Peter Bishop (00:49)
Welcome to the breakdown. I am excited today cause I got an old friend, Owen Briarly here on the, ⁓ on the podcast who is coming to us from the UK, which is a surprise to me. Cause the last time I saw Owen, he was sitting here in Edmonton. So

It's been a hot minute. So I'm using this as a double duty to catch up with an old friend and also talk about obviously AI and potentially some of the AI and gaming, but I think there's a lot to unpack with you Owen. So, hey, welcome.

Owen (01:19)
Thank you. Yeah, when you invited me and said, let's reconnect, I was like, ⁓ fantastic. I haven't heard from Peter in ages.

Peter Bishop (01:30)
We sat together

on the Digital Alberta Board. In fact, you were president and I think I was at some point. ⁓ that ⁓ was the last time we were hanging out, I think.

Owen (01:35)
Yep. Yep.

In fact, I think it was at the ⁓ Gallery, Art Gallery of Alberta, and it was a Digital Alberta Award Show. And I think that was the last time we were in person. Pre-pandemic. So an entire century ago. Yeah.

Peter Bishop (01:49)
Right.

That's correct.

Carrie (02:02)
Whoa.

Almost.

Peter Bishop (02:08)
⁓ do

you mind, Owen, just giving a brief background of Owen, like kind of what brought you to this moment in time, just as far as what, what got you in, like, I know we all have crazy backgrounds, especially when we get a little long in the tooth here, not suggesting you are, but I certainly am. ⁓ so just kind of what brought you to this, this moment in time and maybe, ⁓ even on the AI side, what's gotten you kind of involved on that end. ⁓

Carrie (02:25)
You

Owen (02:35)
⁓ dude, yeah, cool.

Wow. my first career was in theater. I was an actor for 15 years. I went through the musical theater program at McEwen. ⁓ And did you really? Well, there you go.

Carrie (02:45)
wow.

to McEwen. Yeah, I took the professional

writing program there. That's wild. Were you in the orange?

Owen (02:56)
⁓ wow, Prowl. I taught.

Yeah, I was in the old ⁓ in the old ⁓ hub. it's now the orange orange hub. Yeah, Peter Roachie, good friend of mine. And yeah, yeah. So I taught I taught ⁓ on Prowl. I taught the web design course. ⁓ Yeah, it was HTML for writers or something like that. And

Carrie (03:04)
Yeah, I was second year proud with Peter Rocha.

That's so wild!

No way.

Did you teach

me? This is crazy.

Owen (03:26)
I don't know.

Peter Bishop (03:29)
speaking of feeling old.

Owen (03:30)
Yeah, well, maybe. don't know. I did a couple of years, so I don't know what years those were. ⁓ But yeah, I remember, remember proud fondly. And amusingly, now my daughter is teaching at McEwen in the UX ⁓ department.

Carrie (03:44)
Same.

Owen (03:55)
and working with, working with wonderful people at the McEwen Design Program. So yeah, yeah. So theater.

Carrie (03:59)
Small world, small world.

Peter Bishop (04:05)
We digress. ⁓

Owen (04:07)
Uh, started

Carrie (04:08)
you

Owen (04:10)
in theater. Like I said, I was doing that for 15 years.

so fast forward to the middle 90s. And I pivoted because this crazy thing that you mentioned earlier, the World Wide Web came into being. And I was looking for a bit of a career change because we were starting a family.

⁓ And then that led to, man, ⁓ working with multimedia, which then became rich media, which then became new media, and then it became digital media and all the medias. And really it was a long time, for a long time it was my bread and butter was using flash. ⁓

Carrie (05:07)
Mmm.

Owen (05:08)
Fast forward a little further into my time, I eventually found my way into a funny little school called Guru Digital Arts College and ended up taking on the running of the school. And yeah, I ended up owning it at one point and turned it into the Edmonton Digital Arts College.

Carrie (05:32)
Holy smokes.

Owen (05:36)
And we had four amazing programs, a phenomenal faculty of really beautiful people. And yeah, we kind of were doing the whole digital art school thing with a video game design program and an animation program.

So again, fast forward some more time and in 2018, late 2018, a friend of mine who was at the University of Calgary said,

Have you thought about doing a master's degree? And I said, well, I don't even have an undergrad, dude. I'm like, I have a diploma in theater arts. And he said, we can work on that. And so I ended up getting enrolled in this, what's called a trance disciplinary program ⁓ in computational media design.

Carrie (06:23)
Hahaha

Owen (06:39)
I started on the master's program in ⁓ January 2019. ⁓

and spent 11 months in the master's program and they said, ⁓ just do the PhD. And so ⁓ I said, are you sure? And they said, yeah, yeah, yeah, apply, apply, your work is good. And so I applied to get into the PhD program and in January 2020, I became a doctoral student. And then the entire world shut down. How convenient for me that, you know,

Carrie (06:53)
You

Congrats.

