Latest / Azadeh and Guy on AI / Rejecting AI Binaries with Dr. Rachel Horst

Transcript

Rejecting AI Binaries with Dr. Rachel Horst · Azadeh and Guy on AI

• 0:00
  SPEAKER_01: Hello, everyone.
• 0:01
  SPEAKER_01: So this is another episode of Azadeh.
• 0:04
  SPEAKER_00: And Guy.
• 0:05
  SPEAKER_01: On AI.
• 0:07
  SPEAKER_01: In this episode, we have a very special guest, Dr. Rachel Horst.
• 0:12
  SPEAKER_01: And oh, and we need to also introduce ourselves.
• 0:16
  SPEAKER_01: So I'm Azadeh Hassani.
• 0:18
  SPEAKER_01: I'm a first year PhD student at the University of Nebraska-Lincoln.
• 0:22
  SPEAKER_01: And my major is Innovative Learning Technologies.
• 0:27
  SPEAKER_00: And I'm Guy Trenin.
• 0:29
  I'm a professor here at the University of Nebraska-Lincoln,
• 0:32
  Innovative Learning Technologies.
• 0:33
  SPEAKER_00: And Rachel, this is your opportunity.
• 0:36
  SPEAKER_02: Wonderful.
• 0:36
  SPEAKER_02: Well,
• 0:37
  I'm Rachel Horst and I'm a lecturer in the Master of
• 0:42
  Educational Technology program
• 0:43
  SPEAKER_02: at UBC at the University of British Columbia.
• 0:47
  SPEAKER_00: Yes.
• 0:48
  And we are all envious right now because it's super hot here in
• 0:51
  Nebraska suddenly.
• 0:53
  SPEAKER_02: Oh, is it?
• 0:53
  SPEAKER_02: Yeah, it's quite lovely here, as it often is when it's not raining.
• 1:02
  SPEAKER_00: So we went from winter to summer with no break in the middle.
• 1:08
  SPEAKER_02: Oh, yeah, that's hard.
• 1:10
  SPEAKER_02: That's a whiplash.
• 1:14
  SPEAKER_00: But again, that is a Midwestern, the Great Plains area.
• 1:19
  transitional seasons tend to be like that it because it can be
• 1:23
  also awfully cold
• 1:24
  next week and that's just part of the course so right here we
• 1:29
  go all right um as i
• 1:31
  SPEAKER_00: did you want to start
• 1:33
  SPEAKER_01: Sure.
• 1:34
  SPEAKER_01: So Dr.
• 1:36
  SPEAKER_01: Horst,
• 1:36
  SPEAKER_01: I mean,
• 1:37
  SPEAKER_01: her research,
• 1:38
  SPEAKER_01: teaching and creative digital praxis,
• 1:40
  examine meaning making and communicative practices in an era of
• 1:45
  generative AI and
• 1:47
  SPEAKER_01: digital saturation.
• 1:48
  SPEAKER_01: We also know that you have developed an AI literacy hub.
• 1:54
  which is an open educational research or resource for educators
• 1:59
  and also practitioners.
• 2:01
  Do you want to talk more about that or give us more information
• 2:04
  about yourself,
• 2:05
  your research?
• 2:08
  SPEAKER_01: Yeah,
• 2:09
  SPEAKER_02: so I developed that hub,
• 2:12
  which we are in this process of responding to and moving with
• 2:19
  our changing times.
• 2:20
  So that hub is a great place to go and find some really
• 2:26
  sustainable practices with
• 2:28
  SPEAKER_02: using and thinking about AI.
• 2:31
  SPEAKER_02: It's also the ethos of it is to really guide students through the process
• 2:37
  the wide range of perspectives that they will likely encounter
• 2:42
  in their studies,
• 2:43
  SPEAKER_02: both in MET and beyond and just in the world of education,
• 2:47
  with instructors feeling passionately and differently about
• 2:52
  these tools and how
• 2:53
  SPEAKER_02: we're going to use them in our classes.
