Panel: The Importance of AI-Literacy for AI in Education
Summary
TLDRThe panel discussion delves into the critical importance of AI literacy in education, emphasizing the need to understand AI's role in daily life and its impact on learning. Experts from diverse backgrounds stress the significance of teaching AI responsibly, including its ethical considerations, and the necessity for students to be AI literate. They explore ways to integrate AI literacy into curricula, the importance of family engagement, and the challenges of keeping pace with rapidly evolving technology.
Takeaways
- 📚 The panel discussion emphasized the importance of AI literacy in education and the need to understand AI's role in learning and everyday life.
- 🌟 Laura Allen highlighted the comparison between traditional literacy and AI literacy, stressing the need to understand AI's terminology, concepts, and the implications of data usage.
- 👨🏫 Stephanie Adragna discussed the significance of family involvement in fostering AI literacy, suggesting that joint engagement can facilitate a better understanding of AI technologies.
- 🔍 Dru Long emphasized the unique aspects of AI, such as its active decision-making processes, which differentiate it from other technologies and necessitate a certain level of AI literacy.
- 🤖 Tyron Young, from a practitioner's perspective, focused on the ethical considerations of AI and the importance of having a checklist for evaluating new AI technologies.
- 👨🎓 The conversation underscored the responsibility educators have in teaching AI literacy and the need to integrate it with critical thinking and ethical considerations.
- 👥 There was a consensus on the importance of interdisciplinary approaches to AI literacy, combining technical knowledge with social, ethical, and creative perspectives.
- 👶 The panelists agreed that AI literacy should start from an early age, with discussions around AI ethics and understanding integrated into curricula at all levels.
- 🏫 There was a call for more transparency from AI developers and for educational institutions to provide resources that help students understand the implications of AI technologies.
- 💡 The discussion highlighted the need for hands-on, embodied learning experiences with AI to help demystify the technology and make it more accessible for learners.
- 🔗 The panelists provided various resources and publications for further exploration of AI literacy, emphasizing continuous learning and adaptation to the rapidly evolving field of AI.
Q & A
What is the main focus of the panel discussion in the provided transcript?
-The main focus of the panel discussion is AI literacy, specifically how AI is implemented in education and the importance of understanding AI as educators, researchers, and students.
What is the role of Laura Allen in the context of this panel?
-Laura Allen is an assistant professor at the University of Minnesota, and she contributes to the panel with her research background at the intersection of AI literacy, focusing on text comprehension, text production, and developing adaptive strategy instruction and feedback for students.
What does Stephanie Adragna bring to the panel discussion?
-Stephanie Adragna, a PhD candidate at the University of Washington, brings her research on AI literacy education to the panel, emphasizing the importance of guiding young people's interaction with AI and ensuring they are not only consumers but also creators of technology.
What is Drury Long's perspective on AI literacy?
-Drury Long, an assistant professor at Northwestern, discusses the importance of AI literacy as a set of competencies that enable individuals to critically evaluate AI technologies, communicate and collaborate effectively with AI, and use AI as a tool in various settings.
What is Tyron Young's view on AI literacy in the context of education?
-Tyron Young, a senior manager for research and development at Digital Promise, emphasizes the need for a checklist to evaluate the ethical considerations and potential risks and rewards of AI technologies, especially in educational settings.
How does the panel view the importance of AI literacy in everyday life?
-The panel views AI literacy as crucial in everyday life due to AI's active role in decision-making processes, necessitating a basic understanding of how AI works to make informed decisions about its use and the data shared with AI systems.
What is the significance of the panelists' diverse backgrounds in the discussion?
-The diverse backgrounds of the panelists, including academia and industry, provide a multifaceted perspective on AI literacy, covering aspects from theoretical understanding to practical implementation and ethical considerations in various contexts.
What is the panel's stance on the integration of AI in educational technologies?
-The panel acknowledges the potential benefits of AI in educational technologies but stresses the importance of transparency, understanding the implications of AI use, and maintaining spaces for socialization and collaboration among students.
How does the panel address the ethical considerations of AI in the classroom?
-The panel suggests that ethics should not be an afterthought but an integral part of AI literacy, starting with questions about the implications and potential consequences of using AI in educational settings.
What resources or activities does the panel recommend to promote AI literacy?
-The panel recommends various resources and activities, such as the book 'Critically Conscious Computing', which provides teaching units and activities to foster AI literacy, including critical thinking about new technologies.
What is the panel's opinion on the need for AI literacy in university computer science programs?
-The panel believes that AI literacy, including understanding machine learning and its ethical implications, should be a core part of computer science education, not just an elective or add-on course.
Outlines
📚 Introduction to AI Literacy Panel
The panel, consisting of experts from diverse backgrounds, is introduced by the moderator to discuss the importance of AI literacy in education. The moderator emphasizes the ubiquity of AI and the need for researchers, teachers, and students to understand its implementation in learning. The discussion aims to explore the responsibility of making students AI literate, including understanding AI's role in teaching and how their data is utilized. The panelists are given flexibility to share their perspectives, and the audience is encouraged to submit questions during the talks.
🌐 AI Literacy as a Multi-Level Understanding
Laura Allen, an assistant professor at the University of Minnesota, draws parallels between traditional literacy and AI literacy. She discusses the necessity of understanding terminology, concepts, and the multi-level components of AI, similar to language comprehension. Laura emphasizes the importance of not just knowing the individual parts of AI but also interpreting them within one's own experiences and beliefs. Her talk underlines the significance of critical understanding and the interactive nature of language and AI systems.
👪 Engaging Families in AI Literacy Co-Design
Stephanie Adraga, a PhD candidate at the University of Washington, presents her research on involving families in co-designing AI literacy activities. She discusses the transition from digital natives to an AI generation and the importance of guiding interactions with AI. Stephanie's work focuses on family joint engagement, which facilitates access to AI technologies' language and power, fostering critical engagement for designing social futures. Her studies explore various AI literacy activities that families can engage in together, promoting creativity and self-expression with technology.
