Can AI Help Make Higher Education More Equitable?

1
00:00:06,779 --> 00:00:09,989
KiKi L'Italien: Welcome to
Association Chat your online

2
00:00:09,989 --> 00:00:13,079
discussion where we warm
ourselves by the virtual fire

3
00:00:13,079 --> 00:00:15,809
with topics of the day,
welcoming thought leaders and

4
00:00:15,809 --> 00:00:19,259
trailblazers alike to join up in
this online home for the

5
00:00:19,259 --> 00:00:23,249
community. I'm the host of
Association Chat KiKi L'Italien.

6
00:00:23,669 --> 00:00:29,549
And could AI revolutionize
student success? Could it foster

7
00:00:29,759 --> 00:00:35,849
inclusivity? could bridge gaps
and educational equity? We hear

8
00:00:35,849 --> 00:00:39,449
a lot about plagiarism. We hear
a lot about the scary stuff. But

9
00:00:39,449 --> 00:00:42,809
what about the good stuff? Is it
possible that AI could be our

10
00:00:42,809 --> 00:00:46,349
friend? Let's talk about it. I
have some special guests with me

11
00:00:46,349 --> 00:00:50,339
here today to help answer some
of these questions. And so

12
00:00:50,429 --> 00:00:54,149
helping me to answer that
question, Charles Ansel Michael,

13
00:00:54,149 --> 00:00:58,649
Bastien, and we have Audrey
Ellis with us here today. Hi,

14
00:00:58,649 --> 00:01:03,149
welcome, everyone. Hi, thanks
for having us. All right, thank

15
00:01:03,149 --> 00:01:05,819
you for being here. I want to
dive right into it so that we

16
00:01:05,819 --> 00:01:09,869
have a chance to get into this
discussion. You know, ay ay ay

17
00:01:09,869 --> 00:01:15,569
ay ay ay. It's every where you
can't avoid it and the news and

18
00:01:15,599 --> 00:01:20,519
the headlines. And one place
where you start to hear about

19
00:01:20,519 --> 00:01:24,299
complications almost
exclusively, is in higher ed,

20
00:01:24,299 --> 00:01:27,959
people are talking about not
trusting the work that's being

21
00:01:27,959 --> 00:01:31,709
done, they're really dwelling on
that in the conversations that

22
00:01:31,739 --> 00:01:36,959
I'm a party to on a regular
basis. But I think what's really

23
00:01:36,959 --> 00:01:40,109
interesting is that the work
that you all have been doing

24
00:01:40,109 --> 00:01:43,889
together is looking at how it
might actually be something that

25
00:01:43,889 --> 00:01:48,119
is a positive. And I thought,
you know, let's explore it. So

26
00:01:48,119 --> 00:01:52,139
I'm gonna start with you,
Charles, if you don't mind, I

27
00:01:52,139 --> 00:01:58,019
want to ask you about how you
see generative AI impacting

28
00:01:58,049 --> 00:02:02,999
policy and advocacy in higher
ed, especially in terms of

29
00:02:02,999 --> 00:02:05,279
equity and completion rates.

30
00:02:06,330 --> 00:02:11,340
Charles Ansell: Yeah, thanks for
that question. And, you know,

31
00:02:11,340 --> 00:02:13,770
I'll, I'll try to be brief, but
it's actually really hard,

32
00:02:13,770 --> 00:02:16,020
because there's a real lot to
unpack there. Because I think

33
00:02:16,020 --> 00:02:18,930
that very often, when we think
about AI, we think immediately,

34
00:02:19,440 --> 00:02:23,190
you know, oh, wow, this is such
a powerful, new way that we can

35
00:02:23,220 --> 00:02:27,090
do business and get work done.
But unless we really think

36
00:02:27,120 --> 00:02:30,060
problem first about what this
sort of technology is supposed

37
00:02:30,060 --> 00:02:32,940
to solve, we're never going to
use it right. And so, you know,

38
00:02:32,940 --> 00:02:36,210
in that question, you know, how
are we going to use AI to help

39
00:02:36,210 --> 00:02:39,750
with policy and advocacy,
particularly for student

40
00:02:40,110 --> 00:02:44,070
success? And for equitable
outcomes? In post secondary, I

41
00:02:44,070 --> 00:02:46,830
think that we need to like, zoom
back a little bit, right, we

42
00:02:46,830 --> 00:02:49,620
have to start with the
acknowledgement of where higher

43
00:02:49,650 --> 00:02:53,280
education is at where the sector
is at in terms of equity and in

44
00:02:53,280 --> 00:02:56,460
terms of completion rates. And
then what are the policy and

45
00:02:56,460 --> 00:03:00,420
advocacy levers that we're
supposed to, you know, use to

46
00:03:00,420 --> 00:03:03,870
ameliorate the situation? And
really only then how, how can

47
00:03:03,870 --> 00:03:06,420
generative AI help at least
that's how we see that Complete

48
00:03:06,420 --> 00:03:09,030
College America. So just really
quick, running that through.

49
00:03:10,290 --> 00:03:13,560
Higher Ed, in my opinion, is not
really in a great place, right.

50
00:03:13,590 --> 00:03:16,080
And I don't think a lot of
people really know this outside

51
00:03:16,080 --> 00:03:19,320
of the sector. And it does,
indeed need transformation. So

52
00:03:19,590 --> 00:03:22,200
we'll start with some basic
facts, right. So for full time

53
00:03:22,200 --> 00:03:24,810
students attending public four
year colleges and universities,

54
00:03:25,080 --> 00:03:28,230
the number of full time students
who graduate in four years from

55
00:03:28,230 --> 00:03:31,770
four year colleges is 40%. And
that's a number that's buoyed by

56
00:03:31,770 --> 00:03:34,950
well resourced public flagship
universities. For two year

57
00:03:34,950 --> 00:03:37,530
colleges, the two year
graduation rate for full time

58
00:03:37,530 --> 00:03:41,190
students is about 18%. So
that's, you know, less than one

59
00:03:41,190 --> 00:03:44,070
in five. And it's roughly the
same for part time students at

60
00:03:44,070 --> 00:03:47,430
community colleges who graduate
in six years. So these numbers

61
00:03:47,430 --> 00:03:50,610
are lower for students who come
from less resource communities

62
00:03:50,670 --> 00:03:53,220
for students of color. And it's
especially true because

63
00:03:53,910 --> 00:03:57,600
colleges, you know, this, you
know, it's a struggle to work

64
00:03:57,600 --> 00:04:00,120
with working learners, the most
students who work more than 20

65
00:04:00,120 --> 00:04:04,650
hours per week, rarely graduate,
graduate, and when they do, you

66
00:04:04,650 --> 00:04:09,690
know, they, they get saddled
with debt. So the students who

67
00:04:09,690 --> 00:04:13,200
do go to college, the students
who most need to be the

68
00:04:13,200 --> 00:04:15,660
beneficiaries of that
transformative social mobility

69
00:04:15,660 --> 00:04:18,000
that that college can offer,
these are the students who are

70
00:04:18,150 --> 00:04:20,610
most likely not to graduate and
the way the current system is

71
00:04:20,610 --> 00:04:24,900
set up. So you know, this is
happening, of course, at a time

72
00:04:24,900 --> 00:04:28,500
when a college degree has never
had a higher premium. So not

73
00:04:28,500 --> 00:04:31,560
only do most jobs require a
degree, but almost all good

74
00:04:31,560 --> 00:04:35,400
careers do. And by 2031
Georgetown estimates, I want to

75
00:04:35,400 --> 00:04:39,630
say it's 72%, or about three and
four US residents need to have a

76
00:04:39,630 --> 00:04:43,140
post secondary credential of
some sort to meet our workforce

77
00:04:43,140 --> 00:04:46,440
needs optimally. And yet in the
last few decades, you know, what

78
00:04:46,440 --> 00:04:50,730
we've set up is, you know, a
financial system, where the Pell

79
00:04:50,730 --> 00:04:53,190
Grant only has a third of its
original purchasing power in

80
00:04:53,190 --> 00:04:55,710
tuition and student loan debt
has skyrocketed. So in other

81
00:04:55,710 --> 00:04:58,320
words, hold the same Yeah, hold

82
00:04:58,350 --> 00:05:01,500
KiKi L'Italien: up like just
saying that They think, oh my

83
00:05:01,500 --> 00:05:04,680
gosh, I had no idea. First of
all, I had no idea that that was

84
00:05:04,680 --> 00:05:08,040
the case. But that's some.
That's scary. That's really I

85
00:05:08,040 --> 00:05:10,650
mean, that's something that
people really rely on to be able

86
00:05:10,650 --> 00:05:14,670
to go. And just sorry, I was
just distracted by that was

87
00:05:15,630 --> 00:05:19,860
pretty crazy statistic added to
all the other. Well,

88
00:05:21,209 --> 00:05:23,639
Charles Ansell: yeah, and I
think that the big problem,

89
00:05:23,639 --> 00:05:26,279
right is I mean, that's a
problem in itself. But at the

90
00:05:26,279 --> 00:05:29,639
same time, we're also putting
forward this message that's very

91
00:05:29,639 --> 00:05:33,929
accurate and true, that higher
education, you need a credential

92
00:05:33,929 --> 00:05:38,219
now more than ever. So the only
thing worse than being able to

93
00:05:39,509 --> 00:05:42,929
struggle to afford it while
attending, which is something

94
00:05:42,929 --> 00:05:46,229
that, you know, causes student
attrition in a very large

95
00:05:46,229 --> 00:05:49,109
amount, but the only thing worse
is being shut out of the economy

96
00:05:49,109 --> 00:05:51,179
because you don't have the
degree or certificate to be able

97
00:05:51,179 --> 00:05:53,759
to participate in the first
place. And so like, this is

98
00:05:53,759 --> 00:05:57,779
where we get to the those policy
and advocacy levers for

99
00:05:57,779 --> 00:06:01,589
equitable student success, you
know, and, and I, I know, this

100
00:06:01,589 --> 00:06:05,609
is a long lineup, but I am going
to bring it to AI. But I think

101
00:06:05,759 --> 00:06:09,269
you can boil that down to three
things, right? I think first, we

102
00:06:09,269 --> 00:06:11,429
need to get out of our own way.
And we need to get rid of bad

103
00:06:11,429 --> 00:06:14,159
practices and bureaucracy that
prevents students from having

104
00:06:14,159 --> 00:06:17,999
purpose in their major structure
to their academic experience,

105
00:06:17,999 --> 00:06:20,879
momentum to timely completion,
and really having an enough

106
00:06:20,879 --> 00:06:25,109
support to complete college on
time. And, you know, we'll we'll

107
00:06:25,109 --> 00:06:28,079
hear from, you know, a college
president who's doing exactly

108
00:06:28,079 --> 00:06:34,139
those things. Second, we need
targeted interventions to

109
00:06:34,139 --> 00:06:36,959
support students equitably. So
not every student needs the same

110
00:06:36,959 --> 00:06:39,419
thing, right, like, that's the
whole premise. Some students may

111
00:06:39,419 --> 00:06:42,839
need relatively shallow support
some deep, some need different

112
00:06:42,839 --> 00:06:44,969
individualized attention,
depending on what their needs

113
00:06:44,969 --> 00:06:49,799
are. And then the third is we
can't just efficiency our way

114
00:06:49,799 --> 00:06:53,579
out of this, right. There's been
a lot of studies that have gone

115
00:06:53,579 --> 00:06:57,149
on in high schools to show that,
you know, you add a lot of high

116
00:06:57,149 --> 00:07:01,439
touch interventions in a under
resourced school, you can see

117
00:07:01,439 --> 00:07:04,919
incremental gains in high school
completion rates, but it's never

