The Educational Promise of Narrative-Based AI

ABOUT THIS EPISODE

Dr. Jeremy Roschelle, Executive Director of Learning Sciences Research at Digital Promise, joined the po dcast to discuss the next generation of artificial intelligence for educational storytelling and what they’re hoping to build with their $20 million grant from the National Science Foundation.

The technology is coming along to whichan AI assistant can craft stories can enhance. Stories can customize stories.How can we use that to get back to a narrative, censor learning experience, you're, listening to enrolment growth,university from helic education, the best professional development podcastfor higher education leaders looking to grow in Roman at their college oruniversity, whether you're looking for fresh and Roman growth techniques andstrategies or tools and resources? You've come to the right place. Let'sget into the show, welcome back to enrolment growthuniversity, a proud member of the connect E v? U Podcast Network, I'mEric Wilson with Helis Education and we're here today with Doctor Jeremy,Rochelle, executive, director of Learning Sciences, research at digitalpromise, Jeremy, welcome the show thanks Eric it's great to be here,really excited to have you here and talk with you today about theeducational promise of narrative based Ai, but before we dick into that, canyou get the listeners a little background on both digital promise andyour role there sure thing Eric a digital promise is a non profitorganization that fundamentally connects educators, researchers andinnovators to work together and we spend Higher Ed k, twelve preschool,really life long learning within digital promise Eric I lead thelearning sciences, research group and our mission is to conduct fundamentalresearch on the future of learning, but in a very applied way. We want toconduct research, it's useful to educators everywhere so excited aboutthis conversation, because I feel like AI and education gets thrown out a lotand often doesn't really mean a I so maybe tokick us off today. Can you give us just a high level overview of the currentstate? Where are we now? In terms of AI and Education Eric, if I could draw atimeline, it would stretch back about fifty years- it's not new, but we're ata very exciting moment. Let me just talk for a second about that. Really,since the birth of AI, in the L T E S early T S, people have been imagining.What could this mean for education and we have had throughout the years somevery successful applications, but they've been narrow. Those applicationsare in areas of learning that are more puzzle like or more logic like math isa classic example. Chess, of course, is it a classic sort of example. We've allhad early warning systems that many universities will be familiar with thatgiven early indicator when a student may need some additional attention tostay and rolls to engaged where we are now, though, is like a hockey stickturning upwards and that's because the capabilities of ai are suddenlyexploding. We have much more capable possibilities before us and we all seethis every day in our lives. We're...

...start we're talking to home assistance,we're talking to our watches. You know we're used to things that complete oursentences suggest people we ought to talk with, and so it's becomingcommonplace. The underlying capabilities, number of pattens numberof researchers, just expanding, really wildly right now and within the nextfive years, that's going to mean new types of applications, not in education,not just those narrow ones that we've seen for the past. You know thirty orso years, love that overview Jeremy. Now, can yougive us a brief overview of this National Science Foundation Grant andthe really exciting program that you're hoping to build with it right happy todo that? I am really excited. My colleagues are excited to kick off thisAi Institute, the National Fun Science Foundation had a very competitiveawards, program and selected a series of partnerships to come together, andeach partnership had to connect two things: on the one hand, fundamentaland advances in a I itself and, on the other hand, an application area in ourcase learning, so our focus was really unengaged.Learning the new fundamentals are becoming possible. We wanted to takethat opportunity to rethink, engage learning and the place we've chosen tostart for that. Rethinking is to think about US stories throughout. Historyhave engaged learners. How we for the whole history of humanity, have sataround camp fires telling stories and that's been a primary modality oflearning, but somewhere along the way learning became rigid. It lost thestory telling emphasis and it became something that is off putting foreignalienating to too many of our students. So we want to look at the technology iscoming along to which an AI assistance can craft stories can enhance. Storiescan customize stories. How can we use that to get back to a narrative, censor,learning experience, you're blowing my mind a little bit. Ido want you to talk me of this Mental Ledge that I'm going to in my own headnow, because do we think that learning and education evolved from storytelling toward standardized testing, not necessarily because we thought itwas more interesting or better pedagogically, but just in order toscale it has that been the force evolution of education? In a nutshell,I agree with that. Eric you know, standardized testing came up about acentury ago in large part, because the military needed to assign people tojobs in the military they needed an...

