Where Does Artificial Intelligence Fit in the Classroom? Panel One


– Our first panel
comprises a outstanding mix of academics and technology
education entrepreneurs. Paulo let me start maybe with you as the cohost of this panel and event. What trends do you see
from your vantage point as an engineer, as an educator? What are the trends in AI in education and where are we headed? – Thank you Stavros and thanks everyone for coming to the event. So I think one big distinction that we have now in education
that is becoming more and more pronounced is the difference between as Seymour Packard would say, kids programming computers or
computers programming the kids so I think there is a big distinction between technologies that
are meant for teaching and technologies that are meant for kids to learn how
to use to create things. So for example you have
video based systems that are teaching technologies. You have apps that are
teaching technologies and then you have robotics
and then you have coding and all those other tools that
are not for teaching directly but are tools that kids use to build and create and express themselves. So I think that’s one big trend, one big distinction that we have in the world of education technologies and it applies to AI as
well because you can look at AI as a tool to teach again. And in this case even a more
mass-produced kind of way, that you can teach millions of kids. Or you can look at AI as a tool that we’ll give for kids for them to create AI, for them to create software, to create artistic products or inventions. All kinds of things for them to improve their lives, improve the
lives of their communities and learning about AI in the process. So I think that’s one thing
I wanted to highlight. – Yao if I can bring you
into the conversation, you’re a graduate of this institution. And you’re now applying your skills to developing robotics solutions that incorporate AI for education. What trends do you see
and how is AI impacting the classroom or how would you like AI to impact the classroom? – Yes, thank you very much Stavros and it’s really a huge honor and pleasure to be back here and I
do want to correct-ify that I couldn’t put the doctor in front of my name yet because in
all the media interviews, I’m always everything and
dissertation is last step. But especially at alma
mater, definitely don’t wanna have that error but mentioning that I’m definitely grateful for all the great learning experiences here in the economics and
education doctoral program. So that’s how I call myself in my company, Roboterra as a product company that I’m really more from the data side. Actually the main theme when
I started my dissertation in the proposal was
really asking one question that has been fascinating
me for a long time, was economics is about
optimizing decisions with limited resources. But what if, is what if with
technological development some crucial resources
are not limited anymore? And then that question
start to make me think, how can we improve on that? Some very important crucial aspect of resources which are important input for education and with the hope that that can bring in
really important changes to help solving a lot of
problems that we care. So this is the really looking back, the development of my own thoughts of how the training from this very place at Teacher’s College many years ago and then all the work of looking at macrodata, microdata on family level and how companies and countries are investing in education,
what are effective? What are not really? And then those essentially led to the product work if I put it simply is really what if we bring in hardware so that kids according to good
cognitive science researchers that with hands on the learning can be fun and also engaging so that we
don’t lose them immediately. So bringing the hardware into software, combining that, and that
became robotics kit. So that embarked my product
journey five years ago. And I would say of the past five years, proudly enough that working with more than 1000 schools worldwide
in more than 40 countries. Ranging from resource rich schools, international schools, private schools, to a lot of the urban schools in the US or even rural schools in China and India. So the whole experience is giving the Roboterra team a lot of encouragement that we have to follow the good and the right design principles to really design effective learning of whether it’s robotics hardware or the data we’re seeing
from the software side which became part of the
individualized feedbacks to students or the AI feedbacks. So that’s the design principle that we try to follow and
then that’s where I see as a good practice that I’d like to share or explore today. – Thanks Yao, so if we think in terms of how does two
categories, Roboterra deals more I would say with AI technology, robotics as an object of study for kids as well as yes a tool that could potentially enhance learning. Now Shipeng if I understand
what iFlytek does, you are much more on
the side of AI as a tool to replace some of the things that teachers do today
or that educators do. Do you wanna share with us a little bit about what your company
does and how you see the technology developing going forward? – Sure, iFlytek starts
with a speech company. Now expand its area in
a more general AI field. So that’s the technology side. Then later we found that
AI can really can help a lot in the current education. So the first steps we are doing because we are starting
as a speech company, so we are helping a human scorer to do this language learning especially for spoken language test so we work with standard test organization to apply the AI technology right there to replace human scorer for
better consistent score. Then later actually we found
some more important area, actually something always challenging from the start of human being, right? How to enable personalized education? Because limited teachers resource, right? The teacher does not have the time and effort to take care the special need for each individual student. So we are thinking how
