quite moving, actually right. She says,
I know exactly which students need help, and I know
exactly what specific help they need. I don't have enough
of me to go around. The knowledge is there, but
the capacity just isn't.
Speaker 2 (01:05):
It's heartbreaking, really and then you flip it look at
the student side, like Marcus, this kid brilliant apparently, but
started failing algebra okay, And it wasn't because he wasn't
smart enough, is because he needed a different approach. He
needed someone asking him questions socratic style to really get it,
not just memorizing formulas.
Speaker 1 (01:26):
And his teacher couldn't provide that level of personalized back and.
Speaker 2 (01:29):
Forth exactly just physically impossible in that setting, so he
slipped through the cracks. It's not about bad teachers, it's
about a system that's feeling both the teachers and the students.
Speaker 1 (01:38):
The tragedy is we know what good personalized learning looks like,
we just can't scale it up with the current resources.
And that's where MISA comes in, Right, that's a big promise.
Speaker 2 (01:47):
That's the promise technology specifically AI that isn't about replacing Sarah,
but about amplifying her, giving her the tools to deliver
that personalization. And this is teaching students how to think,
not just what to think, okay.
Speaker 1 (02:04):
So amplifying teachers and teaching thinking skills.
Speaker 2 (02:07):
So what's really interesting is they're not basing this on
some brand new, untested idea they're leaning into something ancient,
really proven for what over twenty four hundred years?
Speaker 1 (02:16):
Right, let's pivot to that. The science bit section one
in the material really digs into socratic learning. The idea,
which goes way back, is that you understand something much
more deeply. It sticks better if you discover it yourself
through questions rather than just being told the answer.
Speaker 2 (02:32):
H And the modern research backs us up, like overwhelmingly.
They point to work by doctor Patricia Williams at Stanford.
The numbers are pretty staggering.
Speaker 1 (02:41):
Okay, hit me with the stats.
Speaker 2 (02:42):
All right, So students using Socratic methods a seventy three
percent better retention rate after six months.
Speaker 1 (02:47):
Seventy three Wow, Okay, that's that's not a small difference.
That suggests it's really sinking indifferently. What else you mentioned
emotional metrics, Yeah.
Speaker 2 (02:54):
This is fascinating. They also found an eighty four percent
improvement in critical thinking skills. Yeah, okay, make sense. Yeah,
a ninety one percent jump in engagement also logical. But
then get this, a sixty seven percent reduction in math anxiety.
Speaker 1 (03:10):
Wait, sixty seven percent reduction in anxiety. That feels huge.
That's emotional that changes lives, does the research explain why
how does asking questions reduce fear?
Speaker 2 (03:19):
It does? Yeah, Doctor Williams insight is that when the
student is guided to the discovery, when they build the
knowledge instead of just receiving it, they stop seeing maths
as just being right or wrong. It because a puzzle
they can solve. The fear of failure, the shame it
kind of dissipates because the AI guide is patient, it's
non judgmental, it's just this ongoing, personalized conversation, no social pressure.
Speaker 1 (03:41):
That makes a lot of sense. They shift from being
passive receivers to active thinkers. Like doctor Williams says, it
teaches them how to think. And in a world where
information changes constantly, that how is the skill that lasts right.
Speaker 2 (03:53):
Absolutely, But and here's the catch, Socratic teaching done right
is incredibly I'm consuming. It needs constant, tailored interaction for each.
Speaker 1 (04:04):
Student, which brings us right back to Sarah's problem. One
teacher just can't manage thirty different Socratic dialogues at the
same time. It's humanly impossible.
Speaker 2 (04:13):
And that's the gap MESA is trying to bridge. Their
big innovation is building an AI that is specifically trained
not to just give the answer, which, let's be honest,
is what a lot of current AI tools might default.
Speaker 1 (04:25):
To, like a quick chat GPT search for the solution exactly.