• 2:56
  SPEAKER_02: And at that time,
• 2:57
  SPEAKER_02: I mean,
• 2:57
  things have been shifting and more support has been given to
• 3:02
  instructors within
• 3:03
  SPEAKER_02: their institutions on how they're going to articulate the
• 3:06
  the course requirements or these kinds of documents that
• 3:10
  students need to navigate
• 3:12
  SPEAKER_02: on how they're even allowed to think about or disclose their AI use.
• 3:19
  SPEAKER_02: So anyway,
• 3:20
  part of that hub was just talking about the diversity of
• 3:23
  perspectives and that it's
• 3:25
  okay and that we need to navigate this space together and then
• 3:28
  talking about
• 3:30
  different kinds of practices for using different kinds of tools
• 3:34
  and systems.
• 3:36
  SPEAKER_02: So that's still live and it's still out there.
• 3:38
  SPEAKER_02: I think I, but now I'm doing some research at UB.
• 3:46
  SPEAKER_02: UBC Okanagan,
• 3:47
  a really wonderful group of researchers and students have built
• 3:53
  and are maintaining
• 3:55
  SPEAKER_02: this chatbot called Help Me AI.
• 3:58
  SPEAKER_02: And it's a very interesting platform.
• 4:03
  SPEAKER_02: It's like a RAG platform.
• 4:04
  SPEAKER_02: You can upload a library of course texts and it reads your
• 4:09
  canvas environment and i've been looking at how to integrate it
• 4:15
  so we found that if
• 4:17
  you just put it in the online course graduate course
• 4:20
  environment students don't use
• 4:22
  it and they just revert to the big corporate models out in the
• 4:27
  world and they don't
• 4:28
  understand it can't compete with those models obviously but
• 4:32
  it's an open source
• 4:33
  SPEAKER_02: tool it's got humans in the loop it's collaborative and
• 4:38
  But it needs to be integrated into course doings in a rich,
• 4:43
  critical,
• 4:45
  and thoughtful way.
• 4:47
  So I'm building practices for that kind of integration and
• 4:52
  looking at how students
• 4:54
  SPEAKER_02: are responding and engaging with it.
• 4:56
  SPEAKER_02: And it's been quite exciting.
• 4:59
  SPEAKER_02: So that's sort of where my head is at right now in terms of like learning
• 5:04
  SPEAKER_02: proper research.
• 5:06
  SPEAKER_02: But then I also do a whole lot of creative research on my own,
• 5:12
  playing with and developing and writing with through arts-based
• 5:16
  methods and AI.
• 5:19
  SPEAKER_02: So I have multiple strands of research going on.
• 5:23
  SPEAKER_00: So I have a quick follow-up because you brought this up.
• 5:28
  SPEAKER_00: You talked about art-based research and the way it interacts.
• 5:33
  Are you finding a lot of resistance from people that hear that
• 5:38
  you're playing with
• 5:39
  SPEAKER_00: AI in the arts?
• 5:40
  SPEAKER_00: Because I have.
• 5:41
  SPEAKER_02: Okay, so yes.
• 5:43
  SPEAKER_02: I mean, not just in the arts, but also in education.
• 5:47
  SPEAKER_02: as you well know,
• 5:48
  SPEAKER_02: um,
• 5:50
  SPEAKER_02: people,
• 5:50
  this is an existential moment for people and people are feeling
• 5:54
  very rightly so
• 5:56
  SPEAKER_02: for,
• 5:56
  SPEAKER_02: for many good reasons.
• 5:58
  SPEAKER_02: And I am not discounting that.
• 6:00
  SPEAKER_02: And I,
• 6:00
  and I do build refusal space for refusal and critique into my
• 6:05
  practice and my
• 6:06
  SPEAKER_02: pedagogy,
• 6:07
  SPEAKER_02: but I,
• 6:07
  SPEAKER_02: but I just,
• 6:08
  SPEAKER_02: I myself,
• 6:10
  SPEAKER_02: um,
• 6:11
  I do believe that this technology is much more than any
• 6:15
  specific platform.