🔍 Analyzing AI Education Resources and Family Engagement
Stephanie continues her discussion by sharing her findings on the analysis of AI education resources and the roles parents play in AI literacy activities. She highlights the challenges in finding instructions for using AI resources and the sparse coverage of AI concepts, often ignoring the social impact. Stephanie's work includes developing platforms like CopyMates, which allows children and parents to train custom AI models, fostering skepticism and understanding of AI's limitations. Her research underscores the need for interdisciplinary approaches and the inclusion of social impact topics in AI literacy curriculums.
🤖 The Importance of AI Literacy in Everyday Decision Making
Drury Long, an assistant professor at Northwestern, discusses the unique aspects of AI compared to other technologies due to its active role in decision-making processes. He emphasizes the importance of understanding AI at a high level to make critical decisions about its use, especially given the black box nature of AI decision-making. Drury advocates for public AI literacy to empower individuals to engage with AI systems knowledgeably and critically in various aspects of life, including education.
🏫 Integrating AI Literacy in Educational and Informal Learning Spaces
Drury Long elaborates on his definition of AI literacy as a set of competencies that enable individuals to critically evaluate AI technologies, communicate effectively with AI, and use AI as a tool. He shares his work on designing learning interventions to foster AI literacy, particularly focusing on practical understanding rather than programming skills. Drury also discusses the importance of considering AI literacy in relation to other literacies like data literacy and computational literacy.
👨🏫 Perspective of an Educator on AI and Ethics
Tyron Young, a senior manager for research and development, brings an educator's perspective to the discussion. He reflects on the evolution of digital literacy and its intertwining with AI literacy, emphasizing the cultural context of students and the importance of ethical considerations. Tyron discusses the need for a checklist to evaluate the risks and rewards of engaging with AI technologies, especially in educational settings where students must navigate the rapidly advancing technology landscape.
📘 Final Thoughts on AI Literacy and Moving Forward
The moderator wraps up the panel discussion by thanking the participants for their insights and opening the floor for audience questions. The panelists have shared diverse perspectives on AI literacy, emphasizing the need for understanding, critical evaluation, and ethical considerations in engaging with AI technologies. The conversation highlights the importance of AI literacy in educational and informal learning spaces, as well as the responsibility of educators and developers in fostering this literacy.
Mindmap
Keywords
💡AI Literacy
💡Adaptive Learning
💡Critical Thinking
💡Data Literacy
💡Machine Learning
💡Algorithmic Bias
💡Ethics
💡Human-AI Interaction
💡Educational Technology
💡Interdisciplinary Approach
💡Generative AI
Highlights
The panelists discussed the importance of AI literacy in education and its implementation, emphasizing the need for students, researchers, and teachers to understand AI.
Laura Allen compared AI literacy to traditional literacy, highlighting the need for understanding terminology, concepts, and the ability to interpret AI within one's own experiences.
Stephanie Adragna emphasized the role of families in AI literacy, suggesting that joint family engagement can foster critical conversations and understanding of AI technologies.
Drury Long discussed the unique aspects of AI literacy, noting AI's active decision-making processes and the importance of understanding its workings to make informed decisions.
Tyron Young shared insights from an educator's perspective, stressing the importance of ethical considerations and the need for transparency in AI applications used by students.
The panel explored the idea of AI literacy as an interdisciplinary field, intersecting with data literacy, computational literacy, and the ability to critically evaluate AI technologies.
Adragna shared findings from her research, indicating that children can become more skeptical and aware of AI's limitations through hands-on activities involving AI model training.
Long introduced the concept of 'critically conscious computing', providing a framework for evaluating new AI technologies with a focus on ethics and accountability.
The panelists agreed on the necessity of fostering AI literacy that encompasses understanding AI's social impact, ethical considerations, and the ability to critically engage with AI systems.
Young raised concerns about the rapid advancement of AI and the lack of clear guidelines for evaluating new AI applications, suggesting the need for a checklist to assess AI technologies.
The discussion highlighted the importance of including AI literacy in the curriculum, not as an afterthought but as a core component, starting with ethical considerations.
Adragna and Long both shared resources and activities designed to promote AI literacy in both educational and familial contexts, emphasizing hands-on and interactive learning experiences.
The panelists stressed the need for transparency from AI developers, particularly regarding data collection, algorithmic processes, and the potential implications of AI technologies.
Long and Adragna discussed the importance of machine learning literacy, suggesting that it should be a required part of computer science education to ensure a well-rounded understanding of AI.
The conversation concluded with a call to action for the development of AI literacy programs that are interdisciplinary, ethically grounded, and focused on empowering learners to critically engage with AI.