Speaker 2 (04:29):
MYSA is designed to ask the right next question, the
perfect personalized prompt to nudge the student forward.
Speaker 1 (04:35):
Okay, so let's see it in action. How does this
AI teaching assistant actually work? Because they're very clear, right,
it's an assistant, not a replacement.
Speaker 2 (04:42):
Very clear on that. It's about giving teachers superpowers, managing
that intense cognitive load of individual questioning, but keeping the
human teacher firmly in charge of the relationship and the
overall learning environment.
Speaker 1 (04:53):
So walk us through an example. The source material had
a geometry one right, a ninth grader Jennifer stuck on triangles.
Speaker 2 (04:59):
Yeah, so Jennifer types something like I don't get triangles,
or maybe something about a specific proof. Instead of launching
into a definition the MESA AI might ask something gentle like, Okay,
before we tackle the proof, what do you remember about
the sides of a triangle?
Speaker 1 (05:14):
Okay? Starting with what she knows?
Speaker 2 (05:16):
Right, and if Jennifer says maybe all the sides are equal,
Lisa doesn't just say wrong, it probes hmm, are they
always equal? Can you think of a triangle where there
are different lengths?
Speaker 1 (05:28):
Ah guiding her to refine her own understanding exactly.
Speaker 2 (05:31):
It walks her through its step by step, maybe asking
next how side lengths relate to the angles opposite them,
guiding questioning until Jennifer herself discovers the property or the
step and the proof. It's precisely that patient focused interaction
Marcus needed for.
Speaker 1 (05:46):
Algebra, and you mentioned it's not just for Mather science.
The literature example was quite powerful too.
Speaker 2 (05:51):
Oh yeah. A student asks why did Romeo and Juliet
have to die? A typical chatbot might give a summary.
Speaker 1 (05:58):
Of themes right, fate, family, etc.
Speaker 2 (06:01):
But MESA asks questions like what do you think their
deaths achieved that maybe their lives couldn't, or what does
their tragedy tell us about the society they lived in,
about the cost of hatred? It pushes the student towards deeper,
more profound analysis.
Speaker 1 (06:18):
That's actually critical thinking prompted by AI. But how does
the teacher, say, missus Johnson with her algebra class, keep control?
Does the AI just go off on tangent?
Speaker 2 (06:28):
No, that's where the teaching assistant part is crucial. MISA
integrates with the teacher's curriculum. So Missus Johnson sets the
topic for the day, maybe solving linear equations. Okay, well
she's working with a small group needing extra help. MESA
is simultaneously doing three different things, maybe giving challenge problems
socratic style to the advanced kids, guiding the middle group
(06:49):
through practice exercises with helpful questioning if they get stuck,
and importantly flagging the exact points of confusion for the
students who are really struggling.
Speaker 1 (06:56):
So it's providing instant differentiation within the lesson Missus john
and planned she sets the direction. MISA handles the individual pathways.
Speaker 2 (07:04):
Precisely, which leads us naturally into the data aspect section three.
Because traditionally, understanding how a students learning has been kind
of a black box.
Speaker 1 (07:14):
Right, Yeah, you grade the test, you see the mistakes,
but you don't always know the thought process behind the mistake.
Was it a calculation error, a conceptual misunderstanding something else
entirely exactly?
Speaker 2 (07:26):
MESA changes that because it's capturing the process. It generates
really specific, granular data on how students are approaching problems,
not just whether they got the right answer and.
Speaker 1 (07:37):
What kind of data are we talking about? More than
just pass fail?
Speaker 2 (07:39):
You said, oh, yeah, much more. It tracks things like
the different problem solving strategies students try, common patterns of misconceptions,
like does a student consistently confuse dividing fractions with multiplying them?
Speaker 1 (07:52):
Useful diagnostic very.
Speaker 2 (07:53):
It also attracks confidence levels, engagement metrics, how often are
they interacting, how long do they persist, and something they
call learning velocity.