• 6:17
  ChatGPT isn't the extent of artificial intelligence and what
• 6:21
  this technology is.
• 6:23
  SPEAKER_02: And I think artificial technology,
• 6:25
  SPEAKER_02: sorry,
• 6:26
  artificial intelligence as a term is difficult and problematic
• 6:30
  in all kinds of
• 6:31
  SPEAKER_02: ways.
• 6:32
  So I recently was reading Digital Theory,
• 6:36
  Beatrice Fazzi and a couple of co-authors.
• 6:40
  SPEAKER_02: talking about that,
• 6:42
  about digitality as this sort of ontological condition that any
• 6:47
  one technology or
• 6:49
  SPEAKER_02: platform or whatever is not getting at the root of it.
• 6:55
  And I'm really excited about this space and what these tools
• 7:00
  can do and not do and
• 7:02
  SPEAKER_02: all their problems and all of that,
• 7:04
  SPEAKER_02: but I lean into playing with them
• 7:08
  SPEAKER_02: pushing them, subverting them,
• 7:11
  SPEAKER_02: you know,
• 7:12
  SPEAKER_02: and not and not being contained with one platform,
• 7:15
  SPEAKER_02: but other open source models and all kinds of different things.
• 7:19
  SPEAKER_02: And artists, yes, artists are very worried.
• 7:22
  But there are a lot of artists and I just finished designing a
• 7:25
  course I'm really excited about.
• 7:28
  Each week we look at an artist using the using systems in way
• 7:32
  in radical ways.
• 7:33
  SPEAKER_02: And that's indigenous artists,
• 7:35
  SPEAKER_02: you know,
• 7:35
  SPEAKER_02: looking at data sovereignty and building their own models.
• 7:39
  SPEAKER_02: You know, all kinds of artists just really
• 7:41
  SPEAKER_02: pushing what we can do and think about with these technologies.
• 7:46
  SPEAKER_02: So that's the space I choose to be in.
• 7:50
  SPEAKER_02: I think there's not a lot of voices saying, yes, it's scary.
• 7:54
  SPEAKER_02: Yes, it's horrible.
• 7:55
  SPEAKER_02: But it's also exciting and mind blowing what these systems can do.
• 8:06
  SPEAKER_00: Yeah, that's a great answer.
• 8:08
  And something that I read from your own work is kind of
• 8:13
  rejecting the binaries of,
• 8:15
  SPEAKER_00: yes,
• 8:15
  SPEAKER_00: I'm all in,
• 8:16
  SPEAKER_00: I don't care,
• 8:17
  SPEAKER_00: or no,
• 8:18
  SPEAKER_00: I'm not going to touch it,
• 8:19
  SPEAKER_00: and really finding
• 8:21
  a way kind of weaving through because even the biggest
• 8:25
  proponents that I meet,
• 8:27
  the coders that are like 95 percent of what I do is happening
• 8:31
  by a machine are also
• 8:34
  a little bit like and sometimes it feels uncomfortable and
• 8:37
  sometimes it breaks and
• 8:40
  SPEAKER_00: all of these thoughts.
• 8:42
  SPEAKER_02: It's a both and situation.
• 8:45
  SPEAKER_02: And I was thinking about agency and, you know, and what that means.
• 8:50
  SPEAKER_02: And you had determinism and all that.
• 8:55
  SPEAKER_02: And I think it's a both and situation.
• 8:56
  SPEAKER_02: It's both like, I think it's futile to try and keep it out.
• 9:03
  of you know of education of art of all these and the artists
• 9:08
  are going to take you
• 9:09
  can't tell people you know artists have always used
• 9:12
  technologies in in um wild ways
• 9:16
  that show us ourselves i you know humanity is a troubled um
• 9:21
  project um and so we
• 9:24
  really are in a both and situation um so yeah the the binaries
• 9:28
  are i mean and i but
• 9:30
  SPEAKER_02: i've also said that um that the
• 9:33
  SPEAKER_02: you know, the utopian and dystopian binary, it does make sense.
