Coming Back Full Circle with Nada Dabbag

All right, we are live on DS 106 radio

I'm Meredith Huffman, and I'm joined here with

nada

dabig from George Mason University to talk through

her session on

This this hour of the reclaim radio. I'm also joined with Taylor Jaden as well to introduce as we get started

Great. Thanks Meredith. So I'm nada

Daba I know my last name is a little difficult to pronounce but

That's fine. I

am a professor at George Mason University as

Meredith mentioned I've been at George Mason for a long time too long almost 28 years now, but

I'm also the director of the division of learning technology

So my specialty areas are of course looking at the affordances of learning technologies and instructional design

And how can we?

design

effective and meaningful

learning experiences using

technology and so I

Was interested in this DS

106 radio because I do have an account on

Reclaim hosting and I was interested in this summer camp

Thing because I've never done something on radio before and so, you know, I'm

You know sort of old-fashioned. So when you guys said radio, I'm like sure, you know

why not get my voice out on a

Radio which you know in my mind thought about it more like a podcast kind of thing

But you know, yeah the word radio sort of, you know rings, you know old

media to me so I was I was intrigued by that and

You know, they asked okay pick a theme so I picked a theme

You know about storytelling and I know the gym is very much into storytelling and so

I kind of wanted to look back at the last academic year and sort of share

my highs and and lows and so

I

decided to

title my session

You know looking back so sort of coming back for full circle

so, you know, I'm at a point in my career where I'm looking at our field and it's just really difficult to

to

Define our field if you will, so I'm gonna I'm gonna share on my screen now and

Just show this one slide, you know when I graduated from Penn State

in

1996 my degree said

Instructional systems design so at the time

If you see, you know

Down here at the bottom of this slide. I'm going to highlight it

instructional systems design or ISD was really the

Kind of the name that my field went went by right because we were looking at systems

We looked at designing instruction

Using learning systems. We borrowed from computer science and engineering. So all about systems theory and

You know, how can we design an instructional system? But if you look at the slide again, you'll see that you know

the associations that I

belong to for instructional design like a CT and you know

you learn and

ACE and others they call our field so many different things like here at the top. It's called

Instructional technology, which is a much broader term and it's defined as the discipline in which learning theories become concrete

through the use of tangible

resources such as you know

Technology. It's also called instructional design and it's also called

Right now it's called, you know instructional design and technology

But right now we even are calling it learning design and and technology. So

I'm gonna stop sharing now. So we even are calling it now user experience design

So learning experience design, so there's you know, there is so many different

Notations for our field that it's it's confusing and

I'm trying to you know, look coming back full circle

I'm just trying to figure out how did our field evolved and what what is the cause of this?

Constant, you know evolution and I came to terms

Thinking that you know

We are always piloting change in our field

Why because technology is just he's changing and we have so many different affordances

Look at what we're doing right now

I mean we're meeting on a on a platform that you know

Every one of us is in a different location and we are you know, seeing each other and we are in our homes and we

are

Recording so

I think that our field or my field has always been sort of I wrote down like in in this

Pilot mode or in this beta mode kind of like software, right when Microsoft comes up with a software update

It's you know, why do we have so many updates, you know, why do we have so many like, you know

we can't we keep we keep doing updates because

Technology is always in beta mode

I mean it has to keep evolving and it has to keep

We have you know version one and version two and now we are Microsoft

11.1 and who knows we keep going until version

20 point something in every software that we use and then finally the whole software either dies or something else comes aboard and then

Now we have to learn how to use something else or a different operating system or a cloud platform or whatever. So

One of the reasons why I think our field is in constant beta mode is because you know

instructional design and technology uses technology to support

workforce education and training in all kinds of different contexts and

multimedia and all that so I think that

Yeah, that is one of the reasons why we keep we keep evolving and the name of our field keeps evolving and now we're even

Calling it learning engineering that kind of comes back in circle like yeah, and I see Taylor here

You know nodding and so you've heard the term. Yeah, I

It's probably unfair

Because I always kind of a little bit laugh at learning engineering because to me it always feels and again

I I don't know a lot about where the term comes from but from the outside looking in at it

it sounds like someone felt like

Technologist or instructional designer wasn't legitimate sounding enough. So they were like, well, what if we put math in here?

