GEF: Redefining Teaching with AI

Summary

The roundtable discussion, led by Anel Aurora from Advanced Group, explores the impact of AI on education. The conversation highlights the potential for AI to improve teaching and learning, focusing on how it can support teachers and enhance student engagement. Participants, including educators and entrepreneurs, discuss challenges such as equity, cost, and quality, and emphasize the importance of human relationships in education. They envision AI as a tool to assist teachers in lesson planning, student assessment, and personalized learning, while cautioning against over-reliance on technology and stressing the need for intentional implementation.

Takeaways

• 🌟 The discussion aims to redefine teaching through AI, focusing on improving outcomes for young people by enhancing the teacher-student interaction, which is the core of the educational ecosystem.
• 🤖 AI's potential role in education includes assisting teachers with lesson planning, content creation, and student assessment, thereby freeing up time for more valuable activities with students.
• 🌍 Panelists from various regions highlighted the diverse challenges and contexts in education, such as low-income schools in Latin America and the unique needs of the Middle East and North Africa.
• 📱 The availability of technology in classrooms can be as limited as a single teacher's smartphone, emphasizing the need for AI solutions that work in low-tech environments.
• 🔢 Concerns about the cost of deploying AI in education were raised, particularly for public sector education in countries with limited budgets.
• 👩‍🏫 The importance of not replacing teachers with AI but empowering them was emphasized, focusing on AI as a tool to support teachers in their challenging roles.
• 🚀 The potential of AI to help with the automation of mundane tasks, continuous formative assessment, and personalized learning was discussed as a way to enhance teaching and learning.
• 🌐 The digital divide was identified as a significant issue, with concerns that AI might widen gaps in education if not implemented thoughtfully and equitably.
• 🤝 The group consensus was that AI should be used to support collaboration and project-based learning, rather than as a solitary learning tool.
• 💡 There was a call for intentionality in the use of AI, ensuring it aligns with the broader vision of education and the needs of students, rather than being an end in itself.
• 🌱 The human element of teaching, including happiness, engagement, and relationships, was underscored as irreplaceable and vital for effective learning.

Q & A

What is the main topic of discussion in the provided transcript?

The main topic of discussion is redefining teaching through the use of Artificial Intelligence (AI) and its potential impact on improving educational outcomes for young people.

Who is Anel Aurora and what is her role in the discussion?

Anel Aurora is from Advanced group and works closely with SE on various projects. She is the moderator of the discussion and sets the stage for the panelists to explore the role of AI in redefining teaching.

What is the role of Wise in the context of this discussion?

Wise, represented by Stavros Yanuka, is a global think tank of the Kar Foundation that studies innovation in education. It is implied that they might be exploring how AI can innovate in the education sector.

What challenges does the education system in Latin America face according to Jose Rafa?

Jose Rafa highlights challenges such as low-skilled teachers, low-tech environments, lack of internet connection, and the need for teachers to balance multiple jobs due to low pay, which affects the quality of teaching and learning.

How does Unice Bensma view the potential of AI in education in the context of public sector education?

Unice Bensma discusses the potential of AI to reduce deployment costs and make AI tools more accessible in countries where the cost of subscription models like Chat GPT is too high, by using smaller models and innovative data collection methods.

What is the main concern that John Morak expresses about the use of AI in education?

John Morak expresses concerns about the quality of AI, the rush to use it without proper quality controls, and the ethical implications of using AI without a clear understanding of its impact on education.

What does Fernando Castillo suggest as a potential role for AI in classrooms?

Fernando Castillo suggests that AI could support collaboration, project-based learning, and inquiry-based learning by capturing live classroom discourse and providing insights to teachers about student misconceptions and ideas.

What is Glenn Kimman's view on the role of AI in classrooms of the future?

Glenn Kimman sees AI agents as another element in the classroom, playing various roles to support teachers in their tasks, while emphasizing the importance of maintaining the human aspects of teaching and learning.

What is the potential risk of using AI in education that Janet Vbe highlights?

Janet Vbe highlights the risk of widening the digital divide, where AI might be used as a substitute for human teachers in lower socio-economic schools, potentially leading to less engagement and a poorer learning experience.

How does Ian Martinez relate the use of AI in education to student engagement?

Ian Martinez suggests that student engagement comes from relationships with teachers and classmates, and warns that using AI as a delivery system for personalized education could lead to a lack of engagement if it replaces human interaction.

What is the key takeaway from the discussion according to the closing remarks?

The key takeaway is the importance of intentionality in using AI in education. Educators should focus on the big challenges they face and consider how AI can be used as a tool to help meet those challenges, rather than just adopting AI for its own sake.

