YT YouTube2Text

YouTube2Text → Transcripts → Unlock the Power of AI for Academic Research

Unlock the Power of AI for Academic Research — Transcript

by Alchemy · 10,026 words · 1,423 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:03we're going to focus on unlocking the
0:05power of AI for academic research and
0:08I'll introduce our featured speaker for
0:11the day and our agenda in just a
0:14moment uh first I'll start with my
0:16introductions and then turn it over to
0:18Tracy uh myself I'm VP for educational
0:21Innovation and inclusivity at Alchemy
0:24and I've been in that role for lose of
0:26four years now and I work with a team of
0:29my colleagues to really do what we can
0:31to push educational Innovations to
0:34provide more tools and resources that
0:36can enable Educators and to enable
0:38Student Success and during the closing
0:41of our session today we'll be sharing
0:43some exciting resources that we've
0:44launched recently and our platform and
0:47give you access to many of those
0:48resources and those do include AI in
0:51many of them as well and we'll talk
0:53about that briefly at the end I want to
0:56introduce Tracy mandelia Moore um Tracy
0:59is manager of edtech Innovations at
1:02Western University of Health Sciences um
1:04Tracy is also currently working on her
1:07PhD at University of North Texas um and
1:10that research emphasis is on learning
1:12technology so very well connected to our
1:15overall topic but in getting to know
1:17Tracy over the last few weeks which has
1:19been fabulous um she's been doing such
1:22phenomenal work related to generative Ai
1:25and all sorts of different tools and
1:26applications and doing so with a very
1:28thoughtful mindset which which I really
1:30love and we're going to be applying that
1:32thoughtful mindset as we talk about the
1:34different tools and applications today
1:36but also how to do that appropriately or
1:38most ethically um Tracy a little bit
1:40more about
1:41yourself sure so I'm really excited to
1:44to be here with you all and and share a
1:46little bit about what I know um this
1:49isn't a a full uh everything but just a
1:54touch and AI for me is just overwhelming
1:57and exciting and it's just a great
1:59opportunity
2:00um the research piece is also wonderful
2:03so I'm really looking forward to to
2:05sharing what I know with you guys today
2:07and and hopefully getting some of your
2:08feedback as well on this process so
2:10thanks for having me great thanks Tracy
2:13yes absolutely your feedback and also
2:15continually in the chat make comments
2:18we'll try and keep track but with 300
2:20and something of you that's obviously
2:21going to be hard to do in real time but
2:23we do take the chat and we consolidate
2:26that and we share it out to the group
2:28we'll also be sharing the slides so
2:30don't worry if you miss something or we
2:32goes too quickly or a link doesn't get
2:34dropped into chat you'll get a copy of
2:36the slides and you'll get a copy of the
2:37recording and also all the other
2:39resources that we point out to you today
2:41so go ahead and just sit back and
2:43hopefully enjoy learn with us and share
2:46with us as we go
2:48forward all right uh we've been doing a
2:51series we do many webinars and we do
2:54also leadership panels but we've been
2:55doing a series on AI since May and we're
2:59at session eight or nine here in the one
3:01you have today this is to give you just
3:03a glance of the different topics that
3:05we've covered um at the end of the
3:07session we'll give you a link to where
3:08you could access any of those recordings
3:10or any of the resources that we've cated
3:12from each of those sessions so we have
3:15one page that contains all of that
3:16information for
3:18you okay and we'll continue to do these
3:20webinars going forward so we'll make
3:22more announcements as we
3:24go so our session outline we're going to
3:27do a a starting check-in with ethical
3:30considerations uh then we'll focus on
3:33steps for brainstorming the research
3:35question performing the L review the
3:37research instrument data analysis and
3:40visualization uh drafting conclusions
3:42recommendations and then again the
3:44resources that we're going to be able to
3:45send you away with today that are
3:47related to uh the session as we go
3:50through the outline here it's going to
3:52be primarily Tracy else switch to an
3:54assist role in in just a moment here um
3:57she will be doing actual demonstrations
4:00talking you through it and giving you
4:01screenshots and also some links to some
4:03of the chat conversations that you can
4:06explore um in a followup um but I look
4:09forward to having her share some of
4:11those out because she's really prepared
4:13some exciting demonstrations so let's go
4:16to the padlet here first of all to do a
4:19check-in with
4:20you and I'm going to back out the link
4:23is in the chat for
4:25you and then you'll see me on my screen
4:28go to that as well let me catch catch up
4:30on chat let's get that into the chat
4:32there we go
4:33perfect and now that you've got the link
4:36I'm going to head there as
4:38well and what we're looking for here if
4:40you're new to padlet you would click a
4:42plus under any of those column headers
4:45and then just share out any of the
4:47benefits of using I in research efforts
4:49that you see any of the issues that you
4:53see uh ways maybe you've used AI in
4:56research already and if so any tools
4:59that you would
5:01recommend increasing speed of work
5:04absolutely um we're today talking about
5:08how you can use it to help you shortcut
5:10some steps get to things quicker but not
5:13sacrifice quality or Integrity of your
5:16efforts so there's that balance of
5:18keeping things in check using it to help
5:20you take a bulk of material and get
5:23forward more quickly but still being the
5:25one that's in charge yeah brainstorming
5:28it does a great job of that
5:30efficiency time many times we see that
5:32on here um yeah propping up an outline
5:35of research ideas we're going to cover
5:37that for
5:41you background on the research
5:44generating future ideas yeah you can
5:46analyze the existing literature and look
5:50for suggestions to add to that that's
5:52always what you want to try to be able
5:54to do is contribute something
5:57unique okay developing fr Frameworks
6:00ideation love these um we often do
6:03padlets because it's great for INS
6:05session sharing but this link that we
6:07gave you you can come back any time and
6:09lean some of these ideas if they were to
6:12be helpful to you as an individual or if
6:14you're wanting to lead a similar session
6:16at your campus okay for time purpose let
6:19me go over to the second column and just
6:20scan ethics and plagiarism properly
6:23sighting transparency we'll be talking
6:25about things like that people do abuse
6:27it um people might be abusing it and
6:30using it for research or people may have
6:33abused um AI or what they've put into Ai
6:36and so you also have to look for abuse
6:38at different angles outdated or
6:40incomplete information yes some tools
6:42more than others hallucinations are
6:45always an issue for sure um plagiarism
6:49so and also intellectual property we'll
6:51talk about that data security and
6:53privacy oh yes we're going there for
6:57sure uh reliability and we'll talk about
7:00validity
7:02awesome uh can we get a reshare on the
7:05padlet link for
