Popular in Professional

Business

The Allegory of the Favela: The Multifaceted Effects of Socioeconomic Mobility

University of Pennsylvania Leandro S. Pongeluppe

Architecture

Connecting Cities and Nature

University of Virginia Timothy Beatley

Human Rights Law

Human Rights and Corruption: Problems and Potential of Individualising a Systemic Problem

Max Planck Institute Anne Peters

Finance

Chinese Exchange Rate Policy: Lessons for Global Investors

University of California San Diego Michael Melvin

Law

A statistical approach to law school citation rankings

University of Virginia Joshua Fischman

Education

Firm Performance Pay as Insurance against Promotion Risk

Stockholm School of Economics Alvin Chen

An initial interpretation of the U.S. Department of Education’s AI report: Implications and recommendations for academic libraries

View more insights in Professional - Librarianship Studies

View more figures from University of New Mexico

View the page of Leo Lo

Read the Study

Next in Librarianship Studies

Impact of school librarians on elementary student achievement in reading and mathematics

Lois Wine, Shana Pribesh, Sue Kimmel | Old Dominion University

Leo Lo’s analysis of the US Department of Education’s report on AI and academic libraries highlights AI literacy, educator involvement, and preparing for AI-related issues. A roadmap for libraries in the evolving AI landscape.

Image courtesy of interviewee. December 4, 2023

Generating Workspace...

Export as PDF

This video video discusses the rapid rise of ChatGPT since its public release in November 2022 and its significant impact on various sectors, particularly education. It highlights the challenges posed to educators, as students began utilizing AI tools for homework and tests, necessitating a reevaluation of pedagogical strategies.

The speaker emphasizes the importance of understanding and addressing the implications of AI, particularly in educational settings, advocating for transparency in AI decision-making processes. Key themes include the necessity of human involvement in AI's use, the potential for AI to support educators by handling routine tasks, and the role of educators in maintaining creativity, judgment, and empathy.

The report released by the Department of Education in May 2023 is cited as an important resource, stressing the need for AI literacy among educators and students, which extends beyond mere usage to understanding ethical concerns and the broader impact of AI technologies.

The speaker advocates for collaboration between libraries, faculty, and tech companies to effectively integrate AI tools and to address fundamental issues such as data privacy, biases, and the digital divide. Recommendations include fostering AI literacy, developing guidelines for AI tool evaluation, and encouraging experimentation with AI in educational contexts. The overall message is one of proactive engagement with AI technologies to ensure their ethical and effective application within education.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.155 --> 00:03.695
So let's go back to November, 2022.

00:04.485 --> 00:08.135
When chat, GPT first became available to the, to the public,

00:08.795 --> 00:10.535
it was really astonishing

00:10.555 --> 00:13.215
how rapidly it became a, a sensation.

00:13.795 --> 00:17.375
Um, I think it reached a million users faceted than any

00:17.375 --> 00:19.575
other software, a platform before it.

00:19.955 --> 00:23.895
So the speed at which it permeated several, you know,

00:24.575 --> 00:27.055
multiple sectors was truly unprecedented.

00:27.595 --> 00:30.255
And so it reminded me of how personal computers,

00:30.255 --> 00:33.775
or the internet or smartphone revolutionized our world.

00:34.235 --> 00:37.175
But chat, GPT was moving at the even more

00:37.175 --> 00:38.695
breakneck, speed.

00:39.395 --> 00:43.095
And one of the, the sectors that felt the, the, this wave

00:43.115 --> 00:45.735
of disruption was, education.

00:45.995 --> 00:49.095
It was in the news, all over suddenly,

00:49.535 --> 00:52.895
students had at their fingertips, an AI tool capable

00:53.075 --> 00:55.295
of doing their homework for them, taking the test for them,

00:55.295 --> 00:57.175
and doing even better than they could have.

00:57.715 --> 01:00.775
So the challenge for educators was really significant,

01:01.355 --> 01:02.975
and there wasn't really a clear

01:03.075 --> 01:05.015
or straightforward way to deal with it.

01:05.235 --> 01:07.055
So this was a game changer.