Let's call it engineering and I'm like, I'm not sure how much math is in that role on an average day

But whatever, you know

I'm being a little bit hypercritical. It's probably an unfair take of that term, but I know it kind of makes me smile, you know

Yeah, thanks for for for weighing in so I'm gonna

share my

screen again because

Yeah, I'm trying to grapple with this term as well

Because

The the definition and as you could see now on this slide

It's really you know, what characterizes engineers and I think it's like design is the glue, you know

Engineers are designers, you know, they invent this from Wikipedia. They design they analyze they build they test

We do the same as instructional designers and learning designers. I mean we are

We're designing

learning interactions, we're designing learning experiences and so

We have to you know, look at our audience look at the context and design and do a needs assessment and all that

So it's I guess the scientific method which is sort of you know, we hypothesis testing

I don't think we do much of that

But we do definitely have to problem solve because we go into an organization and we're like, okay

Your performance is down and you think you need

to get your employees to learn something new or to upskill or to reskill or to retool and

Therefore, you know you need us instructional designers to create some, you know learning interactions for your employees

And so that is a problem-solving approach, right? Like

That is where we yeah, we need to go into that organization and figure out you know

What is what what skills do they need to improve their performance?

And and you know, what type of audience do we have here? Are they adult learners? Are they you know?

K-12 learners are they students? So yeah, so going back to learning engineering, I think

The definition in Educause 2018 is the application of engineering

methodologies to develop learning technologies and infrastructures to better support learners and learning

So well, that is the official I guess definition. So it kind of makes sense

but again, you know our field keeps evolving and now for example at

What is this the Carnegie Mellon Carnegie Mellon

I only know of two universities that have adopted a learning engineering curriculum. So Carnegie Mellon has a

metals program

That you know Masters of Engineering in technology

and

I forget the actual what the acronym stands for but these are the

you know the outcomes if you were to get that master's degree and then I'll also

up in

Boston College

They also have a new learning engineering curriculum and you know, I guess there again

they're trying to integrate more the

the hypothesis testing the

the the infrastructure the systems thinking and

You know

making connections between technology and developers

Because you know, this is what really bothers me is the software developers. I'm sorry

I'm not trying to offend anyone but the software developers

you know, they're very smart people and they they just like develop all the software that we use but without

Taking our input they don't come to me as a faculty member at George Mason University and say, okay, how do you teach?

What's your pedagogical approach? You know, do you use experiential learning with your students?

Are you a learning by doing person or do you just lecture and then give your students midterms and finals?

so

software developers who are developing platforms for learning or for educators need to speak with educators and

really understand what educators do and how do we teach or how do we do research and

you know, for example, you know, they've come up with all these learning management systems blackboard and canvas and

You know who knows so many of them, but they're just built in my opinion those software developers

They did a great job. I mean the university does need a centralized

System, I guess because we're still mired into that credit our system FTE and the courses

what they do is they just like

You know

Have a bunch of courses connected and it's all the same format and it's all based on the faculty

uploading their content to that platform and then you know

Stereotyping their lectures and uploading their lectures and they think that that's it. No, it's not it

That's not how I teach. I don't want my students to be just you know, listening to my lectures and just then

Taking a test test test and then just you know, I don't want them to memorize the content

I have learning outcomes and that's where instructional design comes in

I mean and I try to align the learning outcomes in a course that I'm teaching to the learning activities

To the assessment. So that's the backwards design model in instructional design is when you really have this

pedagogical alignment right where you're really trying to

Be you know align

the the the learning outcomes with the learning activities that you give your students or employees at an organization to

Work through or the learning interactions or the learning experiences, you know to to the assessment. And so

Yeah, so that's why that learning engineering term, you know, I'm trying to again figure out why

There's so much push to move our field again. So instead of calling it instructional design and technology now

We have to call it learning design, but now we have to call it learning engineering. I don't know but

Going back to my coming back full circle one thing I can tell you that I felt really

cool about and I was so like so that that kind of maybe stayed the same and made me think you know, no, there is some

grounding here because

Lately and as I as I reading back that submission that I did for this for this radio summer camp

It said, you know looking back at this past academic year

You know, I engaged in two grant writing experiences that took me back full circle to my dissertation study