Outlines

00:00

🌟 Introduction to Redefining Teaching with AI

The script opens with Anel Aurora warmly welcoming participants and setting the stage for a discussion on redefining teaching through AI. Despite last-minute changes, the panel, consisting of various experts, is introduced. The conversation aims to focus on improving educational outcomes for young people by examining the role of AI in enhancing the teaching and learning experience. The panelists share brief introductions of themselves and their organizations, highlighting their diverse backgrounds in education, AI, and global think tanks.

05:01

🚀 Envisioning AI in Education by 2030

Jose from Men Latin America discusses the potential of AI to support low-income schools in Latin America, where technology is limited. He emphasizes the need for AI to assist overworked and under-skilled teachers by providing high-quality educational tools and automating tedious tasks. Unice from Life Working in the Mina region focuses on deploying AI systems nationally while reducing costs, suggesting the use of smaller models and data collection through the systems themselves. The discussion hints at the challenges of scalability, quality, and ethics in AI integration.

10:04

🛠️ The Role of AI in Enhancing Teaching and Learning

John, an education futurist, warns of the dangers of rushing AI implementation without ensuring quality. He stresses the importance of human control over AI tools and the need for ethical considerations. The conversation explores the potential of AI to act as a personal coach for students and teachers, with the goal of improving educational engagement and outcomes. Concerns about the digital divide and the potential for AI to exacerbate existing inequalities are raised.

15:05

🤖 AI as a Collaborative Tool in the Classroom

The panelists discuss the various roles AI could play in the classroom, such as assisting teachers with lesson planning and providing personalized learning experiences for students. Concerns about the potential reduction of the teacher's role and the risk of increasing achievement gaps are addressed. The conversation highlights the importance of intentional use of AI to support, not replace, human interaction and collaboration in learning.

20:06

🌐 Bridging the Digital Divide with AI in Education

The discussion turns to the challenges of deploying AI in different educational contexts, particularly in low-income environments. Panelists emphasize the need for AI solutions that do not widen the digital divide and instead assist in closing educational gaps. The importance of creating AI tools that are contextually relevant, affordable, and capable of enhancing the teacher's role is underscored.

25:08

🌱 The Human Element in AI-Enhanced Education

The panel reflects on the importance of humanity in education and the role of AI as a tool to augment, not replace, human connection and emotional engagement. There is a call for balance, ensuring that AI implementations do not detract from the joy and personal connections that are vital to the learning process. The conversation suggests that AI should be used to free up teachers to focus on the emotional and relational aspects of teaching.

30:08

🔮 Looking Ahead: The Future of AI in Education

In the final paragraph, the panelists consider the future implications of AI in education. They discuss the potential for AI to support collaboration and classroom interaction, as well as the need for careful implementation to ensure that AI serves to enhance, rather than detract from, the educational experience. The conversation concludes with a call to focus on the challenges faced by educators and how AI can be intentionally used to address these challenges.

Keywords

💡Artificial Intelligence (AI)

AI refers to the simulation of human intelligence in machines designed to think and learn like humans. In the video, the discussion centers on how AI can redefine teaching by supporting teachers, automating repetitive tasks, and enhancing the learning experience for students, especially in low-tech environments.

💡Teacher Support

Teacher support involves providing tools and resources to help educators perform their tasks more effectively. The video highlights how AI can assist teachers in planning lessons, assessing students, and managing classroom activities, thereby improving their efficiency and effectiveness.

💡Student Engagement

Student engagement is the level of interest, curiosity, and involvement students show towards their learning. The video addresses concerns about maintaining student engagement in an AI-supported classroom and suggests that AI can help personalize learning, making it more relevant and engaging for students.

💡Educational Equity

Educational equity ensures that all students have access to the same learning opportunities, regardless of their background. The video discusses the potential of AI to bridge educational gaps in low-income areas by providing high-quality resources and support to teachers and students who might otherwise lack them.

💡Low-Income Schools

Low-income schools are educational institutions in areas with limited financial resources. The video highlights the specific challenges faced by these schools, such as lack of technology and trained teachers, and how AI can help address these issues by providing accessible and effective educational tools.

💡AI Tools

AI tools are applications or software that utilize artificial intelligence to perform tasks. In the context of the video, AI tools are used to assist teachers in creating lesson plans, grading, and personalizing student learning experiences, thus reducing the administrative burden on educators.

💡Collaboration

Collaboration in education involves working together to achieve common goals. The video discusses how AI can support collaborative learning by facilitating communication and interaction among students and between students and teachers, thus enhancing the overall learning experience.

💡Vision for AI in Education

The vision for AI in education refers to the anticipated future state where AI is fully integrated into teaching and learning processes. The video outlines an aspirational vision where AI aids in personalizing education, supporting teachers, and making high-quality education accessible to all students.

