Latest / Azadeh and Guy on AI / Beyond the Right Answer

Transcript

Beyond the Right Answer · Azadeh and Guy on AI

• 0:01
  SPEAKER_02: Hi, and welcome to another episode of Azadeh
• 0:04
  SPEAKER_02: and Guy on AI.
• 0:06
  SPEAKER_02: And today we have with us Dr. Chen.
• 0:10
  SPEAKER_00: I'm going to briefly introduce Dr. Chen.
• 0:12
  SPEAKER_00: Dr.
• 0:13
  SPEAKER_00: Chen received a Bachelor in Engineering in Electrical Engineering and Mathematics
• 0:18
  SPEAKER_00: from McGill University and an MA and PhD in Electrical and Computer Engineering from
• 0:25
  SPEAKER_00: the University of Toronto.
• 0:26
  SPEAKER_00: He is a professor in the Department of Electrical and Computer Engineering
• 0:30
  SPEAKER_00: at McGill University and is the academic lead of the Enhancing Learning and Teaching
• 0:35
  SPEAKER_00: in Engineering.
• 0:38
  SPEAKER_00: His research interests include faculty development and the intersections between
• 0:44
  SPEAKER_00: the teaching and learning environments,
• 0:47
  SPEAKER_00: identity and equity, diversity, inclusivity, and accessibility.
• 0:51
  SPEAKER_00: We are really excited to have Dr.
• 0:54
  SPEAKER_00: Chen join us today to share his insights on how AI is shaping teaching and learning
• 0:59
  SPEAKER_00: in higher education and specifically in engineering education.
• 1:03
  SPEAKER_00: Welcome to Azadeh and Guy on AI, Dr. Chen.
• 1:07
  SPEAKER_01: Thank you very much and thank you for having me.
• 1:08
  SPEAKER_01: I'm really happy and excited to be here this afternoon.
• 1:12
  SPEAKER_02: All right, because I thought you said you were anxious, but okay.
• 1:16
  SPEAKER_01: I'm also anxious, but yes,
• 1:18
  SPEAKER_02: happy and
• 1:19
  SPEAKER_01: excited.
• 1:20
  SPEAKER_02: We promise that we are friendly.
• 1:24
  SPEAKER_02: So really,
• 1:25
  SPEAKER_02: I want to start us off kind of Thinking about what are the challenges
• 1:32
  SPEAKER_02: that you see faculty and students face on bringing AI into engineering learning?
• 1:40
  SPEAKER_02: That's an open-ended question, so you can attack it from wherever you want.
• 1:45
  SPEAKER_01: Sure.
• 1:46
  SPEAKER_01: And again, I'm not an expert.
• 1:49
  SPEAKER_01: I provide my opinion just having taught for the last couple of years
• 1:54
  SPEAKER_01: in this new era, if you wish.
• 1:58
  SPEAKER_01: But what I found that has been really helpful has been to join a lot of communities,
• 2:02
  SPEAKER_01: communities of practice,
• 2:03
  SPEAKER_01: to surround yourself by others who share an equal concern and are sort
• 2:08
  SPEAKER_01: of navigating this unknown altogether.
• 2:12
  SPEAKER_01: In terms of the challenges that faculty and students are going to face,
• 2:16
  SPEAKER_01: I think it's a little bit like We lived through a major change in teaching
• 2:22
  SPEAKER_01: and learning when we had the pandemic with COVID and everyone sort of had
• 2:25
  SPEAKER_01: to transition to remote teaching and learning in a relatively short period of time.
• 2:29
  SPEAKER_01: And I think we're living another one of these moments.
• 2:33
  SPEAKER_01: And I think the biggest challenge that everyone faces is really how do we adapt,
• 2:37
  SPEAKER_01: but adapt in a thoughtful manner in face of the AI.
• 2:41
  SPEAKER_01: I know some of us would rather wish that none of this had ever come up.
• 2:45
  SPEAKER_01: But I don't think we can pretend that nothing has changed.
• 2:50
  SPEAKER_01: Things have changed, and I think it's how are we going to adapt.
• 2:53
  SPEAKER_01: So I think education is really in a period of some form of transition.
• 3:00
  SPEAKER_01: AI will certainly change both teaching and learning.
• 3:03
  SPEAKER_01: I think some effects are already visible,
• 3:05
  SPEAKER_01: while others probably won't know the impact for a couple more years.
• 3:09
  SPEAKER_01: And like I said, I think it's something that we saw something similar during COVID.
• 3:13
  SPEAKER_01: There was the immediate things that we can see,
• 3:16
  SPEAKER_01: but how this impacted student learning,
• 3:19
  SPEAKER_01: we didn't see it for a couple of years, right?
• 3:21
  SPEAKER_01: Including the students who are sort of in high school and
• 3:24
  SPEAKER_01: then eventually transitioning their way into university, into higher education.