Speaker 1 (08:01):
Learning velocity Okay, unpack that.
Speaker 2 (08:03):
One for us, sure. Learning velocity is basically how quickly
a student is mastering new concepts, moving from first seeing
something to being able to use it confidently and independently.
Tracking that lets teachers in schools spot kids who are
starting to fall behind way earlier than waiting for the.
Speaker 1 (08:19):
Next big test, and this data actually leads to changes.
The source mentioned and administrator David Thompson.
Speaker 2 (08:26):
Right a great case study his charter school brought in MESA,
and pretty quickly the data revealed this systemic issue. Six
graders were consistently weak on fractions, It wasn't being caught effectively,
and it was basically setting them up to fail in
eighth grade algebra.
Speaker 1 (08:41):
Ah So, Mesa's analysis of thousands of those socratic interactions pinpointed.
Speaker 2 (08:46):
The root cause exactly, much clearer than test scores alone
could show. So the school adjusted the sixth grade curriculum
based on that specific data, and there is improvement, a
big one there. Algebra success rate jumped by thirty four percent,
the next.
Speaker 1 (09:00):
Thirty four percent. That's transformative. That changes trajectories, and I
can see why for charter schools that kind of data
is absolute gold totally.
Speaker 2 (09:09):
It proves learning outcomes, lets them fix gap super quickly
and make curriculum decisions based on real evidence of student thinking,
not just guesswork.
Speaker 1 (09:19):
It sounds like Lisa is positioning this less as selling
software and more is building a partnership very much so.
Speaker 2 (09:24):
They emphasize that they want educators involved in shaping the tool.
Speaker 1 (09:28):
So let's break down the benefits they offer in this
partnership model. For teachers like Sarah, what's in it for them?
Speaker 2 (09:35):
Huge time savings reportedly sixty percent less time on grading
and repetitive tasks, instant insights into student thinking without having
to manually review everything, plus professional development built in it
lets them shift their energy to the human stuff, mentoring, motivating,
connecting well.
Speaker 1 (09:51):
The AI handles the socratic heavy lifting.
Speaker 2 (09:55):
You got it. And for schools, the pitch is affordability,
no need for new hardware where it's web based, it's
customizable to their existing curriculum, and critically, the school keeps
ownership of the student data. They're not selling off insights
into how their kids think.
Speaker 1 (10:10):
That data ownership point is crucial. And what about tutors
you mentioned them earlier.
Speaker 2 (10:14):
Yeah, for tutors, MESA acts as a force multiplier. A
single human tutor can oversee say five or six students
using MESA simultaneously. They can provide that expert guidance when needed.
But MESA handles the constant back and forth and offers
that two hundred and forty seven support between tutoring.
Speaker 1 (10:31):
Sessions, which is often when students actually need the help,
right late at night doing homework exactly.
Speaker 2 (10:37):
And to get schools on board early, they have this
founding partner status offer for the first one hundred schools.
Speaker 1 (10:43):
Okay, what does that involve?
Speaker 2 (10:44):
A permanent fifty percent discount on all pricing, which is
pretty significant, plus direct access, monthly calls with the MESA
development team, real input into how the product evolves. They
genuinely seem to want early adopters to help build it
with them.
Speaker 1 (10:58):
Okay, theory ruckshire partnership model check. Let's talk real world results.
Speaker 2 (11:04):
Yeah.
Speaker 1 (11:04):
You mentioned Lincoln Charter Academy, a Title I school.
Speaker 2 (11:08):
Yeah, a powerful example. Many students there simply couldn't afford
tutors or homework help once they left the school building.
They were kind of on their own if they got stuck.
Speaker 1 (11:16):
So MESA filled that gap.
Speaker 2 (11:17):
The principal, Maria Rodriguez called it the tutor Our students
couldn't afford that two ing force. Faith access was key.