to apply AI technology to enable such a personalized education. So it’s the way we do it is instead of we want to many other
people have tried before, trying to change the behavior of the teacher and the students. We let them use things, the current tools, the current education star
they are much better ways, uncomfortable ways, still
using pen and paper right? So what do we do, we digitize
the whole process of learning. Including the classroom,
including we teach with the test paper,
teacher has the homework, and we do OCR, we do scanning. So the students send in the paper. Where they wrote their
problem solving solved, and then we analyze each steps
how they solved this problem. We identify where the problem
or where the challenge the student has, then we map that into a knowledge graph for that subject. Then the teacher can clearly see how each student has progressed and also in related area
which is student for analyze. Then we use AI to be able to assign individual teaching contents
and individual homework for each student so each student can have different homework. So in doing so, both the teacher can increase their productivity only teaching the common problems that the whole class is facing. And the lazy students
working on their unique, personalized weakness with
the personalized contents. And also with personalized homework. So the whole process is
iterated again and again. So we have applied this technology to about 6000 different
middle schools in China and we can see the
efficiency and the product really improved a lot and the student can save time to do
something more appropriate and in other areas so
that’s roughly the idea. – Thank you, thank you Shipeng. Andre can I bring you
into the conversation now? As a former education you ran a number of charter schools I believe. Now you’re working for
one of the preeminent policy think tanks in this
country and arguably the world. – The world.
– The world, okay. (laughs) As you listen to Shipeng and Yao, it all sounds very benign and positive. What are though some of
the risks with applying this kind of technology to the classroom and what do you worry about? – Yeah and first of all
thanks for having me. Clearly technology is a tool to be used to optimize performance within systems. And it’s used also as a
tool to create products that can enhance our lives. But the true promise, I believe, and I think we, and you
touched on this specifically, is to understand how we learn. How students learn and how teachers learn. Unfortunately we’ve all learned how to be racist and how to be exclusive. Part of the problem,
one of the reasons why AI has not been as transformative as it should have is
because we’re still working within the confines of
a racist environment. And so when I advise startups
and other technologists who are working in this
area, I always remind them of the demand and particularly
in the American context. The demand is to close
the racial disparities that are occurring
across the United States. By region and obviously
along the lines of race. And so for me I’m always saying hey, when you’re developing your product, who’s on your team, who
are you testing it with? Are you inclusive along that pipeline? Because if you’re not inclusive in the development of the tools, you are essentially going to exacerbate many of the problems we currently exist. And that is around racism and so for me, let’s not, technology is a tool. It’s all, that’s what
it is and we’re seeing incredible opportunities
to enhance learning. But as we’re currently seeing now, there’s incredible divergence, economic divergence,
educational divergence. And technology and AI has
a lot to do with that. – Thanks, thank you Andre. Paulo I feel like you wanna come in here? – Well I mean–
– No go ahead. – I just wanted to
comment what Andre said. I think that one important thing to do in any discussion of
education technologies is to… There is a mythology that
the Silicon Valley created about technology being disruptive and benign and their CEOs
being just good people that want to help others, you know? And they are, I want
to hire their PR people because they convinced the entire planet that they are selfless
people running companies for the betterment of the world. And meanwhile just give us all your data, everything you have,
and just leave it to us. Because we are the good
guys, we are hoodies, are are cool and all of that. And I think that has
been, history will tell how an entire civilization was fooled by 10 30-something CEOs
that caused enormous harm to our societies and I don’t want that to happen in education yet again. So I think when we talk
to those companies, when I think about AI in education, we have to be aware that those kind of companies that come to districts, come to teachers and say
no we’re gonna help you with our technology, just
give us all your data and we’re really well-meaning. We know they are not, we know
this is extremely dangerous. I think when I hear things
like move fast and break things or fail fast, fail often, and I say, what the hell are you talking about? When did that become a good thing? Especially in education,
you’re gonna move fast and break things in our schools? Are you gonna fail fast and fail often? It’s all these collective delusion that we got into that all
of those things are great and now we are at the brink
of offering our schools to this, they like to
call themselves a mafia. So it is a mafia of technology companies that are not interested in the improvement of society as we now know. – Well I think that’s
a timely intervention. The only qualifier I would make there is, I don’t think you need to
have sinister objectives or plans for world domination
to get things horribly wrong. You can just leave it to the law of unintended consequences.
