How AI Will Impact Education and Beyond with Julian Moore 

Full Transcript

Dr. Cristi Ford (00:00):

Welcome to Teach and Learn, a podcast for curious educators, brought to you by D2L. I’m your host, Dr. Cristi Ford, VP of academic affairs at D2L. Every two weeks I get candid with some of the sharpest minds in the K-20 space. We break down trending educational topics, discuss teaching strategies, and have frank conversations about the issues plaguing our schools and higher education institutions today. Whether it’s ed tech, personalized learning, virtual classrooms, or diversity inclusion, we’re going to cover it all. Sharpen your pencils, class is about to begin.

Dr. Cristi Ford (00:37):

Many of us have been talking and thinking about the ways we may be able to use AI, either in our work, in our life, or in the classroom. We’ve had some great introductory conversations in season one of Teach and Learn, and today you’re in for a real treat, as this episode will surely take those initial conversations to the next level with my guest today.

Dr. Cristi Ford (00:59):

Joining us in this virtual classroom is a self-proclaimed technology geek who not only has a deep understanding of the AI world, but also works tirelessly with associations all over Australia, helping them to harness the power of this new technology. He’s a leading not-for-profit sponsorship practitioner, and AI implementation specialist, and the director of Strategic Membership Solutions based in Australia. A dear colleague, and someone I’m glad I can now call friend, Julian Moore, welcome to Teach and Learn.

Julian Moore (01:33):

Hey, Cristi. Hello, everyone listening. I’m very excited to be here. It’s wonderful to be invited and really great to see you again, Cristi. Last time we were together in Sydney, and now it’s virtual.

Dr. Cristi Ford (01:48):

Yes. So, a little bit of a backstory for our listeners. Back in September, Julian, you and I had the pleasure of connecting in good old Brissie, a term that I now know. For those who are not familiar, it’s what locals used to refer to Brisbane. And I really enjoyed talking with you and the ways that you were engaging on the bleeding edge of where AI is going. And so, when I got home and started to talk to my producers, I thought to myself, “We’ve got to have Julian on the show.” So, fast-forward, my producers reached out and asked if you would talk with us about AI. And I got a quote from them that you said that nothing is off the table today in this conversation.

Julian Moore (02:31):

I’m here for it. Bring it on.

Dr. Cristi Ford (02:34):

And you gave a really great example of an AI mirror for a hairdresser. Can you unpack that a little bit here and just help our listeners understand how this new opportunity really has implications for the educators of today?

Julian Moore (02:52):

Since we last caught up, the advancement of AI has been exponential. It’s grown at such a rate that we really should consider it the next industrial revolution. Because there is no role that’s going to be untouched by AI as it develops, and it’s developing at such an incredible rate.

Julian Moore (03:15):

So, we know when GPT-3 came out. It’s now a year and 21 days old. So, that’s when we all got into using AI in our daily roles. Before then, it was really, “Oh look, it can play chess.” I don’t play chess, so it was irrelevant, quite honestly. So then they went, “Oh, hang on, there’s this chatbot, and it’s useful. It can do some things.” Fast forward to today, it can now see, it can now talk. It’s got an IQ of 155, which is way more than most of the population. It’s considered in the highest echelon. It’s considered gifted.

Julian Moore (03:59):

So, now it can do these things, we can suddenly start predicting some future. And so, we can see that there is nothing. So, I always talk about the hairdressers because the most common question I’m asked is, “Well, how would it affect me?” Or I hear the people going, “Ah, I’m in this industry. It will never affect me.”

Julian Moore (04:24):

So, with that, I always say, well, imagine sitting in the hairdresser’s chair, and you are looking at a mirror, and that mirror has now scanned your face for your tone, your complexion, your likeness. And then it’s decided, here are the different hair options, because now it’s got your hair color, and you can just swipe across with your hand gesture, and that mirror will now give you all the different hair options for your, I guess, your appointment. And that will really give the most basic of understanding.