Transcripts
so thanks everybody for joining today
super excited to have a really great
group of panelists here
um we are going to be talking a little
bit about uh AI literacy uh we have uh
four great panelists from all different
backgrounds we have Laura Allen who's an
assistant professor at the University of
Minnesota
um Stephanie adruga a PhD candidate at
the University of Washington uh Drury
long who's assistant professor at
Northwestern and Tyron young who is a
senior manager for research and
development at digital promise so super
excited to hear from everybody I'll just
sort of frame the discussion a little
bit uh before we get started and kind of
let everyone know how it's going to work
um this panel sort of came out of uh
increasing uh need for thinking about
how AI is implemented in education and
AI sort of as an ubiquitous
concept all around us
um a lot you know this conference is
themed uh empowering learning Learners
in the age of AI so we also kind of need
to think about how we're teaching AI as
well and hey how AI literate we are as
as researchers teachers and students
um and so you know a lot of the talks
we've seen so far in the conference and
the panels and so forth have really
focused on uh the use of AI to improve
um improve learning like so how can we
build Ai and Technologies you know to
make it more adaptive for students and
so forth
um to prove collaboration
um sorry
um my dog needs to be part of the panel
as well
um
but you know I think one of the equally
important questions to ask is what is
our responsibility
um to sort of make you know our students
literate in AI as we you know try to use
it to improve learning
um so on the one hand
this relates to how we we teach students
about AI
um and and how you know it's ubiquity in
their everyday lives but on other it
really relates to kind of empowering
them to understand how AI itself may be
used to teach them and how their data is
being used
um so uh with that I want to invite all
the panelists so everybody's going to
have about seven to nine minutes but
pretty flexible to kind of give their
take on AI we've left it super open so
I'm hoping to get some variety of of
perspectives here and then uh we'll open
it up for questions at the end I will
encourage everybody who all the
attendees to go ahead and type your
questions in the um there's a q a
section here and it would be great if
you typed your questions in as the
speakers are talking so we can go ahead
and kind of start thinking about them
before we sort of reach that q a part uh
after everyone's done talking so with
that I'll we're just going to go in
alphabetical order so I'll hand it over
to Laura to go ahead and speak first and
then
um we'll go Laura Stephanie Ed uh durian
then Tyron so Laura whenever you're
ready
all right
um let me just share my screen
can you uh oh no
my mouse has disappeared
someone can y'all just see the panelist
list yeah probably we can see a black
background
it would give you popped up so there we
go oh yeah can you we're good now
alrighty
um so hi everyone
um thank you Caitlin for inviting me to
be on this panel
um so for those of you who don't know me
probably everyone uh I um am an
assistant professor at University of
Minnesota and most of my research has
been sort of at this weird intersection
between AI literacy
um where I both study literacy so I've
spent a lot of time studying text
comprehension and text production how
people
um we can best support students in
learning how to read and write
um and then I've also worked on
developing technologies that can produce
provide adaptive strategy instruction
and feedback for students
um where I haven't done as much work is
actually on AI literacy itself um so
um hopefully my take on this is at least
a little bit useful
um so when I was thinking about AI
literacy I thought about again my
backgrounds in literacy so I thought
about you know what we think about in
terms of regular literacy or you know
the traditional view of literacy
um there's if you if you look up the
definition of literacy online you will
find many of them and they're they're
quite different
um but generally right when we think
about literacy we think about some sort
of ability to read write speak listen so
using language
in a way that allows us to communicate
with others and to make sense of the
world right and so when I was thinking I
was thinking about how we can apply that
same sort of perspective and what we
know from teaching students in that
realm of literacy to this sort of new
way of thinking about literacy
so
um so when we think about literacy from
a text perspective we think about the
fact that we have some sort of language
that is generally you know connected in
some sort of way right so we have
individual words that make up the text
and then they're combined you know in
special ways to make sentences and then
we can sort of connect those
um in broader ways to make some sort of
document and if we want students to
really understand
um you know sort of how to make meaning
out of text then they really need to be
able to have knowledge at multiple
different levels so first they need to
understand those individual words right
so all those different words and what
they mean they don't need to know every
single one of them but they generally
need to know most of the words in a text
they're reading they need to know what
they mean and what they might be
associated with in the world
they also need to know the rules for how
those words can be combined right so we
have a lot of different you know words
um but there's a very finite set of
rules that we sort of operate under in
order to combine those words so not
every single word can sort of just be
combined with itself right we need to
know the ways in which they move
together and interact with each other
and then once we have these sentences
right we also need to be able to combine
them with our knowledge of the world in
order to make sense of that language
right so sort of a Hallmark property of
language is that it's sort of
it's all symbolic and so it doesn't
really mean anything in the world
without us interacting with those
symbols in the world right so there's no
reason that the word chair means chair
it just does and we sort of associate it
with certain things to write our
interactions with the world in order to
make sense
so if we try to think about those sort
of same principles and we're thinking
about AI literacy
um I think at least we can draw some
insights from that
so it similarly is going to require us
to sort of break down what these what
this form of literacy means into
multiple different components right and
and this mapping is not perfect so bear
with me but but generally we need to be
able to First think about something like
the words right in the in the analogy
that I'm giving so we need to know what
the terminologies are what are the
concepts right what are the things that
are representing whatever we're using so
they need to know
um what types of data is available to
them and how they can sort of use that
data what are the different ways that
data can be stored or could be modeled
they need to know how that data can then
be used in some sort of model in order
to do something functional right so what
are the ways in which I can take data
from language data from you know click
stream data data from facial recognition
and sort of combine those in different
sorts of ways
um to develop some sort of model and
then finally students need to not only
understand and all of those different
sorts of pieces which I think is a big
Focus right in digital literacy and AI
literacy and they need to know those
things but they also need to know how to
interpret those things within the
context of their own lived experiences
and their own beliefs and their own
knowledge right and I think this piece
is really really critical and I think
something that I is really important
that doesn't get missed
um so similar to research and literacy a
lot for a long time people sort of
focused on students knowing vocabulary
and knowing grammatical rules and
probably you were a kid in school at
some point and you were taught a lot of
vocabulary and a lot of grammar rules
and we don't we do that often while
sacrificing teaching students how to
construct meaning right and how to
interpret things through their own lens
and through their own knowledge and I
think that we need to sort of foster
that in a similar way within AI literacy
um so just to wrap up
um I think that sort of the big ways
that we should support students to
become more literate is both teaching