Kids doing homework at ten PM or kids too shy
to ask in class could get that patient socratic help
right when they needed it. And the results pretty impressive.
Math proficiency scores went up twenty eight percent in just
one semester, and they also stressed the visible growth in
(11:41):
student confidence, which is harder to measure but just as important.
Speaker 1 (11:44):
Twenty eight percent in a semester is fast.
Speaker 2 (11:46):
Okay, what about the tutoring service example elite tutoring.
Speaker 1 (11:49):
Right, they were hitting a wall in terms of how
many students their human tutors could handle. MESA allowed them
to basically double their capacity. Tutors could effectively manage more
students because MESA handled the between session support and practice, so.
Speaker 2 (12:03):
They doubled revenue while improving outcomes. That's a strong business case.
Speaker 1 (12:06):
Seems so. And then there's the story of Tom Bradley,
the rural science teacher. This one really highlights the differentiation power.
Speaker 2 (12:13):
Also, he had.
Speaker 1 (12:14):
Kids in the same classroom with reading levels spanning from
third grade all the way up to tenth grade, an
incredibly tough teaching challenge. Wow, yeah, how do you even
teach a single lesson like that?
Speaker 2 (12:26):
Well, with MISA, he could teach the same core topic,
climate science in this case, but the platform delivered the
background reading and the follow up socratic questions tailored to
each student's actual reading and comprehension level. Everyone learning about
the same thing, but at a level they could access
and be challenged by.
Speaker 1 (12:45):
That's genuine differentiation at scale within one classroom. That's pretty remark.
Speaker 2 (12:50):
It really is, which kind of brings us to the
bigger picture, the urgency they talk.
Speaker 1 (12:54):
About, right, The source material argues this isn't some far
off future thing. The AI revolution and education is happening now.
Speaker 2 (13:01):
Yeah, the choice for educators isn't if AI will be used,
but how will they be the ones shaping its use,
leading the charge, or will they just get swept along
by it?
Speaker 1 (13:11):
And that connects to the whole issue of students already
using AI. Doesn't it like using chat GPT as a
shortcut exactly?
Speaker 2 (13:17):
Kids are smart, they find the loopholes. Many are using
current AI as a cheat code, just grab the answer,
paste it in no real thinking.
Speaker 1 (13:23):
Involved, undermining the whole point of the assignment.
Speaker 2 (13:26):
Right, VISA positions itself as the productive alternative. It uses
AI not to shortcut thinking, but to deepen it. It
teaches students to see AI as a thinking partner, a
tool for discovery, not just forgetting answers.
Speaker 1 (13:42):
So it's about shifting the relationship with the technology itself,
making it a tool for learning, not a way to
avoid it precisely.
Speaker 2 (13:48):
And that leads to the ultimate takeaway from the source material.
MESA aims to democratize socratic learning, to make sure every student,
whether they're like Marcus needing specific socratic probes or as
student at a tidalized school needing help late at night.
Speaker 1 (14:02):
Has access to that kind of personalized, patient effective tutoring,
regardless of their background, their school's budget, or their zip code.
Speaker 2 (14:10):
It's a big vision, but it's focused on what really
matters in education, because, let's face it, years later, students
rarely remember specific facts, true, but they do remember if
they were taught how to think, how to ask questions,
how to figure things out for themselves, that's the skill
that sticks.
Speaker 1 (14:25):
Okay, So that leaves us and you the listener, with
a final thought to chew on. Building directly on this
deep dive, we heard that stunning statistic a sixty seven
percent reduction in math anxiety. So think about this. If
AI like MISA can consistently chip away at that fear,
that debilitating anxiety around subjects like math, what's the long
(14:46):
term impact? How might that change student's willingness to even
try challenging fields, And could that, a generation down the line,
actually reshape the diversity and make up of the global
STEM workforce.
Speaker 2 (14:56):
It's a powerful question, something.
Speaker 1 (14:58):
I'm all over. Thanks for diving deep with us today,