• 9:42
  SPEAKER_02: Utopia,
• 9:43
  both utopian imaginings and dystopian imaginings are completely
• 9:47
  rational at this
• 9:48
  SPEAKER_02: moment,
• 9:49
  SPEAKER_02: right?
• 9:49
  SPEAKER_02: Like,
• 9:50
  SPEAKER_02: so I don't want to shut down that kind of intensity,
• 9:54
  SPEAKER_02: but I think there's room for all the space in between as well.
• 10:01
  SPEAKER_01: So our next question is,
• 10:04
  we know that your work kind of challenges deterministic
• 10:07
  narratives about AI and the
• 10:09
  SPEAKER_01: future.
• 10:10
  So what do you think educators most often get wrong when they
• 10:14
  talk about AI in education?
• 10:17
  SPEAKER_02: Yeah, so I think it comes back to...
• 10:21
  SPEAKER_02: You know, thinking that what you see is what you get.
• 10:24
  SPEAKER_02: So we have this tool, this is how it works, this is how you use it.
• 10:28
  SPEAKER_02: And,
• 10:28
  SPEAKER_02: you know,
• 10:29
  and most people that aren't mind blown by these tools are
• 10:33
  people that aren't using
• 10:34
  SPEAKER_02: them in a rich way.
• 10:36
  SPEAKER_02: So they'll just put in something they'll get,
• 10:38
  SPEAKER_02: you know,
• 10:39
  SPEAKER_02: they'll get a really biased or crappy answer back and they'll be like,
• 10:42
  SPEAKER_02: it's broken.
• 10:43
  SPEAKER_02: It's I'm not going to use this tool and they'll just walk away.
• 10:46
  And I see this and I'm just like kind of blown away by it
• 10:52
  dominates the conversation.
• 10:54
  Like we can't let,
• 10:56
  you know,
• 10:57
  we talk about AI outputs as if they're a thing.
• 11:00
  SPEAKER_02: But,
• 11:01
  SPEAKER_02: but we,
• 11:02
  SPEAKER_02: we are the ones that bring meaning to to our practices,
• 11:06
  SPEAKER_02: we need to be engaged,
• 11:07
  SPEAKER_02: like,
• 11:07
  SPEAKER_02: whoever thought that you could just put in an answer,
• 11:10
  SPEAKER_02: get out a put in a question,
• 11:12
  SPEAKER_02: get out a text,
• 11:13
  SPEAKER_02: and then that's,
• 11:14
  SPEAKER_02: that's it.
• 11:15
  SPEAKER_02: It's, it doesn't.
• 11:16
  SPEAKER_02: So, so
• 11:18
  what I think that people are getting wrong is that it,
• 11:21
  it all is still work.
• 11:23
  Like anyone who says that this is taking away and making things
• 11:27
  easier is just wrong.
• 11:28
  SPEAKER_02: Like I,
• 11:29
  SPEAKER_02: when I create an image or a diagram or write a paper or do anything,
• 11:33
  SPEAKER_02: it still takes work.
• 11:36
  SPEAKER_02: Um, and you're, you know, you're, it's a back and forth.
• 11:39
  SPEAKER_02: It's a,
• 11:40
  it's a building and a digging and a wrestling and all that
• 11:42
  stuff that we always
• 11:44
  SPEAKER_02: have been doing.
• 11:45
  SPEAKER_02: Um, and now we just have, um,
• 11:48
  um I I feel like it's extended my reach so um so the final
• 11:53
  outcome is much is much
• 11:56
  better than I could have done within that time frame on my own
• 11:59
  um but it's still
• 12:00
  work it's still um all that messy stuff I think that um that
• 12:06
  that is part of um
• 12:07
  SPEAKER_02: expressive um work um
• 12:11
  But I think what people get wrong is that they don't see beyond
• 12:15
  this individual
• 12:17
  SPEAKER_02: system or the individual prompt or the individual arrangement of things,
• 12:21
  SPEAKER_02: and they just take it at face value.