💡Challenges

Challenges refer to the difficulties and obstacles in implementing AI in education. The video mentions several challenges, including the high cost of AI deployment, the need for quality control, and ensuring ethical use, particularly in diverse and low-resource educational settings.

💡Implementation

Implementation is the process of putting AI technologies into practice within educational systems. The video emphasizes the importance of intentional and context-specific implementation to ensure AI effectively enhances teaching and learning without widening existing educational disparities.

Highlights

Introduction of the roundtable focused on redefining teaching through AI.

Panelists share their backgrounds and organizations, emphasizing their work in AI and education.

Jose discusses the challenges in Latin America with low-income schools and the potential of AI to assist low-skilled teachers.

Eunice highlights the focus on reducing deployment costs for AI in public sector education and the importance of making AI affordable for widespread use.

John raises concerns about the quality and scalability of AI in education, emphasizing the need for human oversight and control.

Fernando emphasizes the need for AI to support collaborative and project-based learning in addition to individual learning.

Discussion on the potential of AI to help teachers manage classrooms and provide personalized feedback to students.

Glen warns about the risk of widening educational inequities with AI, especially in less affluent communities.

Panelists agree that AI should be used to enhance, not replace, the role of teachers in the classroom.

Jose and Eunice discuss the importance of local context and culturally relevant content when deploying AI in education.

Discussion on using AI to automate administrative tasks, allowing teachers to focus more on high-value interactions with students.

Importance of ensuring AI tools are designed to support the emotional and social aspects of learning, not just academic outcomes.

Concerns about AI being used predominantly in low-resource settings, potentially leading to a lower quality of education for those students.

AI's role in capturing classroom interactions and providing teachers with insights to improve teaching strategies.

Final thoughts on the necessity of having a clear purpose and intentionality when integrating AI into education to address existing challenges effectively.

