Latest / Azadeh and Guy on AI / Navigating AI in Graduate School

Transcript

Navigating AI in Graduate School · Azadeh and Guy on AI

• 0:00
  SPEAKER_00: Hi and welcome to Azadeh on AI and this is our,
• 0:05
  SPEAKER_00: according to our counting, this is our sixth episode.
• 0:09
  SPEAKER_00: Episode five we actually had a problem with the audio so we're going to redo that episode.
• 0:15
  SPEAKER_00: I'm pretty sure Jeremy will come by and talk us through his work
• 0:20
  SPEAKER_00: because it is fascinating and important work.
• 0:23
  SPEAKER_00: And so that being said we're moving to the main topic today and today we're going to talk about
• 0:29
  SPEAKER_01: uh three advice that we can give to graduate students so we're gonna alternate
• 0:34
  SPEAKER_00: and we're gonna give a piece of advice if you're a graduate student
• 0:37
  SPEAKER_00: and you're thinking about where does AI fit into my world maybe the first piece
• 0:42
  SPEAKER_00: of advice that I would give it doesn't count I get my three after that
• 0:46
  SPEAKER_00: but do not ignore AI give it serious consideration and then if you choose not
• 0:53
  SPEAKER_00: to use it for a variety of reasons absolutely but do not ignore it it does exist in
• 0:57
  SPEAKER_00: the world
• 0:58
  SPEAKER_00: So what's your first one?
• 1:00
  SPEAKER_00: As a graduate student who's using AI,
• 1:02
  SPEAKER_00: who's studying in AI in multiple ways,
• 1:05
  SPEAKER_00: if there's a graduate student just coming into graduate school this fall,
• 1:10
  SPEAKER_00: what would you say?
• 1:11
  SPEAKER_01: So I would start with a very short introduction.
• 1:15
  SPEAKER_01: My major is Innovative Learning Technologies and Generative AI emerged
• 1:19
  SPEAKER_01: as I started my PhD program.
• 1:22
  SPEAKER_01: So it's been with me all of these years.
• 1:25
  SPEAKER_01: The first advice that I can give to graduate students is that maybe they don't want
• 1:32
  SPEAKER_01: to expect AI to give them a kind of publishable work.
• 1:37
  SPEAKER_01: I mean, if I want to say it in a short form,
• 1:40
  SPEAKER_01: you need to treat AI as a research partner, not as an author.
• 1:45
  SPEAKER_01: and the biggest shift for me was that AI is not
• 1:49
  SPEAKER_01: the author it's my research partner I use it to brainstorm ideas
• 1:54
  SPEAKER_01: to evaluate my thinking to organize literature or sometimes to critique arguments
• 2:03
  SPEAKER_01: but all of those interpretations final decisions and conclusions are should be mine
• 2:11
  SPEAKER_00: Absolutely.
• 2:12
  SPEAKER_00: And that's a responsibility.
• 2:14
  SPEAKER_00: As an author,
• 2:16
  SPEAKER_00: as the person who's taking charge,
• 2:18
  SPEAKER_00: you have got to make sure that it is first and foremost yours so AI is here to help.
• 2:23
  SPEAKER_01: And I think it hugely impacts your confidence if you just copy and paste what you get from AI.
• 2:31
  SPEAKER_01: Because you may have a ready-made presentation which looks nice,
• 2:35
  SPEAKER_01: but when you are going to present it yourself in a conference or in front
• 2:39
  SPEAKER_01: of your colleagues,
• 2:40
  SPEAKER_01: then you may not feel confident enough and you may feel that it's not mine,
• 2:46
  SPEAKER_01: it doesn't belong to me.
• 2:48
  SPEAKER_00: Yeah, there's a certain alienation when you see that and you're like, wait, what?
• 2:53
  SPEAKER_01: Yeah.
• 2:53
  SPEAKER_01: All right.
• 2:56
  SPEAKER_00: My first one,
• 2:58
  SPEAKER_00: and we're working on our policy for graduate student,
• 3:01
  SPEAKER_00: my first one in that as well is talk to your advisor.
• 3:05
  SPEAKER_01: Mm-hmm.
• 3:06
  SPEAKER_00: That is,
• 3:07
  SPEAKER_00: get a sense of what your advisor and every instructor,
• 3:10
  SPEAKER_00: by the way,
• 3:11
  SPEAKER_00: in any class you are,
• 3:13
  SPEAKER_00: if there isn't an official AI policy or they don't start a conversation about AI,
• 3:18
  SPEAKER_00: start that conversation.