• 12:23
  SPEAKER_02: And I think things get more
• 12:26
  SPEAKER_02: meaningful and interesting when you dig in.
• 12:28
  And then also when you go to the periphery and look at what
• 12:31
  artists are doing and
• 12:32
  SPEAKER_02: look at what,
• 12:33
  SPEAKER_02: you know,
• 12:33
  SPEAKER_02: radicals,
• 12:34
  SPEAKER_02: radicals are doing in these spaces of refusal and,
• 12:37
  SPEAKER_02: and all,
• 12:38
  SPEAKER_02: and all that,
• 12:39
  SPEAKER_02: that richness as well.
• 12:41
  SPEAKER_02: So just accepting an ed tech tool from,
• 12:45
  SPEAKER_02: you know,
• 12:45
  SPEAKER_02: that packages,
• 12:46
  SPEAKER_02: uh,
• 12:46
  SPEAKER_02: an LLM into this,
• 12:48
  SPEAKER_02: you know,
• 12:49
  SPEAKER_02: tidy thing.
• 12:49
  Yes,
• 12:50
  that's that we can all agree is a bad idea for education,
• 12:54
  but that's not what,
• 12:55
  um,
• 12:56
  SPEAKER_02: what I think these systems offer us.
• 13:00
  SPEAKER_00: Yeah.
• 13:01
  SPEAKER_00: Yesterday I was talking to somebody from one of the big companies.
• 13:04
  SPEAKER_00: I will not say that it's Google, but maybe it is.
• 13:09
  SPEAKER_00: And he was saying,
• 13:10
  SPEAKER_00: you know,
• 13:11
  you need to find one model and then stick to it and ask that
• 13:14
  model to do everything
• 13:16
  SPEAKER_00: that you need to do.
• 13:16
  SPEAKER_00: And I'm like,
• 13:18
  SPEAKER_00: I'm not sure.
• 13:19
  SPEAKER_00: I'm not sure I agree here.
• 13:22
  SPEAKER_00: Lovely people,
• 13:23
  SPEAKER_00: I happen to use their model quite often,
• 13:27
  SPEAKER_00: but I use also three or four other things that are maybe LLMs,
• 13:32
  SPEAKER_00: but also small language models and just messing about,
• 13:36
  SPEAKER_00: which I find delightful.
• 13:39
  SPEAKER_02: Yes, agreed.
• 13:41
  SPEAKER_02: Yes, I agree.
• 13:42
  SPEAKER_02: Although I do.
• 13:43
  One thing that I sort of not agree with,
• 13:46
  because I don't think you should just use one model.
• 13:49
  SPEAKER_02: But do you kind of agree if you're going to be if you're interested,
• 13:52
  SPEAKER_02: you should be using the bigger the models themselves and not the packaged
• 13:58
  um you know platforms where um i really think um and and this
• 14:04
  is part of my
• 14:05
  teaching too i think building your own tools with the models is
• 14:09
  where it's at and
• 14:10
  and a lot of people are selling these tools to others to just
• 14:13
  ready as ready-made
• 14:14
  SPEAKER_02: platforms and i think it's that's
• 14:18
  SPEAKER_02: I think we need to build our own tools and that's where we learn.
• 14:22
  And what we learn actually in building these tools is about our
• 14:26
  own capacity to
• 14:27
  SPEAKER_02: learn,
• 14:28
  SPEAKER_02: which is why like I've never been a fan of,
• 14:31
  SPEAKER_02: I mean,
• 14:31
  SPEAKER_02: some people are and it's great,
• 14:33
  SPEAKER_02: but like,
• 14:34
  SPEAKER_02: you know,
• 14:35
  SPEAKER_02: taking on somebody else's methodology or teaching plan or whatever it is,
• 14:38
  SPEAKER_02: it's in the building to come to know what you're doing.