118
00:07:04,919 --> 00:07:07,589
going to be as strong as that,
well, resource high schools,

119
00:07:07,589 --> 00:07:10,139
it's the same thing. In
colleges, students success

120
00:07:10,139 --> 00:07:12,329
interventions can create
incremental change. We've seen

121
00:07:12,329 --> 00:07:14,459
it in the last decade in the
college completion movement, but

122
00:07:14,459 --> 00:07:17,669
we'll hit a ceiling soon. And we
have to fund higher education

123
00:07:17,669 --> 00:07:20,459
like we care about it. And so
that's where generative AI comes

124
00:07:20,459 --> 00:07:22,679
in, not just on that funding
piece, of course, but

125
00:07:22,679 --> 00:07:25,049
essentially on those three
areas, right. So like, first, we

126
00:07:25,049 --> 00:07:28,949
have to have a stance towards
generative AI as something that

127
00:07:28,949 --> 00:07:32,399
is a power for good, not just a
power for plagiarism, like to

128
00:07:32,399 --> 00:07:37,049
your point about all the
controversy in AI around higher

129
00:07:37,049 --> 00:07:40,229
education, you know, it's been
so much of the attention has

130
00:07:40,229 --> 00:07:43,319
been on this negative side of
things, but we have to dive

131
00:07:43,919 --> 00:07:47,009
headlong into its promises. The
second is that we have to use it

132
00:07:47,009 --> 00:07:49,679
to solve the problems that we
see in student success, you

133
00:07:49,679 --> 00:07:52,649
know, talking about having
equitable interventions, the

134
00:07:52,649 --> 00:07:54,749
need to innovate and schedules
and credit for prior learning

135
00:07:54,749 --> 00:07:56,969
for part time students need to
get all students on semester by

136
00:07:56,969 --> 00:08:00,329
semester academic plans, and
really liberate advising

137
00:08:00,329 --> 00:08:03,599
functions, you know, where
counselors meet with students,

138
00:08:03,659 --> 00:08:06,449
so much of that is manual
transactional conversations on

139
00:08:06,449 --> 00:08:09,449
like, well, what course at what
time, and there's not enough

140
00:08:09,449 --> 00:08:12,419
time to have those rich
conversations where advisors are

141
00:08:12,419 --> 00:08:15,839
able to give actual advice. And
finally, and I would say, most

142
00:08:15,839 --> 00:08:19,499
importantly, we need to make
sure generative AI isn't the

143
00:08:19,499 --> 00:08:22,769
sole purview of rich colleges,
right, the MIT's the Stanford's

144
00:08:22,769 --> 00:08:25,949
the pens, the colleges, whose
presidents are important enough

145
00:08:25,949 --> 00:08:28,919
to get yelled at by Congress, we
need to make sure that community

146
00:08:28,919 --> 00:08:32,309
colleges, rural colleges, HBCUs,
PDAs, and other minority serving

147
00:08:32,309 --> 00:08:35,939
institutions aren't working off
hardware infrastructure routers

148
00:08:35,939 --> 00:08:38,729
and switches from the 90s that
are woefully inadequate to the

149
00:08:38,729 --> 00:08:41,129
horsepower required to optimize
generative AI for student

150
00:08:41,129 --> 00:08:46,139
success. So, you know, I know
that was a bit of a long answer,

151
00:08:46,139 --> 00:08:49,139
but I think it's good to like at
the outset, just talk about, you

152
00:08:49,139 --> 00:08:51,119
know, the problems that we're
trying to solve, because that's

153
00:08:51,119 --> 00:08:53,309
the only way we switch
orientation or generative AI.

154
00:08:53,730 --> 00:08:55,560
KiKi L'Italien: I mean, I think
it was the perfect setup,

155
00:08:55,560 --> 00:08:58,680
because I think, you know, you
provided a really great

156
00:08:58,680 --> 00:09:00,870
background and foundation for
why we're having this

157
00:09:00,870 --> 00:09:05,580
conversation in the first place.
So thank you for doing that. You

158
00:09:05,580 --> 00:09:09,810
know, you talked about the
promise of AI and the problems

159
00:09:09,810 --> 00:09:13,470
with success for many students
why they can't be successful or

160
00:09:13,470 --> 00:09:17,550
find success and higher ed, and
then not leaving all of this to

161
00:09:17,550 --> 00:09:21,630
the rich colleges that can
afford to be left alone when

162
00:09:21,630 --> 00:09:25,050
they're when they're
experimenting, and trying some

163
00:09:25,050 --> 00:09:27,780
new things. It's fair, it sounds
scary, let's say trying new

164
00:09:27,780 --> 00:09:31,440
things. I want to actually bring
this over to someone who

165
00:09:31,440 --> 00:09:34,860
probably has very interesting
perspective to share on it as

166
00:09:34,860 --> 00:09:40,770
well. And so so let's talking to
you president Michael Basta. And

167
00:09:40,770 --> 00:09:44,430
let's talk about this, about
from the perspective of a

168
00:09:44,430 --> 00:09:48,210
community college president.
What are the practical

169
00:09:48,210 --> 00:09:52,020
challenges and the opportunities
of integrating AI into the

170
00:09:52,020 --> 00:09:54,840
curriculum? What are the
challenges that you're seeing?

171
00:09:55,350 --> 00:09:58,590
Michael Baston: Well, first and
foremost, is to desensitize the

172
00:09:58,590 --> 00:10:02,970
faculty who have to invigorate
and integrate AI into the

173
00:10:02,970 --> 00:10:07,230
curriculum, because as Charles
mentioned, the idea that some

174
00:10:07,230 --> 00:10:12,930
think about plagiarism, and that
sort of becomes the color by

175
00:10:12,930 --> 00:10:16,830
which they look at the
conversation. But the fact is,

176
00:10:16,860 --> 00:10:20,730
AI is not going anywhere, the
fact that students will actually

177
00:10:20,730 --> 00:10:25,050
have access to the technology.
And so it is not feasible to

178
00:10:25,050 --> 00:10:27,780
just simply say, Well, we're
going to put in our syllabus,

179
00:10:27,780 --> 00:10:32,070
you can't use any AI, because
they're going to use grammerly,

180
00:10:32,070 --> 00:10:35,820
which is AI, they're going to
use all of the other kinds of

181
00:10:35,820 --> 00:10:38,520
technology and tools that are
going to be helpful, and that

182
00:10:38,550 --> 00:10:41,730
actually, the faculty do want
them to actually have that

183
00:10:41,730 --> 00:10:44,340
support. So I think it's
important for educational

184
00:10:44,340 --> 00:10:47,730
institutions to resource and
provide professional development

185
00:10:47,730 --> 00:10:52,860
for faculty, so that they can
best integrate the opportunities

186
00:10:52,860 --> 00:10:56,400
with AI. And particularly at the
community college, when we focus

187
00:10:56,400 --> 00:10:59,940
heavily on thinking about the
career exploration of our

188
00:10:59,940 --> 00:11:04,440
students and the kinds of career
aspirations that they have, are

189
00:11:04,440 --> 00:11:07,440
thinking about preparing those
students to go into the world of

190
00:11:07,440 --> 00:11:11,880
work for those who will go right
into it, you want them to be

191
00:11:11,880 --> 00:11:16,770
prepared for the expectations of
employers who when they get on

192
00:11:16,800 --> 00:11:19,770
in the workplace are going to
expect them to work smarter and

193
00:11:19,770 --> 00:11:24,060
not harder to be producing at a
higher level. And so you know,

194
00:11:24,090 --> 00:11:29,370
even the basic premise of
generative at the prompt, sort

195
00:11:29,370 --> 00:11:33,000
of the ability to get students
to understand prop

196
00:11:33,030 --> 00:11:36,570
characteristics and, and how to
use that effectively, will

197
00:11:36,570 --> 00:11:40,500
actually separate those out in
the workplace who don't have

198
00:11:40,500 --> 00:11:44,010
that orientation and background
and skill set. And so, you know,

199
00:11:44,010 --> 00:11:47,670
we do a disservice to our
students, if we are not arming

200
00:11:47,670 --> 00:11:50,940
them with everything they need
to effectively navigate the

201
00:11:50,940 --> 00:11:54,720
realities of a changing
landscape, whether it's in

202
00:11:54,720 --> 00:11:58,320
education, or in the world of
work. And also listen, we have

203
00:11:58,320 --> 00:12:02,250
turned it in. So you can get
your students to actually look

204
00:12:02,250 --> 00:12:06,000
at, you know, the content of
Turnitin tells you not only

205
00:12:06,000 --> 00:12:09,810
about plagiarism, but it also
tells you about AI generated

206
00:12:09,960 --> 00:12:13,530
content. So it is not that
faculty members are completely

207
00:12:13,530 --> 00:12:19,620
devoid of any kind of ways to
ensure originality and

208
00:12:19,620 --> 00:12:23,730
authenticity of student work, it
is to show students you know,

209
00:12:23,760 --> 00:12:28,200
use this for inspiration, use
this for feedback, use this to

210
00:12:28,200 --> 00:12:31,860
help you sort of, you know, I
can't be with you 24/7, you're

211
00:12:31,860 --> 00:12:34,230
at three o'clock in the morning,
and you're doing your essay, I

212
00:12:34,230 --> 00:12:38,250
can't be there with you to do
that essay. But you can use this

213
00:12:38,250 --> 00:12:41,640
technology to help you spark
ideas, but at the end of the

214
00:12:41,640 --> 00:12:45,240
day, you're gonna get credit for
your originality, you're gonna

215
00:12:45,240 --> 00:12:47,850
get credit for your
authenticity, and we shouldn't

216
00:12:47,880 --> 00:12:51,180
act like they're not going to
have access to what they already

217
00:12:51,180 --> 00:12:52,980
have access to. Right.

218
00:12:53,010 --> 00:12:55,620
KiKi L'Italien: I mean, I
remember this back when we

219
00:12:55,620 --> 00:12:58,920
started looking at things like
social media and, and in the

220
00:12:58,920 --> 00:13:04,080
workforce, what we saw was this
desire to control by pretending

221
00:13:04,080 --> 00:13:07,800
that if we suppress this, if we
said this technology is not

222
00:13:07,800 --> 00:13:12,120
allowed to be used in the air,
what happened, you know,

223
00:13:12,120 --> 00:13:15,630
everybody's over here doing
that, you know, like, you

224
00:13:15,630 --> 00:13:19,440
couldn't suppress it, you
couldn't outline and when you're

225
00:13:19,440 --> 00:13:21,690
talking about something like
generative AI, I think the

226
00:13:21,690 --> 00:13:26,220
challenge is that, to your
point, people are focusing on on

227
00:13:26,220 --> 00:13:29,520
weaponizing it in a way that it
can be used for things like

228
00:13:29,520 --> 00:13:32,700
creating malicious code, or
breaking the, you know, breaking

229
00:13:33,000 --> 00:13:37,320
the rules by like plagiarizing
and things like that. But you

230
00:13:37,320 --> 00:13:41,100
could also be weaponizing it on
the other side by not allowing

231
00:13:41,100 --> 00:13:44,730
people to learn how to
effectively use it in day to day

232
00:13:44,730 --> 00:13:48,690
life, when it's such an
effective tool. So

233
00:13:50,460 --> 00:13:52,200
Charles Ansell: can I just add
something on that point?