• 3:20
  SPEAKER_00: and ask what is allowed,
• 3:22
  SPEAKER_00: what is not.
• 3:23
  SPEAKER_00: If you are experienced with AI
• 3:25
  SPEAKER_00: or if you are using AI in your workflow,
• 3:28
  SPEAKER_00: be honest about it
• 3:29
  SPEAKER_00: and tell them how you're using it
• 3:31
  SPEAKER_00: so you know what the boundaries are.
• 3:34
  SPEAKER_01: Can I ask a question, Guy?
• 3:36
  SPEAKER_01: And it could be very different
• 3:38
  SPEAKER_01: for graduate students
• 3:39
  SPEAKER_01: but sometimes
• 3:40
  SPEAKER_01: it's a kind of hard conversation.
• 3:43
  SPEAKER_01: People are hesitant
• 3:44
  SPEAKER_01: when they want to ask questions
• 3:47
  SPEAKER_01: about AI
• 3:47
  SPEAKER_01: and sometimes they feel that
• 3:49
  SPEAKER_01: advisor may think they are lazy they don't want to do the work so how
• 3:55
  SPEAKER_01: so what is your suggestion how do we need to start that conversation so
• 4:00
  SPEAKER_00: I have two things to say one is this is why I'm working with faculty to make sure
• 4:05
  SPEAKER_00: that we have a structured way to do that so a student doesn't have to start
• 4:09
  SPEAKER_00: the conversation because the right way to do it is for faculty
• 4:13
  SPEAKER_00: to bring it up however if you are a student and you're thinking about using AI
• 4:17
  SPEAKER_00: or you're using AI in your workflow every day because as more generation
• 4:22
  SPEAKER_00: of graduate students come
• 4:24
  SPEAKER_00: and they have practices that they bring from the workplace from their in that time
• 4:29
  SPEAKER_00: that they're working on their own or
• 4:31
  SPEAKER_00: from undergraduate I mean eventually we'll have ones that have already used it
• 4:36
  SPEAKER_00: in high school so if nobody else initiates you need to understand the boundaries
• 4:42
  SPEAKER_00: because you don't want to be in that conversation
• 4:44
  SPEAKER_00: where somebody says this is not your work have you used AI and you're like
• 4:50
  SPEAKER_00: yes i have used ai i always use ai you don't want to have that conversation after
• 4:56
  SPEAKER_00: the fact you want to have that conversation early you want to be gentle and
• 5:01
  SPEAKER_00: so what you're saying is it feels dangerous to do that yeah and i agree
• 5:06
  SPEAKER_00: that it is it's more dangerous to get caught caught is the wrong word
• 5:11
  SPEAKER_00: the consequences are yes for somebody to challenge what you've been doing and to say
• 5:16
  SPEAKER_00: in my mind anything that you've done from the beginning is completely null
• 5:20
  SPEAKER_00: and void you don't want that either yeah so you need to gently ask what is
• 5:27
  SPEAKER_00: the policy about ai how do you feel you may actually want
• 5:31
  SPEAKER_00: to ask you know do you use ai in what ways do you use ai and start
• 5:37
  SPEAKER_00: the conversation carefully yeah
• 5:40
  SPEAKER_00: Engage the response.
• 5:41
  SPEAKER_00: You don't want to overshare.
• 5:43
  SPEAKER_00: And what I mean is not necessarily that you want to hide your AI practice,
• 5:49
  SPEAKER_00: but if you're the student who initiated that conversation and you're getting
• 5:53
  SPEAKER_00: a very negative response, you know that
• 5:57
  SPEAKER_00: The boundaries are now clear.
• 5:58
  SPEAKER_00: You're not supposed to use anything and you probably shouldn't, right?
• 6:03
  SPEAKER_00: At least in any way that is directly connected.
• 6:06
  SPEAKER_00: And this is what people care about.
• 6:07
  SPEAKER_00: They care about analysis, especially qualitative analysis, and they care about the writing.
• 6:12
  SPEAKER_00: you can use it in other ways you can use it to smooth out the writing from
• 6:15
  SPEAKER_00: a grammar perspective nobody challenges that you can use it and I'll talk
• 6:21
  SPEAKER_00: a little bit about this but there are lots of small things you can do that aren't in
• 6:26
  SPEAKER_00: that realm of you're being dishonest
• 6:31
  SPEAKER_00: And those are the places you definitely want to use it.