01:07.755 --> 01:11.375
So as I observed this unfolding, I realized

01:11.375 --> 01:14.095
that we're witnessing a big shift in, in

01:14.155 --> 01:16.775
how this technology will transform education.

01:17.235 --> 01:19.175
So the implications were, were immense,

01:19.315 --> 01:21.615
and not just for teaching in the classroom,

01:21.715 --> 01:23.455
but for libraries as well.

01:24.005 --> 01:27.055
Libraries has always been at the forefront of,

01:27.055 --> 01:29.775
providing resources and support for learning and research.

01:30.155 --> 01:33.095
So with something like, chat, GPT

01:33.095 --> 01:36.895
or any of the generative AI tool, I saw a pressing need

01:37.155 --> 01:38.255
to really understand and,

01:38.255 --> 01:40.895
and address this, these new challenges.

01:41.555 --> 01:45.255
So in my role in leading AI and ai, literacy initiatives

01:45.255 --> 01:48.535
and my own university, I also observed a really,

01:48.935 --> 01:53.015
a pressing need for guidance in navigating this new AI

01:53.115 --> 01:54.295
hub driven landscape.

01:54.635 --> 01:55.775
And educators, students,

01:55.875 --> 01:58.415
and particularly my colleagues in the library, were,

01:58.415 --> 02:00.375
were trying to figure out how to deal with it.

02:00.795 --> 02:04.295
So, when the Department of Education released a report on,

02:04.355 --> 02:07.135
artificial intelligence and the future of teaching

02:07.135 --> 02:09.375
and learning, it was really a significant moment

02:09.755 --> 02:10.855
for the first time.

02:11.275 --> 02:15.055
We had what you might call an official guidance on how AI

02:15.585 --> 02:18.255
could and should inter intersect with education.

02:18.955 --> 02:23.095
So, it was a critical resource, but it's 71 pages long.

02:23.495 --> 02:25.735
I figure that most people don't have the time

02:25.835 --> 02:27.495
or the interest to really study it.

02:27.635 --> 02:30.535
So I, that's really the impetus of it.

02:30.735 --> 02:34.455
I feel a, a responsibility given my role at the university

02:34.455 --> 02:36.375
and in the library world to, to kind

02:36.375 --> 02:38.615
of analyze it a little bit and, and disseminate

02:38.615 --> 02:41.575
and interpret this information for my library colleagues.

02:41.735 --> 02:44.055
I want to analyze the key points of the reports

02:44.195 --> 02:46.615
and maybe share some recommendations,

02:46.955 --> 02:49.535
and really just to offer some kind of easy

02:49.635 --> 02:51.135
to understand guidance and,

02:51.315 --> 02:54.895
and roadmap for, you know, if you will, for, for,

02:55.235 --> 02:56.775
for my peers, basically.

02:57.035 --> 03:01.125
So the report was released in May, 2023, is,

03:01.525 --> 03:04.325
I would say it's a comprehensive exploration into

03:04.345 --> 03:06.845
how AI intersects with education,

03:07.305 --> 03:10.005
and it's trying to provide kind of nuance understanding

03:10.065 --> 03:13.645
of both the potential and the challenges that generative AI

03:13.645 --> 03:17.925
or AI in general brings to the education, landscape.

03:18.345 --> 03:21.245
So, essential theme of the report is the significance

03:21.345 --> 03:24.845
of transparency and understanding in AI

03:24.845 --> 03:26.325
systems used in education.

03:26.985 --> 03:30.845
Um, it emphasizes that AI should not be a black box.

03:31.305 --> 03:34.085
Uh, instead is decision making process should be clear

03:34.545 --> 03:36.925
to all the stakeholders, including teachers,

03:37.125 --> 03:38.245
students, librarians.

03:38.785 --> 03:40.525
So when we can understand

03:40.705 --> 03:44.045
or comprehend how decisions are made by these AI tools,

03:44.605 --> 03:46.445
I think it facilitates a more ethical

03:46.465 --> 03:48.565
and effective use of these technologies.

03:48.905 --> 03:51.725
So the understanding is not just technical,

03:51.825 --> 03:55.205
it really encompasses in, in, awareness of

03:55.205 --> 03:57.125
how AI might impact,

03:57.155 --> 03:59.325
educational practices and, and outcomes.