234
00:13:52,410 --> 00:13:53,040
KiKi L'Italien: Yeah,

235
00:13:53,160 --> 00:13:55,620
Charles Ansell: I think, you
know, what seems particularly

236
00:13:55,620 --> 00:13:58,140
crazy to me, you know, the
social media thing is a really

237
00:13:58,140 --> 00:14:02,100
good point. There's a I forget,
there's a random philosopher,

238
00:14:02,490 --> 00:14:04,770
the Salk random, I just don't
know the person's name. But I

239
00:14:04,770 --> 00:14:07,320
heard it on another podcast who
said, when you invent

240
00:14:07,320 --> 00:14:10,020
electricity, you invent
electrocution, when you invent?

241
00:14:10,410 --> 00:14:12,600
You know, anything good, you
have not the bad thing, too.

242
00:14:12,840 --> 00:14:16,260
And, and sometimes we don't
acknowledge that we invent the

243
00:14:16,260 --> 00:14:18,510
bad thing. And in this case, I
think it's the opposite. It's

244
00:14:18,510 --> 00:14:21,540
like we just ran to Oh, no, we
invented electrocution, without

245
00:14:21,570 --> 00:14:23,940
even bothering to see what sort
of power we had in our hands

246
00:14:23,940 --> 00:14:26,820
that we could harness. And and I
think, you know, to President

247
00:14:26,820 --> 00:14:30,780
bassins point, I think it's very
telling that certain things

248
00:14:30,780 --> 00:14:35,610
that, you know, students with
way more resources have been

249
00:14:35,610 --> 00:14:39,270
able to do forever, like pay
people to write essays for them.

250
00:14:39,870 --> 00:14:43,530
Now that everybody can do it,
it's democratized. Oh, no, now

251
00:14:43,530 --> 00:14:46,110
it's bad. And I think that's a
big talent, what our priorities

252
00:14:46,110 --> 00:14:46,350
are.

253
00:14:48,390 --> 00:14:50,010
KiKi L'Italien: I think that's
really I mean, it's a great

254
00:14:50,010 --> 00:14:53,340
segue into the next question
that I had because I want to

255
00:14:53,610 --> 00:14:58,770
reach over to Audrey and give
Audrey a chance to speak to this

256
00:14:59,310 --> 00:15:02,610
great play. But the work that
you've been doing, you've been

257
00:15:02,610 --> 00:15:05,670
thinking about this quite a bit,
you've been thinking about this,

258
00:15:05,700 --> 00:15:10,890
this topic and the issue at
hand, which is, you know, what

259
00:15:10,890 --> 00:15:16,230
are we doing to actually take
the power of some of this

260
00:15:16,260 --> 00:15:20,550
generative AI? And some of the
advancements in AI to use it for

261
00:15:20,550 --> 00:15:24,210
good, right? And how can we, how
can we allow higher ed to

262
00:15:24,210 --> 00:15:28,020
actually, you know, make the
make use of this in a better

263
00:15:28,020 --> 00:15:32,100
way. So you have a playbook, you
outline in this playbook, you

264
00:15:32,100 --> 00:15:35,340
outline strategies for
leveraging AI to increase

265
00:15:35,340 --> 00:15:41,040
college completion in equity.
And so, I would love if you

266
00:15:41,040 --> 00:15:45,120
could just elaborate on one or
two of the key strategies that

267
00:15:45,120 --> 00:15:49,530
would be particularly relevant,
or association professionals who

268
00:15:49,530 --> 00:15:52,080
are listening to this now, or
maybe they're listening to it

269
00:15:52,080 --> 00:15:55,410
later, for them to pay attention
to on this.

270
00:15:56,250 --> 00:15:59,250
Audrey Ellis: Sure. And it is a
fun thing to think about right

271
00:15:59,250 --> 00:16:03,060
to treat every problem as an
opportunity. And so we really

272
00:16:03,060 --> 00:16:06,450
did try to take that mindset and
kind of flip the problem on its

273
00:16:06,450 --> 00:16:10,380
head. While we were working on
building out this playbook. So

274
00:16:11,010 --> 00:16:13,560
grateful for Charles and
Complete College America, who

275
00:16:13,560 --> 00:16:16,890
really gave us the platform to
facilitate this discussion,

276
00:16:16,890 --> 00:16:20,250
which was built by so many
thought leaders in the higher ed

277
00:16:20,250 --> 00:16:24,570
space. And so I'm going to go a
little bit more broad in terms

278
00:16:24,570 --> 00:16:28,110
of the practices or strategies
that I suggest, because we're

279
00:16:28,110 --> 00:16:30,960
still in the early stages. And I
don't think it makes a lot of

280
00:16:30,960 --> 00:16:35,160
sense to get too tactical right
now. But But kind of building

281
00:16:35,160 --> 00:16:39,600
off of Charles and President
bassins points, the first

282
00:16:39,630 --> 00:16:43,020
suggestion that I have is to
think about how you can create

283
00:16:43,020 --> 00:16:47,820
learning opportunities for your
higher education partners. In

284
00:16:47,820 --> 00:16:51,930
whatever industry you're in. I
know that today we're speaking

285
00:16:51,930 --> 00:16:56,070
to industries, really from
across the the workforce. And so

286
00:16:56,490 --> 00:16:59,430
someone actually I saw from
Facebook said, higher ed should

287
00:16:59,430 --> 00:17:02,460
partner more with industry
partners. And I couldn't agree

288
00:17:02,460 --> 00:17:05,850
more. And I think that's what's
so exciting about this podcast.

289
00:17:06,660 --> 00:17:10,980
You know, I spoke with a college
Provost recently who was working

290
00:17:10,980 --> 00:17:14,520
with their welding corporate
partner to go learn all about

291
00:17:14,520 --> 00:17:20,130
how they are using AI in their
welding work right now, and to

292
00:17:20,130 --> 00:17:24,450
then revamp all of their
curriculum to have, you know, AI

293
00:17:24,450 --> 00:17:27,150
embedded so that they're
preparing their welding students

294
00:17:27,150 --> 00:17:30,720
to graduate and with those
competencies instead of having

295
00:17:30,720 --> 00:17:35,220
to get them out in the field.
And as much as higher ed is a

296
00:17:35,220 --> 00:17:39,120
source of learning and
knowledge. We're on campus,

297
00:17:39,120 --> 00:17:42,510
right, like we're working and
doing research and working with

298
00:17:42,510 --> 00:17:46,710
our students and community
colleges, especially, are a

299
00:17:46,710 --> 00:17:51,450
really amazing intersection of
the workforce space and the

300
00:17:51,450 --> 00:17:55,470
campus environment. And we with
such a rapidly changing

301
00:17:55,470 --> 00:18:00,840
technology, we can't possibly
know exactly how it's impacting

302
00:18:00,840 --> 00:18:03,930
your industry and less, the
lines of communication are wide

303
00:18:03,930 --> 00:18:08,610
open. And I think that because
this is also a cost prohibitive

304
00:18:08,640 --> 00:18:12,480
technology, right now, there's a
lot of upfront costs with

305
00:18:12,810 --> 00:18:16,140
building your own AI
infrastructure, getting your

306
00:18:16,140 --> 00:18:19,320
data even ready to the point
where you could do that. The

307
00:18:19,320 --> 00:18:23,760
other I guess this is kind of a
sub strategy, one recommendation

308
00:18:23,790 --> 00:18:27,210
would be to work with your
industry or your higher

309
00:18:27,210 --> 00:18:31,560
education partners to support
them in their learning, even if

310
00:18:31,560 --> 00:18:35,280
that looks financial, because
it's an investment in your

311
00:18:35,280 --> 00:18:39,780
future workforce to ensure that
they're at an institution that

312
00:18:39,780 --> 00:18:42,840
is preparing them for the
expectations you're going to

313
00:18:42,840 --> 00:18:46,830
have for your future employees.
And the second piece that I'll

314
00:18:46,830 --> 00:18:51,660
touch on, is around thinking
about things in phases. We kind

315
00:18:51,660 --> 00:18:55,710
of talked about this. Charles
has a really interesting analogy

316
00:18:55,950 --> 00:19:00,180
of electrocution and
electricity. But we do tend to

317
00:19:00,180 --> 00:19:03,210
think of the ends of the
spectrum, right? Like what's the

318
00:19:03,210 --> 00:19:09,000
most AI futuristic vision
possible? And what's right now,

319
00:19:09,540 --> 00:19:12,960
but there really, in my mind are
kind of three steps that we need

320
00:19:12,960 --> 00:19:15,750
to be thinking about, which
might look different for every

321
00:19:15,750 --> 00:19:19,140
institution, depending on if
you're an association, or if

322
00:19:19,140 --> 00:19:23,700
you're an institution of higher
education, which is right now,

323
00:19:24,090 --> 00:19:28,200
the middle term, and the future.
And I think that we can't miss

324
00:19:28,200 --> 00:19:31,950
that middle term, because we're
in this process of building

325
00:19:31,950 --> 00:19:36,060
towards that potential future
moment. And if we're not kind of

326
00:19:36,060 --> 00:19:40,110
planning and thinking about all
of those stages, we're going to

327
00:19:40,110 --> 00:19:43,620
do ourselves a disservice when
we get to that middle stage. We

328
00:19:43,620 --> 00:19:48,030
also need to be thinking really
creatively and big picture like

329
00:19:48,030 --> 00:19:51,900
Charles suggested around what is
the ultimate problem we're

330
00:19:51,900 --> 00:19:55,410
trying to solve? Not just how
can we efficiency ourselves out

331
00:19:55,410 --> 00:19:59,910
of this situation in this exact
moment, so keeping all three of

332
00:19:59,910 --> 00:20:04,110
them those timeframes as top of
mind, I think applies to every

333
00:20:04,110 --> 00:20:08,070
industry right now. Because AI
is such a ubiquitous tool that

334
00:20:08,070 --> 00:20:11,550
can apply to really any element
of our lives.

335
00:20:12,390 --> 00:20:16,770
KiKi L'Italien: Yeah. And we had
a comment from someone who is

336
00:20:16,770 --> 00:20:20,310
watching live who said, Remember
when electronic calculators

337
00:20:20,310 --> 00:20:25,140
weren't allowed in the
classroom? I do. I do, actually.

338
00:20:25,470 --> 00:20:29,370
And, you know, we laugh about it
now, but it definitely wasn't

339
00:20:29,370 --> 00:20:33,660
that long ago, relatively
speaking. And, you know, it does

340
00:20:33,660 --> 00:20:38,010
feel like that, where, when you
have these tools, why wouldn't

341
00:20:38,010 --> 00:20:41,220
you figure out how to use them.
And almost it's a hindrance if

342
00:20:41,220 --> 00:20:46,710
you don't teach how to use them
accurately and and in the best

343
00:20:46,710 --> 00:20:50,160
possible way to help people to
do what they need to do better.