– But it helps. – And things like that will happen. But Yao as someone who’s
spent a good amount of time in the valley rubbing shoulders with some of the folks that Paolo just took a big shot at. (laughs) What’s your thinking of this? How do you, presumably you want Roboterra to grow and be as big as
Facebook and Apple and Google. How will you address some of the concerns that Paulo and indeed Andre raised. – Well first I regretted
of not wearing my hoodie and jeans today here.
(audience laughs) Which I rarely do actually this, but just out of respect alma
mater dressed formal. (laughs) I definitely can feel and connect and relate to actually
the really the passion and the emotional hope from both of the gentlemen just made the point. As again, someone who actually choose to study the general study about education in my training here and
from institution here and then went to Silicon Valley. Started not one but two companies
from software to hardware. Again I’m not going to
defend Silicon Valley. It’s just we should be cautious about anything that’s
stereotyping a group or a place. In general I would say data privacy and protection related
to children’s behavior, whether it’s learning behavior, playing behavior or the smarter and smarter cameras that’s being able to capture the facial expressions in order to analyze who is
falling asleep, who lost interest to give feedback to
teachers or it’s teachers actually looking at who
is a little bit slower than the average of the classes so that the teacher can give
more individualized caring and support, yes of course
those are all benign goals. But indeed technology of
AI in the applications for education or learning just
like any typical technologies most of the time it’s double sorted. So that’s how, and also
another aspect I agree is that with the, don’t jump on the word of disrupt fast and fail fast thing. I honestly, and here’s
definitely a confession here. When I just started Roboterra, already it’s the second
company I was working on. I still wanted to go be very fast mode. It took me actually a while,
not just intellectually, not just emotionally but
really intellectually in developing products,
working with all our teachers who’s using the technology
give us feedback so that we can really understand the diversity of student experiences. And how the teachers are really using the hardware software kits we’re providing in their classrooms and
also in evaluations. A lot of our students
got into top universities but a lot of students just in order to get into vocational school. So what are the implications
as the product designers? Comments actually similar comments happens within our products meeting all the time. We wanted to be cautious,
not to unintentionally led to harmful consequences to
certain groups or populations but it takes time, also it’s a process. It’s not that from the very beginning, there is a guideline, we know everything, we’ll be safe for sure so
that’s why I’m definitely on the side that I
believe certain committees or conventions should allow
different stakeholders, product people who have decision powers within organizations,
especially large folks. And also real educators, both academics and also from the
classrooms to be together so that we can identify.
– Yeah thank you. And there may also be a broader principle at play here which is that the incentives that drive commercial behavior do not neccesarily lend themselves to determining what is
in the public interest ultimately with things
like black education. But Shipeng can I bring
you into the conversation because as you listen
to what is essentially at the moment a very
US-centric conversation and you come at this
from the vantage point of China where it seems to me at least, but please feel free to
correct that impression, that the Chinese population in general has a more relaxed view
around data privacy and the use of its data
for things like education or the social capital scoring
and various other innovations that are taking place in China. How do you start, how do you think about some of these issues
and concerns that are raised? What is your company doing with all the data you’re collecting? – I think that’s a very good question. And from the outside world,
looks like China people, Chinese in China, are
actually more like… I’ll say less overall is
the privacy issue of data. And that largely depends on
it’s a larger population. It’s not up to the speed of the rest of the Western world yet. In recent years, that’s erroneous. We’ve been seeing that they have, actually the government had
very strict data regulations to protect users privacy and actually, I know there is a government agency and standard organization study the rest of world like GDPR. A lot of companies nowadays if they want to do business in Europe,
at least one of them, they have to follow
the GDPR rules as well. And iFlytek as a company
especially actually, we follow all the laws
locally and internationally. So from as early as 2015 we
are among the first companies in China have this, join this association of data protection, privacy protection. So I also lead standard
team inside iFlytek, joined this international
standard for data protection. So you can see that
although we also viewed as not care that much about data privacy, but today it’s becoming