Julian Moore (05:01):

But now, imagine a camera. And this is a true one. So, imagine where, if you now have a massage therapist with their phone, so with an iPhone or whatever they’re using, and they can take a picture of your back, see which muscles are in spasm, and the AI will guide the massage therapist on how best to treat it or manipulate it. The osteotherapy is already using these type of techniques.

Julian Moore (05:32):

When you look at surgery, realistically, AI is already better at diagnosing all kinds of conditions, because what human can actually ever try to keep up with the world’s knowledge and medical history, all in one place, that you can chat with?

Julian Moore (05:53):

So, of course, it’s just fantastic. Have a look at doc U-S, so docus.ai, and it’s an amazing diagnosis tool. Just incredible.

Dr. Cristi Ford (06:07):

So, I really appreciate the way that you are painting the pictures of the future, and where we are going from a hairdresser to a massage therapist to some of the highest kinds of professions.

Dr. Cristi Ford (06:21):

I guess I’m wondering, as I’m thinking about the educator lens, if these tools and technologies are available for a massage therapist or a hairdresser or a legal aide, how do we need to really start to engage our educational colleagues to think about how we need to make sure that we’re creating a workforce that is prepared to be able to come in and apply for a job, apply for a new position, and be well prepared in these technologies?

Julian Moore (06:53):

It’s a real challenge at the moment, because we’re at the starting line. This is a year-old technology. The only difference is the rapid uptake.

Julian Moore (07:05):

So, we must keep in mind they’re tools. They’re AI tools that each person within their role can use to make themselves more successful, more productive, more creative. And they’re like superhuman tools. They will give you your hours in the day back. They will allow educators to be more engaging. Microsoft just launched a tool for teachers that will plan the year ahead, and it will plan their teaching calendar, it’ll produce the handouts, and it’ll give them the outdoor experiments to do with the kids. So, teachers are very well equipped.

Julian Moore (07:51):

Then, when we look at the other side of education, we start thinking about CPD courses. There is now course.ai. You can go on and you can produce a course of one week, three weeks, five weeks, eight weeks, with all the quizzes, with all the tests, with all the identification and scoring all the way through that will be generated.

Julian Moore (08:15):

But what’s missing, of course, is the educator. Without the educator, it’s just a piece of software that I could pick up and use, and not do a very good job educating, quite honestly. But if I was an educator, I could overlay all my specialist knowledge. I could overlay all my skills as an educator, and just create the most incredible courses in the very short space of time, and really get away from this idea that we would love to do this, but can’t afford to do it.

Dr. Cristi Ford (08:53):

As I listen to you talk here, one of the things it reminds me of, and you talk about the technology and this really being a super tool, one of the things at D2L we really pride ourselves on as we think about AI advancements, is always thinking about the human being in the loop. Always making sure the kinds of advancements that we engage in our technologies, the ways in which we allow these technologies to help educators, that it’s really…

Dr. Cristi Ford (09:19):

To me, I call it a superpower. It gives the educators something else in their tool kit to help them be successful. And one of the things that you talked about, I think I remembered seeing some work that you shared around a final year of high school students. And I think the study was around a study assistant that was available for teachers. Can you share a little bit more about that?

Julian Moore (09:44):

Yeah, by all means. This was not a client. So, my best friend is the director of science of a very highly regarded school, and he’s a full science nerd, like off the charts. He’s awesome.

Julian Moore (09:58):

So, I was chatting with him on a regular Friday evening catch up, and we probably had a glass of wine too much. But we were saying, well, hang on, we’ve got year 12 students now who are about four weeks away from their final exams. So, the most important exams for university entry. It’s a big deal here in Australia. And really, revision, so they need to be able to ask questions of their educator at any time. But you can’t do that because, well, school’s out and the teacher’s not at work.

Julian Moore (10:39):

And I said, well, why don’t you just create your own assistant? Now, we have to keep in mind, right now programming is becoming less and less required. All the tools I’ve spoken about and all the tools I will speak about are no-code tools. So, the only tool you need is language, and the only language you need is English. So, as long as you can speak English, you can create these tools.