• 3:29
  SPEAKER_01: And I think some of the differences, though,
• 3:33
  SPEAKER_01: between AI and sort of the COVID pandemic is that while COVID sort of thrust us
• 3:39
  SPEAKER_01: into new
• 3:42
  SPEAKER_01: It's a new way of perhaps teaching and engaging students in an online manner.
• 3:47
  SPEAKER_01: I had a sense that at least certainly in my institution,
• 3:50
  SPEAKER_01: a lot of my colleagues,
• 3:50
  SPEAKER_01: they sort of reverted back to their familiar practices when sort
• 3:55
  SPEAKER_01: the pandemic worked its way through.
• 3:57
  SPEAKER_01: So the first couple of years, And so the challenge is not...
• 4:24
  SPEAKER_01: Whether or not to respond, but how to respond in a valuable and a meaningful way.
• 4:29
  SPEAKER_01: And so I think the students have to learn with AI.
• 4:33
  SPEAKER_01: They have to learn their fundamentals, but they also have to learn about AI literacy.
• 4:38
  SPEAKER_01: And I think some students Even those that I speak to and teach are dead fast against AI.
• 4:44
  SPEAKER_01: They just know we want to learn all of this on our own without resorting to AI.
• 4:47
  SPEAKER_01: And then there are a couple of students who are just going to jump and use AI sort
• 4:50
  SPEAKER_01: of blindly and at will.
• 4:52
  SPEAKER_01: And so I think students will have to learn how, you know, they have to learn their fundamentals.
• 4:55
  SPEAKER_01: They have to learn the key concepts that you can't get around that.
• 5:00
  SPEAKER_01: But at the same time, they're going to be using AI at work.
• 5:05
  SPEAKER_01: And so I think they need to learn how to use it responsibly.
• 5:10
  SPEAKER_01: And so they need that fundamentals part to be able
• 5:12
  SPEAKER_01: to challenge whatever AI might be generating for them when they're going
• 5:15
  SPEAKER_01: to be using it at work.
• 5:17
  SPEAKER_01: So, like I said,
• 5:19
  SPEAKER_01: I think students need to learn engineering,
• 5:20
  SPEAKER_01: but they also need to learn how,
• 5:22
  SPEAKER_01: I guess, to engineer in a world where AI is just going to be everywhere.
• 5:27
  SPEAKER_01: And so I think adaptability is going to be one of these big skills that
• 5:30
  SPEAKER_01: the students will need to be able to cope with.
• 5:33
  SPEAKER_01: But I think that that's something that the faculty will also need to be able to do.
• 5:36
  SPEAKER_01: I don't think that they're going to be able to go back to teaching and
• 5:38
  SPEAKER_01: to assessing students in the same ways that they were prior to the appearance
• 5:42
  SPEAKER_01: of Gen AI.
• 5:45
  SPEAKER_01: And that might mean,
• 5:46
  SPEAKER_01: you know,
• 5:46
  SPEAKER_01: faculty need to rethink things like learning outcomes,
• 5:50
  SPEAKER_01: their learning objectives for their course.
• 5:55
  SPEAKER_01: I don't think they need to abandon their assessment practices either,
• 5:58
  SPEAKER_01: but maybe just think a little bit differently in terms of where the focus should be.
• 6:01
  SPEAKER_01: Is the focus really on the final product,
• 6:03
  SPEAKER_01: the paper that they're going to turn in,
• 6:05
  SPEAKER_01: the assignment they're going to turn in,
• 6:06
  SPEAKER_01: the project,
• 6:06
  SPEAKER_01: the lab report, or is it going to be on the process to get to that final part?
• 6:11
  SPEAKER_01: And I think that that's something that faculty will need to look at.
• 6:17
  SPEAKER_01: And I think the biggest challenge is
• 6:20
  SPEAKER_01: I think there's a lot of uncertainty still.
• 6:22
  SPEAKER_01: I don't know where things are going to head.
• 6:24
  SPEAKER_01: You know,
• 6:25
  SPEAKER_01: I sort of joke and I say,
• 6:26
  SPEAKER_01: well, by the time I retire, higher education won't look like it does today.
• 6:31
  SPEAKER_01: But it, you know,
• 6:32
  SPEAKER_01: I'm not retiring for a number of years,
• 6:33
  SPEAKER_01: but at the same time, you know, my projection might come true in two years, right?
• 6:38
  SPEAKER_01: We don't know.
• 6:39
  SPEAKER_01: So I think we have to recognize that for both students and for faculty,
• 6:42
  SPEAKER_01: we don't really know where things are going to be going.
• 6:45
  SPEAKER_01: And I think that that adds to this need for we have to be ready to adapt as necessary.
• 6:51
  SPEAKER_02: Yeah, I love the comparison to the pandemic because it's still,
• 6:55
  SPEAKER_02: for some of us, it's still visceral, right?
• 6:57
  SPEAKER_02: And we still have overhang with students who,
• 7:01
  SPEAKER_02: as you said, were in middle school or in high school during that period.
• 7:05
  SPEAKER_02: In some places, it lasted longer.
• 7:07
  SPEAKER_02: In some places, it lasted shorter.