Right.

Owen (07:18)
we had to go into lockdown so that I could do my research.

Also happened to be a really interesting time to start paying attention to what was going on in the world of AI, especially with neural networks and a lot of the reinforcement learning, machine learning type of work. And I mean, this is all pre-Chat GPT and even before there's now a very common thing that we call generative AI, right?

And, um, it was still a lot of, of, in order to get anything AI happening, you had to go into Python or you had to go into, um, a Google workbook somewhere and spin up your own learning system and training whatnot and hope for the best. And so the early days of working directly with stable diffusion and, and getting

diffusion models to work and that whole thing was really, ⁓ it was fascinating. And then I, you know, of course my research was really focusing on the...

the bringing together of ⁓ the arts, performing arts and ⁓ computer science. And where that took me was into the realm of getting really curious about non-player characters in games. Because I don't know if you remember the Truman Show, the film, the Truman Show, but that whole Truman Show loop, that

Carrie (09:03)
Totally.

Peter Bishop (09:06)
Yeah.

Carrie (09:06)
Mm-hmm.

Owen (09:07)
that broken suspension of disbelief where he's like, they're on a loop. Watch. VW bug, bike, flowers. It's a loop. And I thought, you know, isn't it interesting that we have this incredible suspension of disbelief in games and how the sort of the beauty of these

massive worlds that are being made and these complex stories that are playing out and the notions of player agency. And then we have, you know, performers, actors in the games that are kind of wooden and kind of robotic and, you know, and to the point where popular culture refers to NPCs, people, it's a very disparaging thing to be referred to as an NPC.

Right. Um, and so, you know, it was, um, as an actor, I look at it and go, well, that's just an acting problem. Right. Um, the reason that, that the non-player characters aren't working is because they're not acting well. So how can I make non-player characters better at acting? Well, they need to behave in more convincing ways. And so, um, I ended up studying, uh,

Carrie (10:15)
Hmm.

Owen (10:35)
Dr. Richard Zhao, who is at the University of Calgary and he's a ⁓ digital science, or data science, and also an AI for games expert. And while I was studying with him, I happened upon this funky little, what they call reinforcement learning algorithm called Q-Learning. And when I built...

sort of a test in Unreal Engine using Q-Learning, I discovered something that was kind of interesting to watch from a performance perspective. ⁓ These non-player characters started moving from goal to goal, and they would sometimes get it right, they'd sometimes get it wrong, and then they would, you know, move on to another goal.

And while I'm watching this, I'm like, isn't that interesting? They're kind of browsing. It's kind of like they're shopping. Hmm. It's like they're picking something up, looking at it and putting it back on the shelf, trying to find the thing that they're looking for. So I decided on an experiment that would be sort of the cornerstone of my dissertation, which was to do a very humble thing. And it was to take a bunch of ⁓ non-player characters.

Carrie (11:40)
Hmm.

Owen (12:03)
and send them grocery shopping.

Carrie (12:05)
my gosh.

Owen (12:06)
And grocery shopping is one of those lovely open-ended problems that there is a successful outcome, but there's no one right way to do it. And when you have a group of non-player characters who are moving about a grocery store in a very deterministic way, ⁓ it becomes kind of obvious that they're wandering around in a very deterministic way. I wanted a little better non-determinism.

but I also wanted some intention in there. The flip side, of course, is you get the, they call it stochastic behavior, which is very random, right? And I wanted something that was more than ⁓ just random. I wanted something that was browsing, that was shopping, that was that sort of thing that was there. And Q-Learning has an algorithm.

looking at it from a performance perspective was super powerful and really wonderful and gave me exactly the kind of behavior that I was looking for. So I set up these experiments and began looking at this sort of process of digging into making these non-player characters go shopping.

Because it's one thing to go shopping, right? And not have a clue where you are. But I know what I'm like when I go shopping. Sometimes I'm in a hurry and I go to a place that I'm familiar with and I expect things to be where I saw them last. And if I'm in a hurry, I get really cranky when the duck isn't where it's supposed to be. I already have a map in my mind of what's going on with that shopping trip.

And, and so I thought, wouldn't it be interesting to be able to take all of that data that we're generating with these grocery shopping trips and put them somewhere? And so I, I discovered this, ⁓ this whole notion of, of something called a graph database. Graph database is a really interesting, databases because typically databases are really good at storing, ⁓ data and kind of tabular.

even if you've got relational databases where you can associate one set of table data with another and specific fields and whatnot. Graph databases have this thing that is, ⁓ they refer to data as nodes. And these nodes have relationships that you can build between them. So let's say I have a node Bob and I have a node Sally, right?