• 14:44
  And so I feel that way with systems and programs and tools as
• 14:48
  well with AI.
• 14:50
  SPEAKER_00: Okay.
• 14:53
  SPEAKER_00: So in that world,
• 14:55
  how do you see teacher agency shifting when the AI systems are
• 15:00
  shaping knowledge,
• 15:01
  SPEAKER_00: assessments?
• 15:03
  SPEAKER_00: And you talked about this, even creativity.
• 15:06
  SPEAKER_02: Mm-hmm.
• 15:07
  I feel,
• 15:08
  I mean,
• 15:09
  agency shifts,
• 15:12
  you know,
• 15:13
  over time,
• 15:14
  over the course of a day.
• 15:15
  SPEAKER_02: So it is shifting,
• 15:16
  SPEAKER_02: but it's not,
• 15:17
  SPEAKER_02: but there is a lot of agency there,
• 15:19
  SPEAKER_02: like we still have agency and,
• 15:22
  SPEAKER_02: and to the extent that
• 15:25
  SPEAKER_02: that AI is changing knowledge practices.
• 15:28
  I mean,
• 15:29
  I think that AI is intensifying an epistemic landscape that has
• 15:35
  been always there.
• 15:37
  SPEAKER_02: I think of AI literature,
• 15:38
  SPEAKER_02: I think AI literacy is making us see what has been there all along,
• 15:43
  SPEAKER_02: all the fissures,
• 15:44
  SPEAKER_02: all the possibilities,
• 15:46
  SPEAKER_02: all the broken parts of our systems.
• 15:49
  SPEAKER_02: It's just amplifying everything.
• 15:52
  SPEAKER_02: And so I think we have agency, we need to take on our agency.
• 15:59
  SPEAKER_02: It's very easy to just
• 16:01
  SPEAKER_02: move with the tides and have things forced into the classroom.
• 16:06
  But teachers have always resisted that and done good work to
• 16:13
  protect that space.
• 16:15
  SPEAKER_02: So I think we need to be thinking about why we do things in classrooms,
• 16:20
  SPEAKER_02: what we ask students to do,
• 16:22
  SPEAKER_02: make them care to do these things and understand and feel.
• 16:27
  SPEAKER_02: I think we need to also feel what it feels like
• 16:31
  SPEAKER_02: to use these systems and talk about it.
• 16:33
  SPEAKER_02: And in my experience, I am working with graduate students.
• 16:36
  SPEAKER_02: So,
• 16:38
  SPEAKER_02: so,
• 16:38
  SPEAKER_02: so it could,
• 16:39
  SPEAKER_02: it is a different place than like K to 12 or something,
• 16:43
  SPEAKER_02: but even my son who's 12,
• 16:45
  SPEAKER_02: he knows when an idea is his and when it's his words or,
• 16:49
  SPEAKER_02: and when it's not,
• 16:50
  SPEAKER_02: he feels that tension just as graduate students do.
• 16:53
  SPEAKER_02: And I, and I really feel that we want to express ourselves and, um,
• 16:58
  And so I think what we need to do is pull back on the hardcore
• 17:03
  judgment that we're
• 17:04
  SPEAKER_02: bringing into these spaces and let people experiment.
• 17:08
  SPEAKER_02: It's also very new still and we don't know how to do it.
• 17:13
  So I think we need to allow for all of us to feel and
• 17:18
  experiment and do without
• 17:19
  SPEAKER_02: punishing through,
• 17:21
  SPEAKER_02: you know,
• 17:21
  SPEAKER_02: grading practices and stuff like that.
• 17:23
  SPEAKER_02: And that also goes back to, you know,
• 17:28
  SPEAKER_02: When we ask students to not use the LLM, why are we doing that?
• 17:32
  SPEAKER_02: And then not to punish if it's not perfect.
• 17:35
  SPEAKER_02: You know, there's there's there's all kinds of good reasons to.