Transcripts

good afternoon everyone um thank you

thank you all for being here I know

there have been some changes last minute

on some of the folks um that are going

to be here but a very very warm welcome

um my name is anel Aurora from Advanced

group um work closely with with SE on on

a couple of projects um today we're

going to talk about redefining teaching

through AI um and with AI and I know

there's been so many topics over the

over the course of the morning already

and so many discussions around AI um we

have conversations at assist Sy level we

have conversations um on leadership we

have conversations on curriculum on data

but ultimately if we think about the

purpose of what we're trying to do we're

trying to improve outcomes for young for

young people and the interaction between

a teacher and a student is the

centerpiece and front you know of of all

of that broader ecosystem so I to me in

a bi maybe I'm biased I think this is

the most important topic which is the

teaching and learning experience and how

that gets fected and potentially

Amplified by AI um we have a wonderful

group of of panelists um and Roundtable

members here I thought we could just

start with a very very brief

introduction if you could just please

share your name um the organization um

that you're from and maybe just one

sentence um on the organization um and

maybe stav we could start with you if

that's all right okay just because to

keep it really short stavos yanuka I'm

CEO of wise which is a global think tank

of kar Foundation

that uh studies Innovation

education hi Fernando Castillo Nando I'm

the chief learning officer at mentu we

are a startup that's building AI um

tools for

teachers um hello I'm Jose Rafa esposa

I'm the CEO of

men good morning everyone Ian Martinez

head of school ATK International School

Riyad good morning I'm Janet vbe I'm a

director of a symphony zor Academia and

uh I'm also working for operation

education in the

Netherlands I'm John morac and I know

operation education uh pretty well in

the Netherlands uh but I'm an education

futurist I do a lot of Consulting and

research mostly for the United Nations

system um and also edited a journal on

the horizon which is the international

Journal of learning

Futures hi everyone I'm Glenn kimman I'm

a senior adviser at the Stanford

University accelerated for learning and

also here as part of the Coen Consortium

on school networking uh group thank you

I'm Unice bensa Mori I'm an adjunct

lecture of computer science at Stanford

and also founder of life. working on AI

education for public

sectors thank you thank you all um so

the way we were going to organize the

round table is really in three parts um

we're going to start a little bit with a

visioning um of what folks think um a

vision for redefining teaching could

look like um a second is a little bit

around the challenges that might be

there and the last is potentially a path

for educators in schools today what

could be activities and and undertakings

to kind of go on the positive path um

towards that Vision so in starting with

that Vision um I was speaking with a

friend this morning who was like you

know I go to lots of panels but I don't

know if I get some Earth shattering

Insight so to put you on the stage um

maybe we have two entrepreneurs here who

are working on trying to redefine Vision

um of of how AI can be used in teaching

and I thought maybe we could start with

the two of you Jose and Eunice um

sharing a little bit of how you see

let's say if we were to look five years

out maybe 2030 um what could be the

vision of what teaching and learning

would look like if AI was used in a best

case kind of scenario um would either of

you like to open you can go if you want

um okay thank you um so we we just a

little bit of context we are um Latin

America Focus organization we work

mainly with lowincome schools private

both private and public schools but uh

low income um and we that that means

that we have several constraints right

we we work in lowtech environments the

only uh technology available or device

available um in the school in the

classroom is the teacher smartphone and

probably they don't even have internet

connection

um or it's likely um and uh

unfortunately there are there are low

skilled teachers um probably they are uh

teaching math even though they were not

Tre trained in math um so that that to

give you just a little bit of context

according to the regulation they should

be spending 70% of their time working in

class with students and 30% of their

time preparing lessons reporting is the

truth is they do spend 70% of the shift

working with their students in class but

they spend another

70% uh of a regular shift working on

other um and probably they also have a

second job uh because the pay is not

good and probably that second job is not

necessarily in teaching but as a taxi

driver or or an Uber or or a restaurant

right so so that is in that context a

teacher needs to prepare a great lesson

uh needs to be able to understand very

well how their students are doing they

they she needs to um curate and select

the best materials for tomorrow's lesson

uh and probably she has a bit of um some

challenges in the classroom students

with differentiated needs um conflict

cognitive challenges motivational

challenges so then the question is what

is the role of teaching today in that

sense it's a very challenging role but

what if technology can help a little bit

with that right uh what if um we can

help that

low-skilled teacher in a loic

environment curating best in-class tools

um to help teachers actually in in very

in a short period of time be able to

bring those best-in class tools to their

students um when we talk about best-in

class we're not necessarily talking

about Stanford or we are working with

Stanford but we are also talking about

best-in-class tools developed locally

that have a lot of evidence uh that that

that can work precisely through that

context um what if we can can help

teachers automating things that are

cumbersome even boring and so they can

dedicate more time to um more high value

activities with their students and what

if we can help them assessing their

students um er quicker using their

phones to to understand how each of

their students are doing on reading

fluency uh 70% of Latin American 10y 10

year-old students don't understand a

simple text when they write when they

read so what if we can help teachers run

formative assessments on a continuous

basis um and so on and so on we we we

believe that AI has this if done right

it has a lot of challenges and we I'm

here to discuss that as well and I have

a list of concerns that at some point

keep me awake up night but at the same

time I think there's a lot of potential

if we can use AI to bring like this type

of support that can help the type of

teachers that we want to work with to

actually Empower them make them better

teachers have a better work style a