• 6:34
  SPEAKER_00: But anything that has to do with writing or analysis,
• 6:38
  SPEAKER_00: you want to make sure that you and your advisor or any instructor have
• 6:43
  SPEAKER_00: a clear understanding of the boundaries.
• 6:44
  SPEAKER_00: And then you make your choices, but you want to understand where you stand.
• 6:49
  SPEAKER_01: So maybe your first advice is for the faculty members.
• 6:53
  SPEAKER_01: I mean, they need to develop an AI policy and then...
• 6:57
  SPEAKER_00: Absolutely.
• 6:58
  SPEAKER_00: But you as a student, you don't get to choose that.
• 7:01
  SPEAKER_00: So if there isn't one, so the first thing is to ask, can you direct me to what the AI policy is?
• 7:08
  SPEAKER_00: And
• 7:09
  SPEAKER_00: then if you don't get a response or if you say we don't have a policy then you start
• 7:15
  SPEAKER_00: a careful conversation and again you're asking okay so what are
• 7:18
  SPEAKER_00: the boundaries what is acceptable and not acceptable if you have ways to use it
• 7:24
  SPEAKER_00: and you're willing to share share if you feel a lot
• 7:26
  SPEAKER_00: of negativity just you don't have to kind of dig yourself a hole yeah yeah
• 7:32
  SPEAKER_00: With your advisor,
• 7:33
  SPEAKER_00: if you see,
• 7:34
  SPEAKER_00: because I know some of my colleagues are very negative about AI and you're like,
• 7:38
  SPEAKER_00: okay, I understand where I stand now.
• 7:40
  SPEAKER_00: And then you have choices.
• 7:42
  SPEAKER_00: One of the choices is maybe,
• 7:44
  SPEAKER_00: especially if you're interested in the way AI interacts,
• 7:47
  SPEAKER_00: maybe you need to work with somebody else.
• 7:49
  SPEAKER_00: That is a choice that you can make as a graduate student in most places.
• 7:53
  SPEAKER_00: And the second one is obviously listen.
• 7:57
  SPEAKER_00: because that's going to have an impact.
• 7:58
  SPEAKER_00: If somebody is having a negative reaction when you just mentioned AI,
• 8:03
  SPEAKER_00: imagine if they found out that you were using AI.
• 8:06
  SPEAKER_00: That's not going to be good for you.
• 8:08
  SPEAKER_00: So the way to do that is to have conversations and to build that into
• 8:12
  SPEAKER_00: the conversation in a way that's productive.
• 8:17
  SPEAKER_00: And again, enter the conversation, carefully practice with AI having that conversation.
• 8:24
  SPEAKER_00: and tell AI to be a supportive faculty member or
• 8:28
  SPEAKER_00: a faculty member who's really resistant to AI you can try that
• 8:33
  SPEAKER_00: and see how it plays out and find a few paths to be careful and protect yourself
• 8:38
  SPEAKER_00: because you don't want a faculty member to think oh this is a person who just wants
• 8:43
  SPEAKER_00: to find an easy path and use AI for everything and not do any work you don't want
• 8:48
  SPEAKER_00: that because if
• 8:49
  SPEAKER_00: If that's what you're going to do, don't go to graduate school.
• 8:52
  SPEAKER_00: Just talk to ChatGPT a lot.
• 8:54
  SPEAKER_00: It's the same thing.
• 8:55
  SPEAKER_00: Yeah.
• 8:57
  SPEAKER_00: And you don't have to pay or you pay $20 a month.
• 9:00
  SPEAKER_00: So that's what I would be careful about.
• 9:03
  SPEAKER_00: Yeah.
• 9:04
  SPEAKER_00: But you do want to have that conversation.
• 9:07
  SPEAKER_01: Yeah.
• 9:07
  SPEAKER_01: Yeah.
• 9:08
  SPEAKER_01: That's very good.
• 9:08
  SPEAKER_01: All right.
• 9:09
  SPEAKER_00: Your second piece of advice.
• 9:11
  SPEAKER_01: Let me see.
• 9:14
  SPEAKER_01: The second advice is that AI response is as good as the question that you are asking from AI.
• 9:23
  SPEAKER_01: For example,
• 9:24
  SPEAKER_01: just don't ask explain regression give AI some context explain the problem
• 9:30
  SPEAKER_01: and regarding my experience the more specific you are the better response you get
• 9:36
  SPEAKER_01: from AI and the other thing is
• 9:39
  SPEAKER_01: that AI I think AI has made me think more I mean clearer sometimes I go through
• 9:45
  SPEAKER_01: the AI response
• 9:47
  SPEAKER_01: And just by looking at the response,
• 9:49
  SPEAKER_01: I figure out that, oh, this is the part that I don't understand at all.