My dissertation study was back in 1996 and it was about personalizing

Instruction to students most preferred and personally relevant interests

So that was personalization at the time and what I did at the time is

You know

Students at Penn State just didn't like statistics. Nobody liked to take a course in statistics

They all had to take a course in statistics and the way it was taught was very

conventional like, you know, they wanted to teach them how to compute means and medians and averages and

Like that scales and they did it through

because it was a gen ed course that everyone had to take and so they did it using lectures and

I was like, why do students they don't like the way this course is designed. So let me try to figure out

How to design it in a way that is more relevant relevant to their

Experiences at Penn State as a student. So actually what I did was is I had them

Collect data like I would ask them how many beers do you drink a week or you know?

And I will have them create data sets based on their life

experience like how many how many parties did you go to this week or

How many you know?

Dances, did you go to or how many girlfriends have you had?

since you started your freshman year or so I was trying to relate the

The learning of statistics to their personal

Relevant interests and I thought at the time that you know

If I connected numbers to what they're doing in their daily lives as students

Maybe then so then they would build data sets as groups

And so we would have all these data sets of the number of beers

That they drink at a certain party the number of this or that and then I would ask them

Okay. Now, let's let's compute the average of

the number of beers that you drink let's compute the the mean or the median of

The the number of parties that you went to or something like that

So then they can get a sense of how this data, you know is transformed through statistics

To to give us some, you know results that they can work with or evidence-based

Results or empirical results so they can understand the value of statistics. So anyway, I'll stop in a minute

But this was this was my dissertation and now I find myself

After 28 years. I just published a book

called a personalized learning experiences with one of my doc students and it's called designing personalized learning experiences a

framework for

Higher education and workforce training so circling back, you know

28 years since my dissertation

I'm back to personalized learning. So that was really cool for me because I was like, whoa

It's it's a hot topic now and it was a hot topic when I did my dissertation eons ago. So maybe that is something

That's really cool

In my field that I can talk about I have another you know concept that I also want to talk about

But let me know if you want to add something or ask

Sure. Well, um, there's a there's a lot of

Discussion going on in the discord. So there's there's plenty of shared

Disappointment or frustration with the LMS just want to let you know

So you're definitely not alone there and I will say as a technologist who spent

About eight years in in working in higher ed

I don't know that our my motto at my previous job was literally no one likes LMS

There isn't a person right? It's designed for nobody because it's attempted to be designed for everybody

That's the problem for sure and and you know one thing I wanted to kind of add because I think you're completely right about

The way their design doesn't meet

The needs of teachers or students and I think it's it's kind of a simple problem

And that's just most people don't know that the field of education

Educational psychology exists or what it says

Just most unless you unless you've been through teacher education training or you've encountered this or you've studied teaching

most people just aren't aware of that stuff and so I think

you know folks who are

Planning what features an LMS should have and what what they should provide

They're not really thinking in that light from that angle there there plus there's all of course also business motivations

Why you know they have other?

Objectives and goals than you or I would someone who's using the system

unfortunately

Because I'm gonna share my screen again because you are spot-on in terms of

Our field what might be an instructional design? I have a slide here

It borrows from at site like you said educational psychology is right here the social sciences

So we you know

We are steeped in the learning sciences or if you want to call it the cognitive sciences how people learn and of course at site

educational psychology and learning theory and then of course we borrow from the information sciences communication theory

computer science and now engineering and

Information management and in the old days it used to be audio visual technology and then at the bottom here

I have the management sciences the systems theory the systems analysis and design the operations research and

project management, so you're spot-on by saying

You know most software developers wouldn't know that if they're

Developing a learning system or platform to you know think about how people learn and so

you're spot-on and so

The the new however the problem the problem is because I do have a

sort of a solution and Jim is aware of

my

My other research interest and we personalized learning, but I'm also totally into what we call

Harold your jarci, I mean personal learning environments the problem with and I'd like students to

Create sorry. I'm scrolling through here to find the slide on PL ease, but the idea is to allow

You know students to create their own personal learning environment right using whatever tools they want, right?