• 7:10
  SPEAKER_02: But there was definitely disruption.
• 7:12
  SPEAKER_02: And you're right that the difference is That this is not going away,
• 7:17
  SPEAKER_02: although I have a sense that some of our colleagues are hoping
• 7:20
  SPEAKER_02: that they will wake up one morning and it's gone and then they can go back.
• 7:25
  SPEAKER_02: And I think we are all in agreement.
• 7:28
  SPEAKER_02: It's not going back.
• 7:29
  SPEAKER_02: There's no way to put that back in the ball.
• 7:31
  SPEAKER_02: Whether anybody should have opened that ball is out of our hands, but we're facing that.
• 7:39
  SPEAKER_02: So thank you for that.
• 7:41
  SPEAKER_02: Yes, that is a real challenge.
• 7:44
  SPEAKER_00: And I think we have seen a lot of that,
• 7:46
  SPEAKER_00: a lot of what Dr. Chen just said in our React-E project.
• 7:50
  SPEAKER_00: I mean, so we didn't even talk about that.
• 7:53
  SPEAKER_00: We started collaborating with Dr.
• 7:55
  SPEAKER_00: Chen via the project that we started with React-E.
• 7:58
  SPEAKER_00: Do you want to talk about it a little bit?
• 8:01
  SPEAKER_02: Well, sure, we can bring that up.
• 8:04
  SPEAKER_02: So React-E is our collaboration with now seven institutions of higher education.
• 8:11
  SPEAKER_02: All of them are doing work around STEM and especially engineering,
• 8:15
  SPEAKER_02: and we're really trying to see what's happening.
• 8:17
  SPEAKER_02: We started with a window into our own institution,
• 8:20
  SPEAKER_02: and that's interesting,
• 8:21
  SPEAKER_02: but it's really interesting to see it across institutions in different places,
• 8:27
  SPEAKER_02: And with different students.
• 8:31
  SPEAKER_02: And so the challenge that we have is really learning very quickly
• 8:37
  SPEAKER_02: because I had somebody come to my office
• 8:42
  SPEAKER_02: I didn't ask him, so I'm not sure I'm going to share his name.
• 8:47
  SPEAKER_02: But anyway, he came to my office and said, how are you all?
• 8:50
  SPEAKER_02: He's coming from the field and he's feeling viscerally the change in AI.
• 8:54
  SPEAKER_02: And he's like, how do you even do research in this world that is changing so fast?
• 8:59
  SPEAKER_02: And I'm like, we're still figuring it out.
• 9:02
  SPEAKER_02: Just like the teaching part, this is moving very fast.
• 9:06
  SPEAKER_02: And the idea of forming that connection
• 9:08
  SPEAKER_02: across universities is we can pull all our thinking together,
• 9:12
  SPEAKER_02: but also try to get a picture of what's happening on a cycle that's faster
• 9:17
  SPEAKER_02: than we used to.
• 9:18
  SPEAKER_02: So usually research takes three to five years from idea to a publication, sometimes even more.
• 9:26
  SPEAKER_02: We need to be shorter somehow.
• 9:28
  SPEAKER_02: I'm not sure what the how is,
• 9:30
  SPEAKER_02: but I'm sure that, I mean, we are trying to find different patterns.
• 9:34
  SPEAKER_02: Some of it is,
• 9:35
  SPEAKER_02: and engineering has actually been very good with that,
• 9:38
  SPEAKER_02: is publishing through conferences,
• 9:40
  SPEAKER_02: work in progress,
• 9:41
  SPEAKER_02: papers, things like that, gives people at least a hint of what's happening.
• 9:45
  SPEAKER_00: Mm-hmm.
• 9:46
  SPEAKER_02: But that's not enough.
• 9:48
  SPEAKER_02: So we need to keep thinking about what does it look like when we're trying to do that.
• 9:52
  SPEAKER_02: So that's Reacte.unl.edu.
• 9:55
  SPEAKER_02: Exactly, yes.
• 10:00
  SPEAKER_02: All right.
• 10:00
  SPEAKER_00: So my question is about undergraduate engineering students.
• 10:04
  SPEAKER_00: So much of your recent work, Dr. Chen,
• 10:07
  SPEAKER_00: is on how undergraduate engineering students perceive failure and how
• 10:11
  SPEAKER_00: to reduce stigma.
• 10:13
  SPEAKER_00: AI can help students avoid failing, for example, by catching errors, by smoothing out debugging.
• 10:20
  SPEAKER_00: And all sorts of that things.
• 10:22
  SPEAKER_00: So does that undercut the productive struggle that builds engineers
• 10:27
  SPEAKER_00: or can AI be designed to make failure safer and more instructive?
• 10:33
  SPEAKER_01: It's a great question.
• 10:36
  SPEAKER_01: I don't think AI alone,
• 10:38
  SPEAKER_01: you know, the presence of AI is going to eliminate the productive struggle.
• 10:43
  SPEAKER_01: I think it's more not its presence, but how it's used.