Julian Moore (11:07):

We sat down, and over an evening and another bottle of wine, we then created a study assistant. We just used a no code-tool like Droxy, and we put all the year’s curriculum into this tool. So, drag and drop, PDFs. And then we gave it a personality, but we gave it the teacher’s personality. And then literally, typing this, “Lean in, be helpful, never give an answer.” This is explicit. “Only talk about what is on the content of the forms. Don’t speak outside of this.” And then, once we’ve had it and tested it, we then literally gave every child in that chemistry science group their own link.

Julian Moore (12:01):

And they could go in and use it as a study tool, and it was like having one-on-one tutoring while they’re studying for their exam. And it was so popular that within three weeks, there was one of these assistants for every single exam that was going through. It then went on to TV, it then went viral through the state, and before you know it, I think every school now uses a very similar method.

Dr. Cristi Ford (12:33):

Interesting. It’s really interesting to hear the comparison and the contrast between what is happening in other places on the globe.

Dr. Cristi Ford (12:42):

I think the thing I’d love to move us to, and I’m thinking about your working with associations, you’ve been doing a lot of work and talking a lot about engagement. And I will say to you, my dear friend, A week ago you put out an AI video clone of yourself speaking in different languages that blew my mind away. And so, if you can talk with our listeners a little bit about the ways in which educationally, in associations, we can now start to think about having opportunities for different kinds of engagement using AI.

Julian Moore (13:19):

How many times have we ever sat down and gone, “Wow, my workload. If I could only clone myself.” And so I did. And there’s a tool called HeyGen, and on HeyGen, you talk to your camera, just as we’re doing now, and it takes two and a half minutes. It’ll copy your likeness and copy your voice. And it’s super secure. Once it has that, you can then put any text into there and you will present that text. So, what we’re finding is, it was really useful because the future of AI is tailoring and making it bespoke.

Julian Moore (14:00):

So, what we have to be able to do is, for educators particularly, is understand we’re going from this size of classroom, like 30 children, or 100 association delegates, or 1,000 different CPD students, and we’re turning that into a one-on-one relationship.

Julian Moore (14:25):

So, now you have a clone. What it means is, we can actually do the old-fashioned Outlook merge function, and we can merge the data through, and the outcome is now a video.

Julian Moore (14:40):

So, for instance, if you have an email, and we’re sending Christmas wishes to everyone, we could change the name and we could change the business name and then send a video. And so, we could send 1,800 videos in 10 minutes, and now everyone gets Cristi going, “Happy Christmas!” instead of an email that comes through.

Julian Moore (15:04):

So, when you put that into education, it now means everyone gets a one-on-one experience. So, “Welcome to the course. I’m going to be your educator for this term.” And it’s always that one-on-one experience. So, instead of just the assistant, you now have the ability to engage with your educator at any time, in any place, answering your things with your name. And these experiences, these tailored personalized experiences, are really the boom of AI, I believe. Because whatever you look at, people want that one-on-one. We do better with one-on-one. We learn faster, we understand more, we feel better. Our ego gets massaged. We walk away with a smile on our face. And that one-on-one kind of tailoring is really the big increase, or I think the big win, for education.

Dr. Cristi Ford (16:07):

I completely agree. One of the things that we have been talking about in conceptualizing for several decades now is personalized learning. And so, what you’re offering is a solution to be able to combat some of the challenges we’ve had when we’ve had 30 children in a classroom, in higher education, K-12. When you’re really trying to do a little bit of classroom management, you don’t have the opportunity to do that personalized learning approach.

Dr. Cristi Ford (16:33):

But I guess I want to ask you, for those who may be listening that are maybe a little cynical, who may say, “Okay, HeyGen’s great, but the paid version is really the version you need.” Are there some ethical considerations? Are there some opportunities to really figure out, how are we reducing the barriers of cost to serve for educational providers, or maybe even very small associations? How are we combating some of that?