They could use they could use a wordpress. They could use wikis

They could use you know, I mean, of course, they're gonna have to create a cast the problem with that though

And then we as faculty we become like facilitators for their personal learning

Environments and they try to then create their own content. They create their own learning interactions

I mean, that's huge

I'm a constructivist at heart meaning that I want students to create their own learning experiences

The problem is in higher education if we don't break away from that credit-bearing FTE core system

mean how is the institution going to

Demonstrate analytics or you know, if we don't go into X API and I have to figure out across all of that

The systems talking with each other. I mean that that's that's a huge impediment for

for that's why the LMS is still a staple because you know

It has everything in one place and people can go back and see the courses that they took and we can have surface learning

Analytics and figure out this or that and copy courses and the students are used to like a design

But ultimately you're right

Like if we you know, I don't want students to be tethered because when they leave the university then what I mean, you know

They're kind of done. What about lifelong learning? So then, you know, there are there are systems now that support lifelong learning, but again

like the Defense Acquisition University the ADL lab are considering a

Platform I forgot what I used to call it

I mean where where every one of their employees would create their own personal learning environment and then they can go learn something

Via YouTube or MOOC or they want to learn something new using some other system

but then there needs to be documentation and so the

Organization needs to document that my employee got a certificate or learned this skill or that skill so they need you know

Performance measures on that so that that is a problem, but I think if we can eventually get to that personal learning environment

Approach where every student is empowered to create their own digital learning space their own personal learning

environment using the the technologies that they are comfortable using

But then the teacher's job will become much more difficult because you're gonna have to kind of you know

Go into that personal learning environment for every student that you're teaching and then figure out, you know

How they're doing and are they actually processing the content and are they actually achieving?

so

So yeah that that thing but personalization is huge. I would say circling back to my dissertation

Has been really cool in terms of you know, thinking about

Personalization in a different way and then the other thing that I wrote in my abstract even more surprising

Was the recent connection I made between generative learning and generative AI

I mean that was something huge to me when I was studying at Penn State back in

1996

What's his name?

Whitrock, I believe was his name. He was I studied we read we read articles

for an author called Whitrock's in

1990 generative learning theory. He it's a theory

There was no AI at the time or at least not the AI that we know and now

So there was generative learning theory and then there was Weinstein's learning strategies model

well

I mean in the age of generative AI this is now making a lot of sense to me because I think about the concept of

Mind tools and I like to use technology as a companion as a mind tool

I like to use AI as a mind tool and to help me

To help me figure things out. So

My mentor David Jonathan who passed away in 2012

He was the one who introduced me to technology as a mind tool so that we can you know

Just like we use other tools in our

as humans to build things

Technology can also be looked at as a mind tool to help us learn and so

generative AI

You know, it's making me think of generative learning just like just like generative AI

Generates text from large language models. We as humans, you know can generate learning

from our experiences and from

our learning experiences with the help of

Generative AI so if we can customize if we can even you know

Build a customized version of chat GPT or co-pilot or whatever or Gemini or whichever, you know?

Gen AI tool you want if we can if we can teach our students how to

Create a personalized version of that Gen AI tool

To help them learn the way they learn best and their in corporate their interests sort of going back to you know

Personally relevant interests. I think that would be huge. I think it would be huge to have students and learners, you know

Understand how to customize the use of Gen AI tools for their own

Benefit in terms of you know using it as a mind tool to learn and as a lifelong learning tool

So I was engaged in this

Research proposal that we will have a center at George Mason University where we will

You know

We will we will use Gen AI we you know

We will customize Gen AI tools for students to become self-regulated learners because it's not easy

What I'm saying is not easy. I mean not every student or learner is able to

Figure things out on their own. That's why they need us as faculty and scaffolds to help them

But it's not easy

But if we can get them started to become self-regulated learners and you know

Evaluate their personal learning environments as they go through and figure out whether this is working for them. I think this would be

huge

Yeah, it's amazing to think through what could be possible there there's so much

You know

Personalized learning is such a big area

obviously of of interest I mean, but

what I'm kind of always curious about with

In the context of that in AI is you know right now?