344
00:20:50,940 --> 00:20:51,390
Yeah,

345
00:20:51,420 --> 00:20:53,580
Audrey Ellis: and if I could
just add one piece to that, I

346
00:20:53,580 --> 00:20:59,430
think that AI the calculator is
a really convenient analogy or

347
00:20:59,430 --> 00:21:03,150
kind of mental reference point.
I've heard some interesting

348
00:21:03,150 --> 00:21:07,470
pushback on that, because AI is
applicable to so much more than

349
00:21:07,470 --> 00:21:12,270
just math, which, of course, is
not just math. But I think that

350
00:21:12,420 --> 00:21:16,020
we need to really not just teach
people how to use AI, but also

351
00:21:16,170 --> 00:21:20,070
how to decide if AI is necessary
to use and you can't really make

352
00:21:20,070 --> 00:21:23,610
that distinction if you've never
used it. So that's why we need,

353
00:21:23,850 --> 00:21:27,210
it might be good to learn your
times tables, memorize your

354
00:21:27,270 --> 00:21:31,800
multiplication figures without a
calculator. But if you know that

355
00:21:31,800 --> 00:21:33,840
there's this thing out there
that you've never been allowed

356
00:21:33,840 --> 00:21:37,410
to try, you might be very
tempted. Instead, we should be I

357
00:21:37,410 --> 00:21:41,130
think, convincing people and
giving them that kind of ability

358
00:21:41,130 --> 00:21:44,460
to judge for themselves when
it's appropriate to use and when

359
00:21:44,460 --> 00:21:45,660
it isn't, that's a

360
00:21:45,660 --> 00:21:47,610
KiKi L'Italien: great point,
because my daughter is getting

361
00:21:47,610 --> 00:21:51,210
ready to learn how to drive and
she doesn't know life without

362
00:21:51,210 --> 00:21:55,680
GPS existing. But we also know
that sometimes it doesn't always

363
00:21:55,680 --> 00:22:00,150
work. And so Doesn't she need to
know how to use an atlas if she

364
00:22:00,150 --> 00:22:03,840
needs to. Right. So these are
things that I think it's a

365
00:22:03,840 --> 00:22:06,270
really good point, you know,
when should we use it? When

366
00:22:06,270 --> 00:22:09,420
Shouldn't we? And then actually
is a thank you very much, Andre

367
00:22:09,420 --> 00:22:12,630
leading me into my next
question, which has to do with

368
00:22:12,630 --> 00:22:16,140
some of the ethical
considerations, when we're

369
00:22:16,140 --> 00:22:20,130
thinking about and specifically
thinking about associations in

370
00:22:20,130 --> 00:22:23,610
this case, but it really could
be any organization, when you're

371
00:22:23,610 --> 00:22:27,780
thinking about enhancing
engagement, enhancing learning?

372
00:22:28,410 --> 00:22:31,740
What are the ethical
considerations that you should

373
00:22:31,740 --> 00:22:35,580
be mindful of? So this is an
open question. I don't know if

374
00:22:35,580 --> 00:22:39,090
Michael, if you would like to
answer it, or Charles or Audrey,

375
00:22:39,090 --> 00:22:43,980
but it's open to anyone who
feels particularly passionate in

376
00:22:43,980 --> 00:22:46,560
their answer. Clearly,

377
00:22:46,740 --> 00:22:48,990
Michael Baston: recognizing the
importance of due diligence is

378
00:22:48,990 --> 00:22:53,400
critical. From my perspective,
you ultimately have to

379
00:22:53,400 --> 00:22:56,280
understand that people who put
the information in the

380
00:22:56,280 --> 00:23:00,930
algorithms may not necessarily
have cultural sensitivities, you

381
00:23:00,930 --> 00:23:04,140
may not have accurate
information. And so there's a

382
00:23:04,140 --> 00:23:08,520
certain amount of expectation
and whoever is going to use the

383
00:23:08,550 --> 00:23:13,200
AI, that there is going to be a
need to engage in the level of

384
00:23:13,200 --> 00:23:18,750
due diligence. Because, quite
frankly, you know, when we don't

385
00:23:18,750 --> 00:23:22,650
use this work, and we don't see
it as a tool, and we think that

386
00:23:22,650 --> 00:23:26,610
it is a solution and not a tool,
then when ultimately could be

387
00:23:26,640 --> 00:23:30,510
spreading false information, we
could be, you know, actually

388
00:23:30,570 --> 00:23:34,440
perpetrating a lot of frauds out
there with, with information. So

389
00:23:34,440 --> 00:23:37,440
it's very important from the
ethical perspective that you as

390
00:23:37,440 --> 00:23:41,820
a person who's going to utilize
the AI that you actually do your

391
00:23:41,820 --> 00:23:45,330
due diligence, so that you're
not kind of spreading out

392
00:23:45,540 --> 00:23:48,720
misinformation and bad
information, quite frankly.

393
00:23:50,040 --> 00:23:52,620
KiKi L'Italien: Well, so
Charles, were you about to say

394
00:23:52,620 --> 00:23:53,160
something?

395
00:23:53,250 --> 00:23:55,500
Charles Ansell: Well, I was
actually going to you know, I

396
00:23:55,500 --> 00:23:59,580
hope Audrey doesn't mind but I,
you know, I, I think of you know

397
00:23:59,580 --> 00:24:02,370
your story, and I think you
should be the one to share the

398
00:24:02,400 --> 00:24:05,940
one that the Boston Globe picked
up on where you know, it really

399
00:24:05,970 --> 00:24:08,490
like if you think of this
example, and what it could do in

400
00:24:08,490 --> 00:24:11,880
the classroom or for
policymakers. If AI hallucinates

401
00:24:11,880 --> 00:24:15,900
these answers and incorporates
these biases, I think it's

402
00:24:15,900 --> 00:24:18,480
emblematic of the real dangers,
but I don't want to tell Audrey

403
00:24:18,480 --> 00:24:19,380
story for her so

404
00:24:21,090 --> 00:24:24,150
Audrey Ellis: sure, I think it's
a pretty funny story. And it is

405
00:24:24,180 --> 00:24:28,830
like a kind of low stakes way to
illustrate how wrong these

406
00:24:28,830 --> 00:24:34,830
situations can end up. So
backstory I am on my last year

407
00:24:34,830 --> 00:24:38,910
of playing fantasy football. I
don't have time for it anymore,

408
00:24:38,910 --> 00:24:44,160
but I was on my husband's league
for about eight years. We played

409
00:24:44,160 --> 00:24:48,570
on Yahoo fantasy, and my team
name because I was the only

410
00:24:48,600 --> 00:24:52,770
female on the whole league was
play like a girl, which is a few

411
00:24:52,770 --> 00:24:58,200
years old by now but it's it was
like the Nike slogan for women's

412
00:24:58,200 --> 00:25:02,370
soccer and I love that and
thought, you know, why not? So I

413
00:25:02,370 --> 00:25:05,100
did terribly this year, which is
another reason for the record, I

414
00:25:05,100 --> 00:25:08,820
have one before, but this year
was not my year. And I kind of

415
00:25:08,820 --> 00:25:14,490
just gave up and I got a season
recap. That was chat GPT

416
00:25:14,490 --> 00:25:21,210
generated by Yahoo. And chat GPT
really just took my team name

417
00:25:21,240 --> 00:25:25,170
play like a girl, and the fact
that I lost and ran with it. So

418
00:25:25,170 --> 00:25:30,930
this summary was absolutely full
of suggestions that like a girl,

419
00:25:30,960 --> 00:25:35,010
I lost and had a terrible season
and like a girl, I don't know

420
00:25:35,010 --> 00:25:39,750
anything about football. And it
was honestly, this. I think,

421
00:25:39,750 --> 00:25:42,690
Charles, you're misremembering,
the Boston Globe to not pick

422
00:25:42,690 --> 00:25:47,910
this up. But I did submit, you
know, feedback to Yahoo. And I

423
00:25:47,910 --> 00:25:51,810
think that, you know, as funny
as it is, and generally, people

424
00:25:51,810 --> 00:25:54,690
are playing fantasy football as
a hobby, and hopefully not in

425
00:25:54,720 --> 00:25:58,470
many professional contexts. But
people can put kind of all kinds

426
00:25:58,470 --> 00:26:01,470
of things in their team names,
and in their smack talk and all

427
00:26:01,470 --> 00:26:04,560
of that. And the fact that all
of that was potentially

428
00:26:04,560 --> 00:26:11,190
informing this use case for an
AI bot. Unchecked is a pretty

429
00:26:11,190 --> 00:26:16,500
risky application so early in
this in the time of generative

430
00:26:16,500 --> 00:26:21,150
AI. So hopefully, they learned
their lesson, but hopefully also

431
00:26:21,150 --> 00:26:22,140
brought some laughs.

432
00:26:22,620 --> 00:26:25,140
KiKi L'Italien: Oh, my gosh,
that sounds that sounds like it

433
00:26:25,140 --> 00:26:27,780
could have been a nightmare,
would it? It is such a great

434
00:26:27,780 --> 00:26:30,570
example, though. Charles, I'm so
glad you brought it up. Because

435
00:26:30,900 --> 00:26:34,980
this is it is you can easily see
where that could have gone

436
00:26:35,040 --> 00:26:39,810
really, really, really terribly
wrong. If that had been at a

437
00:26:39,810 --> 00:26:45,030
larger scale. So excellent,
excellent example. Well, you

438
00:26:45,030 --> 00:26:48,480
know, I was wondering about, for
associations, there's an

439
00:26:48,480 --> 00:26:52,080
association for everything, as
we like to say. And so

440
00:26:52,080 --> 00:26:56,610
certainly, there's always this
concern about where can things

441
00:26:56,610 --> 00:26:59,910
go wrong? And how can we
mitigate risk? I'm sure, in

442
00:26:59,910 --> 00:27:02,820
higher ed, you're very familiar
with this concern about

443
00:27:02,820 --> 00:27:09,330
mitigating risk. So are there
ways for associations to

444
00:27:09,330 --> 00:27:16,680
advocate for or against certain
uses of AI in education, to

445
00:27:16,680 --> 00:27:21,720
ensure that it benefits a
diverse student body, because we

446
00:27:21,720 --> 00:27:25,980
have, we have associations that
are looking for its members that

447
00:27:25,980 --> 00:27:30,180
are going into its industries,
starting all the way back, going

448
00:27:30,180 --> 00:27:34,650
through school, very concerned
about making sure that this

449
00:27:34,650 --> 00:27:39,330
science or this industry is
receiving the type of education

450
00:27:39,330 --> 00:27:44,130
and protection and concern that
it needs needs to have, and

451
00:27:44,130 --> 00:27:47,160
certainly with a new technology,
or something that's advancing so

452
00:27:47,160 --> 00:27:52,140
quickly? There's a lot of
concern about like, Well, how do

453
00:27:52,140 --> 00:27:55,710
we how do we make sure it
benefits our people, our

454
00:27:55,710 --> 00:27:58,200
students, the ones that are
coming into our field?

455
00:28:00,870 --> 00:28:03,360
Charles Ansell: I can start on
that. You know, I think that

456
00:28:03,360 --> 00:28:06,510
there's two things that come to
mind for me when you say that

457
00:28:06,510 --> 00:28:11,640
cakey. So one is, I think it's,
you know, important to also just

458
00:28:11,640 --> 00:28:15,420
use our discussion as a sort of
like, analogy or model or

459
00:28:15,420 --> 00:28:18,000
whatever word you want to use
for your own Association, what

460
00:28:18,000 --> 00:28:19,680
you're trying to represent for
your members. Because the

461
00:28:19,680 --> 00:28:22,590
problems that we're grappling
with in higher education, poor

462
00:28:22,590 --> 00:28:26,130
to, you know, real estate, they
poor to health care, you know,

463
00:28:26,130 --> 00:28:31,290
are you being, you know, problem
first, in terms of what this

464
00:28:31,290 --> 00:28:35,400
technology can do? And then are
you taking a comprehensive look

465
00:28:35,460 --> 00:28:39,120
on all the issues that your
association is grappling with?