• 10:49
  SPEAKER_01: So if the students use AI to challenge their understanding,
• 10:55
  SPEAKER_01: to validate or verify what they know and what they don't know,
• 11:00
  SPEAKER_01: then I think it could be useful.
• 11:02
  SPEAKER_01: But if the students are using it just to provide answers
• 11:05
  SPEAKER_01: and they're not questioning anything,
• 11:07
  SPEAKER_01: so they've got problems to solve and they entered into whatever platform they want
• 11:11
  SPEAKER_01: to use and they get an answer and they're not going back and checking anything,
• 11:15
  SPEAKER_01: that's when they're not having that productive struggle.
• 11:18
  SPEAKER_01: But if they're not sure about a topic and then they ask for some clarification
• 11:23
  SPEAKER_01: with AI and they get into an argument,
• 11:25
  SPEAKER_01: for instance,
• 11:26
  SPEAKER_01: with AI,
• 11:26
  SPEAKER_01: just like they can engage in a discussion with their TAs or with the instructor,
• 11:31
  SPEAKER_01: then I think they're going through that process of productive struggle.
• 11:35
  SPEAKER_01: So just the presence of the AI tool alone, I don't think will eliminate the productive struggle.
• 11:41
  SPEAKER_01: And if it can help with increasing confidence for the students,
• 11:45
  SPEAKER_01: I think that that's a really useful I think that that's really useful.
• 11:51
  SPEAKER_01: You know,
• 11:51
  SPEAKER_01: in my own research on failure,
• 11:53
  SPEAKER_01: I think the biggest issue has always been,
• 11:56
  SPEAKER_01: you know,
• 11:57
  SPEAKER_01: there's this one big type of assessment and,
• 12:01
  SPEAKER_01: you know,
• 12:01
  SPEAKER_01: it could be a summative type of assessment,
• 12:03
  SPEAKER_01: whether it's a paper or a final exam and whatnot.
• 12:07
  SPEAKER_01: And it's really, you know, there's a lot of pressure and the stakes are very, very high on it.
• 12:11
  SPEAKER_01: And the students find or the students have, you know, highlighted that, look,
• 12:17
  SPEAKER_01: If we fail that particular assessment, that might only capture one part of our learning.
• 12:24
  SPEAKER_01: And it doesn't capture everything else.
• 12:26
  SPEAKER_01: And it's the same thing.
• 12:27
  SPEAKER_01: I come back to,
• 12:28
  SPEAKER_01: I think I made that point earlier,
• 12:30
  SPEAKER_01: it's the final product is not necessarily the thing that we should all be assessing
• 12:34
  SPEAKER_01: at the end of the day.
• 12:35
  SPEAKER_01: It might be the whole process that took you there.
• 12:37
  SPEAKER_01: And so AI can produce a final product,
• 12:41
  SPEAKER_01: but hopefully we're not only just measuring that final product.
• 12:44
  SPEAKER_01: We're also taking a look at what goes into that final product and the whole process
• 12:48
  SPEAKER_01: that went into the final product.
• 12:50
  SPEAKER_01: So I think overall,
• 12:52
  SPEAKER_01: AI can be useful for the students and can help them engage with productive struggle.
• 12:58
  SPEAKER_01: And it might help increase their confidence in preparing
• 13:03
  SPEAKER_01: for traditional summative types of assessments.
• 13:09
  SPEAKER_02: A thing that I've noticed in teaching,
• 13:13
  SPEAKER_02: and again this goes wider to your work on productive struggle and the ability
• 13:17
  SPEAKER_02: to deal with failure and all of that,
• 13:20
  SPEAKER_02: and that is there's often a need to help Students who may be,
• 13:29
  SPEAKER_02: because many of our students have been fantastic throughout their schooling,
• 13:34
  SPEAKER_02: so they're not used to failure.
• 13:35
  SPEAKER_02: They were the students who did really,
• 13:37
  SPEAKER_02: really well throughout their schooling and suddenly they're facing this situation
• 13:41
  SPEAKER_02: where for the first time in their lives,
• 13:43
  SPEAKER_02: they're like, I'm not sure I know the answer.
• 13:45
  SPEAKER_02: This is really difficult.
• 13:47
  SPEAKER_02: And helping them process those emotions and find a way to find strength in that
• 13:54
  SPEAKER_02: and not just a sense of, I'm no good, I need to drop out.
• 13:58
  SPEAKER_02: And I'm always worried about the students who are most vulnerable,
• 14:01
  SPEAKER_02: who are most likely to get overwhelmed with the sense of,
• 14:07
  SPEAKER_02: Of a failure and suddenly disappear on us when we actually want them to stay.
• 14:13
  SPEAKER_02: We say, you know, we've all failed many, many times.
• 14:16
  SPEAKER_02: I'm still failing every day.
• 14:18
  SPEAKER_02: And what I worry about,
• 14:21
  SPEAKER_02: and this just came from the conversation you were having,
• 14:24
  SPEAKER_02: but I've had these conversations many times,
• 14:26
  SPEAKER_02: for example, about basic chemistry classes.