Julian Moore (17:04):

We haven’t got a solution at the moment. We are so new, and we are a year old. So, we’re not even a toddler, let alone matured. So, we’ve got us a new technology that’s having the most incredible effect across every industry. And when we start looking at mature challenges, like barriers to entry, we haven’t got there. We are just not there yet.

Julian Moore (17:31):

But what we can do is see where this type of radical shift has happened before. So, as you can see, my gray betrays me. I’m old enough to remember when we first had a CD in the desktop tower, and we put it in, and then suddenly, “Oh, hang on, I could drag my music onto my desktop and have –

Dr. Cristi Ford (17:55):

I remember that, too.

Julian Moore (17:56):

WAV files. I’m like, “Wow, check it out.” And then I discovered Napster.

Dr. Cristi Ford (18:03):

That’s right.

Julian Moore (18:05):

And then I had the world’s music. And of course, it meant that all the music, I guess, production houses, all that side of things, all started shaking, going, “Well, hang on, all our revenue has just gone. Everyone is illegally copying music.” And there’s huge platforms to do it on. So, the cure for that became Spotify, and the solution was Amazon Music, Apple Music. There’s lots of platforms where the artist now gets paid for their work, and we pay a subscription service. And this was the early subscription services, because the barrier to entry was brought down to a monthly easy-to-manageable fee.

Julian Moore (18:53):

What we’ll see with AI is something similar. At the moment, everything, just everything in Silicon Valley, just everything is AI. I mean, I can’t begin to say, it is completely AI based. Stick an AI, two letters, after your business name, and now you’re going to be BC ready.

Julian Moore (19:17):

So, what we’re going to see now is platforming. So, HeyGen on the video, more than likely it will get acquired by one of the big video creation companies. So, where you can type text in and get video out, being able to include your avatar into there, and create movies or input video into coursework.

Julian Moore (19:40):

And then, when we’re looking at OpenAI with ChatGPT, it started as a chatbot, but now it can create images. You just type it in. Now it will do data analysis. Now you can talk to it, and it will literally answer you back. They’ve created one that my kids love where it’s called Santa. So, now you can genuinely have a conversation with Santa.

Julian Moore (20:06):

And so, we’re going to see this platforming come through, and that’s good news. Because when we have large platforms, we’ll have educational pricing. When we have educational pricing, we’ve got a level of maturity coming through.

Julian Moore (20:21):

So, my main concern for any of this will become not the haves and have-nots. It’s not going to be that kind of… It’s going to be a knowledge divide, and we can’t allow that to happen. What we have to make sure is everyone has access, in the maturity, the educational pricing.

Dr. Cristi Ford (20:48):

No, I completely agree with you there. As you were talking, I’m even thinking about teacher preparation programs, helping teachers to understand how to do prompt engineering appropriately for lesson planning.

Dr. Cristi Ford (21:01):

And the new skills and technologies that it will take, it is sometimes concerning to me. I had a great opportunity to go to a conference in ASU on AI research, where Candace Thille from Stanford was speaking, and we started to talk about the considerations around these black boxes, and not understanding the human and ethical resource, the power, how many GPUs it takes to power these larger systems. And we’re talking about ChatGPT and some of the open AI systems.

Dr. Cristi Ford (21:36):

We have not talked to any on the ones that are proprietary or closed-box systems that are really trying to work to commercialize LLMs in a very particular way.

Julian Moore (21:47):

Yep. So, we’ve got a couple of things there. The first thing is, when you look at the large language model, there’s a couple of ways of doing this. You can go on to GitHub, or you can go onto a range of places, and just download one of these and fine-tune it and turn it into your proprietary one.