a very serious issue. So it’s a situation improving, and the companies are taking measurement to highly guard user data so that are things happening in China. – I also read recently that
there’s been some pushback, for example against the use of facial recognition
technology in classrooms. In China, specifically on this issue. – Yes so there are, especially recreation, actually regulate that
part of the technology, the use of that technology. But again, there are other ways, other technology to get around. Like instead of using visual information, you can use other sensor results to really reveal the
identity of the students but still get a similar measurement. So people are always developing and inventing new technology around that. – Andre as you again
listen to this exchange, again listen to this exchange. Again from a policy, someone
who advises policymakers, how closely do you watch
developments outside of the US? Because again this is one case
where we may find ourselves going down a path that’s
being set outside the US because again education is not, is never insular, it’s never as local maybe as we would like it to be. Could we find ourselves in some sort of AI arms race as it
relates to education? – Well I’ll say it this way. One, racism is a global
phenomenon unfortunately. And it’s evidenced by the algorithmic bias that is evidenced all around the world. We know that when it comes
to darker-skin people, we’re not using technology in a way to enhance black and brown people’s lives. We’re using it to regulate
and sometimes punish people in that context and so for me, globally if we see a reduction, for instance a reduction in workforce in black and brown communities due to AI with a promise that technology will deliver better
teachers then that’s a sign that we need to regulate an industry for those kind of biases. Again, technology is a tool. Teaching requires higher order skills that will not be replaced by AI. We should be enhancing
those opportunities. But we have too many bad actors who have essentially
created a level of distrust. It’s a little bit different
but the virtual schools for instance in Ohio
were a horrible example of how not to scale up
education using technology. Because it ultimately bad practices will hurt those communities even further. So for me I do think we need an arms race for technology and for AI in particular. I just think we need one with a value of inclusion and equity
that I don’t see present at this time.
– Thanks Andre. And I think we’ll,
we’re gonna circle back. I wanna circle back on some
of the ethical questions around the use of this technology. Race being one of them but I
think there are others too. Paulo I just wanna pick up on something that you said in your remarks earlier, that the fail fast and fail often and break something ethos really doesn’t, has no place in education and by and large I would agree with you in systems
that are well functioning. What do we do though, you’re
originally from Brazil and Wise, a lot of the work that we do and that we look at deals with countries where the formal education
system is very broken. How do you think about the use of technology in these
kinds of situations? And do you think about it differently in a situation where the system is broken versus one that’s imperfect but still more or less functional? – Well I think that’s a great question and when I go around the world visiting the best schools in the world, you never see those
schools firing teachers and bringing AI to their classrooms. You see them hiring more teachers, reducing class size,
creating robotics programs, creating literature programs,
creating afterschool programs. So that’s what those
schools are doing right? If replacing teachers with technology in that sense of teaching
machines was so good, then the best schools should be doing it theoretically right? But they’re not, they’re
hiring more humans. And when you go to low-income districts in the US and in other places, a lot of times you see this discourse which I think is a fallacy which is, this is so broken,
there’s no way to fix it. Let’s try something crazy
because we’ve been trying for years and nothing happens. And I think that’s a
fallacy because first, it has actually no one really tried very hard to deal with those places because there’s no incentive,
there’s no interest in many governments to
deal with the actual cost of education, right, so I think in a way I see in developing
countries a lot of discourse that this is so broken, let’s
try something crazy right? And normally it’s a company
that has a vested interest in getting a big government contract and they say we come from the US and the US is perfect
so you should trust us. And here is, I need
several million dollars to fix your schools with
this software that we, and I think that’s really dangerous because I think we educators,
we need to stop telling public officials that education is cheap, that education can be
done at a discounted rate, that we can just fix it with some kind of miraculous solution. It’s like health care, if
society wants education, that society needs to
know there is a price to pay and it has to pay. Otherwise we will end up
in this dystopian world of elite schools that are amazing and that have all these programs and then these poor districts
that are constantly being used as lab rats, the kids are lab rats, and we experiment one
solution after another without ever facing the fact