them the meaning of those different sort
of component pieces so where the data is
you know what types of data they have
where it's getting collected what type
of models there are and what are the
rules for how they work and then also
how to integrate that with their own
understanding
um and so oh this is coming out weird
sorry uh and so like when we think about
language we think about the fact that I
am a user of language and I'm producing
language right now right but critical to
my the success of me producing language
is that you understand what I'm saying
and that we have some sort of shared
understanding right so if my goal is
misaligned or you don't interpret my
goal in the same way then I have failed
as a speaker right and so I think we
should think about literacy from this
interactive way between in students and
AI systems so that they both are aware
of how they work but also aware of how
different sorts of goals whether those
be explicitly embedded within the system
are sort of more implicit in terms of
biases and the data and things like that
can all sort of govern how those things
work and how students can leverage those
in order to make decisions and to
function in society so
I will end there
thanks Laura
um super interesting uh I think Connie
mentioned that your screen might have
been black at the end but I don't think
we missed anything uh on your last slide
I think it's okay we got the got the uh
we understood your language
um so with that I'll go ahead and turn
it over to uh Steph who who do you have
slides to share okay
cool
can you hear me
yes hi uh let me try to get this on the
big screen
hopefully you can also see my slides now
great uh I'm Steph uh I'm a PhD
candidate at the University of
Washington in Seattle I've been doing
research in the space of AI literacy
education since 2016 until now and I
started doing this kind of research
because I noticed that uh young people
are now growing up with AI in 2016
things like Syria Alexa voice assistants
were just starting to enter kids uh
homes and kids lives we've seen like a
rapid expansion of these Technologies
and the ways in which they're being
exposed to young people ever since so I
realized that we're going from the
digital natives or you know a digital
generation to an AI generation where
young people will probably learn how to
interact with a voice assistant before
they even read to learn how to read and
write and via voice assistant they can
search the web they can have access to
any type of information so
for me it was important to to figure out
like how do we guide this interaction
with AI how do we make sure that
um young people and families are not
only consumers but also creators of
technologies that are being developed in
the space and applications
um and more importantly to make sure
that families are being protected by
from our great mean bias or all sorts of
other issues that are being introduced
with these Technologies becoming
commonplace so what we know until now is
that in general like when parents are
involved in kids learning and exposure
to technology it has beneficial impact
um that family joined media engagement
was also something that was studied over
the years when they were looking at how
kids and parents consume like TV
together or interact with mobile apps um
it has beneficial impacts in terms of
developing and fostering critical
conversations in the family
um and in our in my own work and also in
the work of from other people here in
the panel from Durie and others we've
seen that there are many different ways
in which we could introduce young people
to to Ai literacies and all the
competencies that are part of that so
what I wanted to learn was how could I
actually involve not only kids but also
their families in co-design co-designing
AI literacy activities
um how do kids and parents learn
together about Ai and what are all the
types of Designing that we could do to
support family literacy and you heard me
use this term a lot so I wanted to also
kind of find like how I'm thinking about
it my most also disappears so I cannot
uh minimize the video but um for me I
actually really include the ability to
read work with analyze and author with
AI and they Foster a critical
understanding of this technology so the
critical part is quite important and
what I found throughout the studies I've
been doing
is that family joint engagement in AI
literacy is really facilitates access to
an evolving language of AI Technologies
it also facilitates access to the power
in the community this technology can
bring
um that family joint engagement in AI
literacy Fosters critical engagement
necessary to design social Futures that
are imagining meaningful uses of AI in
the home and we also discovered that
kids and parents like to learn together
with AI how to engage in Creative coding
and imagine like how to
well basically we imagine Computing
Norms at home and enable much more
self-expression family expression with
these Technologies
um so I've done several studies that
were formative like analyzing the
existing Ai curricula and see what works
and what doesn't analyzing how AI coding
impacts and understanding uh analyzing
what are all sorts of different family
co-design activities for AI literacy
that
um both parents and kids like doing
engage in jointly and last but not least
a longitudinal longitudinal study
looking at how families develop AI
literacies over time
and
um the tldr I'm not going to have the
time to go into detail uh for all of
these studies I'll share a link with all
the Publications so you if you're
interested you can learn more later
um right now like there are
um uh growing Corpus of resources for AI
education and literacy
um we found together with my co-authors
um a corpus of 51 resources that are
still active and available online and we
analyze them these include like
activities demos curriculum
um lots of different types of resources
and we analyze them
um to really see like to which extent
like what AI Concepts they
um Talk they
um demonstrate uh what big ideas for a
literacy they cover what age groups are
suitable for and so so on and so forth
um and what we found actually is that
there's really the instructions for how
to use this uh both for families or in a
classroom or for educators or informal
learning spaces are quite uh hard to
find a news that there's a sparse
coverage of AI concept and primarily
ignoring social impact that AI has that
very often for some of these like demos
and tools and activities like their
prohibitive costs or Hardware
requirements and there's
um General lack of consideration of
prior knowledge like what do people
know when they're coming to these
activities to this curriculum to these
demos
um how could we could we support them to
to engage in these activities
um and also limited opportunities for
self-assessment and reflection basically
it's the wild west uh very early stages
there's a lot to do and there's a lot to
opportunities for growth uh in in the
space of AI education resources that are
available online and recommended for
teachers for example
uh in in the second study I mentioned
um where I wanted to try to address some
of these issues that I found
um I use the platform that I built
called copymates which is free it's open
source it expands scratch to actually
allow kids and their parents to train
their custom models with images with
text with sounds and once they have
their custom models they can build
scratch like like games
and what we found in this study we had
57 children ages 8 to 12 from public
private schools after school programs
and we got them to engage in three
different
um activities where they could do a text
training activity an image training
activity and a smart home activity and
after they engaged in this activities
what we found is that they became more
skeptical of the way they were
describing the intelligence of smart
assistant they're already having the
home like Alexa or Google home or smart
robots so it really allowed them to not
only develop ways to test like what
these Technologies can do but also have
uh I realized that this intelligence is
not magical it doesn't come you know
like from The Ether that is actually
done by people who record the data who
train an algorithm
um that there's like
people Behind These Technologies and
that there are limitations that come
with this technology so it allowed them
to become much more skeptical of these
Technologies
um so like I mentioned that yeah
oh yeah no I just wanted to make sure
your slides weren't progressing I just
wanted maybe if you unshare and re-share
that would that would help sorry to