03:59.905 --> 04:02.045
So the, and this transparency, I think,

04:02.045 --> 04:03.405
is key in building trust.

04:03.745 --> 04:06.565
You know, ensuring AI tools are used effectively

04:06.585 --> 04:08.925
and ethically and empowering users

04:09.105 --> 04:10.485
to make informed decisions.

04:11.145 --> 04:13.485
Um, another critical point is the importance

04:13.485 --> 04:16.245
of human involvement in the use of AI in education.

04:16.635 --> 04:19.245
When this, technologies first appeared

04:19.545 --> 04:22.005
and when, when it was in the discourse of, you know,

04:22.005 --> 04:24.925
public discourse, a lot of people are very worried about,

04:25.105 --> 04:28.885
you know, jobs or people's roles being replaced by ai.

04:29.025 --> 04:30.085
So,

04:30.265 --> 04:33.605
so this report basically talks about AI's role in education

04:33.865 --> 04:38.205
can be envisioned as a supportive partner, taking on tasks

04:38.275 --> 04:41.125
that people should not, don't really need to do.

04:41.365 --> 04:44.365
A lot of the time consuming a routine tasks like

04:44.365 --> 04:47.085
administrative duties, like even grading

04:47.085 --> 04:50.085
or providing detailed analytics on student performance.

04:50.345 --> 04:51.805
Now, we may not be there yet,

04:51.985 --> 04:54.405
but it's very likely we'll get there at some point.

04:54.945 --> 04:57.245
So the advantage here is really significant.

04:57.305 --> 04:59.965
You know, by handling these tasks,

05:00.305 --> 05:03.765
AI can free up our time allowing us educators

05:03.765 --> 05:06.565
to invest our time and energy in more impact for, areas.

05:07.065 --> 05:08.205
So educators

05:08.205 --> 05:11.445
and focus on direct, student engagement,

05:11.515 --> 05:14.925
fostering a creative and, and dynamic learning environment,

05:14.945 --> 05:17.685
and nurturing the kind of interactive experiences that,

05:17.945 --> 05:20.725
you know, it's really at the heart of good education.

05:22.035 --> 05:23.215
But I think this kind

05:23.215 --> 05:25.215
of human-centric approach goes even deeper.

05:25.795 --> 05:27.775
Um, it really, really,

05:27.995 --> 05:30.055
the report highlights the inherent value

05:30.075 --> 05:32.695
of human qualities in education, judgment,

05:33.345 --> 05:34.895
creativity, and empathy.

05:34.995 --> 05:37.215
So that's something we need to pay attention to as well.

05:37.995 --> 05:40.735
Um, at the same time, it talks about, you know,

05:40.995 --> 05:42.735
AI shouldn't be making decisions for us.

05:42.765 --> 05:46.255
It's for us to, interpret

05:46.635 --> 05:50.095
or, apply the insights provided by ai.

05:50.275 --> 05:52.215
So that's another important part.

05:52.835 --> 05:55.615
At the same time, I think this technology can,

05:55.895 --> 05:57.445
democratize learning opportunities

05:57.585 --> 06:00.285
and can provide a lot of personalized learning

06:00.305 --> 06:03.125
and teaching so that the students with different abilities,

06:03.125 --> 06:06.285
different learning styles, can actually have that kind of,

06:06.545 --> 06:09.365
tailor made learning opportunities for them.

06:10.065 --> 06:12.045
But at the same time, we need to, you know,

06:13.025 --> 06:14.605
pay attention to the digital divide.

06:15.385 --> 06:18.325
Not every, not everyone may have the opportunity

06:18.425 --> 06:20.965
to use these technologies of how can we make sure

06:20.965 --> 06:22.805
that there's, equity

06:22.905 --> 06:26.165
or equality in this kind of, using the technology.

06:27.105 --> 06:30.525
Um, so, and then it talks a little bit more about more

06:30.525 --> 06:34.405
complex issues like, the ethical implications

06:34.465 --> 06:35.725
and equity concerns.