• 14:29
  SPEAKER_02: And that is that if the summative assessment is happening at the end,
• 14:34
  SPEAKER_02: There's no one for them to work out these sense of failure and what to do about it.
• 14:41
  SPEAKER_02: Because class is over and you go on winter break or summer break or whatever it is,
• 14:46
  SPEAKER_02: like everybody disappears,
• 14:48
  SPEAKER_02: including the instructors,
• 14:49
  SPEAKER_02: including anybody that can help them process through what does it feel like.
• 14:54
  SPEAKER_02: And so that samity, if it's happening at the end, can be A major stumbling block.
• 15:01
  SPEAKER_02: I haven't thought about it quite that way in a while, but I'm a little bit worried there.
• 15:07
  SPEAKER_01: I think overall,
• 15:08
  SPEAKER_01: at least our institution has...
• 15:13
  SPEAKER_01: Are
• 15:18
  SPEAKER_01: the support and resources and mechanisms and processes and procedures
• 15:23
  SPEAKER_01: and policies all there to help the students along or not?
• 15:28
  SPEAKER_01: You know,
• 15:28
  SPEAKER_01: a lot of the students that I've interviewed and talked to with their experiences
• 15:32
  SPEAKER_01: of failure have all said, well, it's just such a defining moment.
• 15:35
  SPEAKER_01: And if they fail a class,
• 15:36
  SPEAKER_01: it's there and it sort of is there with them for the rest of their lives.
• 15:39
  SPEAKER_01: It's on the transcript.
• 15:40
  SPEAKER_01: There's no erasure of it.
• 15:42
  SPEAKER_01: And, you know,
• 15:43
  SPEAKER_01: there's been some places that have talked about, well,
• 15:45
  SPEAKER_01: maybe you don't replace the grade, but then it doesn't count towards their GPA.
• 15:49
  SPEAKER_01: So they still have it.
• 15:51
  SPEAKER_01: But if they get a better grade,
• 15:52
  SPEAKER_01: well, it doesn't impact, let's say, their grade point average, at least.
• 15:56
  SPEAKER_01: Whereas, you know, some institution will just say, well, you get a failure and that's it, right?
• 15:59
  SPEAKER_01: It appears there and it's going to impact your grade.
• 16:03
  SPEAKER_01: And then you have the other extreme
• 16:04
  SPEAKER_01: where I think you can even just replace your grade by if you take the course again,
• 16:09
  SPEAKER_01: for instance, and you get a better grade.
• 16:10
  SPEAKER_01: So it's sort of erased.
• 16:12
  SPEAKER_01: You know,
• 16:13
  SPEAKER_01: I don't know what's best,
• 16:15
  SPEAKER_01: but I think anything that winds up being punitive or the student's view
• 16:19
  SPEAKER_01: as being punitive won't help, right, with the development.
• 16:21
  SPEAKER_01: And so if you have a policy that you fail the course, well,
• 16:25
  SPEAKER_01: you can retake it,
• 16:27
  SPEAKER_01: but the impact is still there on your GPA,
• 16:30
  SPEAKER_01: that tends to be, for me,
• 16:32
  SPEAKER_01: it's more of a punitive approach than really, okay, well,
• 16:35
  SPEAKER_01: I can really be able to learn From that experience,
• 16:39
  SPEAKER_01: because you wind up being afraid of failure.
• 16:41
  SPEAKER_01: And I think what the students focus on at that point is they don't focus
• 16:44
  SPEAKER_01: on learning, they focus on the grade.
• 16:46
  SPEAKER_02: Yeah.
• 16:47
  SPEAKER_02: And that's where AI suddenly becomes a lot more attractive if you're lost,
• 16:52
  SPEAKER_02: which is not what we want.
• 16:53
  SPEAKER_02: So you teach an introductory engineering class,
• 16:57
  SPEAKER_02: which I'm sure your colleagues are very happy because in many,
• 17:00
  SPEAKER_02: many places, people don't like teaching the introductory course.
• 17:04
  SPEAKER_02: It's like everybody all in one class and it's this...
• 17:08
  SPEAKER_02: But I do know that the people who are doing that are doing amazing work,
• 17:13
  SPEAKER_02: especially if that's something they choose,
• 17:15
  SPEAKER_02: because sometimes it's the punishment that you get.
• 17:18
  SPEAKER_01: It's your turn to do this thing.
• 17:20
  SPEAKER_02: But it seems to me that this is exactly your jam, and correct me if I'm wrong.
• 17:25
  SPEAKER_02: And in your ethics in engineering assignment,
• 17:29
  SPEAKER_02: we're getting to the details,
• 17:30
  SPEAKER_02: but I think it's important to say,
• 17:32
  SPEAKER_02: You built an assignment to talk about ethics in engineering.
• 17:36
  SPEAKER_02: You used Gen AI as a second phase into the discussion.
• 17:41
  SPEAKER_02: So what are your goals for this activity and how does it really play out?