interrupt
no worries uh let me try again
are they progressing now okay we're good
yeah okay
um yeah so the platform is available
online feel free to to try it uh it's
translated in lots of different
languages and it's being used worldwide
um uh so moving forward
um
I uh I mentioned that we did this
longitudinal studies with families at
home especially during covet because it
was very hard to do this in person
um what was interesting there for for
the long-term study with families we
actually developed lots of different
activities around image classification
machine learning
um games with voice assistance design
and analyze AI these sessions and what
we found here uh was that parents
actually played lots of different roles
when engaging in AI literacy activities
learning activities with their kids
the the range the roles range from being
a cheerleader to being a mediator a
mentor student a teacher an observer
um and the activities that supported the
most The Joint roles where the parents
would Tinker together with their kids
and would collaborate with their kids
where activities that involved like a
physical component like a Hands-On
component a crafty component so
for example like the teachable machine
activities worked very well uh and um
the game for like testing the voice
assistance the design activity on paper
worked very well so the takeaway here
was really when we're designing or using
existing AI resources for family AI
literacies like thinking like how could
we support both kids and parents to have
discussions around these activities to
build on each other's ideas to be able
to test and Tinker
uh with these activities and not think
that because it's AI everything has to
be digital that actually providing
Hands-On and crafty ways of engaging
with these Concepts goes a long way
um so that was like a big takeaway from
from this from the study
um and kind of like moving forward you
know it was important to see how
families of different backgrounds we had
families that spoke several languages
different ethnicities different uh parts
of the world uh different family
configuration
uh we're such a diverse range of
families from 10 different states in
North America
um what we found is that they engage in
multiple forms of AI literacy which is
why we we use this theory in
um multiliteracy uh lens
um and the way we situate like Ai
literacies and not only Literacy for for
a diverse set of families is like the
ability to engage in the following
practices being able to engage in
multi-model and in body situated
practices in their home being able to
learn about different AI Concepts being
able to
engage in a critical framing of AI and
being able to design future meaningful
uses so this this approach Builds on the
theory of multiple literacies uh from
the New London group and for each of
these facets we have different
activities that are also available
online that you could try out if you so
want to
um so to end I wanted to to share some
of my current ideas for like what would
be cool to do in the space of learning
about AI uh developing AI Literacy for
families and I think it's very important
to take an interdisciplinary
approach for the curriculum we're
creating in this space to focus much
more on social impact
um topics that are relevant for both for
kids and parents
um we know we all live through covet
like what can we learn about that Nai at
the same time
climate change is big on on youth's
Minds how could we use AI to do
something about that and of course as a
form of personal expression creativity
um especially when we've seen everything
that is happening in the space of the
diffusion model stability AI Delhi like
it's like creativity on steroids it's
like really really uh it's been
exploding so creating a rich collection
of projects and lessons that stimulate
families in various ways and allow them
to engage and have lots of interactive
um playgrounds where like uh both kids
and parents could learn by doing so
I took here an example from of the
spaces interactive spaces that are on
hugging face which is the largest I
think we might need you to
I think we might need it on share and
re-share one last time okay
that is annoying uh let me try again
can you see it now
yeah yeah
so like looking at the spaces from
hugging face which is the largest open
source Community for natural language
processing machine learning uh where
people build like all of these
interactive spaces to to play with AI
and thinking like what would that look
like if we have something similar that
is customizable demos that allows for
easy integration in curriculum for youth
curriculum for families
um and then also thinking about what is
this generative AI that is exploding
right now what does that mean for for
the future of AI literacy like can we
support automatic asset creation and
meet kids where they're at right like if
they spend a lot of time on Minecraft
like what does it look like to actually
automate like assets and worlds in
Minecraft and allowing them to do that
and customize that
um how do we allow them to learn from
their own data right like I showed you
how uncognits they can train and test
their own models like
um I I noticed that kids and parents
have much more intuition about Ai and ml
Concepts when they work with their own
data
then like how do we integrate some of
these AI libraries plugin programming in
platforms that are widely successful
with kids like Roblox
um again hugging face demos
um and yeah chats are big like maybe we
could use a smart assistant to get to
teach kids like how to train their own
assistant
um and there are many many things that
we could do in this space so
uh I'm gonna stop here because I'm I
think I'm a time and we also have the
technical challenges but I'm looking
forward to the discussion later and
thank you for your attention
thanks Steph uh really interesting stuff
thanks for sharing um during you ready
you're up next
yes I will share my screen just a sec
and uh I'm not sure about this the
slides issue will continue I'll try to
um either like wave my hand or put it in
the chat if they get seem to get stuck
again
yeah just let me know can y'all see my
screen
yep
great
um well I've really enjoyed the talk so
far thanks so much for having me today
um I obviously didn't come up with the
world's most creative title for this
talk but
um just to give a little context about
who I am
um I am an assistant professor in the
department of communication studies at
Northwestern
um and I'm a human centered AI
researcher so sort of half of my work is
focused on designing uh learning
interventions that can help people
better understand AI Technologies
um so like Stefani I've worked with
families and kids and I'm also
interested in looking at other sort of
informal learning spaces like helping
people understand AI in the workplaces
and elsewhere in their everyday lives
and then the sort of flip side of my
research focuses on designing AI that
can better understand people
um so AI That's better able to engage
with us in sort of social creative
environments and explain itself to us
but today I'll talk about the the AI
literacy side of things
um
so in the the spirit of sort of the
um uh topic of the session
um I wanted to talk today sort of about
why I think AI literacy is important and
in particular um you know in the
educational context why I think AI is
sort of different from other types of
technologies that we interact with
um and why this this literacy is
necessary I'm going to talk a little bit
about how I Define and conceptualize AI
literacy
um it's uh you know a very like emerging
field and so I'm really excited to
engage in discussion about sort of what
fits in the definition of AI literacy I
think that's very much uh sort of still
forming
um and then I'm going to give one
example of how I have fostered ai
literacy in my own work uh just an
example of a project that I've worked on
um
so to dive into it
um I'm going to talk first about why I
think AI literacy is important
um and I I think this is important to
distinguish because there's a lot of
Technology that's become integrated into
our lives and even though we use it
every day we usually don't really need
to know much about how it works so I can
like screw in a light bulb and I usually
don't need to know a lot about
electricity except to understand you
know how not to get shocked
um I don't need to you know understand
in depth how my microwave Works in order
to use it and so something I get asked a
lot in my work is like how is AI
different from these Technologies uh you
know it's integrated into our everyday