466
00:28:39,150 --> 00:28:42,090
Are you tracking those issues in
a mutually exclusive manner,

467
00:28:42,300 --> 00:28:46,530
such that you know, exactly
where AI could plug in, or any

468
00:28:46,530 --> 00:28:49,950
technology for that matter? And
what the limitations are? If you

469
00:28:49,950 --> 00:28:55,620
just don't do that, in that way,
it's going to be, I think, a

470
00:28:55,620 --> 00:28:59,040
lost cause. The second piece is,
is really getting directly to

471
00:28:59,040 --> 00:29:01,380
your question, which is, you
know, how can associations

472
00:29:01,380 --> 00:29:04,620
interact with higher education
on this? You know, I'm curious

473
00:29:04,620 --> 00:29:10,950
what others here, you know,
Audrey President passed and

474
00:29:11,550 --> 00:29:16,020
think about this. But I think
that we need to find a way to

475
00:29:16,440 --> 00:29:22,920
very quickly get AI learning
outcomes into the heads of all

476
00:29:22,920 --> 00:29:26,550
of our students in all
disciplines. Because what I fear

477
00:29:26,550 --> 00:29:30,240
is going to happen is that a lot
of social mobility ladders that

478
00:29:30,240 --> 00:29:32,610
exists to really good paying
jobs, like really strong

479
00:29:32,610 --> 00:29:34,890
careers, like thinking like
major accounting firms, major

480
00:29:34,890 --> 00:29:37,350
law firms, major consultancies,
lots of these things start at

481
00:29:37,350 --> 00:29:41,490
the associate level. And we're
beginning to see the Inklings

482
00:29:41,520 --> 00:29:44,700
that a lot of those jobs that
start these career ladders could

483
00:29:44,700 --> 00:29:48,690
go away, because AI can sweep
these knowledge bases and do

484
00:29:48,690 --> 00:29:51,330
these things that end up in
forming, you know, the the

485
00:29:51,330 --> 00:29:54,390
lawyers or the consultants of
the, you know, engagement

486
00:29:54,390 --> 00:30:00,810
managers. And so, and yet, if if
we the way At higher ed works,

487
00:30:00,810 --> 00:30:03,750
it's kind of slow, right? Like,
we have to, you know, to get new

488
00:30:03,750 --> 00:30:05,940
programs approved and all this,
you have to go through this

489
00:30:05,940 --> 00:30:08,430
accreditation process. And you
know, sometimes these are multi

490
00:30:08,430 --> 00:30:12,630
year things. And so I think, you
know, there's, there's players

491
00:30:12,630 --> 00:30:16,410
in the field of post secondary,
that associations, you know, and

492
00:30:16,410 --> 00:30:19,110
I think complete college,
America is one of them. And I

493
00:30:19,110 --> 00:30:21,510
encourage folks to, you know, if
they want to reach out after

494
00:30:21,510 --> 00:30:27,870
this, maybe they could have my
contact info. But that can look

495
00:30:27,870 --> 00:30:31,650
at ways to make sure that the
course by course level, and at

496
00:30:31,650 --> 00:30:34,980
the certificate level, that
these competencies as the field

497
00:30:35,010 --> 00:30:39,810
is just starting, that they get
swept in right away. So then

498
00:30:39,810 --> 00:30:44,070
that way, students are ready in
their existing programs of

499
00:30:44,070 --> 00:30:47,010
study. So we don't need to wait,
the several years that higher ed

500
00:30:47,010 --> 00:30:48,810
always takes to catch up to
these things, if that makes

501
00:30:48,810 --> 00:30:51,030
sense. You know, I think for
people who are not in higher ed,

502
00:30:51,360 --> 00:30:54,990
it, you know, so often, we think
of the 7% of colleges that are

503
00:30:54,990 --> 00:30:57,840
selective enrollment that get
hauled in front of the Senate to

504
00:30:57,840 --> 00:31:00,540
get yelled at and stuff, because
that's where like, most of the

505
00:31:00,540 --> 00:31:04,020
media went to college, but like
93% of students are at, you

506
00:31:04,020 --> 00:31:06,120
know, non selective colleges,
and most of them are in a

507
00:31:06,120 --> 00:31:08,610
resource, and lots of them
through no fault of their own

508
00:31:08,610 --> 00:31:10,770
have to move a little slower
than they could otherwise. And

509
00:31:10,770 --> 00:31:13,920
so we need to find ways to fund
that, but also disrupt it so

510
00:31:13,920 --> 00:31:17,520
that way, students are ready,
not just in terms of the AI that

511
00:31:17,520 --> 00:31:19,980
is used to deliver higher
education, which has been our

512
00:31:19,980 --> 00:31:24,450
discussion so far, but also
being a successful lung, which

513
00:31:24,450 --> 00:31:30,000
is the point of higher ed is
including being very competent

514
00:31:30,000 --> 00:31:30,570
in AI.

515
00:31:32,100 --> 00:31:34,950
KiKi L'Italien: I mean, well
said, and actually, I want to

516
00:31:34,980 --> 00:31:38,880
give a shout out to a friend of
ours in the association space

517
00:31:38,880 --> 00:31:42,270
six degrees of associations that
says love this discussion, so

518
00:31:42,270 --> 00:31:46,860
many aha moments in this one.
And I have to say, first of all,

519
00:31:46,890 --> 00:31:51,420
check out their podcast to also
can't can't help but agree. I

520
00:31:51,420 --> 00:31:54,330
mean, I think that there are a
lot of aha moments I've

521
00:31:54,330 --> 00:31:58,500
certainly heard in this one. And
and you know, Charles, when you

522
00:31:58,500 --> 00:32:01,620
were talking about it, I was
thinking, absolutely.

523
00:32:01,620 --> 00:32:04,650
Associations are always thinking
about like, what's the future

524
00:32:04,650 --> 00:32:08,580
for the industry? What's the
future for our members? How do

525
00:32:08,580 --> 00:32:12,510
we best prepare them, like
higher ed, higher ed is also

526
00:32:12,510 --> 00:32:15,270
thinking about this. And when
you want to talk about some

527
00:32:15,270 --> 00:32:20,160
really great partnerships, all
you know, higher ed, and the and

528
00:32:20,160 --> 00:32:23,040
the associations that are
working to represent some of

529
00:32:23,040 --> 00:32:26,880
those industries, that we're all
concerned about the students

530
00:32:26,880 --> 00:32:29,250
that are going through the
individuals who are going

531
00:32:29,250 --> 00:32:33,390
through, and then forming the
workforce and and really the

532
00:32:33,390 --> 00:32:37,230
society that's around us. So
that all leads me to this next

533
00:32:37,230 --> 00:32:40,140
question that has to do with
what are some examples where

534
00:32:40,170 --> 00:32:44,550
it's already beginning to work?
Or we're starting to see, okay,

535
00:32:44,580 --> 00:32:50,070
we're using AI in a meaningful
way that is successful. That's

536
00:32:50,070 --> 00:32:53,670
something that we would look at,
and we would say, okay, AI, your

537
00:32:53,670 --> 00:32:58,230
we have you in this higher
educational setting? What

538
00:32:58,230 --> 00:33:02,700
lessons can we learn from some
of the early success stories

539
00:33:03,420 --> 00:33:05,190
that we're hearing, we

540
00:33:05,190 --> 00:33:09,960
Michael Baston: do have faculty
right now who are fully engaged

541
00:33:10,080 --> 00:33:14,460
in piloting a number of
innovative utilizations of AI.

542
00:33:14,460 --> 00:33:18,570
So I think that, you know, while
there might be this suggestion,

543
00:33:18,750 --> 00:33:24,480
that higher education writ large
is sort of concerned about and

544
00:33:24,480 --> 00:33:26,760
don't want to get involved in
AI. That's not the case

545
00:33:26,760 --> 00:33:29,550
everywhere. There are a lot of
places that recognize that we

546
00:33:29,550 --> 00:33:33,060
can't unilaterally disarm
opportunity for students. The

547
00:33:33,060 --> 00:33:36,090
fact is, if we want graduates
who are going to effectively

548
00:33:36,090 --> 00:33:39,300
navigate the workplace
environment, and this technology

549
00:33:39,300 --> 00:33:43,410
is now an expectation in that
environment, if we don't look at

550
00:33:43,410 --> 00:33:47,130
our learning objectives, and now
reassess them, and actually

551
00:33:47,160 --> 00:33:51,630
invigorates our syllabi, in ways
that allow them to develop the

552
00:33:51,630 --> 00:33:54,690
skill sets are, they're not
going to be as successful and

553
00:33:54,690 --> 00:33:57,180
we're not going to be as
effective in the work that we

554
00:33:57,180 --> 00:34:00,450
do. So there are a lot of folks
all around the country who are

555
00:34:00,450 --> 00:34:04,530
in higher education spaces that
are already starting to really

556
00:34:04,560 --> 00:34:09,240
examine this work. You look at
Maricopa, that, that system

557
00:34:09,240 --> 00:34:13,770
actually has degrees now in a
brand new degree. So So while it

558
00:34:13,770 --> 00:34:18,570
often takes a long time for some
in higher ed to move, not in a

559
00:34:18,570 --> 00:34:22,230
non degree space, you can run
certificate programs are pretty

560
00:34:22,230 --> 00:34:25,680
flexibly, you can build those
stackable credentials into

561
00:34:25,800 --> 00:34:29,550
existing programs. So so it's
not that we are completely

562
00:34:29,550 --> 00:34:33,030
handcuffed from being
innovative. What we have to do

563
00:34:33,030 --> 00:34:36,990
is understand the importance of
balancing AI with really the

564
00:34:36,990 --> 00:34:41,010
human interaction because it's
not gonna replace all of the

565
00:34:41,010 --> 00:34:46,020
humans as well, but we've got to
learn how to best integrate the

566
00:34:46,020 --> 00:34:47,010
efforts together.

567
00:34:47,760 --> 00:34:50,910
KiKi L'Italien: Right, that
balanced approach, I think is is

568
00:34:51,060 --> 00:34:55,410
it's full of nuances. And as
with anything that has nuances,

569
00:34:55,410 --> 00:34:58,890
it's very difficult when when
something's not black or white,

570
00:34:58,920 --> 00:35:04,560
it's very difficult To know what
the right thing is, and the best

571
00:35:04,560 --> 00:35:09,030
way to navigate it. But we have
to we have to figure out the

572
00:35:09,030 --> 00:35:12,480
best way forward. Andre, do you
have examples? What are you

573
00:35:12,480 --> 00:35:15,990
hearing? I know you're pulling
together stories and and talking

574
00:35:15,990 --> 00:35:16,770
with everyone.

575
00:35:17,940 --> 00:35:21,000
Audrey Ellis: Yeah, we are
creating a council on equitable

576
00:35:21,000 --> 00:35:24,060
AI or we have created and we're
meeting later this week in

577
00:35:24,060 --> 00:35:27,720
person for the first time, which
is really exciting. And so we

578
00:35:27,720 --> 00:35:31,860
hope that that will really help
us kind of gather and collect

579
00:35:31,860 --> 00:35:34,830
these practices, because I
couldn't agree more President

580
00:35:34,830 --> 00:35:40,650
Bastien that this is happening
in incredible scope and scale,

581
00:35:40,680 --> 00:35:43,830
unfortunately, what I'm seeing
is that it's happening in a

582
00:35:43,830 --> 00:35:47,730
really decentralized way, which
is not bad for innovation, don't

583
00:35:47,730 --> 00:35:51,390
get me wrong, but makes it
really hard to learn from what

584
00:35:51,390 --> 00:35:55,500
others are doing and replicate
without starting from scratch in

585
00:35:55,500 --> 00:36:01,590
every pocket. And so you have
institutions that are maybe have

586
00:36:01,590 --> 00:36:05,910
some pockets of enthusiasm, and
some pockets of terror, or

587
00:36:06,390 --> 00:36:08,250
however you want to call it,
whatever you want to call it.