lifestyle like work uh and eventually

improve learning outcomes so so Jose if

I can play back simplistically what I

took away you have a vision that a

teacher could have a set of content and

tools which are from the best in the

world

with them in a classroom to make their

life more efficient and to make their

teaching more effective with a student

is that a fair summary um wonderful

Unice if you can if you could build on

what you think the vision is a few years

out and feel free to be bold as you like

thank you definitely agree with what

Jose uh mentioned just a brief

background uh we're mainly focused on

the Mina region uh North Africa you know

Arabic speaking uh countries and I

personally come from a background doing

Ms so usually deploying in millions of

Learners in when doing such systems and

our main focus is specifically public

sector education from the government

side so uh for example in the US uh when

you look at the ratios of administration

versus Educators you you tend to find

around the 50/50% ratio uh in private

sector you tend to find 75 25% ratio

more Educators than admins which makes

things more efficient

in terms of costs you look at uh it's

almost 50% more expensive in terms of

raw costs for someone to go through the

public sector just because of all of the

you know overall you know operations

that are extra over there and one of the

main things that uh we've been looking

at is how to deploy such systems on a

national scale by reducing the costs and

one thing that's always like to say when

you look at chat GPT subscriptions for

example they cost $20 a month when you

go to any country even though like we're

working with Saudi Arabia or Qatar or

Kuwait or different entities you just

cannot charge $20 per user per month

because it's just too expensive and the

most you can charge at least from you

know going into the market and seeing

what things are like it's probably

around $10 per year which is around less

than a dollar per month when you look at

Raw costs of gpus and all it just

becomes almost impossible to do that so

one of the things that uh we've been

focusing on is how can you reduce uh the

deployment costs for such countries how

do you you know uh could use smaller

models uh that could potentially run on

the browser how could you uh uh you know

reduce the cost on the operating side of

the admin to educat your ratio and we've

seen a tremendous amount of you know low

hanging fruits uh when it comes to uh to

to that side the other thing is of

course different countries tend to have

different uh uh problems you can't just

use a plug andplay a large language

model so you need to actually do the

fine tuning you need to do the data

collection mechanism and when you look

at open ai's you know costs uh main

source of of funding you see like they

have like $30 billion in Investments

most of it actually goes in data

collection then compute power so you

cannot expect a government that you know

their entire budget is a few billion

dollars to go and build an equivalent

large language model so one of the

things that we've done is we deploy such

systems to help us collect the data sets

and think of it as two business models

you can either have like the wayo the

self-driving car of Google where you

hire someone that collects the data sets

for you or you can have Teslas where you

sell the the car and then you can have

three million people driving and

collecting the data sets for you which

is a much more efficient way and use of

resources for these countries so that's

another angle that's you know we've been

anchoring on to allow these countries to

go below the $1 per user per month you

know price point to be able to to deploy

such systems that Jose was was talking

about so when I think about the vision

that Jose is sharing which was very much

in a classroom or a school and you think

about it how do we get that kind of

quality to a much wider audience across

the system by bringing down costs and

looking at a data um let me open it up

to the rest of the group um John I know

you you're a futurist um I was I was

looking last night at the books you've

written and thinking about the future of

of what learning might look like um do

you have any thoughts and comments of

when we and feel you know look few years

out what could teaching look like well

you know I think um especially speaking

to the Latin context right Latin

American context no

yeah no it's just a hammer right like a

tool right with hammers you could I mean

the best I can use a hammer for is

hanging pictures that's about the extent

of my skill in using it you can make

beautiful stuff with a hammer right you

can kill somebody with a hammer

on purpose or by accident too and so we

have to view these these things as tools

for what they are um the real danger

here and you're you're just touching on

it I think is a question of quality I

think that there's a there's a rush to

think that this stuff is effective and

is cheap AI does not scale very well in

National context it gets really

expensive really fast and then there is

the question of quality and I've been on

so many meeting need so many calls um

largely with companies uh in the Latin

American Market are like okay let's

create create create create content

create content create content and

literally say don't worry about quality

you have people who are working on that

yeah right so if there's a push to use

there's a push to create but the the

quality controls the human controls just

quite aren't there yet so as we look

towards the future because AI tools

right now are incredibly dumb they they

know literally nothing there is no

knowledge in there it's just you know an

algorithm's best guess as to what you

want to get out of it but you know once

they they once we start to develop

machines that actually build knowledge

or can create new knowledge with you I

think there's some possibilities but we

have to look at at the scalability

questions uh that were that are totally

related to this right right now we rely

on a handful of companies who have you

know Monopoly on the entire conversation

right and you see this with uh if you're

using chat gbt or other other things you

asking questions about using AI in

education and no matter what what it

spits back in you always includes

something about yeah you have to be

ethical in the use of AI right this is

because a company wants to offload the

responsibility of Ethics onto US onto

each of us rather the companies

themselves and so the question is when

we get responses like that are we

programming the machines or are we just

simply program

ourselves um I feel like we jumping deep

into some of the the challenges right

which is around quality which is around

ethics which is around cost I'm just

going to hold us back just for a few

minutes before we look at the challenges

if we were to say okay in the years to

come we're going to find certain ways to