that what those districts need, what they need is actually real investment and hiring and training teachers. The old good solutions
that have been denied for them for a long time. – Thank you for that, and I’m inclined to agree that if you want good education you have to pay for it,
although I think again, if it were just about the money
then this country would have the world’s greatest education system. We know that it falls
short in many respects and so the question then is it’s also about how you spend that money that makes a difference
and maybe we can get into that a little bit later. Yao, again I think
there’s general consensus on this panel that the role
of the teacher is integral. Now you mentioned in your earlier comments that you work alongside teachers. It took you a little while to make that emotional and intellectual
shift, as you said. But how have you found
working with educators? As a technologist what are some of the dos and don’ts
of this collaboration between the technology
space and educators? – Yes, so just really using
Roboterra as an example of adopting AI technologies
into education. So from the beginning the way we innovate, we’re bringing in hardware kit, microcontrollers, mechanical bars, sensors into classrooms and then a lot of our, of course the founding team, a lot of us had been teaching as well. So we were engineers, software engineers, hardware designers, data person, and then who taught as well. That was the beginning and
then the interesting think, I think here is where some of
the Silicon Valley’s typical agile work method, I wouldn’t call it just be fast and fail fast only, right? If agile method was, our ultimate goal is we wanted to provide
a good learning solution that actually solved the problems. But then we are aware of potential issues such as the designs might
have issues affecting different groups of students because of the design essentially. So from the beginning
we, but we would start by having the hardware
out, the software out. And then along the way,
while we were promoting, introducing our technology solutions to different schools we keep hearing again and again and again from the schools that oh well it’s wonderful
but we don’t have teachers who actually understand coding and also engineering or who are just in STEM or who can use this at all. Whether it’s super resource-rich schools or actually very disadvantaged schools. So very fastly, because
we kept the mind open, we heard that the issue in the field, even in the field we wanna do is really, okay you can have innovative
kits and software. But the issue of all the schools are facing is there
are simply no teachers. If you wanted to help some of the schools to hire teachers who knows coding, you know the teacher, we then the schools are competing with the big companies who’s paying two or three
times more higher salaries. So again the law in the labor market of teachers kicks in again. So that’s how we realized
there’s actually no teachers. Then the problem we try to solve as a product company, technology company, we actually get to focus on okay, so how the things we developed are able to help schools who
just have current teachers who don’t really, who never coded, who don’t understand
mechanical engineering. But they’re passionate enough. They care about students
from different backgrounds. How can we help them, so that became the central design question for us and then so that’s how besides
the hardware and software our designing goal actually
became clearer and clearer. And of course along this way,
it’s the teachers’ feedback. Whether even without being
a software programmer but how they can be effectively still guiding classroom,
helping the students start user kits, helping
the students work in groups and interact with each other
and then support the students with that with different strengths. Working groups, how to design
contextualized competitions. So a lot of the good feedbacks and the good things that’s
today on our technologies are actually feedbacks from the teachers. So that’s how our design
principle design group, users experience group,
around learning lingo, teachers from the schools
that we’re working with and also other colleagues on our team with education background
plays very important role and then that guides us a lot. – Yeah.
– Pick up on that. Teachers are wasting their
higher order thinking skills on things like scheduling
and other mundane things that if we applied technology and AI, teachers would have the
time to learn other things like coding but instead they are spending inordinate amount of times trying to determine what subject
level each student belongs in. And other mundane things and so for me, it’s always been how
can we use the promise of this work to make the lives of hardworking teachers better? Because at least in the US, we know that working conditions are the reason why they’re striking and walking out. Technology can be the tool that
makes teachers’ lives easier but again, we have to give them the time to learn technology ’cause the other issue that is unsaid but you mentioned it. Most teachers do not have the skills that are required to train
students in this next economy. And so we need to build it in somehow. But again, that’s where
technology is supposed to make that time but instead, more often than not
it’s burdening teachers with extra responsibility. – Yeah, please.