lives it's showing up in our homes our
workplaces our schools and even in a lot
of the same contexts as the Technologies
I showed on the previous slides it's
being used in our TVs it's being used in
our light bulbs it's being used in our
cell phones so what sets AI apart from
these other technologies that are sort
of we sort of take for granted and we
don't necessarily dig into how they work
and and I think in my opinion one of the
things that really sets AI apart from
these other Technologies is that it's
engaged in active decision making
processes with and for us in a way that
other Technologies are not so AI is
deciding um you know what content to put
in front of us when it's selecting what
to show on our news feeds selecting what
to display when we type in a search
query it's providing us with
recommendations that it decides to show
us when we're engaging in entertainment
uh searches and shopping and an even
more consequential domains it engages in
decision making uh sort of with and on
behalf of us in terms of deciding people
to hire deciding when someone should be
released from jail deciding who and who
who is and who is not a criminal and I
think that this really makes AI
um much sort of more consequential and
more important for us to be able to
understand at least at a high level
um how it works so that we can you know
make critical decisions about where to
use it when it's appropriate to use it
um and I think this is sort of
exacerbated by the sort of Black Box
nature of the decision-making process
which has been shown in Prior work to
um
sort of lead to issues with trusting AI
so we you know we see the input that we
put in we see the you know search result
that we type into the query we see the
uh sorry we see the query we type in we
see the results that come back but we
don't always know what's happening in
between and this can lead to issues with
trust people not even recognizing that
they're engaging with algorithms
um and um sort of confusion over why
algorithms are making the decisions that
they are
um confusion over you know when it's
appropriate to share data
um and so I for all of these reasons I
think it's really important that we have
and Foster broader uh public AI literacy
and I think this is particularly
important within an educational context
um you know to to draw on the title of
the the conference to empower Learners
um to sort of critically engage with
when they're sharing their data with AI
algorithms understanding what that data
is being used for making sense of like
you know Technologies that they might be
using to help with their own research
and learning processes and understanding
you know what's put in front of them if
it's like recommended by an AI system so
I think this is equally as important in
an educational context as it is in other
contexts
um so with that context in mind I'm
going to talk a little bit about how
I've defined AI literacy in my own work
and how that sort of informed the work
that I've done
um
I've defined AI literacy in my work as a
set of competencies that enables
individuals to critically evaluate AI
Technologies communicate and collaborate
effectively with AI and use AI as a tool
online at home and in the workplace so
this is sort of a very practical
definition I'm thinking a lot about like
how this can be useful to people
and in particular the framing that I
take on AI literacy is focused on things
that can be useful to people even if
they don't know how to program AI so
they may not know how to develop or
build AI systems but what are some high
level ideas that they can sort of take
away that can help them as they engage
with systems in their everyday lives and
I take this Focus because I think it it
lowers the barrier to entry a little bit
for people and also helps to I think
there are useful things that people can
understand about AI that do not
um even if they may not have an interest
in learning how to program or build AI
um and so I situate AI literacy sort of
in respect to some other literacies that
are commonly talked about I see AI
literacy as sort of intersecting with
data literacy especially within the sort
of subfield of machine learning which
engages a lot with data
um I see other forms of literacy as
being able to inform AI literacy so for
example computational literacy uh
understanding how to like program and
um using computers to build and create
things I see this as something that can
certainly help understanding AI but it's
not necessarily uh required
um and
um but of course I see like
understanding how to use and interact
with devices as being more required for
AI literacy so this is sort of how I've
been conceptualizing this I'd love to
engage in more discussion about other
people's um you know thoughts on this
um and in my work I've engaged in a
review of uh literature
um sort of in the space
um relating to AI literacy so looking at
literature related to AI education human
AI interaction Computer Science
Education
um research on
um
sort of how people perceive and make
sense of AI and from that I've
synthesized a set of competencies and
design principles so the competencies
are sort of high level ideas uh for
people to understand about Ai and the
design principles are aimed at like
researchers researchers Educators and
designers who might be interested in
designing AI literacy learning
interventions some uh sort of principles
based on the literature that
um Can facilitate effective learning
interventions and I'm not going to talk
about all of these today but I just want
to highlight a couple
um to give you an idea of what I'm
talking about so one of the AI literacy
competencies
um that I defined based on prior work
was the ability to recognize the
computers often learned from data
including one's own data so this is the
type of sort of high level idea I'm
talking about
and one of the uh design principles is
to consider designing interventions with
embodied simulations of algorithms or
Hands-on Physical experimentation with
AI technology I know this is something
that both Stefania and I have found in
our work and that has sort of
corroborated each other is that this
sort of like tangible embodied nature of
AI literacy learning interventions can
really help concretize this like
abstract concept of AI for learners
and um I just want to close with um sort
of how I've fostered AI literacy in my
own work I'm just going to give one
example
um for context this was developed as
sort of an at-home activity
um for families with kids of all ages to
interact with we are focused mostly on
like Middle School age learners but we
were developing activities for like the
whole family to engage with
and in this activity uh Learners were
tasked with creating data sets to teach
an AI how to recognize and distinguish
birds from non-birds and they were given
a set of cards to create data sets with
so the cards contained creatures some of
the creatures uh were Birds others were
things that might be mistaken for Birds
such as like a turtle or an alligator or
a bat and each card contained a picture
of the creature and the list of features
describing the creatures with things
like it's Habitat it's color its size
and they were tasked with creating a
positive training data set and a
negative training data set with like a
very limited number of examples so they
placed like three cards in the positive
training data set three in the negative
training data set and they were supposed
to sort of engage in thinking about how
to balance even that very small training
data set so thinking about how to
represent
um you know all birds with just three
examples
um and then they could put weights on
different birds to put little bit more
weight on that example if they thought
it was more emblematic of being a bird
so if you put like three tokens on a
sparrow and one token on a goose the
algorithm would be trained with three
sparrows and one goose and then they
could take a picture of the board they
created upload it to a website we
developed and see how well their
algorithm did at classifying different
animals as Birds versus non-birds and
they could sort of iterate on this and
test out different outcomes and sort of
explore how different types of data sets
resulted in different outcomes
and this was intended to communicate AI
literacy competition competencies
relating to understanding the steps of
machine learning understanding how AI
uses knowledge representations to make
decisions and understanding that AI
systems learn from data and it used
design principles like contextualizing
data providing opportunities for social
and embodied interaction making AI more