7:07Carla thank you Carla okay and then ways
7:10you've used Ai and research so just
7:12scanning the group here help with coding
7:15um assistance writing interview and
7:16survey questions we're going to talk
7:18about forming questions a little bit
7:21brainstorming keyword um yeah I'm not
7:24sure I guess I guess your target words
7:26for your liter literature search and
7:28review
7:31great refining your write up yeah using
7:34it like a grammarly to give feedback on
7:36your write up to help check for
7:38understanding summarizing PDFs powerful
7:41there generating topics love all of
7:44these things that you've put in here
7:46teaching yourself python
7:49great instrument
7:52design all right there's so much here I
7:54won't get to all of yours I
7:56apologize always love revisiting this
7:59okay uh that's research rabbit that's
8:01popped up very identifi identifiable uh
8:04perplexity great we'll be talking about
8:06some of those tools already uh rhetorica
8:09that's new to me okay awesome academic
8:12Insight lab they've been doing some
8:13great um work related to Ai and research
8:16and they've got some nice briefs as
8:18well um chat PDF yes site we'll be
8:23talking about uh kenas that's a new one
8:25to me great thank you elicit we'll cover
8:30that perplexity again awesome uh Magic
8:33School you need I need to revisit that
8:35one consensus will be covering site yes
8:38s CIT will be covering site. and
8:42elicit awesome and then there's a thumbs
8:45up yeah feel free to thumbs up on any of
8:47these that you would
8:49recommend okay sspace we'll talk about
8:52very
8:53briefly and many of these that you've
8:56seen here as recommended tools many of
8:58those are in a table at the end of our
9:00session today where we're going to focus
9:02on the tools that we're talked about and
9:04then additional tools that are
9:06extensions of those for those that want
9:07to go deeper into specialized research
9:10because it's it's quite a bit to cover
9:12in you know 45 minutes or so left here
9:15okay let me jump back to our deck thank
9:18you so
9:19much great new information
9:22there okay and back into slideshow and
9:25again you can come back to this padlet
9:27anytime all right I'm just framing this
9:30I've been thinking about you know I'm
9:32later in my career I've been around
9:34quite a while and I'm thinking back to
9:36when I was doing research early on as a
9:37grad student in the early 90s and and as
9:40a young faculty member it was this card
9:43catalog experience and going from the
9:45card catalog to the stacks or to the
9:48microfish and the copy machine and just
9:50having a table full of stuff and being
9:53there all day long and doing all of this
9:55you know manual sorting and and finding
9:58and then having that roll of quarter
10:00stack of dollar bills to be able to you
10:02know make those copies and having these
10:04massive notebooks that to then cart home
10:06and then go back and forth through that
10:08process it was just so much work and it
10:10could get exciting once you got to a
10:12certain point for sure um but in just
10:15reflecting on that and where we are now
10:17and by the way I realized the
10:18stereotypical image there and I
10:20purposely didn't tell chat GPT to filter
10:22the image I just said here's what I'd
10:24like is something about doing research
10:26you know 30 years ago card catalog
10:28microfish blah blah blah and this is
10:30what it gave me I tried to improve that
10:33it did some things better on the second
10:34image but didn't quite do what I asked
10:36for but now if you think about what's at
10:39your fingertips from your desk whether
10:42at work or at home wherever and what you
10:44can do to get information sort
10:47information analyze doing all of that
10:50and with very little cost in some cases
10:54no cost at all other cases minimal cost
10:56or part of the overall budget you can
10:59come up with some powerful research
11:01models and Powerful results and then
11:04outputs in a much more streamlined
11:06fashion but again using those safeguards
11:09of your integrity and ethics that you're
11:12going to apply along the way so this was
11:14just sort of reflecting on how far
11:17things have come um pre- internet to now
11:19internet and now geni and we're just at
11:23the beginning stages as many people have
11:24said as far as where AI is going so
11:27it'll be interesting to keep tracking
11:28this of course
11:30okay um so let's just talk about basics
11:33of using chat GPT this is not meant to
11:35be a basic session by any means but just
11:38to prop it up we focus on um generative
11:41AI as a creative companion that it's
11:44helping you with tasks you're still the
11:47expert whether it be subject matter
11:49expert or research expert whatever the
11:50task is you are in charge you're just
11:53using this to enable your efforts and to
11:55get you from point A to point B maybe
11:58more efficiently and also to maybe
12:00expose you to more things that you
12:01wouldn't otherwise have access to but
12:03you're still the one um leading the
12:06effort and these are the types of
12:08efforts you might engage in and you can
12:10see some of the things on the left were
12:11mentioned already related to research in
12:13that
12:15padlet okay so I'll just move forward
12:17very quickly there are a lot of
12:19prompting Frameworks out there um many
12:21of them that I've seen and some of them
12:23are as many as you know eight characters
12:25long as far as an acronym we try and
12:27keep ours pretty tight to if you think
12:29about the role that you want the Gen to
12:32act as the task you want it to complete
12:35and what are the requirements related to
12:37that task and what type of output are
12:38you looking for and you can think
12:40through the different scenarios that
12:41you'd be doing and then whenever you can
12:44through these four stages or within if
12:46you can apply some learning science at
12:48least if you're creating teaching
12:49resources and learning experiences for
12:52your students how can you embed elements
12:54of learning science to make sure that
12:55you're going to get the best results um
12:58how can you speak to in a way where
12:59you're going to get uh more likely to
13:01get better results um you know take a
13:04moment to think it over you can even
13:06follow up with it and trac's going to do
13:08a little bit of that with some of her
13:10steps
13:11today okay quick checkin on ethical
13:14considerations just at a high level
13:17scan um you always want to keep a
13:20lookout for deep fakes or misinformation
13:23um within generative AI itself it's
13:25going to be more about misinformation
13:27and any of the hallucination that might
13:29come up from that um deep fakes are
13:31going to tend to be more out in the
13:33public domain um but still something you
13:36definitely want to be aware of um
13:38keeping a lens or lookout for bias and
13:40discrimination when you're looking at
13:43different um results it's giv you and
13:45the sources it may have used to give you
13:47those results Trace back a little bit
13:49and see what it's gleaned and are there
13:52any biases in the way it's interpreted
13:55those or in the studies that it's been
13:57drawing from
13:59um copyright and intellectual property
14:01so as you're using these tools you want
14:04to be able to know where the
14:05information's coming from and be able to