• 9:53
  SPEAKER_01: So I need to go back,
• 9:54
  SPEAKER_01: study more,
• 9:55
  SPEAKER_01: analyze it,
• 9:56
  SPEAKER_01: and then go back to AI and ask a more specific question or a clearer question.
• 10:03
  SPEAKER_01: AI has helped me in that regard a lot, yeah.
• 10:07
  SPEAKER_00: And do you upload documents into it?
• 10:10
  SPEAKER_00: I do.
• 10:10
  SPEAKER_00: If there's an output or a chapter or something like that?
• 10:14
  SPEAKER_01: Yeah, I do that.
• 10:15
  SPEAKER_01: It helps a lot in terms of the context.
• 10:18
  SPEAKER_01: And I think I talked about it once, even for answering IRB questions.
• 10:25
  SPEAKER_01: So IRB forms are very long and many of those sections are interconnected.
• 10:32
  SPEAKER_01: Sometimes you forget the details.
• 10:35
  SPEAKER_01: So in that case,
• 10:36
  SPEAKER_01: it helps a lot if you ask AI to help you with formulating the responses or
• 10:43
  SPEAKER_01: for showing you what are the conflicts in different parts of that.
• 10:50
  SPEAKER_00: And so that's a way to maximize what you get, the quality.
• 10:55
  SPEAKER_00: So what you're saying is the quality of the response is connected
• 10:58
  SPEAKER_00: to how much context and how much detail you give AI.
• 11:02
  SPEAKER_01: Yep, yep.
• 11:04
  SPEAKER_00: All right.
• 11:04
  SPEAKER_00: My second one is to play.
• 11:07
  SPEAKER_00: To play lots.
• 11:09
  SPEAKER_00: And what I mean by that is use AI as much as you can for things that are not high stakes.
• 11:18
  SPEAKER_00: So, for example, I had to do something today.
• 11:22
  SPEAKER_00: I had to add a few things to my resume.
• 11:25
  SPEAKER_01: Okay.
• 11:26
  SPEAKER_00: So usually it's like, okay,
• 11:29
  SPEAKER_00: find the conference presentation and then copy from the program into the conference,
• 11:36
  SPEAKER_00: into my CV and then make sure all the details are right and blah, blah, blah.
• 11:41
  SPEAKER_00: And the copy and paste doesn't work.
• 11:43
  SPEAKER_01: Oh my God.
• 11:44
  SPEAKER_01: Yeah.
• 11:45
  SPEAKER_00: Yes.
• 11:45
  SPEAKER_00: So what I did today is I said, here's the program.
• 11:50
  SPEAKER_00: Okay.
• 11:51
  SPEAKER_00: You know, on a whim.
• 11:52
  SPEAKER_00: So while I'm trying to do this, I'm like, wait, before I do this, can AI do this well?
• 11:58
  SPEAKER_00: And I said, here's the program.
• 12:00
  SPEAKER_00: Can you find all the presentations I made and quote them or create an APA style citation?