• 17:46
  SPEAKER_01: Yeah, that's a good question.
• 17:47
  SPEAKER_01: I think it all started with when,
• 17:52
  SPEAKER_01: you know, Gen AI came about and people were trying to say, well, what do you do?
• 17:56
  SPEAKER_01: Do you engineer?
• 17:57
  SPEAKER_01: We're designing AI in, do you design it out?
• 18:01
  SPEAKER_01: And I was sort of looking at, well,
• 18:03
  SPEAKER_01: you know,
• 18:03
  SPEAKER_01: this is a one credit class, this Introduction to the Engineering Profession course.
• 18:07
  SPEAKER_01: It covers all topics about engineering that are non-technical.
• 18:10
  SPEAKER_01: So things like ethics,
• 18:12
  SPEAKER_01: things like professional values,
• 18:13
  SPEAKER_01: sustainability,
• 18:14
  SPEAKER_01: and all these other topics that are, you know, the students don't necessarily...
• 18:20
  SPEAKER_01: I don't know, maybe they go, oh, well, we already know this in terms of ethics.
• 18:24
  SPEAKER_01: I've got good morals.
• 18:24
  SPEAKER_01: I'm going to make ethical decisions.
• 18:26
  SPEAKER_01: I'll be able to respond to professional dilemmas.
• 18:28
  SPEAKER_01: No problem.
• 18:30
  SPEAKER_01: So in the past, I've always sort of given scenarios.
• 18:33
  SPEAKER_01: And what I like now with Gen AI I can use it to actually help me modify some
• 18:37
  SPEAKER_01: of these scenarios.
• 18:39
  SPEAKER_01: But the students would have to address the scenarios.
• 18:43
  SPEAKER_01: But with the advent of GenAI,
• 18:45
  SPEAKER_01: you know, it was just the idea, well, they might not engage in discussion at all.
• 18:48
  SPEAKER_01: And so even if they might be discussing this in class or in a team,
• 18:52
  SPEAKER_01: they might just say, okay, well,
• 18:53
  SPEAKER_01: here's the scenario,
• 18:54
  SPEAKER_01: feed it into,
• 18:55
  SPEAKER_01: you know, Claude or into chat, GPT, and then let's see what it spits out.
• 18:59
  SPEAKER_01: And I said, well, I'd like to get them away from that.
• 19:01
  SPEAKER_01: I'd really like them to engage in the discussion first,
• 19:05
  SPEAKER_01: take advantage of the fact that generally in the team,
• 19:07
  SPEAKER_01: there's a bit of diversity there.
• 19:09
  SPEAKER_01: Not all students are coming, in my case, from Montreal.
• 19:12
  SPEAKER_01: Not all the students are domestic.
• 19:15
  SPEAKER_01: There's a mix between international domestic students,
• 19:18
  SPEAKER_01: especially when we get into the cases regarding EDI and accessibility.
• 19:21
  SPEAKER_01: Not all students are, let's say, men.
• 19:23
  SPEAKER_01: There might be women.
• 19:24
  SPEAKER_01: So there's a good mix.
• 19:26
  SPEAKER_01: And take advantage of that.
• 19:27
  SPEAKER_01: And everyone comes into that class with lived experiences.
• 19:32
  SPEAKER_01: And so when you're faced with one of these scenarios,
• 19:34
  SPEAKER_01: those are lived experiences that you can use to guide your decision-making,
• 19:37
  SPEAKER_01: to guide how you're going to respond to that particular dilemma.
• 19:39
  SPEAKER_01: That might not be...
• 19:40
  SPEAKER_01: You know,
• 19:41
  SPEAKER_01: immediately embedded within whatever Gen AI platform we use to try to have
• 19:45
  SPEAKER_01: a response, to generate a response.
• 19:48
  SPEAKER_01: So my thought with the activity was,
• 19:51
  SPEAKER_01: okay, I'm not going to tell the students what they have to submit.
• 19:53
  SPEAKER_01: I'm just going to say, here's the scenario.
• 19:55
  SPEAKER_01: Go and talk about it.
• 19:56
  SPEAKER_01: Go and figure it out.
• 19:57
  SPEAKER_01: Go and think about what you want to do with it.
• 19:59
  SPEAKER_01: And by and large, they all do.
• 20:01
  SPEAKER_01: And they start debating, discussing, and they share their opinions and viewpoints.
• 20:06
  SPEAKER_01: And then at the end of the class, I introduce, okay, well, this is what you have to do now.
• 20:10
  SPEAKER_01: So now I'd like you to...
• 20:13
  SPEAKER_01: You know, as a team, you've come up with an idea and maybe you don't agree with the team.
• 20:15
  SPEAKER_01: Maybe, you know,
• 20:16
  SPEAKER_01: as an individual, you still have a disagreement with how the team wants to proceed.
• 20:20
  SPEAKER_01: And that's fine.
• 20:21
  SPEAKER_01: But go in now and engage with Gen AI.