overcome these challenges what's the

dream what's now that you you you know

you're being a teacher in a classroom

and uh if I think there there's a

conversation yesterday with Brookings

and they talked about how young people

need to be anchored in the sense of

purpose and ethics

they need to be able to build great

relationships they need to have joyful

learning which is all the softer side

and on the other side there's all the

learning stuff right so share like what

could be an aspirational vision of what

teaching looks like with AI a few years

down the line do you have a thought or

do others

health I I can think of two different

ways of using AI for teaching one is uh

from the perspective of the teacher in

most in line with what Jose has been

saying supporting the teacher in the

educational tasks that the teacher needs

to do not in the classroom uh planning

classes reviewing students work finding

information to teach creating content

for students that's a lot of the work a

teacher does and it takes a lot of time

and I can see a a large role there for

AI to support all of those tasks uh to

help us create better content to help us

find better content that is already

created by others and use it maybe

adapting it locally to our context

through the AI and uh I can see a a

great uh future there for both teachers

and students and uh then the other part

is actually using the AI to learn and in

the classroom and I feel that the first

part is mostly focused on individual

learning and uh me using the AI as a

teacher to be more efficient to or as a

student to grasp Concepts more quickly

to have personalized information and

things I'm interested

in but I think the other piece that

we're not asking enough about is how can

we use

AI to support collaboration to support

Project based learning to support

inquiry based learning and I see another

very interesting route there where the

AI is more like a collaborator with both

the teacher and the student and it can

help both it can help the teacher manage

a little bit of that messy classroom

that comes out of those more modern

pedagogies um and uh on the other hand

uh it can support students by helping

the teacher be in places they cannot be

so if I am teaching in a pbl classroom I

cannot listen to every student and uh

some of the things the students say get

lost like like things get lost here

we're not writing anything down so as

soon as I stopped talking this has gone

away and that's why we invented writing

in humanity that's one thing that I

think we can do with AI capture the live

classroom discourse and make it

persistent and something something that

you can talk about so the teacher can

ask the AI what miscon ceptions did my

students Express in their thinking uh

what interesting ideas did Anu have that

I can connect with others uh what

questions are still pending and I think

those are supports for collaboration

that are more at the group level than at

the individual level thank you Nando you

know um what you're sharing makes me

think of a conversation I had with

someone leading Google for Education

about six months ago and he said you

know we've been working with all types

of of online learning over the last you

know couple of decades and he said what

I think is the most powerful here is

we've never had a personal coach for

students or for teachers and we have the

ability to have a one-on-one personal

coach and amongst the many things that

you're say sharing for him that one

stood out as potentially the most

powerful how do people Rea react to that

is that the most powerful or you think

there's others that are the most

powerful glenties I I worry about

calling one the most powerful but I

agree a lot with what's been said and

what what Fernando just said but I think

if we really looking down the road you

know classrooms have always been there's

a teacher there are students there may

be teachers AIDS there are textbooks

Maybe videos there's a set of elements

that are in interaction and I see it as

down the road will'll have another

element that are AI agents and they can

play many many roles and certainly when

I talk with school folks I think the

starting point is helping teachers

develop lesson plans personaliz lessons

but there are so many different things

AI can do and our challenge is to figure

out how to maintain good quality or

create good quality in the classroom

with these agents how to reinforce and

further the role of teachers in doing

the many many things that only teachers

can do we were in a classroom with the

coing group yes they a wonderful

Elementary classroom and in each

classroom there were interactive things

going on they were doing robots they

were doing all kinds of things and there

was this joy and there was this Joy

between the students and the teachers

and that's not what AI creates and there

was trust and there was

inspiration so I think we really need to

focus on another element in the

classroom that allows the teacher to be

a more powerful teacher and spend more

time with the students but I don't know

if do I have to hold off on the negative

view of that I flip between the two um

you know on the other side that there's

this idea that AI will just to to

students that the teachers role will be

reduced rather increased which I think

would be dis disasterous um and I

particularly worry this is the US

context right now applies elsewhere

around the equity and inclusion issues

um you know AI can be a tremendous asset

for students in our country who are

English language Learners can do having

the immediate translation it can provide

feedback it can do all kinds of things

but I worry we could see in the future

that the students in the more

impoverished environments are getting a

lot of time with AI basically as that

teacher or tutor and they're being

prepared for tests and to do sort of

very standard stuff where in the more

privileged environments they have high

quality human teachers and they're

learning how to how to be in control of

the AI and use the AI so I think as we

look ahead we really have to keep in

mind the dangers of this driving our

opportunity and achievement gaps to be

bigger rather than smaller how do how do

others react to that is this a potential

equalizer where we can get you know for

example these modern pedagogies that we

could help teachers better less plan

these modern pedagogies when they don't

have that training and therefore it's

beneficial to lowcost environments or is

it something that's going to be a

greater you know great greater divide

what do what do others

feel just J real quick I I really see

these as a c as assist of Technologies

uh more than anything um one one thing

we're piloting out this is with the

World Bank is is using uh AI to to help

measure the development of hard to

measure skills and competencies um this

related to the fol program so it's like

uh entrepreneurship or your creative

Visionary these sort of things are