I mean and maybe someday we'll see a different approach to these tools that will

Work differently, but AI is real good at reflecting back sentiment to you

or

Kind of what I would I cut I like to call like a creative mean where like it's sort of like

It takes it sucks all this content up and will deliver you some middle percentage thing something that's kind of

Acceptable and average and is you know, whatever that means obviously that is its own

can of worms but

I'm really curious to see

like I because I can see like the idea of

Personalized learning is like sure

Yeah, AI like a large language model can spit out like some questions or different variations on things

But what I'm really curious to hear about from and maybe I'm just not aware of the research too because I'm not necessarily looking at

journals every day, but

Has anyone been able to successfully

Use a tool like that to to personalize learning in a way that students find helpful

Or is it just is it a theory right now? It's been unproven so far. You know, I don't

Kind of yeah, I'm thinking about right now

Yeah, that's a good question. I mean, I

Don't have data to support. I know that

people in the corporate and workplace settings are

using it more like

Like a performance tool like to help them make their job more

Efficient or effective right? So we do have data on

on that I was at a

So so for example if you want to send

An email communication you can ask chat GPT to help you write that but I want to hear more about

your thoughts Taylor about

You said something that really resonated with me here is AI as a creative mean mean

But more AI as a reflective tool

how how do you see that because I think that ties a lot to my research on personal learning environments and

Self-regulated learning because if we can get learners to become

reflective and

evaluate their learning space and their digital learning space and you know figure out if they are actually

Achieving the learning outcomes and the goals that they want to achieve to become

Better employees or more efficient effective whatever or meaningful learning, you know

What how did you come up with that AI as a reflective?

- well to me, it's something that I'm interested in as a possibility. So one of the things

and and I'm it's on my brain a lot right now because

Later today. I have a session where I'm gonna be playing back a project where basically I interviewed an AI

I

About AI and education. So it's it's weird

it's a weird project but I wanted to do this because I wanted to come to terms with the tools a little bit and I

am

Interested in the possibilities of these things but also, you know

pessimistic about the business realities that tie them and and there's you know, there's plenty of things to be

You know or or or like the realities of how artists and musicians and there's a million things right

You don't necessarily need to go in

It's at it's at two o'clock one Eastern I believe

There'll be a recording of it too, but what I what I figured out though is

because I went into it thinking it would be a

Exercise in me asking kind of probing and critical questions and it being as an AI

Positive right about AI and its uses and it really wasn't I think

Because it picked up on my tone of the questions. I was writing immediately and mirrored that back to me

You know, I mean

Yeah, so but that's that's just this tool in the way. I prompted it, right and my point is though

Is so that was one thing that the more I did this project the more I was like, you know

It's it's reflecting my tone back to me

It's not it's it will and that's what I mean is it's not in that way right because I'm feeding it text that is

Tinge negative in some ways and so it's gonna respond to me negatively

And that that's that to me is something that could change right? This is how the tools work right now

it's how I use them, but that's sort of the my point is that like

the quality of the response is so reflective or so

Dependent on how you prompt it and I'm curious to see where

These things will go in a way where people can use them as reflective tools

They can because because I know people already do this though the layout an argument and say hey

Criticize this for me right and that can be a useful way to use a large language model

My problem is I didn't tell me I didn't tell it. Hey, I'm gonna ask you kind of negative questions

I want you to to go back at the questions and try to try to criticize my arguments

Because I didn't specifically tell it to do that. It's sort of commiserated, you know

it just seemed because I I was at the learning ideas conference at Columbia University in June and

The speaker I'm trying to remember which speaker

It was they think that prompt engineering is BS. Sorry, but I mean like they don't believe in prompt engineering

I mean since the AI came out everyone's talking about prompts and prompt engineering and it all depends on your prompt. So this

This guy I don't think it was Chris DeeDee, but it could have been he was the keynote speaker

You know, we are a Chris DeeDee from Harvard. Yeah Meredith has so um

He may have been him in prompt engineering is a marketing ploy the prompt changes every day

So the thing the same prompt you engineer will not yield the same answer

even if you typed in the same prompt and so

There's a ceiling on LLM's is what I guess they're trying to say, you know

it's not like you know, every time you you change the prompt you're gonna get a

different answer but going back to your point about

Yeah, I mean about it changing its tone that's that's pretty

That's that's I guess that's where that reflective maybe part comes in

for when we're using it as a mind tool with students or as a companion where

Students can really maybe use it to have it assess their learning. I don't know that yeah

You know, I think it could be used in that way, but I think folks would have to know

about the

The by the biases is not really the right word

But like the tendencies of the tool and how to set it up in a way

I agree. I I don't like I don't know much about like

Creating good prompts, but but you know your point on prompt engineering

Would you ever engineer something where you can't actually have visibility into what's below what you're engineering?