06:36.145 --> 06:39.605
Um, we know we've, we have heard in the news,

06:39.665 --> 06:42.605
and, you know, there are a lot of research on,

06:42.945 --> 06:47.165
the learning data that these, models use may not be

06:47.725 --> 06:49.125
accurate or fair,

06:49.225 --> 06:51.365
or even they may have a lot biases in them.

06:51.985 --> 06:54.285
Um, so that's something we need to pay attention to

06:54.705 --> 06:57.245
and also be respectful of data privacy.

06:57.825 --> 07:00.925
Uh, how, how do we use these models without

07:00.925 --> 07:02.845
compromising, you know, student privacy?

07:03.505 --> 07:06.445
So in summary, underlines of the importance of transparency,

07:07.095 --> 07:10.165
human environment, and the need for strategic,

07:10.265 --> 07:11.685
and I think ethical approach

07:11.745 --> 07:15.645
to integrating this AI in education, it's really important

07:15.645 --> 07:18.805
to recognize that we are at, at the crossroads of, you know,

07:18.865 --> 07:22.805
how technology intersects with our roles as educators and,

07:22.865 --> 07:23.885
and librarians.

07:24.425 --> 07:27.925
So AI literacy as outlined in the report, there's,

07:28.235 --> 07:31.205
it's not just about knowing how to use the tools.

07:31.755 --> 07:33.085
It's about understanding

07:33.265 --> 07:35.285
and, understanding the broader picture,

07:35.785 --> 07:37.165
how these technologies work.

07:37.345 --> 07:40.125
Yes, the applications in the educational context,

07:40.465 --> 07:42.125
the ethical implications,

07:42.785 --> 07:44.805
and being able to engage in a meaningful

07:44.805 --> 07:46.125
discussions about ai.

07:46.465 --> 07:50.525
So this kind of literacy is becoming increasingly crucial,

07:51.025 --> 07:53.445
not only for protecting educators

07:53.445 --> 07:55.005
and students from potential risks,

07:55.425 --> 07:58.645
but also for empowering them to really harness these,

07:58.985 --> 08:01.925
the immense power of these,

08:01.925 --> 08:04.885
technologies in a very innovative and transformative ways.

08:05.305 --> 08:09.045
So for librarians, this new age

08:09.065 --> 08:12.845
of AI literacy means going deeper into how these tools

08:13.365 --> 08:14.685
function and can be used.

08:14.955 --> 08:18.245
It's about, bringing them into the library,

08:18.245 --> 08:20.525
environment, using it for better, for example,

08:20.805 --> 08:23.405
resource management and enhancing user experience

08:23.545 --> 08:27.405
and supporting really complicated or complex research needs.

08:27.985 --> 08:30.725
And understanding the, like I mentioned

08:30.725 --> 08:34.285
before, the data privacy part, the, the inherent biases

08:34.745 --> 08:38.805
and, and, a lot of the ethical issues within it.

08:39.585 --> 08:43.325
Um, and that's another one, another thing that the,

08:43.625 --> 08:47.245
report talked about is the involvement in the design

08:47.305 --> 08:50.285
and imple implementation of this AI technology.

08:51.585 --> 08:54.285
And it talks about librarians as, you know, a lot

08:54.285 --> 08:55.445
of the web users

08:55.545 --> 08:57.605
of these educational tech, technologies.

08:57.705 --> 09:01.125
And we can provide really valuable insight into the

09:01.125 --> 09:02.645
design of these tools.

09:02.825 --> 09:06.765
So I think that's a opportunity for us to really think about

09:06.785 --> 09:10.085
how we can partner with both tech companies

09:10.105 --> 09:11.965
and even researchers,

09:12.365 --> 09:15.805
computer science researchers on campus, on how to,

09:15.955 --> 09:18.685
make these, technologies even more, you know,

09:18.785 --> 09:19.925
effective.

09:20.505 --> 09:24.565
Um, that's a, a ano another broader institutional context

09:24.825 --> 09:28.805
to consider as AI becomes more integrated into our,

09:28.985 --> 09:32.725
you know, educational landscape, libraries need

09:32.725 --> 09:35.885
to ensure the AI strategies align with the overall goals

09:36.065 --> 09:38.445
and initiatives of our institution.