– It’s really one of, I just wanna add the successful cases from a lot of the teachers
that we work with. So from the very beginning,
at the beginning, teachers might be resisting of I have to spend time on learning something new and then they will
clarify, the vocal ones, they will clarify that okay we would love to start to work with us
because we really wanna provide new learning contents and
robotics to our students. But as long as the training
and does not require me stay in my office longer because I have to pick up my kids too, right? Take care of home things
or even some teachers in rural schools they’re like, their days, teachers
have to work on farms. So that’s how if you are
really truly open-minded as technology designers and listening to the need so that you know the goal of designing is not to add in burden through whatever methods, right, from top-down level
administration whatever. But really addressing the need, make sure that no extra burden is added to teachers but they can be effective and then that’s how you really are seeing a lot of successful results as we did. – So the, I hear what you guys are saying and I think in principle it’s hard to argue with a thesis that if we get technology to take away all this boring and mundane and repetitive work and we free up the teacher’s time all our
problems will be fixed. The cynic in me says we’ve
been down this path before with the personal computer
and the enterprise software solutions and to a certain extent, the reason teachers in this country and others are disgruntled
despite record expenditures on education is because a
lot of the money’s been going to the technology companies
and not to the teachers. Why is it gonna be different this time? And Shipeng, maybe let
me ask you that question. Why is iFlytek a better investment than one laptop per child
type of initiatives? – Okay I think you are
asking a very great question. I’m not in education background but I’m in technology background. First of all I think today there’s a lot of hype on how AI can do, okay? So people always think oh your
AI going to replace teachers. No, that’s not going to happen in 10 years, even 20 years I would say. There are many aspects human
can do, machine cannot. So we have to lower our expectation of AI technology can do or
can help with education. So that’s number one. The second thing is if you want to using the current AI technology to revolutionize education is another hype because technology is not ready and you are not going
change more dramatically anything currently going on. You would rather using new technology to do a more incremental improvement based on the current education system. For example, you are not going to make the students buy new equipment, learn new things, instead
you use technology to help both the teacher and the students in a backstage way, in a hidden way. They don’t change the current behavior, but they will fill how the
technology will have them in the long term and the idea is you’re not going to ask the robot to take place of the
teacher in any aspect. So one concept actually in our company is we’ve always promoted it’s a human interlope for the education policy. Human teachers is always going to be in the AI education loop. So that’s the principle we want to stick with for a long time. Only in that way, without giving people too much hype and you’ll be doing realistic things, very
practical things step by step to help to
improve their education. That’s our take on this. I think step by step you
will change a lot of things. For example, the thing I’m
holding on is a notebook. And this thing can do live transcribing of whatever the teacher said into text. So it will help the students to reveal whatever the teacher has taught at school. At class it’s very easy to research, to do other things, and to notate. So that’s a small step
is to have it in today but without changing
too much of the behavior of students, you are already helping them. – Over lunch you gave me a good example of your technology being,
or your company’s technology being used to grade and
to mark essentially tests. And you said you, which
is maybe not surprising, but you achieved greater consistency versus human graders or markers. Is that, again, an area
where the technology can indeed help improve productivity? – Yes indeed.
– Of teachers. – That’s another area we are working. Help the teachers doing easy job in scoring all the homework tests and the exercise so they can save time to do something else so only in doing so, you let the teachers know
the benefit of technology without actually changing too
much their behavior in theory. – I’m mindful of the
time, I was just given the 15 minute mark so we’re
about five minutes away from Q and A however I’m actually minded to see if we can just
continue the conversation. So unless there’s a massive reaction from the audience we might just carry on if that’s okay with everyone. I’m not getting… (laughs) But let’s, okay let’s do a quick round now of what are some of the ethical concerns that we have about the
use of AI in education and how do we go about addressing them? So Paulo maybe start with you? – I think just before
getting to the ethical part, there is this amazing
opportunity that we have that goes back to what
I said in the beginning of creating technologies to empower kids to create stuff to express their ideas. And I think today we
have a climate strike. Today that’s fully
organized by young people. We adults, we are like
ruining their future and young people are now taking things, matters into their own hands. And I think that’s a role
for our education systems is also to empower them
instead of saying we have to wait 10 years until
you can say anything. So I think creating technologies that give kids voice, that gives them ways to express their ideas, I think
that’s an ethical issue too. I think the other part of the ethical– – [Stavros] I was gonna say,
you’re beginning to sound like the Silicon Valley evangelists. (audience laughs) Wasn’t the internet supposed–
– It’s coming back to me. (audience laughs) I think the other part
of the ethical thing that’s really, it’s the whole
data privacy, safety thing that I’m sure we will talk more about. But I think there’s another piece of this which is who decides
what is good education, right? And I think it’s very common that the technology drives our thinking about what good education
is if we let it happen. So if we say what good
education is scoring well on tests then we will