Like you wouldn't build a bridge on ground or

Over ground you haven't surveyed right or build a building is a better example, and I think that's a that's a good point

Yeah, and and I'm interested in that the the it you it's used as a reflective tool

But so far I've been a little underwhelmed and I am

curious about

You know if the large language model thing even has a path forward because it does seem like

You know two years ago. We're like wow imagine how good these things will be in a year, but I don't really think this

Line of generative AI is gonna get better in certain ways at least it will get better at certain things

But it's probably always gonna be bad at some of the things that's bad at right now at least yeah

But it's impressive that you were able that it had a tone as you're mentioning

And yeah, I guess yeah, that would be something we should do research on because you know when we're gonna use it more and more

It's important that you know to

Give you back a positive tone because if it gives you a negative tone it could you know it could affect your well-being it could

affect your mental health right and so

Yeah, so Chris Dede at that learning ideas conference. You know

Try to make the

the

Association between you know well AI is more like

You know the brain versus the mind like we humans

You know have the mind and the brain, but AI is kind of like just that brain the neural circuitry

So it's pulling from large language models, and he was like you know

Humans make are better at making judgments and decisions whereas AI is better at reckoning

I you know he made that distinction. I had to look up really what reckoning means, but I mean it's murky like you know

But he was trying to make a distinction because a lot of people are you know just saying you know

That AI is just gonna Wow like it's just it's just gonna

Disrupt our lives, and I don't know but he was like this is like the ninth

Hype cycle that we've had and they've been wrong for nine times

I mean you know what I forget what the other hype cycles were maybe I remember

Remember the 2000 where we had the Y to everything what was gonna happen then the whole Y2k?

Because everything was gonna change because we were gonna have two zeros at the end of the year and then everything was gonna like

All of our data systems were just gonna like and so he talked about we've gone through nine

Hype cycles for technology, and they've been wrong nine times. I I'm not sure I

Agree with him, but I'm just sharing and then he says you know LLM solve the natural language

processing problem

But AI is like a parrot capable of processing, but not comprehending

I mean you know that that's why I'm opting to think that maybe

You know that maybe

We can use AI as a mind tool because it's not gonna overpower us or it's not gonna like have all the answers for us

Because I want my learners

You know and when they when they design instructional systems. I want them to be able to you know

Generate their own content. I want them to be able to think through the design processes

I don't want AI to tell them

What to do so it's it's it's complex because when I like I'm teaching a course that starts in two weeks about

you know cognition and technology the intersection between the affordances and capabilities of learning technologies and and

our pedagogical sequences and cognitive

states of mind and learning preferences, etc

And so I asked actually

Chat GPT to generate learning outcomes for the course and it did an amazing job

I mean I've talked to score several times and I was like totally impressed with the learning outcomes and generated for the course

So so honestly, I I don't know. I mean, I don't know how powerful

these generative AI tools are but one thing that you said Taylor is

I

Think I also heard on NPR radio that

that

Large language models even here in Virginia where I live and you know

because of the processing speed when you ask chat GPT something and it

Processes it in like five milliseconds

the data centers that are pulling from those large language models are like

using so much power and electricity that it's not sustainable apparently and so

That is why like you said Taylor

We may need to move away from large language models and they talk about small language models

I don't know what small language maybe small language models are more like

more like

Customized, you know like by topic or customized so you it's only pulling from a limited

Dataset versus pulling from from large language models, right? So, you know anything about that?

I mean, yeah, I mean I know a little bit but

Like my understanding with the with the smaller language models is that essentially have less

Connections in their data set is one way to think about it. And so their their responses are

Average more to the again, I'll call it mean the center right the most likely response. You're less likely to get

Like interesting or outlier kind of responses in in those in those is one part of it and another part is

context window so like if you're if you're having a

if you want to use a small smaller language model and you're

Conversing with it right which and and one way I've heard this described to me

way like chat works in large language models is think of it like a

Like a like a pachinko machine like the ball with the pins, right?

you drop a ball down and it goes all the way to the bottom and

The model is where the pins are placed and the ball is the text that you've your your prompt

basically what you've put into it and when you you put in your prompt and you drop the ball down and something comes out right and

it's

Influenced heavily by the pins right and then when you ask it to follow up

So you say, you know, blah blah blah and it it responds and then you ask a follow-up question

You're submitting the entire history again. So you're submitting your first question its response and your follow-up question

And the longer you do that and so it's it's it you can think of it like autocomplete in that way, right?