14:07track forward and site forward those
14:10resources um not just a simple using gen
14:13getting the output and then holding that
14:15as your own and then moving forward and
14:17adapting that um privacy and data
14:20security we'll talk about that when
14:22you're inputting data and Tracy will
14:24cover that so looking out for that and
14:27I'm just going to do a scan because you
14:28we'll come back to of those within
14:29specific sections academic integrity and
14:33then checking with your institution or
14:34academic Journal as to the policies when
14:36you're doing research or scholarship
14:39related to your um findings and then
14:42we'll talk briefly about the
14:43environmental environmental impact or
14:45sustainability related to generative AI
14:48okay so at this point I'm going to
14:50transition and stop share on my end and
14:53turn it over to
14:56Tracy super thank you let me get mine
15:00all set up
15:02here and while she's doing that I see
15:04deloria's question adapt that rtro or
15:07retro framework as you choose Gloria um
15:11and we also have some resources that'll
15:12lead you to that as a
15:25handout all righty so to start us off we
15:29are going to discuss brainstorming that
15:31research question what does it look
15:34like so for this example here and I'm
15:37going to go through a series of examples
15:38with each one of these phases um we're
15:41going to go with the old standard chat
15:43GPT The Prompt that I'm using here or I
15:46guess the model that I'm using here is
15:48um for most of our examples today is
15:50going to be the 4.0 model which is the
15:53newest uh paid version of chat GPT which
15:56has been updated through December of
15:572023 um 3.5 is still on the older
16:02training data but 4.0 is now up to up to
16:05date through December of last year um so
16:08we're gonna find that research question
16:09and you'll see there's a a theme here
16:11through our examples of Puck Satani Phil
16:14and his weather forecasts it was just a
16:16fun way to kind of keep the theme
16:18throughout our presentation today so my
16:20chat GPT prompt says develop a research
16:23question for an academic paper on Foxon
16:25Phil's annual weather prediction
16:27ensuring it's clear focused and aligned
16:28with the objectives okay this is my
16:31result that I got with this simple one-
16:35sentence prompt here it's rather robust
16:38it gave me a great research question
16:40which you can just see right here to
16:42what extent is poxon Phil's annual
16:44Groundhog Day prediction align with you
16:45know yada yada so that's
16:48wonderful um I'm not a person to take
16:50most of these things at face value so
16:52what I want to do next is that follow-up
16:55prompt that you can see down here at the
16:57bottom and I want it to identify five
16:59additional questions for me and it has
17:03so I don't have just one research
17:05question I have five and then from here
17:07what I can do is I can take these
17:09various research questions maybe pick
17:11them apart and create something that's
17:13more specific to exactly what I
17:16need so we've got our research question
17:19few ethics on this one um Brett do you
17:22want want to talk us through this one
17:24yeah briefly so I'll be coming in to
17:26just do ethics along the way just give
17:28Tracy a in voice um so yeah bias and
17:31fairness so with the large large
17:32language model they may introduce or
17:35perpetuate any existing biases um you
17:39know AI itself is not necessarily biased
17:41it's everything that's been put into it
17:43and continues continues to be put into
17:44it those patterns of bias so how can you
17:47look for those identify those and
17:48mitigate those as much as possible and
17:51then looking at originality and
17:53authenticity so um any of those as AI
17:56assisted questions how can those be not
17:58novel or original um how do they
18:01contribute something meaningful to the
18:02field again that's always the step as a
18:04re of the researcher is to maybe confirm
18:07and validate but then extend to the
18:08research the next step to
18:13contribute so we've got our research
18:16question or questions and now we're
18:18going to work on our literature
18:21review so this can sometimes be time
18:23consuming we all know and um the way
18:27we're going to phrase this next section
18:29of the presentation is through these
18:31four steps so we're going to first start
18:33with our outline of our literature
18:34review we're going to do the actual
18:36search for the literature then we're
18:38going to understand what it was that we
18:40actually uncovered and we're going to
18:41write and these are some of the tools um
18:44that we're going to talk through in each
18:45one of these steps and again there's
18:46links and everything at the end of your
18:48um session today on all of these so for
18:52our first one in creating that outline
18:55my prompt here again going straight
18:57through gp4 create a literatur view
19:00outline for a study about how poxon
19:02fails weather prediction affects the
19:04economy it gave me this wonderful result
19:07here that you can see on the screen it
19:08talked about the introduction of what
19:11that might look like and some of the
19:12subtopics within that it wanted to make
19:14sure that I have the historical
19:16background and the cultural significance
19:18of our groundhog friend um the economic
19:20impact of Groundhog Day itself the
19:23predictions of um the influence of
19:25weather predictions on those economic
19:27activities consumer Behavior Behavior so
19:29I gave it a simple prompt and it gave me
19:31this really nice outline that then I can
19:34take and continue on to the next step
19:36which is going to be our search
19:38component and this is where we're going
19:40to spend some time we've seen some of
19:42these apps pop into the padlet there
19:45again if you have more apps that you
19:47would recommend to our community please
19:49put them in that padlet so that way we
19:50can have this really great repository of
19:53of content and resources there so elicit
19:57is wonderful one of my favorites um alic
20:01it it finds papers it uh compiles some
20:05of the papers it'll find the top four
20:07we're going to show you this in just a
20:08second um but Alysa it's one of the top
20:10ones that I use on a regular basis
20:12research rabbit is another one that I
20:14use quite frequently research rabbit
20:16does a wonderful job of finding
20:19connected papers and additional
20:21resources that maybe you wouldn't have
20:22uncovered otherwise consensus is another
20:25popular one um I personally am not too
20:27familiar with this one um but consensus
20:30you can ask yes ask yes or no questions
20:33to the research and the content so that
20:35could be beneficial for you does a few
20:36more things than that um lit Maps again
20:39those connected Maps same with connected
20:41papers it is able to find those
20:44additional resources for you that you
20:45wouldn't have found
20:47otherwise I know site was mentioned in
20:50our our padlet board as well um it could
20:54produce inline citations which is nice
20:57um one of the things I like about that
20:58particular app is that if you have a
21:01statement and you know this statement to
21:03be true you've read it you know it's
21:05true you need a resource for it so you
21:07can put that statement into site and it
21:09will help you find a resource to support