explainable and providing opportunities
to like teach or program AI so that's
just one example of how I've done this
in in my own work and I look forward to
the panel discussion
thanks Jerry that was super interesting
I love love the video at the end
um we'll just in the interest of time
just pass it right over to Tyron if
you're if you're ready to get going
uh yeah let me go ahead and
share my screen and hopefully
fingers crossed
all right
um so I guess in for context uh I am a
been the educator for many years now and
and recently stepping away from the
classroom but as I kind of think about
digital literacy I was thinking of it
really from this practitioner standpoint
and so even in preparing for this it was
really looking at a lot of the work of
of various people
um but really looking at the work that
dirty long actually I did so kind of
what she was just explaining uh in some
of the competencies I began to kind of
dig into that and then think about some
of this research of what does this mean
from a practitioner's standpoint and it
was really around somebody's
competencies of Ethics
um in where we have designers that think
about some of these ethics pieces
um but it's helping students think about
the outside world and think about how do
we interpret what is the ethics of the
people who created this should I be
questioning how am I engaging with this
when they see something at school
because of the particular space or
spirit that they're in in that moment
there is a automatic assumption that
this is good for me and this is safe and
this is okay and sometimes students
bring that over into their home life and
to say if a friend has it or someone
shows this to me some of these things of
Ethics it must be okay and it must be
good for me and so as we get to think
more about this type of work I began to
think well what many of us probably grew
up with was this realm of digital
literacy or this term in the sub of
digital literacy and thinking about well
what's the culture around this and how
should I be saved and what's a practical
thing
um so for those of you who can remember
the days of like aim and AOL chats and
all that good stuff you would really
begin to think that was the initial push
for about the safety piece
um and so we didn't see anything wrong
with it Myspace was a thing at one point
where we just didn't see anything wrong
with chatting in certain ways and then
there became this like larger social
Consciousness around this term digital
literacy so I put the tape there to say
don't forget it it is really tied and
interconnected into this and
understanding what is a part of the
culture of
um the students that we're kind of
working with what is in the culture of
families and what is in the culture of
um this generation
and so thinking about in the culture of
the generation just
in that same P same vein uh we have to
think about what is it that we think
about as far as risk versus reward right
um this is what is the constant battle
whether that is an educator trying to
decide what do they bring into their
classroom
um District leaders trying to figure
those pieces out
um but students are are also starting to
have to think about this families are
starting to have to think about this big
risk versus reward and so in that vein
of both digital literacy and AI literacy
um you have to really begin to think
about
what is it that we're putting in front
of each other what are we putting in
front of our families and so this would
be um I'm not sure if anyone has been on
social media in the last like 72 to 120
hours you have probably seen a friend
family member someone who wanted their
morphed themselves into this these
different characters and these different
types of avatars to see what their
identities would look like
um
but in that it seems like one of those
it seems fun it seems entertaining it
seems very now and everyone is doing it
it's like all the cool kids are doing it
so we all want to try to do this try out
the Avatar everyone was trying to figure
out what is that but then
I go back to this and I began to think
about some of these components right
um I can't say inherently that the
developers are doing something unethical
but I
I'm not certain and then I began to
think about well is are they fully
transparent in certain things I then can
begin to think about a checklist for
myself to say do I want to engage or do
I not engage
um and so that is something that youth
are constantly having to deal with and
that Educators in the classroom are
constantly having to deal with to figure
out we want to be hip we want to be one
student to be engaged we want it to be
relevant
um to what's going on but we need a
checklist and I think that's one of the
things that
hasn't been as crystal clear because
technology has been so rapidly advancing
we haven't had a clear understanding of
what are the new checklists that we need
to do what are the new types of
questions we need to be asking to ensure
that both students and families can make
informed decisions about how they engage
with AI
um at that point I think the last I feel
like the best big wave of that might
have been around having Siri and uh
Alexa and saying well they're listening
to all our conversations and that would
was a big point contention for some
people about how much could they have
access to and how much cert how much
data is starting apps maintain and so
now some people just download apps
without even thinking twice about it but
it's that type of checklist that the
public needs to think about
um when we're thinking about what are
the known inputs and outputs of like
what are we actually feeding directly
into it and then what are we immediately
getting back but then also thinking
about what are the possible unknown
inputs and uh outputs that should be
um of what's going to be coming out of
this how is my data going to be used how
do I think about what are some of the
ethical pieces of this and then what are
the potential uh what is the potential
risk and reward for a lot of this so I'm
interested to hear from hear from you
all as we kind of talk about this and
delve into this a bit more
um just to see what people think
thanks everyone I really appreciate the
the perspective from someone who's been
in the classroom with middle school
right in the last few years
um yeah so I'm pretty sure seventh day
eighth graders will are right on this
um wave
um okay well thank you all for these
really really interesting talks I really
appreciate um your effort putting them
together I want to go ahead and uh you
know invite the the attendees to go
ahead and drop any questions that you
have in the the Q a box that you have
um but but I want to start um just kind
of falling off of Tyron while other
folks maybe have a chance to type in
their q a but
um I think you bring up with some really
interesting points about how to
translate this to uh teachers in the
classroom especially when you have this
um you know an age group that does have
access to a lot of devices and apps that
are just sort of readily available and
so I'm curious if any other panelists
have any ideas for what what that looks
like you know if you have any
recommendations for like you know or
what would be like a in a perfect world
how would that work for for especially I
know like you know talking about inside
the classroom or or um Stefania you were
talking about like at home like what are
what are some of the things that people
might pay attention to when we're
thinking about
what are the apps that would promote AI
literacy or what are the what what are
what are some of the things that we want
to consider as we are filled with the
world full of AI
yeah I just want to say I love the idea
of this like checklist of considering
you know what we should be looking at
when we're you know or what children
should be looking at when they're
starting to engage with these
Technologies and some of the things that
like pop into my mind that would be
interesting to have on you know such a
checklist is
um you know who created the data set
like where
um you know was this data like pulled
from
um asking what data like they're
collecting from you
um those would be some initial things
that I think would be valuable to ask
about these applications
um just off the top of my head
yeah
um
sorry about that uh I was about to say
the when you're saying that just now Dr
long
um even just thinking about these pieces
of pushing developers about this piece
of transparency on actually just having
to like blatantly list it out and not
buried in the fine terms of fine print
that could be randomly emailed to you at