...efficient system. So the problem withvery much a scale problem that drove standardized testing. How do we takethis mass of people we're bringing into a particular occupation and assign themto the right roles? It wasn't necessarily for the betterment of theirlearning that the standardized assessments were developed, but oncedeveloped they were quite useful that scale, and so you could see theeducation system start to tilt towards being organized around thoseassessments. I think we've lost. We know. I don't just think we know. We'velost a lot of learners by going too far down that road, and you know no onestarts out with their own child sitting on the Sofa opening a book and sayingHey I'd like you to take this multiple choice test. We know his parents, westart with stories, but it's also true the best CEOS. I know you're laughingbut read the Harvard Business Review About n cultures and huge companies,the best CEOS or story tellers, yeah, Jeremy, that's so fascinating, becauseI think to your point in terms of the learners that this evolution ofeducation works for, isn't fully inclusive, and I dotal we've seen thatonline education today seems to work best for a very, very self motivatedstudents when they do not have that benefit of storytelling. Is thatpartially what your research is hoping to solve for to develop adaptation,specifically through storytelling, to keep more students moving and engagedin the learning process? Yeah, absolutely you know one of the thingsthat's coming along on the foundation side of Ai and it's scary, but it'salso promising is the ability to synthesize things. Synthesize may bethe narrative for how a sports game might play out. You know to take a fewsentences that someone started and create a complete paragraph thatfollows the thoughts they started and sounds vaguely like a human may havewritten it. You know an assistance that can trackwhere you're going in a more narrative sense and play along towards that. Sowhat we're seeing in terms of student engagement is sins are engaged in partwhen they feel they belong. That's a very powerful human thing, belongingand stories are something that make us feel we belong or can make us feel like.We don't belong if they feel like they're, not welcoming our identity, our personality forinstructors, trying to teach hundreds or thousands of students. They don'thave the time to make slightly different variant story, experiencesfor all their students to be part of, but imagine that we are recontextualize.For example, a science lab experience to be really about trying to discoversomething. That's that's fundamental to...

...a problem. You know, maybe there's adisease spreading, that's something we're familiar with right now, andpeople have to address some of the scientific basis in differentcommunities and such a story about disease spreading. We need to vary alittle bit with different groups of students to let them follow what theyare really interested in, while still keeping some of the curricular emphasis.That is the goal of the course and so problem based learning work, learningin teams learning through collaborations, there's intrinsically asense of culture and socialization and story. An intrinsically is verymotivating students to be part of that, and what we're looking at the at is,instead of breaking that to come back to a very instructor driven one sizefits all narrative to try to keep that going a little bit longer stick withwhere the students want to take the learning experience. Jeremy, youmentioned being at the the beginning of this hockey, stick curve that you'restarting to see become more and more visible. What's the easy ask for Ai,what do you think a? I will be for sure, pretty great at in five ten yearsversus the much harder and long term mission that you're helping to startsaw for yeah. That's a that's a fascinating question. You know, we'veseen a lot of basically machine learning typeapplications already, and I think that 'll continue what these applications dois look for patterns or associations between two things, and so I do thinkone thing: that's going to become a lot easier, is kind of coarse recommendations where you mayhave an a B alternative. Should I start with a discussion, or should I startwith a presentation? Should the discussion be like this or be like that?I think we have so much data gathering in platforms that you may seeinstructors get some pretty sensible recommendations from a course we'replanning assistant, for example. That seems very it is very feasible andTrafitta to me. I just want to paint for you what the difference is betweenthat and what we're trying to do, giving this twenty million dollars totry to do something, exciting, yeah and really really Eric. The difference ispretty straightforward. Most of that kind of work is recommending betweentwo things that pretty much already exist or are just very routine, and youjust want to know: Is it better to do more to or more B or do a first or befirst or is a or be better for a particular student, but the as an besare the same things yeah. What we're trying to do here is create a kind ofexperience. This narrative, centered experience. That is very rarely donenow, because it's too hard too expensive. I V, you know to do it anykind of scale. It's been too hard, and...

...so the only place you see it is amassively talented faculty members who can weave their students into a into astory but you're, not every faculty member is going to have that talent,and so we're trying to break through to that really. An experience that cannotbe created today cannot be assigned A or B it's just a different type oflearning experience. Can you help us picture? What that looks like? I thinkthe de sequential based testing is something that's easy for me to wrap myhead around. We think that a you know hour and twenty long learning sessionwith my class could covered these ten things and what sequential order arelearning out, comes maximize, Wab, multivariate t and find out that one Ican get my head around. What's an example of what this looks like whenfinished, what does a I base storytelling? Look like what would aclass look like for one student versus another yeah, so Eric? I have anexample I like to use that. I think it certainly would fit high school, maybemiddle school. We know already that inquiry based science is reallyimportant to translate students sciences, it just facts: it's a processwith a certain quality, and so I like to imagine a group of students beingengaged with the basic outline that a month from now, we are going on a tripto Mars together and we're going there for a scientific project where we wantto collect some data on the Martian soil, and we need to plan this out. Whatinstruments are we going to bring? Where are we going to collect oursamples? How are we going to analyze them, and you know we may have somedifferent interests about what we want to really look at on Mars. Justdifferent SI different questions we may want to look at, but this whole coursetogether. Maybe there's a hundred students together, we're going to be insmall groups we're all going on this big ship tomorrows in a month. So let'sfigure out what you want to do when you get there and let's spend a month ofour class time learning about the fundamental physics or chemistry orwhatever it is that we need to know what the scientific instrumentation is,learn that datas sampling plans, planting out or analysis, and then youknow a month from today. Mission starts we're going to get on that rocket chip.We're going to go on this, of course, now simulated mission to Mars and we'regoing to do it. We're gonna collect our data, we're going to come back to classand talk about what we what we found when we got to Mars. So to me, that's akind of experience that we cannot. We can see how exciting it would be. Wecan't produce it right now today would just be too expensive to pull off,maybe in a museum you could do it today, but doing it in routine online learningto complicated. So that's the kind of thing we hope in five years, we'vereally made some progress towards, and what's the difference between, I I wantto sign up my child for that class...