Yeah, and so this is again something that I kind of learned or I guess saw demonstrated when doing this project

I was working on because I had a hard time keeping it on in the conversation context

until I figured out how to use the tool properly to do that and

So

So the smaller models have a smaller context window, which means they can't keep as much of your prompt in

Memory, and I don't mean memory in a computer sense. I mean memory and whatever AI

I actually don't know from a technical perspective what you call that. I keep hearing it called context window, but

Maybe I guess it might literally be memory from a computer perspective. But so that's that's one thing

I really think though that and I don't think this is great

I really wish the large companies that would do this with that are doing this with this huge environmental impact

Would not have done this

But I think that what they're banking on is that what we call large language models today in three years will be small and

They can run with low energy usage and on

Directly on your phone or something, right? We're seeing we're already seeing this with with

Stuff on Android and Apple is got their thing called Apple intelligence, which is coming out next year

or end of this year next year and it won't be as

Sophisticated as the like chat GPT 4o model

But it will be you know

It can do some it's a smaller version of something like that is my understanding. Obviously. I don't it's not out yet

So I but yeah, I think that's where they're banking on this

I think that is a very Silicon Valley way of doing it and I don't really appreciate it

But I can see their line of thinking

You know

But I mean, can I have like when I teach this course?

Can I can I create I guess I can I create a customized chat GPT?

Based on the content of my course only like I can feed it

Can I feed it the articles and the book that I want the students to read because I'm the expert on that topic

can I can I feed it the the readings that I want them to read and then the

I don't know some other content and and and then have them

Query that can I do that? Can I create a customized?

Chat GPT just for my course

Yeah, you you technically can and I I am less familiar with

the actual chat GPT like that that - of course I've used it but

Most of the stuff I've been using lately has been the open models like llama

And I don't I don't I'm not a heavy user of these things

But that is one that actually can run on your laptop with a program called Oh llama

And it's on my laptop not real fast. It's maybe not quite as good, but it does work

and

You can you know give some of those tools that the if the developers have set it up this way

Essentially access to a bunch of text that sort of gets added

Before your prompt if that makes sense. Yeah, and I my understanding is the chat GPT

Subscription lets you give it a library of documents too, but I don't actually use that

Yeah, I mean I also have to ask my university if they would allow me to do that because you know

Yeah, but doing that with an open model that can run on your computer locally. There are less

Problematic implications of that right because you're not actually sending

Yeah, then that doesn't work yeah, we I will also plug that we did a

Reclaimed TV stream a few weeks ago about this tool called open web UI and it's like that

But it runs on our cloud or or any cloud for that matter really, but it's it's local to that instance

So the the data doesn't leave

So if you could delete it at the end of the semester and it's gone. It's not been sent to Microsoft or whatever

So that could be something that you could use but you know, it's it's expensive to run those

I'm it's very clear to me that

Chat GPT and the alternatives are really heavily subsidized right now by a venture capital and stuff investment

Okay, so in closing and then I want to see if there are any questions in discord or what else is happening

I just want to say that again, you know going back to

Coming back full circle to personalized learning and how I'm trying to in generative learning and generative AI and

trying to figure out how to how to integrate all that and I want to say that you know, I

Just hate it when people say Oh technology is just a tool to me that is really really

not

Correct, because I believe that you know

Technology is

Has affordances and so technology shapes our socio-cultural practices and like I started, you know

Look, look what we're doing here, but it's really you know, we there's a famous saying I forget who said it

No, it's like, you know, we built our tools and then and then these tools in return shape our lives, right?

I mean, so we we you know, it's just like, you know

We make things we are tool builders as humans and we make things that change our lives

And so theory of affordances and Gibson is the big

Communications scholar who said, you know media is the message

So for example, if someone were to listen to me over video

They will get a different

impression or or learning experience than if they said they just listened to my talk over audio right because

The media is the message different media

different affordances and for learning and for teaching and for

Whatever so he was you know

he was big on on the theory of affordances and

So I want to emphasize that I think there is this reciprocal relationship between

you know between technology or the tools that we use and

Our knowledge or cognition and so we don't control technology and technology should not control us

But what was the theory of affordances as we we become what we behold?