• 20:24
  SPEAKER_01: So again,
• 20:25
  SPEAKER_01: this is a first year course and so many of the students
• 20:27
  SPEAKER_01: The students are probably getting the, well, don't, you should not use Gen AI.
• 20:32
  SPEAKER_01: And I can understand that in sort of like a very fundamental math class,
• 20:35
  SPEAKER_01: for instance, or physics class where you really want them to get that fundamentals.
• 20:39
  SPEAKER_01: So here,
• 20:40
  SPEAKER_01: go and use Gen AI,
• 20:42
  SPEAKER_01: ask it and prompt it with the scenario and have it come up
• 20:46
  SPEAKER_01: with several different possible solutions.
• 20:49
  SPEAKER_01: But then what I want you to do is to compare those solutions
• 20:51
  SPEAKER_01: with what your team came up with or even your own personal solution.
• 20:57
  SPEAKER_01: And see where it has strengths and see where it has weaknesses.
• 21:01
  SPEAKER_01: And then once you've done that analysis, try to come up with something even better.
• 21:05
  SPEAKER_01: And so the whole idea is to have them engage a bit with the Gen AI,
• 21:09
  SPEAKER_01: as opposed to just take the Gen AI and blindly use that as an answer,
• 21:12
  SPEAKER_01: but to compare it to what their team discussed.
• 21:14
  SPEAKER_01: And in the team discussion,
• 21:16
  SPEAKER_01: I'm hoping that they brought in their own personal lived experiences
• 21:18
  SPEAKER_01: that Gen AI wouldn't have access to.
• 21:20
  SPEAKER_01: And so they would be able to say, oh, well, okay, yes,
• 21:23
  SPEAKER_01: Gen AI are about these good points, but it hasn't considered this point of view.
• 21:27
  SPEAKER_01: And so by taking those points of view that you get from those rich discussions
• 21:31
  SPEAKER_01: in class, I'm hoping that they can use that and then build a final response.
• 21:36
  SPEAKER_01: And that's what they would have to perhaps submit as an assignment,
• 21:40
  SPEAKER_01: for instance, a short assignment.
• 21:43
  SPEAKER_02: And this connects well with the things I was observing just this week
• 21:50
  SPEAKER_02: when I was writing on my Substack, and that is the idea of AI second, not AI first.
• 21:58
  SPEAKER_02: Seems to be one of these ideas that is emerging throughout.
• 22:02
  SPEAKER_02: It's like first think,
• 22:04
  SPEAKER_02: write or draw or whatever it is, and then ask AI and find out what's going on.
• 22:10
  SPEAKER_02: All right.
• 22:11
  SPEAKER_00: Yeah, perfect.
• 22:12
  SPEAKER_00: So our last question is, there's a lot of critique about the built-in biases of Gen AI.
• 22:19
  SPEAKER_00: You, however, use Gen AI in student assignments about equity, diversity and inclusion.
• 22:24
  SPEAKER_00: How do you see Gen AI having a positive impact on the students thinking in this area?
• 22:31
  SPEAKER_01: It's another good question.
• 22:32
  SPEAKER_01: And I think, you know, everyone's recognized that Gen AI can have bias.
• 22:37
  SPEAKER_01: And I don't know what biases are I think it depends on the models that are being used.
• 22:45
  SPEAKER_01: Again, I come back to what I hope that when
• 22:48
  SPEAKER_01: the students use Gen AI is it presents additional perspectives
• 22:51
  SPEAKER_01: that they might not have already thought of.
• 22:53
  SPEAKER_01: And then they can build from that.
• 22:55
  SPEAKER_01: So again,
• 22:56
  SPEAKER_01: it's this idea of not just taking whatever Gen AI has come up with,
• 22:58
  SPEAKER_01: but providing you with that additional perspective.
• 23:02
  SPEAKER_01: And I think moving forward,
• 23:04
  SPEAKER_01: the next time I teach the course,
• 23:05
  SPEAKER_01: I might actually prompt them Or have them prompt an AI differently and actually say,
• 23:11
  SPEAKER_01: oh, well, here's a scenario, but take maybe a minority position.
• 23:14
  SPEAKER_01: If I was in this minority position, what would you say?
• 23:17
  SPEAKER_01: And if I was in the majority position, what would you say?
• 23:20
  SPEAKER_01: And see if it responds any differently,
• 23:23
  SPEAKER_01: because the students themselves might come from a position of a majority or from
• 23:26
  SPEAKER_01: a position of a minority,
• 23:27
  SPEAKER_01: and they might themselves have some thoughts about what's coming up.
• 23:31
  SPEAKER_01: And then they could challenge again what AI is coming up with.
• 23:38
  SPEAKER_01: I think the biggest thing that I want the students to walk away with is just this realization.
• 23:42
  SPEAKER_01: And this is a fantastic presentation I heard,
• 23:44
  SPEAKER_01: a plenary presentation I heard at the CEFI conference,
• 23:48
  SPEAKER_01: the European Society of Engineering Education Conference earlier this year.