• 17:38
  SPEAKER_02: To.
• 17:41
  SPEAKER_02: you know, cheat, you know, for when it's a mark, for the mark, you know.
• 17:48
  So,
• 17:49
  yeah,
• 17:49
  so I don't see things as radically different than they've been
• 17:54
  before.
• 17:55
  But what it is doing is it's forcing us all to ask ourselves
• 18:00
  questions we should
• 18:01
  SPEAKER_02: have been asking all along about assessment,
• 18:04
  SPEAKER_02: about work,
• 18:04
  SPEAKER_02: about,
• 18:05
  SPEAKER_02: you know,
• 18:05
  SPEAKER_02: about meritocracy and all of these things.
• 18:09
  SPEAKER_00: Absolutely.
• 18:10
  SPEAKER_00: And just recently, I have an open policy about AI in my classroom.
• 18:17
  SPEAKER_00: We've been playing and observing what happens and all of that.
• 18:20
  SPEAKER_00: And it's very clear to my students that they can if they want to use.
• 18:24
  SPEAKER_00: There's a disclosure and they have to also say if they didn't use it.
• 18:27
  SPEAKER_00: And I just got,
• 18:29
  SPEAKER_00: about a week ago,
• 18:30
  I got a set of papers about a topic and I had a few that wrote
• 18:35
  and you could almost
• 18:36
  SPEAKER_00: sense the
• 18:38
  SPEAKER_00: their resistance.
• 18:41
  SPEAKER_00: It's like, I did not use AI in this assignment.
• 18:45
  And all I was thinking quietly is maybe you could have used a
• 18:49
  little bit of AI in
• 18:50
  this case because that would have pushed your thinking just a
• 18:54
  little bit further.
• 18:55
  SPEAKER_02: Yes.
• 18:56
  SPEAKER_02: I have I designed and I published it on LinkedIn.
• 18:59
  It's it's a tree and it has all these different apples and it's
• 19:04
  all it's the the
• 19:05
  SPEAKER_02: the the I call it the orchard of mediation.
• 19:09
  SPEAKER_02: And,
• 19:10
  SPEAKER_02: you know,
• 19:10
  SPEAKER_02: yeah,
• 19:11
  SPEAKER_02: like a fully,
• 19:12
  SPEAKER_02: you know,
• 19:12
  SPEAKER_02: passionate,
• 19:14
  SPEAKER_02: voiceful,
• 19:15
  SPEAKER_02: you know,
• 19:15
  SPEAKER_02: expression of,
• 19:17
  SPEAKER_02: you know,
• 19:17
  SPEAKER_02: bias and rage or whatever like that can be amazing.
• 19:21
  SPEAKER_02: But,
• 19:21
  SPEAKER_02: you know,
• 19:22
  SPEAKER_02: you could temper that with a little,
• 19:23
  SPEAKER_02: you know,
• 19:24
  SPEAKER_02: balance and rigor,
• 19:25
  SPEAKER_02: right,
• 19:26
  SPEAKER_02: by using the system.
• 19:27
  SPEAKER_02: And similarly, you know, and I just give my students this as feedback.
• 19:35
  SPEAKER_02: I say to students,
• 19:37
  SPEAKER_02: and I also get the glossy,
• 19:39
  SPEAKER_02: like,
• 19:39
  SPEAKER_02: I can't even see this thing because it's the reflections.
• 19:42
  SPEAKER_02: It's so glossy and slick, right?
• 19:44
  SPEAKER_02: It's nothing's here, right?
• 19:47
  SPEAKER_02: And I just offer that as feedback.
• 19:50
  SPEAKER_02: And I don't, this is not about punishing students for using the LLM.
• 19:57
  SPEAKER_02: It's that they're just not using it very well.
• 19:59
  SPEAKER_02: And that like, that's okay.
• 20:01
  SPEAKER_02: We can learn.
• 20:04
  And similarly,
• 20:05
  when not using it,
• 20:07
  you know,
• 20:09
  there's a kind of a measure of professionalism.