09:38.825 --> 09:41.765
So that means involve, we, that involves collaboration

09:41.835 --> 09:46.725
with faculty in AI research, ensuring the AI system,

09:46.725 --> 09:48.925
compatible with our campus technologies.

09:49.625 --> 09:51.965
And, so it, it,

09:51.965 --> 09:54.205
it's another opportunity to bring people together.

09:54.705 --> 09:57.245
Um, and so I think those are really great things,

09:58.025 --> 10:00.645
but then there are also potential challenges.

10:01.185 --> 10:04.205
Uh, for example, there are psychological ones, resistance

10:04.205 --> 10:07.965
to change from us, from library staff,

10:08.175 --> 10:09.365
librarians to users.

10:09.515 --> 10:12.445
There's of, there's often a kind, a lack of understanding

10:12.545 --> 10:13.965
of these technologies.

10:14.145 --> 10:15.965
And, it's natural

10:16.225 --> 10:18.685
to fear the unknown, the uncertainties.

10:19.225 --> 10:21.885
And obviously there are a lot of news out there with both,

10:21.945 --> 10:24.285
you know, benefits and, and risk of the, of these things.

10:24.425 --> 10:27.685
So there's something to, for us to think about how

10:27.745 --> 10:29.485
to address these challenges.

10:29.985 --> 10:31.565
Uh, I think transparent

10:32.125 --> 10:33.405
communication is always a good thing.

10:33.405 --> 10:36.805
For example, myself, I like, I would like

10:36.805 --> 10:39.685
to encourage my staff to use it, to experiment with it.

10:40.265 --> 10:43.605
Um, so I always say I use it as well, you know, lead

10:43.605 --> 10:47.325
by example and say, I encourage you, use it, especially in,

10:47.585 --> 10:51.165
in low sticks, tough situation and just play with it.

10:51.265 --> 10:54.805
The, the best way to, to learn something is by doing a lot

10:54.805 --> 10:57.605
of the times, and then people can talk and share.

10:57.665 --> 11:00.125
And then that's that way we can com, you know,

11:00.125 --> 11:04.165
create communities of practice to, further come,

11:04.385 --> 11:06.405
our understanding of these technologies.

11:07.625 --> 11:10.725
So in wrapping up the future of AI in education, in,

11:11.105 --> 11:14.285
we have libraries have a, a unique opportunity

11:14.425 --> 11:18.165
to be at the forefront of this, change or evolution.

11:18.665 --> 11:22.565
And I think AI literacy is definitely the most critical

11:22.565 --> 11:27.205
first step so that people can have a general understanding

11:27.305 --> 11:29.605
of what these technologies are about.

11:30.185 --> 11:33.805
And then, then we can make the informed decisions on how,

11:33.805 --> 11:36.805
whether we want to use it, how we can use it

11:36.805 --> 11:38.925
and use it intelligently and responsibly.

11:40.095 --> 11:43.115
So some of the recommendations would be, you know, I,

11:43.735 --> 11:46.795
you know, echoing the AI literacy part, I think

11:46.795 --> 11:50.355
that is the most important thing, is to make sure that

11:51.715 --> 11:54.565
everyone has a, a basic knowledge of what that is.

11:54.945 --> 11:58.645
And I think it ex extends to everybody in society really,

11:58.645 --> 12:01.165
because it is really changing our world.

12:01.665 --> 12:04.885
And as leaders in the library sector, we, we need to

12:05.725 --> 12:09.765
champion this AI literacy, that, like I talked about,

12:09.765 --> 12:12.365
even both more than just using it, we need

12:12.365 --> 12:15.445
to understand the implications and ethics,

12:15.505 --> 12:18.525
and all of that, because with that, that means we, all

12:18.525 --> 12:22.245
of us can have a voice in influencing how the,

12:22.625 --> 12:25.125
the direction of this technology right

12:25.185 --> 12:26.445
now, like this report.

12:26.545 --> 12:30.565
And, many countries are setting their own regulations

12:30.585 --> 12:34.085
and policies, rules are being set right now,

12:34.705 --> 12:36.885
but they haven't been really been set yet.

12:37.475 --> 12:40.445
Once the rules are set, extremely difficult to change.