...sounds amazing. I want to join thatclass myself. What's the difference between finding that Master Story,teller instructor to develop that sequential story versus relying on Aito help develop what that narrative actually is? Well, I think they'll be ablending of those two. I think I always really when I think aboutlearning theory or instructions is fine. I always want to iterate between soundsresearch principles and what the Bay best instructors do and we want to bereally human centered. I think the way to make this understand the feasibilityof this for A. I is that's the customizations, it's the variationsthat really will be too much for anyone structure to do or to manage, but thatlittle bit it's just what you and I do in conversation. We don't follow ascript we made ahead of time. We listen to each other, we're a littleresponsive. We take it a little bit different direction today, instructors, you know eventomorrow. He structures don't have the time to do all that, but I think that's.What we're looking at is within something that's based on sound wisdomfrom practice practice of great instructors based on sand learningtheory. We know this is the right over all experience. How can we bring in a Imake it feasible to do that right at scale, it's fascinating and so againtrying to translate to myself the layman? Is it still helpful for me tothink about it? Is there is still that sequential experimentation happening,but within a story you can't just take chapter for and put it in front ofchapter one without synthesizing and creating new transitions, and is thatthe missing piece of being able to utilize a I within a story? Yeah,absolutely and then also you know, bringing it to students in experientialformat that that sort of fits too. For example, a voice assistant instead of anew text to read, perhaps am imagery that may weave in some of their owndrawings and so forth, but can textual ize that- and you know just to make- Imake this feel realistic, but not just realistic to get the learning relevantresource in the right place. So I I know for me it would be very excitingif I was part of a scientific mission, and I had made some plans in my notebook. If now, some of those miloo pages were automatically coming together inan image for me that looked like. I was part of just this rocket ship thatwe're getting on, and I would really feel I sometimes these cues really cure you that it's real and it's justthat feeling of it being real. You know what I students complain about abouthigher education. That seems so...

...detached for reality. So what can we doto make this feel like? This is a real experience and you're learning in areal. I think a little bit can go a long way. Jeremy, I'm not sure. I canwait five years to see what your team builds. This is too exciting, but forthose of us who have to stand on the sidelines and cheering you on, but butwant to start thinking better about Ai. Today I mean next steps. Advice forinstitutions who are looking to leverage where I ai is today andprepare for where AI is going to improve their students, learningoutcomes. Where should they start yeah great questions? You know there aresome nice readings out there that just summarize the state of the art of wherea I is that's one thing is just to understand what the technologies areability to do. naturalistic input like voice like making sense of a sketchabilities to sense patterns, abilities to synthesize a constructive action ortext so anyway, some reading about what the capabilities are. I think it'simportant to compliment that with some readings about human seturday. Isometimes it's called responsible. I sometimes called Ethical II. We haveissue to bias because there are a lot of challenges with this technology aswell, and so tracking. Some of that those issues and becoming aware of thatis super important, starting to think about policies and safeguards andreally how they'll be informed. consitutes of anything that mightbecome available and don't just take promises for granted going down thatpath. A little bit more, something I think is important- is just developingsome ability to look under the hood or have someone on your teamwho can,because there are superficial promises being made out there, that somethinghas ai it's good and sometimes what's in it is really trivial and the teamdoesn't really the team that built that thing really hasn't gone, particularlydeep. So as a kick the tires process so anyway, those are the three things I'dsay you know read about these fundamentals that are exploding andopen your eyes, that think about the ethics and think about how you're goingto build a team that has enough capability of really kick the tires.Jeremy thanks. So much for your time today. What's the best place forlisteners to reach out a they have me follow up questions yeah. I would behappy to engage listeners by writing to me at my email address. It's j Rochelleat digital promised that ord J R O S C H, e l, Le At digital promised thatword, and you can also pretty easily find me on link Dan if that's easierfor you, Jeremy thanks. So much for...

...joining us today, best of luck to youand your team we're all counting on you. It's been a pleasure. I hope, to seeyou on our rocket ship error attracting today's new post.Traditional learners means adopting new enrolment strategies. helic educationsdata driven enterprise, wide approach to enrolment growth is uniquely helpingcolleges and universities thrive in this new education, landscape and Helixhas just published the second edition of their enrollment growth playbook,with fifty percent brand new content on how institutions can solve today's mostpressing Androma growth challenges download it today for free at HelosEducation Com play book, you've been listening to enrolmentgrowth university from helic education to ensure that you never miss anepisode subscribe to the show in Itunes or your favorite podcast player. Thankyou so much for listening until next time. I.

In-Stream Audio Search

NEW

Search across all episodes within this podcast

Episodes (223)