21:12that statement so that's wonderful um
21:14and then there are a lot of custom gpts
21:18and plugins that will help you along
21:20with this way um with your research as
21:22well within chat gbt is um and that's
21:25chat gbt 4 not the 3.5 the four allows
21:28you to have that that's the paid version
21:29allows you to have those extra
21:31components to enhance your um your GPT
21:34interface so these are just a few of the
21:36tools few of the most popular tools
21:38let's take a look at elicit and research
21:42rabbit a little more in depth and for
21:45that I'm gonna come over here
21:48and drag my screen in um we should be
21:53able to see that Brett can you see that
21:55extra screen there yes you list it thank
21:58you
21:59so just a quick demo on elicit this is
22:01elicit and this is how it uh you start I
22:06have my research question I'm just going
22:07to pull it from my other screen that we
22:10uncovered uh when we were doing our
22:12research question and identifying that
22:14I'm going to pop that in to elicit just
22:17like that to what extent is poxon spills
22:19annual Grand hod day PR prediction yada
22:23yada and then it searches 125 million
22:26academic papers go figure we've got the
22:29top four though here in just a matter of
22:31seconds it gives me this nice summary of
22:34those top four papers which is lovely
22:37and then if I scroll down it will
22:39actually give me a little brief synopsis
22:41of each one of those papers you can
22:42continue to scroll it will continue to
22:44give you more um
22:46examples over here on the right hand
22:48side you can select additional uh
22:51columns that you might want to have in
22:53this summary breakdown here let's look
22:56at um
22:59I don't know main findings just pick one
23:02sometimes they're not there and
23:04sometimes they are so it gives you a
23:06really good look at what some of that
23:08literature it just took their whole
23:09literature process and were able to find
23:11something quickly so from here what I'm
23:13going to do is I'm going to take let's
23:16see let's just grab this guy I'm going
23:18to copy the DOI on this particular
23:20article and then I'm going to come over
23:22here into research rabbit so research
23:25rabbit love the interface it's very user
23:27friendly let me zoom out of this one a
23:30little bit so we can see it because it
23:32likes to go uh it's a rather long one
23:35and you can see I've already started my
23:37predictions here on this collection
23:40we're going to add our paper I'm going
23:41to pop in that
23:44DOI we've just added that paper to our
23:47collection of paper so we're going to
23:48we're building our repository of content
23:51from here it's taking all of these
23:53papers that you see on this one column
23:55and it's giving us similar work earlier
23:57work later work and so on want
23:58additional papers written by these
23:59authors so wonderful resource I'm going
24:02to click on similar work here and you
24:05can see it gives me this fun little map
24:08of various literature and how these
24:10literature topics are connected the
24:12larger bubbles are ones that are a
24:14little more frequent than the smaller
24:16Bubbles and if I want to continue on my
24:18literature and continue to go down that
24:20rabbit hole right research rabbit we're
24:21going down the rabbit hole of research
24:23I'm going to click on Donovan right here
24:26we've got our selected paper similar
24:28work and it just keeps going and going
24:30and going down that rabbit hole for you
24:31so this is a wonderful Resource as
24:33you're trying to find that literature
24:35view content and pull that information
24:38so very
24:39cool all right again links resources all
24:42of that at the end for you today so I'm
24:44going to move that back over to my other
24:46screen and um and continue on so the
24:50next part of our literature riew process
24:53is going to be understanding those
24:55papers so we've got a collection they're
24:57wonderful what do they mean maybe
24:59there's a few that we just don't fully
25:02understand that's okay so explain paper
25:05is another wonderful application of
25:07course if you have GPT um chat GPT some
25:09of this can be done within that
25:10interface but trying to introduce you to
25:12a few new ones explain paper you could
25:15upload your literature your paper into
25:18this particular platform from there
25:20you'll highlight a particular section of
25:22text that maybe you don't fully
25:24understand or you want it to reward it
25:25in a different way to make sure you're
25:27understanding it correctly and then here
25:29you can I'm going to just kind of circle
25:31this area right here you can say explain
25:34this to me as if I were a middle
25:35schooler or a high schooler and it gives
25:38you this wonderful simplified version of
25:40what it is that you just read without
25:42the attitude for any parents that are in
25:44our audience today you know what I'm
25:45saying um so explain paper is really
25:48really wonderful to be able to take
25:49those larger comp those more complex
25:51pieces and break them down a little bit
25:54smaller another one that I don't think
25:56gets enough credit um that I personally
25:59actually use this I love this one it's
26:01doc analyzer and Doc analyzer again
26:04takes that paper you uploaded it into
26:06their interface and then you're able to
26:08ask questions to the research paper and
26:12when this first came out these types of
26:13tools was thinking why do I want to ask
26:16questions to my research paper but then
26:18as I'm experimenting with it it's
26:20actually really
26:21beneficial so I have this one example
26:23here that on your screen it's one
26:25research paper it's 16 pages I didn't
26:27read it
26:28and what I'm able to gather from this
26:31one I simply ask what are the
26:33independent independent variables and it
26:35tells me what are the researchers
26:37hypotheses and it tells me what are the
26:40study rev the study results and it tells
26:42me so it's a wonderful way to take some
26:45of those papers and highlight those
26:49questions so you can uh sort through
26:51them a little quicker the other great
26:53thing about this particular one is it
26:55has this labels component and as you're
26:58going through your literature View and
27:00you have the chunks you can add labels
27:02to them so instead of asking questions
27:04to just one paper you can ask questions
27:07to say six papers so how are these
27:10papers connected what are the common
27:12themes between these papers what are the
27:13commonalities um where are the gaps in
27:15the literature of this collection of
27:17papers so it's not just one paper it's a
27:19multitude of papers so doc analyzer is
27:24uh another great one to help you
27:25understand the content um and get
27:27through that literature a little bit
27:30quicker so continuing on um write is my
27:35next one and um and I'm I'm pausing here
27:38for just a second because my current
27:40research is focused on academic
27:43integrity and so this particular one
27:46hits me a little a little hard there are
27:48multitude of uh applications that will
27:53write your manuscript for you and um
27:58there's a few here that I have um two
28:01Jenny Ai and yomu are two examples of uh
28:05applications that will write that will
28:07write it for you this can be beneficial
28:10if you're having a hard time getting