12:40.785 --> 12:45.165
So this is the time for us to give our feedback to,

12:45.505 --> 12:48.685
to try to influence, you know, the, the direction

12:48.885 --> 12:51.885
of this technology will be good for society, good

12:51.885 --> 12:54.485
for education, good for information, for example.

12:55.105 --> 12:58.925
Um, so, and I think AI literacy is not just about just,

12:59.245 --> 13:00.925
offering training sessions and workshops.

13:01.145 --> 13:02.645
Yes, we need those, but I think those

13:02.645 --> 13:03.685
are just starting points.

13:04.385 --> 13:07.645
Um, psychologically we need to make sure, you know,

13:07.645 --> 13:10.965
help people understand that, the importance of this.

13:11.465 --> 13:15.685
Um, and yes, it's not perfect. It may never be perfect.

13:15.825 --> 13:17.125
The internet is not perfect.

13:17.265 --> 13:19.605
We still use it in the, hopefully in the right way.

13:19.985 --> 13:21.485
And I think this is the same thing.

13:21.905 --> 13:23.725
The more we under we understand it,

13:23.785 --> 13:25.445
the more we use it correctly,

13:25.985 --> 13:30.605
and that can alleviate some of the, the, the fear of it.

13:31.145 --> 13:34.445
So I think definitely as leaders, they,

13:34.545 --> 13:37.005
we should encourage people to experiment

13:37.065 --> 13:38.445
and also lead by example.

13:38.585 --> 13:39.685
We should experiment with it

13:39.685 --> 13:40.805
and learn a little bit more about that.

13:41.545 --> 13:44.685
Um, and moving, you know,

13:45.395 --> 13:47.325
more than just AI literacy, I,

13:47.445 --> 13:49.765
I talked about getting involved in the design

13:49.945 --> 13:51.885
and development of these tools,

13:52.125 --> 13:55.285
I think this is another great opportunity to reach out,

13:55.285 --> 13:57.885
proactively, reach out to either companies

13:58.065 --> 13:59.925
or, you know, partner up with them, for example,

14:00.425 --> 14:03.765
or reach out to our researchers and students even

14:04.145 --> 14:06.645
and see how we can work together on this.

14:08.105 --> 14:12.005
And during all this, another recommendation I'll give is to,

14:12.265 --> 14:15.405
hey, let's, develop some guidelines,

14:15.715 --> 14:19.885
some guidance in, how we should be u as a profession,

14:20.065 --> 14:23.685
as a sector, how we can use this, technology

14:23.945 --> 14:26.765
and hopefully, have a more of a standard,

14:26.945 --> 14:28.525
not necessarily regulations or policy,

14:28.665 --> 14:30.685
but standards in how we, let's say,

14:30.745 --> 14:32.805
how do we evaluate AI tools?

14:32.955 --> 14:37.005
What are some of the factors in, purchasing using some

14:37.005 --> 14:39.125
of these AI technologies in the libraries

14:39.125 --> 14:43.405
or in the, in the university Right now I'm chairing a,

14:43.705 --> 14:47.325
academic honesty task force in, at my university.

14:47.385 --> 14:50.765
And we're looking at AI detection software. Now, there

14:50.765 --> 14:53.085
Are, you know, a lot of studies saying whether they,

14:53.085 --> 14:54.325
they are useful or not.

14:54.825 --> 14:58.925
Um, so we are trying u try to use this as a exercise

14:59.025 --> 15:02.245
to develop a set of criteria when we look at AI technologies

15:02.545 --> 15:05.885
or software, how we can decide, you know, what are some

15:05.885 --> 15:07.565
of the important things we should look into

15:07.585 --> 15:09.725
before we adopt them or use them.

15:10.165 --> 15:11.565
I think that's the same thing for a lot

15:11.565 --> 15:13.525
of the library, out there.

15:13.825 --> 15:17.125
And, and software designed for libraries.

15:17.145 --> 15:20.725
How can we make sure that the, these software aligns

15:20.725 --> 15:23.245
with our, our, mission and vision.

15:23.665 --> 15:25.685
So those are some of the recommendations.

15:25.785 --> 15:27.525
I'm sure there will be a lot more coming out.

Faculti AI can make mistakes

Research

Policy

Analysis