588
00:36:08,460 --> 00:36:11,040
And then you have institutions
themselves, where you have some

589
00:36:11,040 --> 00:36:13,440
institutions that are just
totally on board and some that

590
00:36:13,440 --> 00:36:17,130
are not. So I think that that
that is, it's really

591
00:36:17,130 --> 00:36:20,130
interesting, because there's a
lot of great examples of AI

592
00:36:20,160 --> 00:36:23,760
broadly, right, and what it's
how it's been used in AI, in

593
00:36:23,760 --> 00:36:27,600
higher education, you know,
Georgia State University, as in

594
00:36:27,600 --> 00:36:30,870
a lot of retention work for
years, and kind of initially

595
00:36:30,870 --> 00:36:36,780
started a lot of that focus.
Recently, John J. Work partnered

596
00:36:36,780 --> 00:36:40,950
with a startup that was, you
know, a nonprofit startup to

597
00:36:41,220 --> 00:36:44,340
figure out how to increase their
retention. But these are our

598
00:36:44,460 --> 00:36:50,430
old, not older, but less, kind
of groundbreaking types of AI,

599
00:36:50,460 --> 00:36:55,380
because AI is a wide swath of
technology. What we're really

600
00:36:55,380 --> 00:36:59,010
interested right now, and
everyone is is generative AI,

601
00:36:59,040 --> 00:37:03,480
chat, GPT, things like that, and
how that can then inform

602
00:37:03,510 --> 00:37:06,930
everything else, even other
types of AI. And that's where I

603
00:37:06,930 --> 00:37:10,290
think we still haven't really
been able to put a pin on

604
00:37:10,500 --> 00:37:14,430
exactly what institutions are
doing necessarily, or anything

605
00:37:14,430 --> 00:37:18,390
beyond pilots, because there
hasn't been enough time, we need

606
00:37:18,390 --> 00:37:22,710
to see how this goes. But we
also can't wait necessarily,

607
00:37:23,070 --> 00:37:27,630
until they're done or until the
academic peer review process

608
00:37:27,840 --> 00:37:30,810
finalizes so that, you know, a
journal article can get

609
00:37:30,810 --> 00:37:34,950
published to talk about it,
because I can guarantee that

610
00:37:34,950 --> 00:37:37,800
other institutions are thinking
about it or attempting it

611
00:37:37,800 --> 00:37:41,370
themselves, and would benefit
greatly from hearing kind of

612
00:37:41,370 --> 00:37:45,600
more intermediate updates on how
things are going. So that's why

613
00:37:45,600 --> 00:37:48,390
we're really trying to
facilitate that more rapid

614
00:37:48,420 --> 00:37:51,540
lesson sharing practices through
the playbook through the

615
00:37:51,540 --> 00:37:56,730
council, so that we're not, you
know, hurry up, then wait, hurry

616
00:37:56,730 --> 00:38:01,020
up and wait type of timeline and
process because AI is not

617
00:38:01,020 --> 00:38:04,050
waiting. It is just like full
speed ahead, whether or not we

618
00:38:04,050 --> 00:38:06,270
get on its timeline? Well.

619
00:38:09,120 --> 00:38:11,400
Charles Ansell: You know, I
think, you know, in terms of the

620
00:38:11,400 --> 00:38:13,500
use cases, just thinking about
the associations in the

621
00:38:13,500 --> 00:38:17,520
audience, I think that it's
important to take a look at the

622
00:38:17,550 --> 00:38:20,880
playbook that I believe was
shared, you know, around or can

623
00:38:20,880 --> 00:38:24,600
be shared in the chat that that
Audrey is referring to. Because

624
00:38:24,750 --> 00:38:26,910
even though it's highly
specific, I think it's organized

625
00:38:26,910 --> 00:38:29,160
into things that all
organizations end up benefiting

626
00:38:29,160 --> 00:38:32,190
from, right, like we split it
by, like, you know, teaching and

627
00:38:32,190 --> 00:38:36,030
learning and student success and
organizational effectiveness,

628
00:38:36,930 --> 00:38:40,080
and, and data. And sure some of
these won't be one to one,

629
00:38:40,080 --> 00:38:41,970
right, especially like the
teaching and learning session,

630
00:38:42,000 --> 00:38:44,820
although most companies have
professional development, right?

631
00:38:45,510 --> 00:38:48,630
Most companies have a knowledge
base. And I think that the big

632
00:38:48,630 --> 00:38:51,450
thing to get those use cases
going at such an embryonic stage

633
00:38:51,450 --> 00:38:54,360
for the industry is to get the
sample prompts into more and

634
00:38:54,360 --> 00:38:58,260
more hands. And and so whether
it's porting these over

635
00:38:58,260 --> 00:39:01,950
yourself, or even asking a
generative AI to like do that

636
00:39:01,950 --> 00:39:05,010
work for you? How can this
document work for my industry, I

637
00:39:05,010 --> 00:39:07,170
think that'd be like a very
worthwhile task and could save

638
00:39:07,170 --> 00:39:08,190
hundreds of hours of time.

639
00:39:08,910 --> 00:39:10,950
KiKi L'Italien: I mean, it
really could. And what I was

640
00:39:10,950 --> 00:39:14,940
gonna say is, and I love I love
when somebody can share their

641
00:39:14,940 --> 00:39:18,870
prompt, and I know that it's
going to get me closer to more

642
00:39:18,870 --> 00:39:22,920
usable information that I can
then apply and customize for

643
00:39:22,920 --> 00:39:27,360
myself. So excellent, excellent
advice there. And for anyone

644
00:39:27,360 --> 00:39:31,140
who's who's listening to this
now or listening to it later.

645
00:39:31,710 --> 00:39:34,740
What I'm going to do is I'm
gonna grab the link to that

646
00:39:34,740 --> 00:39:37,290
playbook. You want to check out
the playbook. I'll put it in the

647
00:39:37,290 --> 00:39:40,980
show notes from the edited
version of the show, and then

648
00:39:40,980 --> 00:39:45,750
you'll be able to download that
there and go check it out. You

649
00:39:45,750 --> 00:39:50,130
know, one thing you said
earlier, Audrey was about you

650
00:39:50,130 --> 00:39:53,580
talked about, you can't wait
necessarily for the peer

651
00:39:53,580 --> 00:39:57,570
reviewed everything to come out
when there's something advancing

652
00:39:57,570 --> 00:40:01,800
so quickly right in the case of
AI, we can't wait until the end

653
00:40:01,800 --> 00:40:03,960
of the day. It's not like it's
going to just stop and say,

654
00:40:03,960 --> 00:40:06,930
Okay, now catch up. Let's wait
for all the peer reviewed

655
00:40:06,930 --> 00:40:10,050
research to be done. We have to
be sort of developing, as they

656
00:40:10,050 --> 00:40:12,510
say, you know, while the planes
in the air, we have to be

657
00:40:12,510 --> 00:40:18,150
building the plane, right, so.
So my point to that, is this

658
00:40:18,150 --> 00:40:24,120
quantifying success? How can we
look at and this is forever, the

659
00:40:24,120 --> 00:40:27,600
thing that I know, my friends in
association land are looking

660
00:40:27,600 --> 00:40:31,800
for. I know, like any
professional, really any anyone

661
00:40:31,800 --> 00:40:34,590
in higher ed, you're looking
for? How can I quantify that

662
00:40:34,590 --> 00:40:37,260
this is something that's
successful? So what are we

663
00:40:37,260 --> 00:40:41,400
looking at? What are some
measures of success that we

664
00:40:41,400 --> 00:40:45,300
might be able to point to to
say, I yeah, here's where it's

665
00:40:45,300 --> 00:40:51,030
working. Here's where we know
that either the efficiency, or

666
00:40:51,450 --> 00:40:54,690
the new ideas and innovation
that's a result of it, the new

667
00:40:54,690 --> 00:40:58,740
products, etc, etc, it can come
into play?

668
00:41:01,050 --> 00:41:02,070
Charles Ansell: Does that for
anybody?

669
00:41:02,430 --> 00:41:03,930
KiKi L'Italien: This for anyone?
And?

670
00:41:10,860 --> 00:41:15,000
You go, Charles, I love it. I
love it answers.

671
00:41:16,200 --> 00:41:17,790
Charles Ansell: Well, I'll just
say that I think that there's

672
00:41:17,790 --> 00:41:19,770
probably two different types of
metrics, right. And I'm sure

673
00:41:19,770 --> 00:41:22,440
that this ports over into other
sectors for higher ed, there's

674
00:41:22,440 --> 00:41:24,690
like your students success
metrics. And then there's like

675
00:41:24,690 --> 00:41:28,260
the operational metrics that you
use to get that done. Right. And

676
00:41:28,260 --> 00:41:32,760
so, you know, when we look at
things like, you know, are we

677
00:41:32,970 --> 00:41:35,550
increasing our graduation rates?
are we increasing retention,

678
00:41:35,550 --> 00:41:37,440
there's usually going to be
leading indicators of that,

679
00:41:37,470 --> 00:41:39,870
like, how many students are on
education plans? And can we

680
00:41:39,870 --> 00:41:42,960
track that like week to week,
because we know that the proofs

681
00:41:42,960 --> 00:41:46,770
out there, that's mastery by
semester education plans that

682
00:41:46,770 --> 00:41:49,500
actually registered towards them
end up graduating at a higher

683
00:41:49,500 --> 00:41:51,600
rate? I'm sure there's like
analogous outcomes, metrics

684
00:41:51,600 --> 00:41:55,170
versus process metrics for all
types of sectors on this call.

685
00:41:55,860 --> 00:41:59,610
And so I think that you probably
get pretty deep into some weedy

686
00:42:01,320 --> 00:42:04,410
AI metrics, right. So it's like,
what percent of students are in

687
00:42:04,410 --> 00:42:07,110
education plans, you go one
layer down? To what extent were

688
00:42:07,110 --> 00:42:09,990
they created by this technology
of that technology and tracking

689
00:42:09,990 --> 00:42:13,770
that like day to week? I think
that we're very nascent here. So

690
00:42:13,770 --> 00:42:16,140
we haven't like made that full
metrics tree as a secretary at

691
00:42:16,140 --> 00:42:18,630
AI, or at least I haven't seen
it, because we need to, like

692
00:42:18,630 --> 00:42:21,420
start getting more of these use
cases. I would also say that,

693
00:42:22,410 --> 00:42:25,020
again, I don't know how this
ports outside of higher ed, but

694
00:42:25,230 --> 00:42:27,450
I'm curious if Audrey and
President Bassett agree with

695
00:42:27,450 --> 00:42:31,500
this. But I think that we can
also like fetishize the

696
00:42:31,830 --> 00:42:35,550
causation versus correlation
thing a little too much. Because

697
00:42:35,550 --> 00:42:37,500
in higher ed, you know, I think
about the things that you know,

698
00:42:37,500 --> 00:42:40,230
Tracy is doing in other colleges
where you're doing a lot of

699
00:42:40,230 --> 00:42:43,440
things at once, and you're not
waiting to say, Oh, it was the

700
00:42:43,440 --> 00:42:46,230
eggplants. Oh, it was making
sure every student's on advisor,

701
00:42:46,320 --> 00:42:49,170
and I think AI is just going to
be in that mix in that way. And

702
00:42:49,170 --> 00:42:52,260
we're never going to be able to,
you know, if you make an

703
00:42:52,260 --> 00:42:54,750
academic scholarly article about
it, that would be helpful for

704
00:42:54,750 --> 00:42:56,850
the field if somebody wants to
do a randomized controlled

705
00:42:56,850 --> 00:42:59,280
trial, but the problem is, if
it's a common sense

706
00:42:59,280 --> 00:43:02,400
implementation solution, that
you hate to have somebody be the

707
00:43:02,400 --> 00:43:05,340
control group. So I hope that
makes sense. It wasn't too like

708
00:43:05,340 --> 00:43:05,790
nerdy.

709
00:43:07,200 --> 00:43:08,220
KiKi L'Italien: I don't think it
was, too.