really hard to tell or really hard for

mentors to work with especially with

written works so just just a quick pilot

just taking inputs from mood and the

form of portfolio pose or essays be

readback and do a lot that really try to

offload that cognitive processing that a

mentor or teacher I mean it takes a lot

of work to to really read and create

these sort of value judgments and

provide feedback does it replace the

teacher the role no but it helps with

that initial review at least and help

offloading a lot of that cognitive work

and the results we had so far are

actually quite promising um so so as you

know I don't mean as much as probably

sound like a contrar a little bit before

on AI though I think the realistic view

is that it's you know that if we look at

as a as an opportunity for collaboration

or co-creative work it really does help

offload a lot of that cognitive work and

let to allow time for teachers to do

things that teachers should be doing

thank you um go ahead so so to your

question on on the digital divide I

think it depends on the type of product

that us as entrepreneurs would be

building right um I was in San Diego for

ASU gsv the largest tech event uh two

weeks three weeks ago and we we were

walking down the aisle asking all of

them H do you need a a student device

for your tool to work uh yes everyone

required a device for their students and

that doesn't work in the Latin American

context if that tool is going to be

deployed then the Gap is going to be

widen um so the question is how and and

you have to do that not only in terms of

device but in terms of the language in

terms of the learning model in terms of

at so many different levels you you need

to be very aware of this is an assisted

tool what is what we are going to be

bringing to the specific context to

close the gap instead of weding it what

El I can just build on that I think I

think it's interesting

that concerns around quality Equity tend

to emanate from uh from the affluent uh

West and I I'm not I'm not minimizing

the the concerns in in any way but I I

think it's interesting that you know you

started by having us sort of paint some

kind of idealized picture of what um

what good you know teaching and learning

you know with AI looks like uh and then

to a certain extent if you start from

that vantage point then you know the

problems kind of fall out naturally what

what should we be worried about what

should we um you know look look out for

if you start though with the premise

that that Jose kind of briefly alluded

to but we kind of then quickly glossed

over it which is at the moment you have

contexts um in in Lam it's probably

worse than subsaharan Africa and and you

know parts of South Asia where with the

current human only let's call it system

we have essentially failure rates you

know of 70 to 80% where you not getting

kids you know age 10 cannot read they

cannot do basic arithmetic you start

from that and then you ask yourself okay

if I deploy this technology can I you

know can I start changing that

picture then you you start maybe

thinking in a slightly different you

know more pragmatic way which is to say

okay you know okay I've got what can I

work with well most teachers have you

know at least an

Android uh phone right so so let me see

if I can do um if I can use AI to make

their life easier we know that they're

not going to you know this is not going

to transform them into super teachers uh

by any stretch of the imagination but is

it going to allow them to do their job

that little bit better so that the kids

can learn to read write and do basic

arithmetic and I think for those of us

who are you know working in in uh uh in

in emerging contexts that's really what

we need to worry about more than can I

you know get to this idealized you know

uh standard of of of teaching and

learning again don't mean to minimize

the the concerns and the dangerous but

but I think the framing of it depends

very much on your on your context yeah

absolutely and you I appreciate that and

at many years ago I was at EDC for many

years and we did a lot of work in

subsaharan Africa and it was radio

education because radio was the only

technology to deliver it so now that we

have a much more interactive powerful

technology they're great things to be

done absolutely agree it's all context

based we also tend to have these Global

conversations and they need to separate

things you know what this means for a

first grader learn to read versus a high

school student studying calculus um and

everything in between it's not the same

and we have to be cautious of making

these Global statements about it

you you know on that GL yesterday there

was a slide put up by Rebecca and the

Brookings team showing student

engagement and the level of happiness

and engagement of students at different

age groups I think this was for the US

context and yeah exactly right which is

you looked at grade three or so and

there was 75% of children who said they

felt very engaged with learning and as

you went for grade to grade to grade it

slowly dropped from 75% to Children

feeling TW average on average 25 % of

students feel engaged by the time

they're in I think it's already eth or

nth eighth grade or so and it stays at

that level so when when you look at that

context and you think is that going to

get affected in any way or improved or

not improved with the kind of

conversation that we're having does

anyone have any thoughts or comments and

feel free to come in at any point as as

you feel fit yes well um for me

artificial intelligence I I I think is

is like any other technology or at least

um it is accelerating things um but what

is most important before we are able to

use it properly is that we use the

humanity the the skills we actually need

to be human uh so I actually uh working

on a program which is opposite from

artificial intelligence um because if we

don't know who we are and what we

actually um doing to ourselves to each

other how how can we then trust on on

the technology so uh for me um and the

other thing it is um I mean most

children are or most people are and also

this events have been on their mobile

phone all the time um whereas that's not

the real world the real world is us you

know looking in the eyes and have fun

and and the energy um which we all going

to lose out when we you know do too much

with artificial intelligence

so um and there's indeed one thing all

people have in common all over the world

is they want to be

happy and happiness is not what is given

by the market at the moment that's

pleasure happiness is something that's

is given to you it's for free

um it's you don't have to do anything

for it and you feel great ful and that

is something which is content and it's

staying with you so you know I'm very

grateful to be here and I take it with

me and I'm happy to share this

information because I hope it's very

important that you're conscious that

artificial intelligence is not going to