reshape the whole system to make that happen and I think that’s an ethical responsibility of educators, schools of
education and all of that is to say well who decides? We as a society should
decide what education is and then the technologies
and everything else should follow that philosophical vision and decision about is education
about direct instruction? Is it about emancipating people? Is it about opening up their minds to new realities, to new things? And that’s a societal
decision and the point that I think is complicated is that we know as I think
we see in our keynote and everything that what AI can do now is very limited in terms
of delivering content or conversational agents
or things like that. So if we make that the
ceiling of education because we want to put AI everywhere, then I think we’ll be
in very deep trouble. – Andre, a comment? – Yeah for me I do a
lot of work on value add of assets in particularly
black and brown communities. I think society must start to demand that there’s added value
to underserved communities. We see economic growth,
incredible economic growth. We’ve seen advances in technologies like you haven’t seen before. But we also see wage stagnation. We see educational divergence. You see problems just going off the rails. And so at some point we’ve got to say I’m not going to incentivize many, particularly the larger companies, and you have the HQ2
situation here in New York where we essentially were giving away the farm to attract
a technology company. I think that the true ethical decision is when are we going to really enforce this notion that we’re
all in this together? That we need our company
to show a public good in terms of not only
advancing people’s learning but also advancing their incomes, their housing, their transportation. All of these things
have to, we have to show that there’s added value and
so from a regulatory standpoint we’ve gotta hold companies accountable for doing what they’re supposed to do and that’s advance society. And I just think that
there’s a lot of evidence that shows growth in
certain economic areas without the societal advancement. – Thank you, Yao and Shipeng, how do you from the side
of a technology company think about some of
these ethical questions? – Yeah I think I can
answer something here. I did look up on Wikipedia
the definition of education. It said education is the process of facilitating learning or
the acquisition of knowledge, skill, values, beliefs, and habits. I think most of what we are talking about today is especially
for our technology company is teaching knowledge or skills at large. But for values, for beliefs,
and for even happiness, that’s something that cannot be taught or cannot even be asked
by AI technology today. So that’s something I think is
very important going forward will be trying to
evaluate education system or technologies, that’s something we want to include in the consideration. So aside from that, I
think today’s technology, AI technologies are more like improvement based on the current framework like Andre mentioned in the beginning. But there are more
interesting research actually, also related to AI research. Through another branch of the AI research related to study how human learns things, how human cognition works, right? So the finding is very interesting. And the finding by Dr. James Heckman at University of Chicago is he found that actually the ability
of problem solving was more or less formed when
the kids was very small. Like before five years old. So that’s a very striking
finding which means your problem solving
skill is there at five. Later in your stage you always learn your theory with new knowledge. It’s just new knowledge, not
the problem solving skill. If that’s true or proved, should we change our whole education system to better for human development rather than how today’s, in today’s education system? I don’t know, but that’s a direction I think especially very
important for education. And other aspects of
ethics like what if AI had promote such ways in our homework, such ways in the process,
what are the social aspects? What’s the face-to-face intention with the fellow students with
the teacher would look like? Would be diminished or
would be further enhanced? That’s another aspect we
would take a serious look in. – Thank you, Yao last word? – As a product person trained here from this great institution,
I guess if I can take this chance to make a call for action. It’s really the technological trend is developing no matter what, right? 10 years ago this, we
all believe we are smart. Who can envision all the apps today that’s affecting our life so AI, the huge amount of data is here. Cloud computation is here, 5G is here. So all of these will
have huge implications on how we’re going to
live and how our children is going to learn so
big issues like today, all we mentioned the disparities, the racial issues and
genders, many issues, cannot be answered by any
one single product company. Yet I’ve had the honor of being invited to represent education person and to be on the AI robotics committee
on the global council. But still it takes a
village to raise a child so it takes all of us together. The academic community and
also the real educators. I really hope that
everyone can start to ask, pay attention and ask questions, what does new tech mean to the new way of us to organize the information, to deliver information,
and to really foster the creativity of the children. ‘Cause I believe as a
product person again, technology should be
tools and it’s really us to decide how do we use them and not harm the children in any way. – Thank you Yao, and I wanna thank all the panelists for their
contributions to the discussion. I’m personally encouraged because over the last several years I’ve observed again from the Wise vantage point that there’s a hell of a
lot more engagement now in education issues from
multiple stakeholders. So I echo Yao’s call to action. I think it’s incumbent on educators to get involved in the conversation and make their voices heard
with the technology companies, with the policymakers, and
hopefully this is the beginning. I apologize that we weren’t able to take questions but I
think there is a coffee break and I think we’re all
around and if folks wanna raise questions with individual panelists, I’m sure they’ll be very
happy to talk to you. So please join me in thanking the panel. Thank you.
(audience applauds)

Leave a Reply

Your email address will not be published. Required fields are marked *