28:12started on some of that content and
28:15maybe you just need some help getting it
28:17going um or maybe you're not uh English
28:21is not your first language so this could
28:23help in that area um but Jenny AI which
28:27is the script screen that you see here
28:29um it you just popped in your um your
28:33prompt it will and it will start writing
28:35you can accept it you can customize it
28:37it gives you real references which is
28:40nice and of course if you pay for the
28:42upgrade then you can collect those
28:43references and the example here they're
28:45all in apa 7 format so a lot of the
28:47formatting is done as well um I do have
28:51to throw out because I am studying the
28:54ethics side of things and the academic
28:56Integrity um for me these are red flag
28:59tools um they do exist it's good to know
29:02about them um but the fun is writing um
29:06and and so we don't want to take that
29:07away from from us as as from academics
29:10and Scholars because the fun is writing
29:12um the uh the little side screen there
29:16that you can see is um yomu and yomu
29:20this is on their homepage it's a little
29:22uh blurb that I pulled out yomu is
29:24advertising itself um if you're failing
29:26a class yomu can help if you're
29:28sacrificing your friends for academics
29:30yumu can help um if you're worrying or
29:32stressing about school yomu can help so
29:34they're really these some of these
29:35companies are poising themselves as uh
29:38tools to help
29:40students get around doing the work um so
29:43it's good to know about them um hard I
29:46don't recommend them but they're there
29:48they're there all that being said um
29:51Brad I'll turn it over to you to maybe
29:53kind of elaborate some of the ethics of
29:55that one
29:58and you're on
30:01mute sorry folks a lot going on and if
30:04you want to peek at the chat while we
30:05take a break feel free there's so much
30:07going on in there and I know people are
30:09asking about um these tools and the
30:12databases they're accessing we're we're
30:14trying to show you the tools that are
30:15accessing these um databases that are
30:20you know the best sources out there um
30:23you know so Google Scholar came up as
30:26one there was a question about that um
30:29so we're we're trying to use these tools
30:31specifically not just chat GPT so that
30:34it's going into these um databases and
30:36the sources are maintained and the
30:38Integrity is maintained as much as
30:40possible um so transparency in your lit
30:43review the fact that AI was a part of
30:45the process um any of the algorithms you
30:48may have used the databases that were
30:50consulted as you gathered the
30:51information so which databases were part
30:54of that which tools were part of that um
30:56again using your l as the expert to try
30:59and focus on bias
31:02mitigation and then using AI as a
31:04complement rather than a replacement for
31:06that human judgment and expertise again
31:08it's that creative companion that
31:10enabler that's hopefully offloading some
31:13of the lift but not taking away the
31:15knowledge not taking away the learning
31:18process um and the Integrity of that
31:21okay and then ensuring that those
31:22literature views respect int
31:24intellectual property rights so again
31:27that's keeping those sources tracked U
31:30managing the databases and I know one of
31:32the um mentions in there is that you
31:34know you might have 70% of the citations
31:37that come from chat GPT that are false
31:39in many cases are dead ends um yes
31:42that's that's why we're not using that
31:44as the primary tool here whatsoever um
31:46it has some options or there might be
31:49some gpts that work well that you can
31:51use that have valid sources but yeah you
31:54have to be careful with chat GPT in
31:56general and using that for getting
31:58citations and tracking research
32:01leads okay um and I also saw more
32:04questions about is this re being
32:06recorded absolutely we'll have the
32:08recording and the slides sent and all
32:10the resources bundled um by early next
32:12week to
32:13everybody so do know that that's
32:16available and you'll be able to retrace
32:17anything you want on
32:20here okay get us to the research
32:25instrument thank you um so for our
32:28research instrument piece we've uh got
32:30our outline we've done our search we
32:33understand we've written it up now we're
32:35going to talk about the
32:36instrument and for this example here we
32:40could simply put into chat GPT create a
32:43research instrument on um poxon Phil's
32:46weather predictions that's a rather weak
32:50prompt so Brett mentioned in the very
32:52beginning the RT prompt and so I'm going
32:56to take you through a series Ser of
32:58pieces and again there's a chat log that
33:01you can go back to and review it later
33:05um so what I'm going to first do in this
33:07example is I'm going to define the RT
33:10prompt you can see that on your screen
33:12simply saying the RT prompt is role
33:16which means act as task create a
33:20requirements which is be sure to and
33:23then the output um turn it into whatever
33:26so that's my prompt so I put that into
33:28chat GPT chat GPT replies and
33:31acknowledges what it is that I just told
33:34it next I'm going to take that and I'm
33:37going to say chat GPT this is what I'm
33:40doing next I'm going to say use this
33:42framework to create a prompt so I'm
33:46using my rtro framework which is my role
33:48my task my requirements and my output
33:51and I'm using it in the context of my
33:53poxon fill example so my um create a
33:58prompt where you act as a researcher to
34:00create a survey instrument because
34:01that's what I want that will measure
34:04adults perceptions and values of poxon
34:06fills predictions the output should be
34:08at least 25 liyer scale questions to
34:11measure this so I've taken that um the
34:14framework rephrased it to do what I
34:17needed it to do and chat gbt because it
34:20knows what this prompt framework is now
34:23it will give me the result that you see
34:24down here which is that very large
34:27robust
34:28prompt and there's you know little tip
34:30if of uh if you're using these AI tools
34:34the more robust and detailed your prompt
34:36is the better response and the more
34:38accurate response you're going to get so
34:40by using chat GPT to create that prompt
34:43we're able to get a very large robust
34:45prompt that is going to give us a more
34:48detailed example of what it is that
34:49we're looking
34:50for so on this next uh image here on the
34:54far right hand side of your screen you
34:56can see
34:58um what we're going to do with that
35:00prompt so I started a new conversation
35:02in chat GPT and I took the old prompt
35:06that it gave me and I pasted it into
35:08that new conversation and this is the uh
35:12survey instrument that it gave me so
35:15it's pretty good it um it's based on the
35:17lier scale questions it's giving me um
35:2125 very unique uh survey questions based
35:25on the topic that I gave it well I
35:27didn't like the format that it gave me
35:29in this particular example so my next
35:32followup was um convert these questions
35:35into uh a table format with the
35:38measurements because I like this this is
35:40just better for me visually makes more
35:43sense when I'm doing one of these
35:44surveys so it simply did that for me but