give us

Humanity

so we should use it as a as a tool um

that's my opinion so is it that what I'm

starting to he hear from the group want

say no to follow up glad to hear that

because any anytime I have students or

anytime I have students or

families discussing something with me in

my office there's an issue there's

anything I always tell them that the

most important for us as a school

Community is to have students who are

happy because when you're happy you're

going to be it's going to be easy to

learn anything uh so the emotions and

going back to what also um igno Martinez

mentioned this morning um it's it's very

important that we feel alive in the

class and we it is proven already that

that uh you learn better when there are

emotions involved you remember things

better when you're feeling this these H

emotions and what you really what you're

learning about resonates with you is

always uh better so that's something

that I just uh wanted to to highlight

however and going back to what you

mentioned of a student

engagement what we're going to see

happening more often in the following

months hopefully only until we're able

to

um um handle the steering wheel and and

move everything to where it should be or

use the artificial intelligence properly

as you mentioned earlier are students

who see in the artificial intelligence

way to do things faster and simpler

maybe with not much quality uh families

who would be worried about it so

probably that's going to be one of the

immediate challenge that we are all

seeing at at schools the role of

teachers in providing um tools tasks

activities experiences for them to uh

use other tools also artificial

intelligence but to have other tools and

set of skills is very very very

important and something completely out

of the but I believe we mention all of

these things just one that I I believe

is very interesting linked with

artificial intelligence and with what

stabos mentioned earlier is that there

are schools with uh members of a staff

fully devoted to analyze data thata

coming from from assessments or

standardized test etc etc to me it's

very interesting how artificial

intelligence could help in analyzing

this data for us and helping us Point uh

to where should we focus our efforts in

terms of learning

wonderful go ahead I I want to talk to

the engagement piece of bit and uh my

sense there is and following up on

something Glenn said there's a danger of

equity in

putting lower socio economic status

students in schools who work with AI to

have an automated teacher and more

affluent students to have a real person

and my sense is that engaging ment even

though I see my

kid very interested in his video games

does not come from your relationship

with the computer but through your

relationship with another person and

engagement in the classroom comes

through that relationship of trust that

you build with your teacher with your

classmates and with the content if we

remove two of those the teacher and the

classmates and leave only the student

and the content I believe there's rarely

going to be engagement and I see a trend

mostly in developed countries like the

US where that's what's happening to less

affluent communities I see schools I

visited a friend in Oklahoma a few weeks

back and he told me his kids in high

school mostly have math class with an

online tutor they rarely have a math

teacher and their math teacher is not

somebody who can

have a good discussion about a

math and that's what really engages you

and gets you interested in math unless

you have that conversation with a person

about the interesting details of that

problem you're not going to get engaged

by watching Can Academy videos or

repeating drill and kill

exercises so to connect finally with my

last part I'd say as long as we use the

AI to to help us collaborate to help us

connect with others that will create

more engagement if we use it as a

delivery system for that view of

personalized education where everybody's

in their own Silo doing their own thing

that's not going to be engaging for

any yeah I'll just mention in support of

that this this is all at the you know

experimental stage but we have a lot of

work at Stanford looking at AI for to

support collaboration and collaboration

learning as a question is is the AI

agent a member of the collaborative team

is it a coach for the team is it a coach

for individuals do you have a something

saying hey stop talking too much or make

sure you picked up that point uh there's

so many different roles it can play uh

but there have been some interesting

things where I'm every once in a while

yeah it just shocks me I think we all

have this experience it did that where

it was coaching a simulated team and it

was very much right on it was chat GPT

4.0 there's also research by ad msky of

AI analyzing classroom interactions so

really serving as a coach to the teacher

and picking up you know are you asking

questions that are closed so you just

get these brief answers from students

are you only engaging certain some of

your students and not others so I think

we take a step back just having AI

contribute to the kind of interactions

we want to see uh could be a very

important asset yeah which which uh I I

know we're on time which I guess comes

to the heart of what intentionality

right what I'm hearing music group which

is if we have a very clear sense of

purpose of what we're trying to do with

young people if they are happy in the

deeper sense of have a sense of purpose

and ethics and grounding and

relationships and we make sure that is

the front and center of teaching and

learning and then very intentionally

think about where can AI as a tool help

with the kinds of things you were

talking about such that implementation

you were saying when you come down to

implementation it veers toward work that

Vision that we really have um I know

this this the time was short but but

thank you um to each and every one of

you and if someone has a a dying last

word there welcome to otherwise we'll

we'll wrap it up uh wrap it up there not

dying but I can bring I think we we hit

our important things when I talk to

education leaders and policy makers

theme is always you know AI is here and

people are going what do we do with this

AI That's not the right question for

educators we know we have the equity

gaps we know we have future Workforce

gaps in future shortages we know we have

problems with student engagement and

learning the list can go on but I think

we have to step back as educated as

these are our big challenges how can we

use AI to help meet them not what the

hell do we do with this AI to just

appear in our classroom yeah thank you

man thank you thank you to each and

every one of you thank thank you

thanks sorry stos if I cut you off at

the end we we've met before was it at

the at Brook

I'm were you the one in

palro yeah that's where it is February

10 yeah well pic like before and