Julian Moore (22:07):

The challenge is, OpenAI, is GPT-4, is by far and away the best. There is a distinct gap between them and everyone else. Now, the black box kind of mystery, let’s keep it shrouded in mystery, to some degree it’s there because the people who create… Like the scientists, the tech wizards who actually create the LLMs, they discovered years back this transformer code. And the easiest way to understand the transformer code, you take the entire internet and shove it on a hard drive. And I’m oversimplifying, but essentially, mass data.

Julian Moore (22:59):

So, in GPT-4s it’s 17.5 trillion documents. That’s the easy way. So, stick that onto a… And now this transformer is such a clever piece of code that, much like our conversation, it finds its way and weaves its way through and picks out the important parts. So, we may forget half the conversation, but we’ll remember that cloning is HeyGen. We’ll remember that education is this. And that’s what the transformer does. It works its way through all the code and all the data, and chooses the relevant parts, the important parts, to create its response.

Julian Moore (23:43):

Outside of there, and when you get really deep into the dense files and the dense nature or the multi styles of use, people don’t exactly know how it works. And when I say people, no one. That’s the challenge. It’s literally no one really knows exactly. LLMs, for those listening, is a large language model. And it really should be called large data model or large document model, because its read the internet, it’s saved all the information, and it’s got rid of all the bits it doesn’t like. Because let’s face it, if it looks at all of Twitter, not everything’s relevant. And then it says, “This is our data set,” and saves all that.

Julian Moore (24:37):

And that’s where it draws all its information from. Because very commonly, people go, “I don’t want to use it because it’s going to take my info that I’ve just put in it and then train itself on it.” And that’s not how it works. It is already trained, and you are not going to train it anymore. It’s not going to use your data. There’s a lot of misinformation out there.

Dr. Cristi Ford (24:59):

Well, thank you for sharing that with our listeners. Sometimes I feel like we’ve advanced this conversation so much, I forget to take a moment and pause and explain what an LLM is. So, thank you for that, Julian.

Dr. Cristi Ford (25:10):

I think the other thing that’s so interesting that we’ve been talking about at D2L is that AI has been around for a very long time, for decades, but this layer of the large language models is another kind of lens, or layer, to some really intentional work that has been happening for a very, very long time.

Julian Moore (25:32):

Look, we are standing on the shoulders of giants is the truth of this. This has taken decades to get to this point, but we’ve all ignored it because, well, only scientists found that a computer playing chess was interesting. Unless you’re a chess player. But outside of that, then building up, “Oh, it can hold a conversation,” but a really basic one at nine-year-old level. And then, it’s just entered public consciousness when it could hold a level of conversation that we went, “Wow, hang on, that’s like conversing with my friend.” And so, it became useful.

Julian Moore (26:20):

And that’s the new tech, this transformer technology, this software code, this transformer software code, and interrogating files. So, what happened was they got some money and they went, “You know, if we really get big data… So, let’s get 30 million documents.” And they went, “Oh, the response using this transformer is pretty good. Maybe we could make this data bigger.” So, they went to 30 billion. And they went, “Oh, hang on, hang on, now we’re getting somewhere.” And it still made errors, huge number of errors, but they could see where it was going.

Julian Moore (27:06):

To the point now where we’re in the trillions, and we’ll be in the hundreds in very shortly.

Dr. Cristi Ford (27:14):

Oh my God.

Julian Moore (27:15):

Yeah, we’re absolutely standing on the shoulders of giants, the scientific community, the tech community, that have been working away behind the scenes for decades.

Julian Moore (27:26):

And we have AI everywhere we look. If you go onto Netflix, it’s got predictive… “Look, you’ve liked these things before. You should watch this.” Same on Spotify, same on Amazon. We are surrounded by AI, this is just a new version. It’s called generative, and it’s called generative because you type in something and it generates it.

Dr. Cristi Ford (27:49):

Before I want to let you go, I have two questions for you. One, as a call to action for our listeners, what would you hope is the future of AI in terms of how people approach it? I think it’ll be my first to last question. And then my last question is, what’s next for Julian?