a later time
I was going to say that my entire lab
actually wrote a book about this which
is available online uh it's called
critically conscious Computing and on
the each chapter has like teaching units
and activities and on the AI chapter
there's actually a unit on teaching AI
critique that actually goes through a
series of questions that you can ask
about each new application or each new
you know like technology model Gadget
that you might be exposed to or your
friends would introduce to you at home
in the classroom or just on the web so
I'll post the links on the chat
um that's like an immediate immediate
like action item or like response
um I think the part that is
I I really like that you talked about uh
lensa and you know stability Ai and
generative AI I think the part that is
tricky is that the technology is moving
very fast and it is for most part pretty
opaque it's a black box so
um it's kind of a chicken and egg
because you want people to understand
more how it works so they ask for their
rights so they understand like oh my
data is being scraped without my consent
or like this is actually using artists
work without uh proper reference or like
you you want to understand how the
technology works just so you can keep
the designers and the companies that
created accountable at the same time
it's hard to understand how how it works
when it's constantly evolving changing
and it's so opaque right so my push in
the space is to to have like more craft
with AI so instead of like okay I have
Dali I have lens I have all of these
things they make super cool images is
maybe like try to look at like where's
the API where's the SDK or like where is
like where can I see what's behind you
know like the scenes like how can I if
we had a physical device we could
actually open it and take it apart right
and we we did this activity with kids
where we asked them to draw what's
inside Alexa and it's fascinating right
so kind of like thinking like how do we
break these black boxes and how do we
constantly try to not be swayed Away by
the
you know the the hype wave and uh the
compelling examples but try to
understand more like how they work under
the hood
yeah I love that I like the the example
of asking them to draw draw it out
like inside
yeah alarms while you're unmuted
oh yeah I just had a question it's kind
of for Tyrone but it's kind of for
everyone but I was just thinking I feel
like we all kind of talked a lot about
teaching how AI works and and like how
the models works and I think that's
really really important but I also think
about like what our responsibility is
since we're talking a lot about this
within the context of Education I think
a lot about
you know a lot of funding is being
poured into educational Technologies
right now that are teaching you know
reading or writing or math or science or
all these sorts of things and a big
driver of the AI
is like automated scoring and adaptivity
and personalization of learning and I
think
I I guess I don't really know how to
form this question exactly Tyrone but
but I'm sort of wondering you know
what's the responsibility in in terms of
that transparency piece you're talking
about of these companies when when you
know you say something can score
something if you have people that don't
necessarily know how well it's scoring
and they're not going to sit there and
read your accuracy rates you know for
every single model and all the different
sorts of
um things but you know if we're saying
oh I can score this piece of writing
that you have right then
and we just put it in your classroom and
I don't know I was just thinking of that
piece and what you think we should do
about that
yeah and and it was it's funny as you
were asking I'm reading from uh the
question from Priscilla Gonzalez where
she is kind of hitting on this thing if
it's changing uh like human interactions
and the whole I guess that essence of
our human lives
um because some of it is you uh with
that adaptive AI I think some of the
frustration pieces that naturally
happened in learning
um
that yeah that naturally happens in
learning uh I sometimes gets eliminated
because students don't have to have some
of that collaboration with one another
and I think in the successful
integration of AI into spaces it's
making sure that there is that space for
collaboration there is some space for
there to even just be a step away from
the technology and it's just pins on
paper with a group with a couple group
members to brainstorm some things and so
I think it is trying to strike the
balance of understanding yes Ai and yes
technology is going to be there to
assist us with teaching but there are
still socialization that there's still
socialization that needs to happen
um in order for students to be able to
develop health and be able to be
productive in various settings
um whether that's within personal or
work settings
all right if no one else else has any
follow-up on that I just want to uh go
to we have about four minutes left maybe
three just so we have some time for
closing and
um we have a question about what AI
literacy looks like in University
programs that are specifically focused
on computer science
um and if any of these programs may be
missing some of the critical AI literacy
skills
whoever can Whoever has any thoughts can
jump in first
I guess I'm almost wondering if this is
some like with this question I guess if
someone could probably take away I
should say if some of this is the
critical thinking critical reasoning
pieces and aspects of it
um because I almost wonder in some of
those places of what does it mean for
some of that critical analysis of things
that we often approach to as being very
linear
um in in having just like a kind of on
off switch or if this is in the realm of
thinking about what's missing from
computer science programs is some of
that
um
what are the kind of ramifications of
what it is that we create
um so thinking about both the intended
consequences and under consequences of
the work and
uh that's why I think that one
competency uh
uh Dory you you all hit on the head when
you when you hit the part about the
ethics and I think it's often a part of
AI that is often missing and where we
are going to constantly need to Circle
back to about what are the ethical
immoral where where is ethics immorality
fall into exposing students to this
yeah I definitely think that ethics is
super important to hit on and I also
think that
um you know machine learning and like a
lot of the like most current AI
Technologies are not even necessarily
covered in a lot of like core CS
curriculum
um there's an interesting uh paper I'll
share here
um on machine learning
um in the undergraduate curriculum and
sort of a call for a need for that to be
incorporated
um you know in a sort of required way in
computer science undergraduate courses
and I think part of that is the the sort
of ethical implications and also
thinking about the I mean I something
I'm personally interested is sort of
like the interdisciplinary potential
um of a lot of these tools and
um you know how can we create
um you know experiences uh in like the
undergraduate curriculum or even in high
school curriculums that sort of help
span disciplines
um and engage people in learning about
AI in computer science but also you know
in the Arts and Humanities where it's
relevant
yeah this is super super interesting
thanks for sharing I see a lot of
resources in the chat so hopefully
everybody has a chance to to Mark those
down um we're just at the top of the
hour so I just want to yep go ahead
I had a quick ad to say that my first
reaction to this question was that
ethics shouldn't be an add-on on the Cs
classes
um and maybe that's the first question
that the Cs instructor should start with
it8 and taught in AI informatics ethics
and we start with the with ethics like
should we do this and what are the risks
and challenges and implications of
starting this project or doing this app
um so I think that should be the first
entry point and not a uh an add-on that
makes people feel good and I would say
the work that Casey fistler has on in
kind of keeping track of all the classes
and good resources for csfx would be a
great great thing to look at
thank you thank you for that Edition
okay well I just want to say thank you
again to all of you this is super
fascinating and and really uh some
interesting things to think about as we
as we move forward so thanks again for
your time I know that you all spend time
preparing for this so I really really
appreciate it and uh looking forward to
more later
5.0 / 5 (0 votes)