• 20:12
  SPEAKER_02: I say that, and at the same time,
• 20:16
  SPEAKER_02: That's the voice and mess of not using LLMs is kind of precious now.
• 20:23
  SPEAKER_02: So sometimes when I see that, I just like it.
• 20:28
  SPEAKER_02: I accept more of that than I would have before, I think.
• 20:32
  SPEAKER_02: But yeah, it's a very interesting time.
• 20:38
  SPEAKER_01: Okay.
• 20:39
  So our last question is you talk about how AI unsettles
• 20:43
  traditional notions of
• 20:45
  SPEAKER_01: authority and expertise.
• 20:47
  SPEAKER_01: So what does that look like in real classrooms?
• 20:51
  SPEAKER_02: Well,
• 20:52
  SPEAKER_02: I mean,
• 20:53
  SPEAKER_02: just right off the top of my head,
• 20:54
  SPEAKER_02: I'm thinking I just recently built a drawing platform.
• 20:59
  SPEAKER_02: Like I have no business making these kinds of tools.
• 21:02
  SPEAKER_02: I don't know what I'm doing and yet it works.
• 21:05
  SPEAKER_02: So, so, um,
• 21:07
  SPEAKER_02: LLMs are bringing, bringing coding and, and building and
• 21:13
  into the hands of students and and that's amazing like the the
• 21:18
  possibilities of
• 21:19
  that are truly astounding and also just making like like I was
• 21:25
  saying it's making
• 21:26
  our reach um quite wider um so that that's impacting the class
• 21:33
  in all kinds of ways
• 21:35
  SPEAKER_02: however I do think that it also
• 21:37
  SPEAKER_02: It's not diminishing the role of expertise and some kind of authority.
• 21:44
  Because we need to be engaging with these outputs and these
• 21:50
  texts and these reams
• 21:51
  SPEAKER_02: of images and diagrams and all this stuff that we can just,
• 21:56
  SPEAKER_02: you know,
• 21:57
  just make with one press of a button,
• 22:02
  we need to be able to navigate all that.
• 22:05
  And I do think that there's a measure of expertise and
• 22:08
  authority that we bring as
• 22:11
  SPEAKER_02: situated people with histories to the table.
• 22:15
  SPEAKER_02: So yeah, I don't know if that answers your question.
• 22:21
  SPEAKER_02: I think it both opens the field of play,
• 22:25
  but also that we need to continue to develop expertise and
• 22:30
  knowledge about the
• 22:32
  SPEAKER_02: world and not just hope that the LLM will say the right thing for us.
• 22:39
  SPEAKER_00: All right.
• 22:40
  So we could go on for hours because this is interesting and
• 22:47
  we're playing and we're making.
• 22:48
  I don't think there's been a period that I've made this many
• 22:51
  things in this shorter
• 22:53
  SPEAKER_00: time in my life.
• 22:54
  SPEAKER_00: So there's a joy in it,
• 22:56
  but also once in a while I wake up in the morning and think
• 22:59
  about my kids and
• 23:00
  whether they're going to have a job and all of these other
• 23:05
  things that are hanging
• 23:06
  SPEAKER_00: over us.
• 23:07
  SPEAKER_00: So I think that that's fantastic.
• 23:09
  SPEAKER_00: But I wanted to say thank you for coming on.
• 23:12
  SPEAKER_02: My absolute pleasure.
• 23:14
  SPEAKER_02: It's lovely to meet you both.
• 23:16
  SPEAKER_00: Thank you.
• 23:17
  SPEAKER_00: It was awesome.
• 23:19
  SPEAKER_00: Yes.
• 23:20
  SPEAKER_00: And we'll see you next time on Azadeh.
• 23:23
  SPEAKER_02: And Guy.
• 23:24
  SPEAKER_00: On AI.
• 23:26
  SPEAKER_02: Wonderful.
• 23:26
  SPEAKER_02: Thank you.