We shape our tools and then our tools shape us in return. That was actually from Marshall McLuhan

Marshall McLuhan is the the huge communications

scholar I believe and

But then Gibson is the one who came up with the theory of affordances is that you know

Look what happened when we built cars and then cars shaped our everyday life and when you look at you know

The Industrial Revolution where we are right now

and so things evolve going back to how our field is evolving our field of instructional design is because

Mainly because of all these technologies that we build whether it's the Industrial Revolution

Whether it's airplanes whether it's cars whether it's instructional systems. All of these things are

Shaping after we use them. They are shaping the way we understand the way we communicate. So I just wanted to

emphasize that I'm a firm believer that technology is not just a tool that it is a

You know, it has its affordances and it it's it impacts hugely

our our instructional designs and therefore when we design

personalized learning experiences or whatever we're designing as instructional designers

we need to be aware of the affordances of the

Technologies or the platform that we're using and that's the course that I teach for my doctoral students

They have to do what we call an affordance analysis and there are you know, physical affordances

cognitive affordances

pedagogical affordances

So many different types of affordances negative affordances positive affordances false affordances

Hidden affordances. There's so many different types of affordances and what I engage my students to do is to do this

affordance analysis of a platform that they like whether it's an LMS or

WordPress or some type of technology and come up with all the list of affordances for that specific technology and then

Figure out what are the best affordances to capitalize on when they're designing instruction?

Okay, so I think

Do you want to ask a question kind of going back to your open?

learning environments or personal learning environments

Shannon mentioned in the in the chat a couple minutes ago about

finding it tricky to

See the line of matching students with their interests and ways they learn

but then also making sure to encourage them to push them outside of their comfort zone and

Wanted to see like in your experience if you have any like tips or tricks on how that might work or

How how you you've seen success in that or how you can kind of frame your your curriculum around that?

Right, so I have not actually

been able to

Have my students create their own personal learning environments because my university

wants me to use the LMS and so I

wish the LMS had a place in it where

So what I do is I have students create their own wikis

Which I believe is sort of a way to

Implement a PLE approach I have them create their own blog spaces within the LMS

So I've done that and that has been

Very successful. I do have some research papers on that

That Shannon might want to read

We did look at blogs and wikis and one more

Discussion boards I think as as technologies within LMS is that can support

students to build their own sort of digital space and

And reflective space where they can, you know reflect on their learning in the course

Yes, I do have a couple of articles where they could use those within the LMS

I've never I've had some colleagues for example create

Facebook pages for their courses. I have not done that but that's another way of implementing a PLE is that if you

If you create a Facebook page for your course and then allow students to create their own

personal learning networks within Facebook

Reach out to other learners kind of that goes back to my personalized learning interactions framework, which is in this book

That was just published in that framework

We have six types of interactions and I think I can overlay on that personal learning environments because students can you know?

Interact with other learners they can interact with coaches. They can interact with AI

Outside of your course, it doesn't have to be all

You know, that's a great question because when we think of a course with in higher ed we think you know

This is a closed space. That's it. That's my course it goes on for 15 weeks

These are the readings that they're doing

These are the students that they're interacting with a personal learning environment and a personal learning network opens it up

to a much larger network and allows

Students to interact have different types of interactions with different technologies and different people

Yeah, definitely

Awesome. Yeah

E-learning techie and the discord chat says that Facebook would so not fly where I am for privacy concerns. So

Can confirm on that side

Well, cool, well, yeah, thank you so much nada for this awesome discussion and Taylor for

Talking through all of the awesome stuff. We've got

Maren up next working through some

music streams for we've got a double hour or two hours of

lunchtime tunes

Coming up. So

Before that so maybe I'll do it at some other point, but thank you. This was great talking with you

I enjoyed having Taylor and you Meredith because that made it more, you know

Interactive and social and you know, we covered a lot of ground

But so this is gonna be recorded now and then I can have like you're gonna send me a link that

that has

the recording

We will share all of that

I am gonna pass it off to Maren and you're listening to DS 106 radio camp from here and we've got the tunes