35:48then I thought well now I got to put it
35:50all into qual tricks or whatever other
35:53survey tool I'm going to do so I simply
35:57said X ported as a
35:59CSV and it did so it wrapped up the
36:01table gave me a CSV and then I can take
36:05that CSV I can download it and then I
36:07can upload it into my qual Trix account
36:09or whatever other survey tool that
36:12you're using so that's a pretty neat
36:14process to be able to go through to
36:15really expedite those survey instrument
36:17pieces of course you want to make sure
36:19that it's accurate before you put it
36:21into your um your uh qual Trex or your
36:24technology that you're using um so do
36:27some some checks and balances along the
36:29way but pretty cool that it does that
36:32and again there's a chat log so it'll
36:34take you through every single one of
36:35those steps you can follow
36:39along I see Dr Harland very timy timely
36:44uh question there because yes we are
36:45going to talk about um you know pilot
36:48testing and and looking for validating
36:51and reliability and all those things so
36:54um the first thing though is informed
36:56consent so if AI was used letting
36:58participants know and really what you
37:00can do is you can put multiple of these
37:02into one consent statement for um people
37:05who might be completing your survey so
37:07that they're aware of these things of
37:09course how their data might be used how
37:11it might be reported you're going to
37:12have that type of statement um any
37:14testing that you've done piloting that
37:15you've done any expert validity applies
37:18statistical analyses that inform this
37:21you can have that available to folks and
37:22maybe it's aze statement you have and a
37:25link to more details but they can do
37:26that check off box as participants or
37:29subjects in your study um and then the
37:33other thing is we're not integrating it
37:35into this session but I mentioned in
37:37chat we're going to do an ethics and AI
37:39panel session leadership panel in the
37:42near future and we'll talk more about
37:44what irbs or institutional review boards
37:46are starting to do to support this for
37:50um faculty who are doing research um
37:52they shouldn't have to discover all of
37:54this on their own or inadvertently
37:58mistakenly do something that may not be
38:00ethical because they use a certain AI
38:02tool so there's more and more guidance
38:04coming um through the institutions and
38:06then also through the IRB or
38:09institutional re review board check off
38:11process there can be these elements in
38:13there as well so look for more there are
38:16some models out there and we'll be
38:17covering that in our near future session
38:19this
38:21spring okay great I'll keep tracking
38:25here we'll go into the next section
38:27we've got um about 15 minutes or a
38:30little bit less Tracy okay perfect thank
38:32you um so next we're going to talk about
38:34working with your data so we have um
38:37we've done our research question we've
38:38done our literature review we've created
38:40our instrument and now and we've
38:42executed our our survey and now we have
38:44data what do we do with
38:46it so in this next example here we have
38:50poxon fills winter weather forecast data
38:53which um I was able to to pull um I
38:56gathered the the data from I think it's
38:581886 and used that as my test data for
39:01the examples I'm going to show you on
39:02our next slide um for here we're using a
39:06GPT plugin called data analyst and data
39:11analyst is powered by Julius AI which
39:13also has an external component as well
39:16and this particular tool it can suggest
39:20next steps confirm um whether or not to
39:22move forward with your particular
39:24research summarize the data it can give
39:27you ideas for broad analysis perform
39:29data analysis steps um give you ideas
39:32and suggestions for further examination
39:34it can create those visualizations that
39:36you might want to include um it can give
39:40uh refine or you can refine and and
39:43correct as needed but it's pretty cool
39:47so let's take a look at some of these
39:49screenshots that we have here again we
39:51can um give you the chat log on on this
39:53at the end so you can follow through
39:55with everything so in this example I did
39:57is I took the uh puxatony Phils data
40:03over the last hundred or so years and I
40:07put it into a Google slide and then
40:09embedded that Google slide or the Google
40:11Sheets within the um chat GPT enabling
40:14the plugin okay and I simply asked it
40:18can you create a visual representation
40:20of your findings you can see here on the
40:23screen that created um correlation of
40:25the findings for the weather um and gave
40:28us this nice
40:29visual in the middle here I simply asked
40:33it based on the predictive model what
40:35trends can you forecast over the next 20
40:38years it analyzed the data and gave me a
40:42um a trend forecast over the next 20
40:44years as you can see here it's going to
40:46get a little bit hotter over the next 20
40:47years according to the data um but it
40:50gave me this nice rundown of explain the
40:52x- axis the y- AIS explain the different
40:54colors and and why it did what it did
40:56which is just
40:58wonderful and then this last example
41:01here um we have these great graphs these
41:03wonderful pieces and I wanted it to
41:06explain the conclusions of the data what
41:09can be concluded um or what can you
41:11conclude with the data that I have
41:13provided you and it gave me this
41:15wonderful rundown of the data analysis
41:18and what it all means so it talks about
41:19the weak correlation about how Phil
41:21might always be right it's
41:24inconsistent um variability and fills
41:26data in Phil prediction accuracy it
41:29talks about the predictive modeling of
41:30the future temperatures comparison of
41:33the model of the predictions um it also
41:35addressed Regional variations which um I
41:38didn't ask it to do so it it uh gave
41:41that to me and it gave me an overall
41:42conclusion of that data so um another
41:47potential tool um that you can use for a
41:50highle data analysis I know this isn't
41:52going to replace anything um like SPSS
41:55or anything really intricate that you're
41:56going to to be utilizing but it's a nice
41:59place to
42:02start great thanks Tracy so some of the
42:07ethics for data analysis of course data
42:09privacy and confidentiality so strict
42:12adherence to those human subjects
42:14policies knowing the terms and
42:16conditions of any of the tools you might
42:18be using and just making sure not to put
42:20in any personally identifiable
42:22information um so keeping that practice
42:25as you're going forward with these with
42:26these tools TOs and not letting these
42:28tools just take it from you and take it
42:30away um and then you know reading the
42:34terms and conditions is something that
42:35people don't often do but in these cases
42:38I would certainly make sure to do that
42:41um I would check with folks on your
42:42campus who might have expertise either
42:44either in the library uh your research
42:47office any of those places it sometimes
42:49depending on what the purpose or the
42:50tool is um as far as any of the
42:54Predictive Analytics definitely be
42:56meticulous invalidating the integrity
42:58and reliability of your results um that
43:01has huge implications if you're going to