• 23:51
  SPEAKER_01: And someone was talking about AI systems are socio-technological systems.
• 23:55
  SPEAKER_01: And as engineers, you have to think about them in that manner.
• 23:59
  SPEAKER_01: And so you really have to understand that when it's coming up
• 24:02
  SPEAKER_01: with technological solutions and other solutions,
• 24:05
  SPEAKER_01: that there's a social aspect behind it that you don't know and that you have
• 24:09
  SPEAKER_01: to make sure that you understand and cover those bases,
• 24:15
  SPEAKER_01: you know,
• 24:15
  SPEAKER_01: when you take whatever Gen AI is going to give you as a particular solution.
• 24:19
  SPEAKER_01: And I think that that's the idea.
• 24:21
  SPEAKER_01: I don't know what models are being used.
• 24:23
  SPEAKER_01: I don't know what GenAI is going to say to the students when it's fed with one
• 24:26
  SPEAKER_01: of these scenarios.
• 24:28
  SPEAKER_01: Unless I've used GenAI to help create the scenario, then it already has its list of responses.
• 24:34
  SPEAKER_01: But I think it's just this idea.
• 24:36
  SPEAKER_01: I think now that I've been thinking about it more,
• 24:38
  SPEAKER_01: I would try to have the students prompted differently
• 24:41
  SPEAKER_01: and use different view perspectives and try to tease out.
• 24:45
  SPEAKER_01: And then to say, look, you've had GenAI come up with a couple of responses.
• 24:49
  SPEAKER_01: Do you think it's Had any bias in its response to just again trying to get
• 24:54
  SPEAKER_01: the students to think more broadly and critically about what Gen AI is producing
• 24:58
  SPEAKER_01: as opposed to just taking it for face value.
• 25:02
  SPEAKER_01: Yeah.
• 25:03
  SPEAKER_02: And in some of our work,
• 25:05
  SPEAKER_02: we've done exactly the opposite,
• 25:07
  SPEAKER_02: and that is we've used the same exact prompt to show how you use
• 25:12
  SPEAKER_02: the same exact prompt,
• 25:13
  SPEAKER_02: and you get somewhat different answers,
• 25:15
  SPEAKER_02: sometimes vastly different answers,
• 25:17
  SPEAKER_02: depending on how good your prompt is,
• 25:20
  SPEAKER_02: but also knowing that AI in general works in probabilistic ways,
• 25:25
  SPEAKER_02: which means you never get the same output twice.
• 25:28
  SPEAKER_02: Absolutely.
• 25:30
  SPEAKER_01: And I think just the idea of conversing with it You know,
• 25:34
  SPEAKER_01: again, you get a response back and you can say, well, I don't quite agree with that.
• 25:38
  SPEAKER_01: What about this view?
• 25:39
  SPEAKER_01: And you just get into that discussion and that could lead you down that,
• 25:43
  SPEAKER_01: what we talked about earlier,
• 25:44
  SPEAKER_01: you know,
• 25:45
  SPEAKER_01: points of productive struggle where you're,
• 25:47
  SPEAKER_01: you know,
• 25:47
  SPEAKER_01: you're not quite sure about this or it's saying something,
• 25:50
  SPEAKER_01: but you don't really believe it.
• 25:51
  SPEAKER_01: Well, you shouldn't believe it.
• 25:53
  SPEAKER_01: Challenge it, you know.
• 25:56
  SPEAKER_01: And engage in that discussion.
• 25:57
  SPEAKER_01: Now, I understand,
• 25:58
  SPEAKER_01: of course,
• 25:58
  SPEAKER_01: you don't want to spend three hours on it because you'll consume a lot of energy
• 26:01
  SPEAKER_01: and there's all those concerns.
• 26:03
  SPEAKER_01: But there are ways that you can engage with it productively.
• 26:07
  SPEAKER_01: And again,
• 26:08
  SPEAKER_01: it's just trying to move away from I'm just going to take the answer
• 26:12
  SPEAKER_01: that it gives me.
• 26:14
  SPEAKER_01: The other great takeaway I had from that Cephi presentation was,
• 26:19
  SPEAKER_01: you know, plausibility is not validity.
• 26:22
  SPEAKER_01: And Gen.AI will always give you something that sounds really plausible,
• 26:26
  SPEAKER_01: but we need to check, validate, question, you know, whatever it comes up with.
• 26:32
  SPEAKER_01: And I think that that's all part of being a good student,
• 26:33
  SPEAKER_01: being a good engineer at the end of the day.
• 26:36
  SPEAKER_02: Yeah.
• 26:37
  SPEAKER_02: So thank you so much for spending half an hour with us.
• 26:41
  SPEAKER_02: We appreciate your time.
• 26:42
  SPEAKER_02: Absolutely, yeah.
• 26:44
  SPEAKER_02: And we'll see you all next time on Azadeh.
• 26:46
  SPEAKER_00: And Guy.
• 26:47
  SPEAKER_02: On AI.
• 26:48
  SPEAKER_02: Thank you.