710
00:43:09,930 --> 00:43:12,480
Audrey Ellis: No, I agree with
you, Charles, I think that it's

711
00:43:12,540 --> 00:43:15,240
just like all of these other
interventions that we talked

712
00:43:15,240 --> 00:43:18,570
about, it's really difficult to
isolate the effects of one

713
00:43:18,570 --> 00:43:21,390
specific thing if you have
people who are benefiting,

714
00:43:21,390 --> 00:43:25,350
hopefully benefiting from many.
So I think that that's a fair

715
00:43:26,250 --> 00:43:29,490
assertion to make. One other
piece that I want to share is

716
00:43:29,520 --> 00:43:34,380
around more of like a systems
level thinking about knowledge,

717
00:43:34,410 --> 00:43:38,700
which maybe sounds really broad
and lofty. But I think that one,

718
00:43:39,060 --> 00:43:43,110
one of the things that we might,
that might be as a hypothesis, I

719
00:43:43,110 --> 00:43:48,180
guess, at the root of some of
the kind of panic around AI is

720
00:43:48,420 --> 00:43:52,860
that AI is a threat to kind of
how we value knowledge, how we

721
00:43:52,860 --> 00:43:55,740
think it can be assessed, or it
should be assessed how we think

722
00:43:55,740 --> 00:43:58,890
it should be measured. And I
think that we're gonna see a lot

723
00:43:58,920 --> 00:44:04,020
of conversation here in the
coming months and years ahead,

724
00:44:04,020 --> 00:44:08,040
because as the purveyors of
knowledge and higher education.

725
00:44:08,460 --> 00:44:11,640
I think there's a feeling that
kind of, we know, we're the

726
00:44:11,640 --> 00:44:16,320
experts on that. But this is
really going to shake up and

727
00:44:16,320 --> 00:44:22,080
shift potentially, how we view
all of that and how we do assess

728
00:44:22,080 --> 00:44:25,620
that. And so one leading
indicator that I'm looking for,

729
00:44:25,980 --> 00:44:29,010
in on the equity space, in
particular on equity

730
00:44:29,010 --> 00:44:33,570
conversation, in particular, is
for new types of knowledge and

731
00:44:33,600 --> 00:44:37,620
voices and perspectives from
from groups of people who have

732
00:44:37,620 --> 00:44:41,700
historically not had their
value, their knowledge valued at

733
00:44:41,700 --> 00:44:46,140
the same level as past maybe
folks who can make it all the

734
00:44:46,140 --> 00:44:49,320
way through an academic journey
all the way to that point of

735
00:44:49,320 --> 00:44:53,460
peer reviewed work. I want to
see new voices and I want I

736
00:44:53,460 --> 00:44:57,930
think AI will be working if it's
helping bubble up and surface,

737
00:44:58,050 --> 00:45:02,400
different perspectives that In
the past, because of kind of the

738
00:45:02,400 --> 00:45:06,240
structure of, of higher
education, those perspectives

739
00:45:06,240 --> 00:45:10,020
might have been kept out. So
that's more on like a societal

740
00:45:10,020 --> 00:45:13,530
systems level. But I think that
will be a really great signifier

741
00:45:13,530 --> 00:45:17,040
that we're doing something right
if we're making new paths for

742
00:45:17,040 --> 00:45:20,460
people to contribute to our way
of thinking through AI.

743
00:45:21,510 --> 00:45:24,660
Michael Baston: And I would just
add, from my perspective, as

744
00:45:24,660 --> 00:45:29,250
well about expanding capacity,
how does the utilization of AI

745
00:45:29,430 --> 00:45:34,230
enable us to expand the kind of
capacity so that we can actually

746
00:45:34,230 --> 00:45:37,950
get more people getting the help
they need when they need it at

747
00:45:37,950 --> 00:45:41,490
times that are convenient to
them. That's why from my

748
00:45:41,490 --> 00:45:45,990
perspective, it is critical for
us to be able to really master

749
00:45:46,020 --> 00:45:49,380
prompt generation, so that we
actually can get the right

750
00:45:49,380 --> 00:45:54,240
quality information. And we can
actually use ways to actually

751
00:45:54,300 --> 00:45:58,530
streamline processes and
actually accelerate the ability

752
00:45:58,530 --> 00:46:01,560
to people to navigate complex
systems that don't need to be

753
00:46:01,560 --> 00:46:05,160
complex. You know, when we think
about the bureaucratic structure

754
00:46:05,160 --> 00:46:08,160
of higher education, it was
supposed to actually not help

755
00:46:08,160 --> 00:46:11,940
the student, but actually help
the bureaucracy of how the the

756
00:46:11,940 --> 00:46:16,740
administration of higher
education Well, if we now don't

757
00:46:16,740 --> 00:46:20,280
need all of that bureaucracy
that supports the actual

758
00:46:20,280 --> 00:46:22,980
structures, but not the
students. And we actually can

759
00:46:22,980 --> 00:46:28,170
utilize some technology that
gets at minimizing all of that

760
00:46:28,230 --> 00:46:32,880
bloat. Wow, what a wonderful
breakthrough opportunity for

761
00:46:32,880 --> 00:46:35,760
students to not have to, you
know, at three o'clock in the

762
00:46:35,760 --> 00:46:38,340
morning, when they're up, and
they need somebody that they

763
00:46:38,340 --> 00:46:41,880
actually can go to a knowledge
base that can resolve issues

764
00:46:41,970 --> 00:46:44,250
without talking to somebody
who's going to send them

765
00:46:44,250 --> 00:46:46,950
someplace else. And then they
play the chutes and ladders game

766
00:46:46,950 --> 00:46:51,870
until they drop out of school.
So here's an opportunity if we

767
00:46:51,870 --> 00:46:54,180
think about it, and we utilize
it. Well, I

768
00:46:54,180 --> 00:46:56,490
KiKi L'Italien: mean, the people
who are listening to this as an

769
00:46:56,490 --> 00:47:00,270
audio podcast later, they cannot
see me right now, but I'm just

770
00:47:00,270 --> 00:47:03,990
like furiously shaking my head,
my hands, anything that can move

771
00:47:04,440 --> 00:47:10,290
in the camera, just in furious
agreement. Because it's so true.

772
00:47:10,290 --> 00:47:13,620
Can you imagine not having all
of the bureaucracy that's

773
00:47:13,620 --> 00:47:17,550
pressing back and making things
so difficult? It's hard. It's

774
00:47:17,550 --> 00:47:20,940
hard for me to imagine, but
you're painting a picture that I

775
00:47:20,940 --> 00:47:26,100
really want to see painted. So
you're the let's go there. And

776
00:47:26,100 --> 00:47:30,150
actually, to that end, my my
last question for you, as I'm,

777
00:47:30,180 --> 00:47:34,470
as I'm wrapping things up, which
by the way, this has been a

778
00:47:34,470 --> 00:47:37,860
phenomenal conversation today.
So thank you so much ahead of

779
00:47:37,860 --> 00:47:43,020
time, but where do you see this
work that you're doing going?

780
00:47:43,020 --> 00:47:46,380
Now all of you know each other,
you're you're paying attention

781
00:47:46,380 --> 00:47:50,490
to what's happening with the
development of AI. And as it

782
00:47:50,490 --> 00:47:55,680
pertains to higher ed, where do
you see what do you see for the

783
00:47:55,680 --> 00:47:58,500
work that you're doing right now
to sort of help shape this

784
00:47:58,500 --> 00:48:02,100
conversation around AI and
higher education as you're

785
00:48:02,100 --> 00:48:07,230
moving forward? What is to come?
All right, and I see you,

786
00:48:07,230 --> 00:48:08,760
Adrienne, go, I'm looking for
you.

787
00:48:09,030 --> 00:48:13,950
Audrey Ellis: Sure. You know,
there's an interesting, popular,

788
00:48:15,240 --> 00:48:17,640
I'm not maybe they're an
association, actually another

789
00:48:17,640 --> 00:48:21,750
thinking about a group in higher
education, EDUCAUSE, they

790
00:48:21,750 --> 00:48:25,980
released a report yesterday, an
AI landscape report, and I was

791
00:48:25,980 --> 00:48:29,130
spending some time with it over
the past two days, and something

792
00:48:29,130 --> 00:48:33,060
that I, you know, observed, and
being the data person that I am,

793
00:48:33,060 --> 00:48:35,850
who loves community colleges,
I'm always jumping to the end,

794
00:48:35,880 --> 00:48:38,910
or if there's a survey and
seeing Okay, well, who are they

795
00:48:38,910 --> 00:48:42,480
talking about? And who was
included in the survey? And what

796
00:48:42,480 --> 00:48:46,800
groups do they represent. And so
this is really interesting

797
00:48:46,800 --> 00:48:50,370
report, lots of great insights
that I think can relate to lots

798
00:48:50,370 --> 00:48:54,030
of industries. A lot of things
we covered today, strategic

799
00:48:54,030 --> 00:48:57,870
planning for AI prepare your
data for aI think about ethical

800
00:48:57,870 --> 00:49:01,650
and equitable use cases. But
what I'm still seeing is, you

801
00:49:01,650 --> 00:49:05,670
know, 13% of the participants
were from the community college

802
00:49:05,670 --> 00:49:10,170
space, and about 21% of the
participants were from minority

803
00:49:10,170 --> 00:49:14,670
serving institutions, which is
not on which is not does not

804
00:49:14,670 --> 00:49:18,720
represent parity, if you look at
the survey, you know, the survey

805
00:49:18,720 --> 00:49:23,520
population and our larger post
secondary space. And so what I

806
00:49:23,520 --> 00:49:27,270
hope will become an emerging
trend is that we can push the

807
00:49:27,270 --> 00:49:33,630
edges of our data collection and
our conversations, our sharing

808
00:49:33,660 --> 00:49:39,000
of knowledge and resources,
right, like funding and legal

809
00:49:39,000 --> 00:49:44,490
support and technical support to
these institutions who right now

810
00:49:44,490 --> 00:49:48,300
aren't even participating in
surveys or and therefore aren't

811
00:49:48,300 --> 00:49:51,270
even being considered part of
the landscape when they are

812
00:49:51,780 --> 00:49:55,980
because we're not making true
truly representative statements

813
00:49:55,980 --> 00:49:59,220
like most colleges are doing
this if we really aren't even

814
00:49:59,220 --> 00:50:02,850
including You know, so many
community colleges that educate

815
00:50:02,880 --> 00:50:07,260
a vast majority of our students
in the country that are in the

816
00:50:07,260 --> 00:50:11,190
post secondary space, and
probably your future workers at,

817
00:50:11,880 --> 00:50:13,980
you know, all of the
associations and the industries

818
00:50:13,980 --> 00:50:16,830
you represent. So that's one
space where I'd really like us

819
00:50:16,830 --> 00:50:21,480
to see the conversation pushed
and, you know, supported

820
00:50:21,510 --> 00:50:22,110
further.

821
00:50:23,010 --> 00:50:24,090
KiKi L'Italien: Absolutely.

822
00:50:25,080 --> 00:50:27,870
Charles Ansell: Charles. Yeah. I
mean, I think what Audrey said,

823
00:50:27,870 --> 00:50:30,990
I mean, I think it's going to be
matter more of seeing where the

824
00:50:30,990 --> 00:50:33,300
conversations going. But like,
more importantly, like moving

825
00:50:33,300 --> 00:50:36,450
the conversation, I think it's
going to be kind of pointless to

826
00:50:36,450 --> 00:50:39,000
speculate as to where it's gonna
go. You gave the example earlier

827
00:50:39,000 --> 00:50:41,340
of social media. And then we
repress that somebody put in the

828
00:50:41,340 --> 00:50:43,710
chat that item about
calculators, like, we never

829
00:50:43,710 --> 00:50:46,230
know, these things, like when
email came out in the 90s. It

830
00:50:46,230 --> 00:50:48,690
wasn't like, you know, what's
going to happen? This, and then

831
00:50:48,690 --> 00:50:51,240
it was like, even remotely
correct, like 10 years later.