43:02make predictives and and um you can
43:05definitely cause certain Panic um and
43:08and there can be certain biases of
43:09course that go forward into your
43:11Predictive Analytics so think about the
43:13impact that those predictives may have
43:16social psychological and
43:19otherwise okay great let's go on to the
43:21next section I'm seeing chat of course
43:24um you know cheating and plagiarism
43:26catching students always comes up this
43:28session is not related to that
43:30whatsoever but it always comes up and
43:31people are sharing that's great um we
43:34have resources related to that as well
43:36anyway let's go into the next section
43:38okay uh drafting conclusions and
43:41recommendations um again I'm going back
43:44to chat GPT on this one I have an
43:47original prompt here that just says use
43:49the following manuscript to identify and
43:50explain the conclusions and
43:52recommendations for the manuscript in an
43:54academic tone um the tool that I'm
43:57displaying for this purpose is something
43:59called prompt perfect and I can take
44:01this original prompt put it into prompt
44:04perfect and it will um give me a more
44:07bloated prompt that will give me a
44:10better result from what it is that I'm
44:12asking um again um to kind of circle
44:15back to some of the ethics on this we
44:17have to keep in mind that the content
44:19that you're putting into these large
44:20language models could potentially be
44:22discoverable so uh just to Echo what
44:25what Brett had said in the last piece um
44:28don't include any personal identifiable
44:30information and um also keep in mind if
44:33you are putting in something that is
44:35intellectual property or something that
44:36you might want to keep um you might not
44:38want to use these tools so keep that in
44:40mind but drafting the conclusions and
44:42recommendations just using chat GPT and
44:44making that prompt a little bit
44:46better um the next one on this
44:49particular topic we have our article
44:51we've written it it's wonderful but what
44:53do we do with it um this example here is
44:55from a program called hey science um yep
45:00and um for this one we are uploading our
45:03manuscript into this model and we're
45:05going to say well what Journal does it
45:08match with it's going to find a journal
45:09for you that matches with the content
45:12that it is that you have put into your
45:14manuscript in this example here it says
45:16that I have an 84% match on my Jour on
45:19my manuscript related to our friend the
45:21groundhog um has a high impact factor
45:24that's
45:24lovely and then it's going to do the
45:28initial review for me so the initial
45:30Journal Review which is really helpful
45:32so from here you can see in the middle
45:34of the screen that it gives me the
45:35strengths of my manuscript based on the
45:38review and it gives me the weaknesses
45:41from here I would take this particular
45:42manuscript I would focus on those
45:44weaknesses and address those within the
45:46manuscript before submitting it to the
45:47actual journal or conference this one
45:49will also recommend conferences as
45:53well awesome and and going go back one
45:56real quick
45:57Tracy so just on that one it's not
46:00suggesting that AI wrote the manuscript
46:03and now you're just going to pump it out
46:05and get it published and have your name
46:07on it AI was part of the process it was
46:09supporting you you've now developed this
46:12manuscript and now you're using AI to
46:15help shop it out there finding the best
46:17match um and then what's some feedback
46:20that might come based on the standards
46:22of that journal based on the focus of
46:24that journal the audience Etc so it's
46:26just another tool to be able to help you
46:28make decisions of where you might go in
46:31those next steps U there was a a great
46:33comment um Kirsten made related to using
46:36AI for scholarship some may some may not
46:40for the research and scholarship um how
46:43do you compare those who do and don't
46:45and how is that identified how is that
46:47transparent in the submissions and and
46:50then how how are the ratings done
46:52differently and so that's something that
46:54um a tenure and promotion committee are
46:57definitely going to be dealing with and
46:58setting their standard is they look at
47:01um evaluation yeah that's that's a big
47:05one okay um other things that were
47:09shared we will be leaning the chat I
47:11know some of those documents that were
47:12shared we'll make sure those get into
47:14our summary of the session so don't
47:17worry if you didn't grab all of those
47:18going by in the chat so just a recap
47:21ethical considerations broadly but also
47:24within each of those steps um you can
47:26see the stages that we covered through
47:29there and how you might use some of
47:31these tools along the way to enable your
47:33efforts but still staying grounded in
47:36who you are and the expertise that you
47:38have the rigor that you need knows to be
47:40done um and the ethics and compliance
47:43and intellectual property um and privacy
47:46and all those things that stay critical
47:48okay let's move on to the last section
47:52here um let's see let's do a quick poll
47:56so so your likelihood of using Genai in
48:00research now if you aren't in a role
48:02that does research but maybe you're
48:04supporting or you're in an
48:06administrative role where the others are
48:08doing research what's your likelihood
48:11you would suggest others consider using
48:14geni as part of their research
48:18process given that they do
48:20it in the ways that we talked about in
48:23being ethical and transparent and all of
48:25that
48:34and you all have brought up so many
48:35great
48:37points love it can't cover it all in 50
48:41minutes or
48:42so uh look for that ethics panel coming
48:48up and feel free to suggest a session um
48:52suggest maybe participating in a
48:55session and we'll do more
48:57more okay and do we have our poll
49:08results okay we're a little bit loaded
49:10toward um already using or very likely
49:13with about 6
49:1564% okay and then likely 20% so we're at
49:20about
49:2185% that are sort of on the positive and
49:24then 11% on the bubble and then two 5%
49:28um less likely or very unlikely to use
49:30it yeah and there's there are some
49:32concerns or maybe it doesn't fit your
49:34role um or maybe you need to know more
49:37maybe you need to see how it evolves U
49:39maybe you're not ready could be one or
49:42all of those things absolutely um
49:44exciting possibilities with all those
49:47considerations that we've mentioned
49:49today and there's also I know the
49:50learning curve um this is evolving so
49:53quickly and we continue to do so so
49:55we've tried to to give you an example um
49:59Tracy you've done an amazing job of
50:01stepping through those things and
50:03showing some of the main tools there and
50:05then coming up in just a moment we can
50:06close the poll now um what we'll do is
50:09go ahead and turn back to me for share
50:15Tracy and then we'll share some final
50:17resources and announcements here and get
50:20you going okay let me get back to my
50:22screen
50:29aha okay I realize what I've done let me
50:32stop
50:34share I did not exit
50:41properly from the last share okay now
50:44I'm ready to
50:52go yeah let me get us all the way down
50:55here great
50:57okay let's go down the final stretch so