• 12:09
  SPEAKER_00: and he did it beautifully well and he said the one thing i'm missing is the date
• 12:13
  SPEAKER_00: because when you do posters you need the exact date and he said what was the date
• 12:18
  SPEAKER_00: or do you want me to go look for it and i went no i know the date here's the date
• 12:23
  SPEAKER_00: and it generated and i cut the whole thing and pasted it straight
• 12:27
  SPEAKER_00: and it saved me probably half an hour of work and probably a few mistakes i had
• 12:31
  SPEAKER_00: to verify that it's correct yeah yeah it's not making things on yes but
• 12:37
  SPEAKER_00: it was so much faster and that's what i mean
• 12:39
  SPEAKER_00: by play is every time you do something especially as you said something that is like
• 12:43
  SPEAKER_00: an irb or something that is annoying and that's not why you're
• 12:47
  SPEAKER_00: in graduate school yeah right this is a thing you have to do to get to
• 12:51
  SPEAKER_00: the real thing yeah that's true that's when you play make something make it do
• 12:57
  SPEAKER_00: the things you don't like doing make it do some things you can't do
• 13:02
  SPEAKER_00: and you're like maybe it can do that for me yeah
• 13:05
  SPEAKER_00: those are the things that you want to try and in that mix also try different kinds
• 13:10
  SPEAKER_00: of AIs until you find one that you prefer the models change all the time
• 13:16
  SPEAKER_00: so some people say just choose one model and focus on
• 13:19
  SPEAKER_00: that I have all three top models three four top models and I alternate between them
• 13:27
  SPEAKER_00: because sometimes some of them do something really well
• 13:30
  SPEAKER_00: and other things not they're also connected in different way to my documents
• 13:34
  SPEAKER_00: I do have my favorite AI I mean I've talked about this before
• 13:38
  SPEAKER_00: but cloud is my favorite AI everything that I own that isn't part of
• 13:42
  SPEAKER_00: the university is connected to my cloud it's not part of
• 13:46
  SPEAKER_00: the institution therefore I can't connect it to everything inside
• 13:50
  SPEAKER_00: the university it's my loss because it complicates my life
• 13:54
  SPEAKER_00: but I understand why there's data security there's privacy there's all
• 13:58
  SPEAKER_00: of these concerns I'm fine with it
• 14:01
  SPEAKER_00: it does do everything I want as good as if not better than other models despite
• 14:07
  SPEAKER_00: the fact that I do like open AI and its different models and those have worked
• 14:12
  SPEAKER_00: for me rather well and those are inside the institution and so they function
• 14:17
  SPEAKER_00: a little bit different for me I've been using Gemini in the last few months
• 14:21
  SPEAKER_00: that has been less fantastic but
• 14:25
  SPEAKER_00: it's still good the last piece about playing a lot is if you can afford it
• 14:30
  SPEAKER_00: and I know graduate students take the vow of poverty so that's hard but
• 14:38
  SPEAKER_00: But I would argue that paying for one kind of AI or another gives you just
• 14:45
  SPEAKER_00: a completely different level of quality.
• 14:48
  SPEAKER_00: And as the field moves forward,
• 14:51
  SPEAKER_00: a paid subscription or API or whatever it is
• 14:54
  SPEAKER_00: that you're using has tremendous benefits, consistency and a better memory.
• 15:02
  SPEAKER_00: And that's something that I think gives dividends in the long run.
• 15:07
  SPEAKER_00: you obviously don't want to pay for all of them and $200 or more or
• 15:12
  SPEAKER_00: an expensive API is probably out of the range for most graduate students so don't do
• 15:17
  SPEAKER_00: that but a paid version does give you a better quality so choose one
• 15:21
  SPEAKER_00: if it doesn't work for you stop paying pickups another thing
• 15:29
  SPEAKER_01: your next one my third advice would be use AI for doing repetitive tasks for you so
• 15:36
  SPEAKER_01: as all of us know graduate school is full of repetitive tasks you need
• 15:41
  SPEAKER_01: to I don't know formatting you need to format references you need
• 15:46
  SPEAKER_01: to check APS style you need sometimes to organize your writing sometimes you need
• 15:51
  SPEAKER_01: a second eye to check your writing to make sure that everything is correct
• 15:55
  SPEAKER_01: so for those tasks which are all of them are important but they are time consuming
• 16:01
  SPEAKER_01: and sometimes you are too tired you don't want to spend I mean that much time
• 16:05
  SPEAKER_01: on such kind of a task you can use AI to do that for you and
• 16:10
  SPEAKER_01: as Guy said you always need to verify before you accept it
• 16:14
  SPEAKER_00: Yeah, and that goes back to your first point, right?
• 16:18
  SPEAKER_00: You are the author.
• 16:20
  SPEAKER_00: You sign your name the minute that you said, this is mine.
• 16:24
  SPEAKER_00: You need to be responsible for every comma, every period, every number.
• 16:30
  SPEAKER_00: So check them.
• 16:31
  SPEAKER_00: But it does save you a tremendous amount of time when you are the editor and not
• 16:36
  SPEAKER_00: the person who wrote it.
• 16:36
  SPEAKER_00: You're also much more likely to catch mistakes.
• 16:40
  SPEAKER_01: Yeah, yeah, that's true.
• 16:41
  SPEAKER_00: all right my third one in this what is becoming a very short not very short
• 16:48
  SPEAKER_00: but short episode yeah which is good yeah we need
• 16:50
  SPEAKER_00: to be short often i talk too long i measured i talked too much good idea
• 16:56
  SPEAKER_00: so okay all right um
• 17:00
  SPEAKER_00: And I just went yesterday to XAI in Omaha, Omaha XAI, Omaha by AI.