832
00:50:51,240 --> 00:50:53,880
And so I think that's what we're
seeing now. So I think like to

833
00:50:53,880 --> 00:50:56,670
Adrienne's point about, like,
how are we going to like hacked

834
00:50:56,670 --> 00:51:00,300
right now, I would consider two
or three things, right? Like one

835
00:51:00,300 --> 00:51:05,250
is that we've got to like stay
vigilantly ahead of where AI is

836
00:51:05,250 --> 00:51:09,120
going and not be, you know,
doing too much punditry or

837
00:51:09,120 --> 00:51:13,620
something. The second thing is
like we have to shift. I don't

838
00:51:13,620 --> 00:51:16,050
know the best way to put this
but like shift the moral urgency

839
00:51:16,380 --> 00:51:20,340
around things like plagiarism to
the problems that we need to

840
00:51:20,340 --> 00:51:24,060
solve. So instead of moral
panics for things, like, you

841
00:51:24,060 --> 00:51:26,730
know, oh my gosh, somebody's
gonna write their essay using

842
00:51:26,730 --> 00:51:29,430
Chechi. Btw, how about we have
the moral panic about a one in

843
00:51:29,430 --> 00:51:32,400
five graduation rate? How about
we have a moral panic about

844
00:51:32,400 --> 00:51:35,220
countries, leapfrogging us in
higher education attainment of

845
00:51:35,220 --> 00:51:38,250
having an innovation gap during
climate change and COVID and

846
00:51:38,250 --> 00:51:41,460
threats to democracy, we need to
be a beacon to the world again.

847
00:51:41,460 --> 00:51:44,100
And that's what Higher Ed was
supposed to be in the post war

848
00:51:44,100 --> 00:51:46,560
era. And that's how they talked
about it. And I think we need to

849
00:51:46,560 --> 00:51:50,190
have that orientation again, and
bring generative AI in that

850
00:51:50,190 --> 00:51:52,680
orientation. And then I guess
the last thing is, this is just

851
00:51:52,680 --> 00:51:56,190
building off of what President
passen said, but like, I think

852
00:51:56,640 --> 00:51:59,760
the lowest hanging fruit is that
bureaucracy is that bloat. You

853
00:51:59,760 --> 00:52:02,850
know, I think that that's even
true with just writing, right? I

854
00:52:02,850 --> 00:52:06,510
mean, you wouldn't trust AI so
much to write you a novel. But

855
00:52:06,540 --> 00:52:09,540
in terms of making some anodyne
policy that gets the job done

856
00:52:09,780 --> 00:52:14,010
like it can we find all the ways
to make life hard for students.

857
00:52:14,010 --> 00:52:16,800
And I'm sure that's true in
health care for patients and

858
00:52:17,280 --> 00:52:20,160
just pick your sector pick the
association here. And I think

859
00:52:20,160 --> 00:52:22,350
you start with the low hanging
fruit and exactly what President

860
00:52:22,350 --> 00:52:25,860
Bastian said about killing the
Bloat killing the bureaucracy,

861
00:52:25,860 --> 00:52:28,440
this could really be a
bureaucracy killer in the best

862
00:52:28,440 --> 00:52:29,070
way possible.

863
00:52:29,880 --> 00:52:33,300
KiKi L'Italien: Oh, gosh, where
is it? The words that speak

864
00:52:33,300 --> 00:52:34,680
right to my heart.

865
00:52:37,590 --> 00:52:40,290
Michael Baston: I would just
add, that we've got to be

866
00:52:40,290 --> 00:52:45,600
careful of false choices. Too
often we decide you can either

867
00:52:45,600 --> 00:52:49,860
be for it or against it. You
know, when the folks that had

868
00:52:49,860 --> 00:52:53,700
radios heard about TVs, I'm sure
they were like, yeah, that TV

869
00:52:53,700 --> 00:52:58,320
thing, it's probably not going
to take off. They got TVs, and

870
00:52:58,320 --> 00:53:03,210
they kept their radios too. So
we have to understand that you

871
00:53:03,210 --> 00:53:06,330
can actually keep two thoughts
in your mind at the same time,

872
00:53:06,450 --> 00:53:10,650
you actually can continue to
advance and be built on the

873
00:53:10,650 --> 00:53:14,400
foundations of really
understanding how the world

874
00:53:14,400 --> 00:53:18,090
works. It's not either or, and
we've got to be able to expand

875
00:53:18,090 --> 00:53:22,140
capacity, we've got to be able
to give people the resources are

876
00:53:22,140 --> 00:53:26,610
particularly under resourced
institutions, the opportunity to

877
00:53:26,640 --> 00:53:31,470
engage and to learn and to grow,
and to get with it, because if

878
00:53:31,470 --> 00:53:35,190
the country is going to move
forward, we we really don't have

879
00:53:35,190 --> 00:53:37,230
any educational institution to
leave.

880
00:53:38,700 --> 00:53:42,510
KiKi L'Italien: Yes, I mean, oh,
my gosh, this is this has been

881
00:53:42,540 --> 00:53:50,940
such a great, great show. I want
to end on a high note, and I see

882
00:53:50,940 --> 00:53:54,660
this great question. Actually,
that comes from Peggy

883
00:53:54,660 --> 00:53:57,780
Lambertson. So if you have the
time, just answer one question,

884
00:53:57,780 --> 00:54:02,220
folks. This is Audrey, do you
think the pool of knowledge

885
00:54:02,220 --> 00:54:06,330
feeding AI is at risk for cyber
attack? There might be pollution

886
00:54:06,330 --> 00:54:10,320
of online knowledge. Okay, so I
get this. This is about talking

887
00:54:10,320 --> 00:54:14,160
about, you know, where is the
data coming from? Can we trust

888
00:54:14,160 --> 00:54:19,110
where it's the pool of knowledge
that that generative AI is is

889
00:54:19,140 --> 00:54:22,530
pulling from for the information
that's giving us?

890
00:54:24,150 --> 00:54:26,340
Audrey Ellis: Sure, I mean, I
think for all just be

891
00:54:26,340 --> 00:54:29,520
transparent. I am no
cybersecurity expert. So take

892
00:54:29,520 --> 00:54:33,840
everything I say with a grain of
salt. But I think that in any

893
00:54:33,870 --> 00:54:40,140
space right now, where we have
documented information that is

894
00:54:40,140 --> 00:54:43,980
feeding into a tool, you said it
yourself very well, garbage in,

895
00:54:43,980 --> 00:54:48,000
garbage out. That's exactly how
I'm thinking about this. And so

896
00:54:48,180 --> 00:54:52,020
really important for us to be
thinking about any the quality

897
00:54:52,290 --> 00:54:57,600
of any data that we're using to
train these systems, and asking

898
00:54:57,600 --> 00:55:01,650
those critical questions as
consumers have, what data are

899
00:55:01,650 --> 00:55:06,510
you using to? You know, give me
this product or service? And

900
00:55:06,510 --> 00:55:10,140
also, what data are you
collecting for me? So really

901
00:55:10,140 --> 00:55:12,900
important question. And I think,
you know, we might get to a

902
00:55:12,900 --> 00:55:16,290
point you might have noticed on
Google Now, you search and

903
00:55:16,290 --> 00:55:21,360
there's like an AI generated sub
summary, before you even you

904
00:55:21,360 --> 00:55:25,020
know, get to the results. And
for all, you know, we're gonna

905
00:55:25,020 --> 00:55:28,380
end up at a time where AI is
training AI is training AI. And

906
00:55:28,380 --> 00:55:30,990
so like, I still think we're
really early in this space. And

907
00:55:30,990 --> 00:55:34,650
we need to think about, again,
how are we shaping it, and also

908
00:55:34,650 --> 00:55:38,730
whose voices stories
perspectives, knowledge is even

909
00:55:38,970 --> 00:55:43,410
included, for better or for
worse, looking at the history of

910
00:55:43,410 --> 00:55:49,950
kind of many disparate use cases
of tech and equity. But there's

911
00:55:49,950 --> 00:55:54,450
a lot of a lot of information
and knowledge to come out here.

912
00:55:54,990 --> 00:55:58,470
And I will just sound the equity
piece. There's a great book,

913
00:55:58,500 --> 00:56:01,980
unmasking AI, if you're
interested in learning more

914
00:56:01,980 --> 00:56:09,120
about equity and AI, especially
in terms of biometric data. So

915
00:56:09,150 --> 00:56:12,420
like facial recognition, things
like that. Highly recommend,

916
00:56:13,470 --> 00:56:15,510
Michael Baston: what once again,
it's all also about due

917
00:56:15,510 --> 00:56:19,800
diligence. So if we tell our
students that AI is like you're

918
00:56:19,950 --> 00:56:23,190
an intern or you know, an
assistant, you get the you can't

919
00:56:23,190 --> 00:56:27,840
turn in to your boss work, you
didn't check. So it is important

920
00:56:27,840 --> 00:56:32,220
for us to make sure that we
encourage everyone that is

921
00:56:32,220 --> 00:56:35,370
utilizing this information to do
the due diligence you got to

922
00:56:35,370 --> 00:56:35,760
check.

923
00:56:37,440 --> 00:56:40,110
KiKi L'Italien: Well, I just
can't thank you enough for

924
00:56:40,140 --> 00:56:44,520
taking the time to talk with us
today. Audrey, thank you. I

925
00:56:44,550 --> 00:56:49,050
first talked, I first spoke with
Audrey, she was able to bring

926
00:56:49,050 --> 00:56:53,580
Charles and bring Michael into
this discussion. So I'm very

927
00:56:53,580 --> 00:56:56,310
thankful for that. I think it
was a really meaningful

928
00:56:56,310 --> 00:57:00,360
conversation. And I hope that
you all will be open to maybe me

929
00:57:00,360 --> 00:57:03,120
reaching out for a follow up
conversation one of these days

930
00:57:03,120 --> 00:57:07,140
to find out as we go along.
What's the story now? What's

931
00:57:07,140 --> 00:57:08,310
happened since then?

932
00:57:08,850 --> 00:57:10,260
Charles Ansell: Absolutely.
We'll have a lot to show

933
00:57:10,260 --> 00:57:10,800
anytime.

934
00:57:12,330 --> 00:57:13,980
KiKi L'Italien: Thank you so
much, everyone. I really

935
00:57:13,980 --> 00:57:17,580
appreciate it. And thanks to all
of you for watching today. What

936
00:57:17,580 --> 00:57:21,540
did you think? Tell me in the
comments. write back to me, tell

937
00:57:21,540 --> 00:57:24,720
me what you thought. What were
your takeaways? What What were

938
00:57:24,720 --> 00:57:27,630
the biggest pieces that stood
out to you that said, Ah, I

939
00:57:27,630 --> 00:57:30,090
never thought about it that way.
But now I'm going to do

940
00:57:30,090 --> 00:57:33,450
something different because of
this. Whether it's something

941
00:57:33,450 --> 00:57:36,570
that's going to help you to
explore further ask more

942
00:57:36,570 --> 00:57:40,470
questions. I hope that you will
stay curious and keep asking

943
00:57:40,470 --> 00:57:44,580
questions every day, especially
when you're scared because as

944
00:57:44,580 --> 00:57:49,980
Joseph Campbell once said, The
cave you fear to enter holds the

945
00:57:49,980 --> 00:57:55,110
treasure you seek. Have a great
rest of the week, everyone