51:00here's the summary of the geni research
51:03tools that were shared and now we've got
51:06this first slide here of the main tools
51:08that we're shared and on the left you've
51:10got the name and the link to them um
51:13just a quick little bit about what they
51:15might be most um applicable to or
51:17specialized for and then on the right
51:20we've got the cost now fremium means
51:22you're going to get a basic version of
51:25it at for free
51:27if you want to go for the higher level
51:28version all access then depending on the
51:32tool you're going to have a cost chat
51:33GPT to do plus you're going to pay $20 a
51:36month and then you get access to more
51:38speed more power more plugins um all of
51:42those things so that's just one example
51:45and then the next slide that we have
51:46here because there's a lot of tools um
51:49these are some additional tools for
51:50consideration this is probably when
51:52you're going to be going deeper and
51:53specializing more in your research and
51:56really put p in things more deeply um so
51:59those some of those we did show some
52:01were just sort of a pass by and again
52:03you can see the cost there on some of
52:05those Okay so we've got those for you to
52:08be able to follow up on and then now
52:11some additional resources that we have
52:13available that are ongoing resources and
52:15then also an announcement is for that AI
52:19series that we've been doing over the
52:20last eight months or so we have
52:23everything um in one YouTube playlist if
52:25you want it in video format we've
52:27actually been getting a lot of views
52:29even though these are you know 50
52:31minutes to an hour um we have been
52:34getting a lot of traffic there but we
52:36also at the second link that's in the
52:38chat is we have our collated AI
52:40resources which is per webinar but also
52:44all of the resources that were shared
52:45through the webinar and then in some
52:46cases shared by attendees related to
52:48each webinar so we've got that there
52:51we've also got prompting resources and
52:53other things as well that are available
52:54so you do have that in the chat great
52:57and let me go on to the next one so this
52:59is a new resource we've developed we've
53:01been doing different prompts and de
53:03demonstrating different prompts through
53:04different sessions now we've created a
53:06resource that's I think 18 Pages what's
53:10but what's great about this resource is
53:11you can get this PDF and the link just
53:14went into the chat you can get the PDF
53:16and on that first page you thumbnail you
53:18see you can jump to any of those
53:19sections or you can simply scroll down
53:21the pages and you can see the text um
53:24The Prompt that we've created and you
53:26would be able to adapt or adopt that
53:28prompt to your context and your goals
53:31and you could also click and view the
53:32chat log that was demonstrated you can
53:34see a live example of that prompt in
53:36action and what the chat gbt results
53:38were so you can scroll through and look
53:40at any of those across those sections
53:42you see that are related to teaching
53:45learning some are student-based some are
53:49co-instructor prompts related to co-
53:51instructional tasks how to help offload
53:54some things that help you to focus more
53:55on teaching if that's your primary focus
53:57or research if that's your primary focus
54:00so look for that um new resource that
54:03we're excited to share
54:05today and then we're giving folks access
54:08to our platform cury which has these
54:11teaching optimizers and these teaching
54:13optimizers you can kind of think as
54:15tools and templates and wizards if you
54:17will and allows you to have a teaching
54:19task and go in and use an Optimizer
54:21that's designed for that and it will
54:24take you through as a Ser through a
54:25series of promps and you can respond to
54:27those prompts and those prompts are
54:29Guided by learning science related to
54:31that task whatever it is let's say it's
54:33an active learning activity that you
54:35want to implement or a group um project
54:37assignment it'll help you develop that
54:39in a matter of minutes and if you want
54:42to you can turn on our AI assist and you
54:44can get that even more quickly and get
54:46that developed you're still able to edit
54:49the results and then able to then
54:51download and put those into your
54:52teaching resources into your learning
54:54management system so through these
54:56optimizers we've been able to take it so
54:58you don't have a blank screen to start
55:00with you don't have to go out on the
55:01broad Internet and try and find the good
55:03stuff and then um consolidate it
55:05condense it and get to your results to
55:07your resources in hours and hours
55:09instead this can be done in a matter of
55:11minutes so by going to the link that's
55:13in the chat or following the QR code you
55:16can actually um jump in and start using
55:19our cury platform and have access to
55:21these optimizers and a lot of other
55:22resources that we have out there as well
55:25so you'll be hearing more and more as we
55:28um launch more of these
55:30optimizers and our next webinar oops uh
55:34we're getting a link error so if Kelly
55:37or others could look into that we'll get
55:39that cleared up for you thanks for
55:41bringing that
55:43up okay um the next webinar that we'll
55:46be doing is February 29th so in one week
55:49and so we are partnered with Open Stacks
55:52and they've been around for 25 years
55:54they have some of if not the best um
55:58open educational resource or opened
56:00textbooks out there and so our
56:03optimizers more and more of them are
56:05connected to the Open Stacks textbook so
56:08you can build activities and learning
56:10experiences for your students that are
56:12based on the Open Stacks content and
56:14will bring in the objectives bring in
56:16the content and focus on that for you
56:18and again in a matter of minutes you've
56:20got some resources that are aligned to
56:22these Oar resources so we'll have this
56:25webinar next week where we're um
56:28demonstrating that for you and giving
56:29you access to some of that okay anytime
56:33you can make your work more efficient
56:35while being effective and saving
56:37students money that's that's the
56:40best um here's just some links to keep
56:43following us on social media if that's
56:45helpful for you um we'll continue to
56:48putting out more of this thought
56:50leadership and thought partnership and
56:51doing what we can to just help us all
56:53going forward effectively as possible
56:56possible in our in our Collective and
56:58individual missions um with and without
57:01AI so again thank you so much Tracy for
57:05the tremendous job you did in this
57:07session all the expertise you brought
57:09thank you participants for all of your
57:10sharing I wish we could have responded
57:12to everything that was there I look
57:13forward to going back through the
57:15transcript and US coating resources and
57:18getting those out to you um early next
57:20week thank you so much

About this transcript

This page contains the full transcript of Unlock the Power of AI for Academic Research by Alchemy, generated from the public captions YouTube serves with the video. The transcript has 10,026 words across 1,423 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.