Julian Moore (28:08):

I think first thing, my hope is that, like every radical… I don’t like the word radical… every revolution that’s ever come through, we’ve had the people go, “Those cars, they’ll never take off. My horse is much more reliable. It’s faster. It’s cheaper.” The cars were more practical, so of course they took off. And then, from there we had television, and the people on radio went, “It’ll never take off. It’s too expensive,” dah, dah, dah. And now we have television.

Julian Moore (28:43):

When the internet first came out, and I was working in London with a chap, I won’t say his name but he was very old, and he was like, “Julian, you have to learn how to write the perfect letter. It must be crafted in this way, you have to engage, and your handwriting needs to be better. And so, we’ll correct your handwriting. We’ll correct the grammar. We’ll correct the things. And we’ll always send these letters. Those emails, they’re too impersonal.”

Julian Moore (29:15):

When was the last time we wrote a letter? So, with AI, we have to get past the doomer kind of prediction, because we’re not there, and we’re a far way off the sentient self-awareness. At the moment, we have tools. The tools are so simple to learn. I mean, so simple. Most of them are free trials, and they will take you half an hour, 30 minutes, to change your entire work process and to be more effective. Lean in, take the 30 minutes, and have fun. Truly go, “Wow, I don’t have to do those hours of work anymore because now I’ve got… For 30 minutes of effort, I can now lose all these hours of work and have a higher quality of output.”

Julian Moore (30:17):

So, that’s what I would say. Lean in. Lean in, enjoy and get creative. It was a chatbot that was created, really. And what the world has done with that is truly remarkable. Just fantastic.

Dr. Cristi Ford (30:33):

I love it.

Julian Moore (30:34):

So, lean in, get creative, enjoy what you’re doing with it, and really get out there. Be early. We’re at the start line. There is no late to the party here. The party hasn’t started, and you’re on the invite list, so turn up early.

Julian Moore (30:55):

What’s next for me? Right now, wow, I’m crashing into Christmas. It’s a huge amount of work going on. We’ve got robots launching next year, so we actually have home assistant robots coming, so they will vacuum your house, polish your tables, stack your dishwasher, unstack, walk the dog, mow the lawn.

Dr. Cristi Ford (31:22):

You can send one to me. Send one to me.

Julian Moore (31:27):

I’ve got a dentist’s group asking how AI can assist them. We have educators all over the place going, “Well, how can we get our foot in the door so that we can be involved?” Agriculture is involved. They’re really all over this. Medicine, oh wow. What AI is doing in medicine is just fantastic. So, at the moment, I am really, really enjoying just the ability to go in and help people go, “You mean, this thing that was going to take us a year, we can do this week?”

Dr. Cristi Ford (32:10):

Yeah, I love that. I love that about you, Julian. I love your growth mindset around this work in AI. It has been an absolute pleasure to have you here with me today. Thank you so much for joining us.

Julian Moore (32:24):

No, thank you for having me. When I saw that come through, I couldn’t wait to be involved.

Dr. Cristi Ford (32:32):

Well, colleagues, if you’re listening to us from down under, or curious to learn more about what Julian does and the services he provides, check out the Strategic Management Solutions website at smsonline.net.au. And a special thank you to our listeners. Remember to follow us on social media. You can find us on the X platform, LinkedIn, or Facebook at D2L, and our YouTube channel is @DesireToLearnInc. Reach out to us on social media and share your thoughts and opinions with us. We’d love to hear if there’s a subject or a trend in education that you’d like Teach and Learn to discuss. Thanks for listening.

Dr. Cristi Ford (33:11):

You’ve been listening to Teach and Learn, a podcast for curious educators. This episode was produced by D2L, a global learning innovation company helping organizations reshape the future of educational work. To learn more about our solutions for both K-20 and corporate institutions, please visit www.d2l.com. You can also find us on LinkedIn, Twitter and Instagram. And remember to hit that subscribe button, so you can stay up to date with all new episodes. Thanks for joining us, and until next time, school’s out.