• 17:06
  SPEAKER_00: This is what they called it.
• 17:08
  SPEAKER_00: And this is my advice.
• 17:10
  SPEAKER_00: Go to professional development.
• 17:12
  SPEAKER_00: Go to any opportunity for something, especially if it's free or if it's reasonably priced.
• 17:18
  SPEAKER_00: Go to that professional development.
• 17:19
  SPEAKER_00: As a graduate student,
• 17:20
  SPEAKER_00: you have a whole university that has more events than anybody can go to,
• 17:24
  SPEAKER_00: more talks than anybody can go to.
• 17:26
  SPEAKER_00: Choose the AI one and go.
• 17:29
  SPEAKER_00: Sometimes you're like, yeah, I know all of this.
• 17:31
  SPEAKER_00: They're talking about Notebook LM again.
• 17:35
  SPEAKER_00: If I hear another talk about Notebook LM, I'm going to blow up.
• 17:38
  SPEAKER_00: But I understand why they're talking about it.
• 17:41
  SPEAKER_00: It's a great product.
• 17:42
  SPEAKER_00: And if you've never used AI, this is one of the most useful ways you can start.
• 17:47
  SPEAKER_01: Yeah.
• 17:48
  SPEAKER_00: And so I understand why they're talking about it.
• 17:50
  SPEAKER_00: But if there's an opportunity, go.
• 17:52
  SPEAKER_00: If there's an opportunity when they're actually showing you things
• 17:56
  SPEAKER_00: and they're letting you practice, go.
• 18:00
  SPEAKER_00: Because it's the time on task.
• 18:03
  SPEAKER_00: It's the playing and trying.
• 18:06
  SPEAKER_00: And when there's somebody else that you can ask questions and say, hey, how do you solve this?
• 18:12
  SPEAKER_00: That is incredibly helpful.
• 18:14
  SPEAKER_00: So if they offer professional development, go.
• 18:16
  SPEAKER_00: I've been doing a lot of AI.
• 18:18
  SPEAKER_00: That's all I think about or 90% of what I think about in my professional life is AI.
• 18:25
  SPEAKER_00: And still I go to almost any opportunities that is out there.
• 18:29
  SPEAKER_00: Sometimes I leave after five minutes.
• 18:30
  SPEAKER_01: yeah
• 18:32
  SPEAKER_01: But I go.
• 18:33
  SPEAKER_01: Absolutely, yeah.
• 18:34
  SPEAKER_01: because I mean the titles are intriguing when you are talking about workshops and
• 18:39
  SPEAKER_01: then you start and you see okay I know 80% of all of it yes and so you learn also
• 18:44
  SPEAKER_00: which ones are good which ones are not but especially if you're still beginning
• 18:47
  SPEAKER_00: and especially if you're wondering about its place as a new graduate student
• 18:51
  SPEAKER_00: Yeah.
• 19:08
  SPEAKER_00: that's true the fact that the federal government in response
• 19:14
  SPEAKER_00: to grant applications has started saying that you're not allowed to submit things
• 19:18
  SPEAKER_00: that were produced by AI I'm not sure how they're judging it and there's a lot of
• 19:25
  SPEAKER_00: Very little clarity about that issue because we know that AI detectors are not very good.
• 19:32
  SPEAKER_00: And we know that they have too many false positives.
• 19:35
  SPEAKER_00: And so we're not sure how that would impact.
• 19:37
  SPEAKER_00: But again,
• 19:38
  SPEAKER_00: if you don't go to that meeting,
• 19:39
  SPEAKER_00: if you don't go to that workshop, you don't even hear that there might be a problem.
• 19:43
  SPEAKER_01: Yeah, that's true.
• 19:44
  SPEAKER_00: So it's a way to think about the work we're doing.
• 19:47
  SPEAKER_00: So say your three kind of quickly, just to remind everybody.
• 19:52
  SPEAKER_01: Using AI for doing repetitive tasks.
• 19:56
  SPEAKER_01: using AI as a research partner and I think the quality of responses that you get
• 20:03
  SPEAKER_01: from AI depends on the questions that you ask and for me it's talk to your advisor
• 20:08
  SPEAKER_00: and any instructor you're working with play lots yeah
• 20:13
  SPEAKER_00: if you can afford it paid version even at the lowest level gives you a lot more and
• 20:18
  SPEAKER_00: the last one is go
• 20:19
  SPEAKER_00: to any professional development true yeah all right we'll see you next time on
• 20:24
  SPEAKER_01: as I did and guy on AI