And these are two, we'll call them people, right? And these nodes, and we can have a bunch of other people, we can have houses and dogs and cats and whatever else we want, right? And these are all nodes. But then when we start drawing relationships between them, that becomes really interesting. And so we can draw a relationship between ⁓ Bob and Sally and say, okay, there's a relationship between these two points of data. ⁓ But now let's put some character to that.

And so we can say things like, well, Bob and Sally have a relationship and they live together. So immediately I'm starting to form a story about, you know, these two nouns and something that connects these two nouns. Right. And, and again, being the theater person that I am like, ⁓ there's, there's something cool here. Now I'm going to add the beauty of some of these graph databases is that you can actually have multiple.

Carrie (15:26)
Hmm.

Owen (15:48)
⁓ pieces of information that go into the relationships. For example, I just said, Bob and Sally are living together. What if I also added in there their siblings? Suddenly the story changes and it has a whole other nuance and context and right and so what I started doing was I started taking my non-player characters who were going shopping and I started recording into this graph database all this information that they were learning.

about where stuff was and the fact that bagels happen to be in the bakery. So maybe bread is somewhere nearby, right? And started playing with what are the nuances in the relationships between the shopper, the thing that they're shopping for, where it is in the grocery store and started to build. And what I discovered I was doing is...

Carrie (16:27)
Hmm.

Owen (16:45)
something well known as a knowledge graph. And knowledge graphs are really powerful things that allow us to, and in fact, they get used in AI as part of something called retrieval augmented generation. So it's not just a generic prompt that we're writing. We're actually building something that is contextualized so that when it gets passed to the large language model, it says, oh, here's the context for this query.

There you go. And that is ongoing. I am continuing to do my research. I'm working with a company in the States that we're at commercializing it in a variety of ways. And so there you go. That's kind of, I skipped a whole bunch, but I'm already like 25 minutes into the...

Carrie (17:29)
Wow.

Yeah, you've lived a full life by the sounds

of it Owen.

You

Owen (17:45)
You were saying something about being long in tooth.

Carrie (17:47)
Hahaha!

Peter Bishop (17:48)
yeah.

Yeah. You know, in the podcast, we're supposed to ask questions, right?

Carrie (17:53)
You

Peter Bishop (17:57)
You know, a couple things there. First of all, ⁓ Kerry, I don't think you've ever used the phrase, fast forward to the mid 90s. But

Carrie (18:09)
was doing that on my tricycle. Yeah.

Peter Bishop (18:11)
Ha ⁓

⁓ So hey Owen, so this is is fascinating to me Because I'm a gamer I play a lot of games so I totally get the idea of NPCs and they tended to be these kind of like you said robotic like in a typical game you'll have a couple like Non-playable characters that have a lot of time and effort put into them and even these days they're having they're being voiced by real actors now And there's a handful of them that kind of are on along the main storyline

Owen (18:15)
Uh-huh.

Peter Bishop (18:42)
And then there's all these peripheral characters that you can interact with, but they all act like kind of rumbas, right? Like they just kind of run their own pattern. They're kind of awkward. Is when you think of like, if this all went the way you're thinking down the road, what could a game experience feel like? Like it feels like open world times 10. Is that true?

Owen (18:49)
Yep.

Yeah,

well my goal is to, is to make, basically to, we're gonna get super nerdy here.

Carrie (19:17)
Let's go!

Owen (19:19)
I'm

going to use a term called LARPing.

Carrie (19:23)
Yeah!

Peter Bishop (19:24)
Nice.

Owen (19:24)
Right?

And so LARPing is an experience where you're dealing with other humans who are playing non-player care.

Peter Bishop (19:25)
This could be the title of the podcast.

Carrie (19:34)
interesting.

Owen (19:36)
So wouldn't it be an incredible experience to go into an open world where every single non-player character felt like another human player?

Carrie (19:43)
Whoa.

Peter Bishop (19:45)
Wow.

Owen (19:46)
where

Carrie (19:46)
So they would

still have limits, obviously.

Owen (19:48)
⁓

And that's the thing, right? Is you have a role to play, which is larping You would have ⁓ an awareness of the world around you. You wouldn't be omniscient, right? No AI. I don't ever want an omniscient AI, right? ⁓ I want an AI that will be potentially a collaborator, ⁓ a...

Carrie (20:08)
Agreed.

Owen (20:17)
signpost that can help me figure out where I need to go next. ⁓ You know, give me some warning signs. But again, I also don't want perfect information, right? I want rumors. I want innuendo. I want superstition. I want, you know, and then the joy of it for me is, and really, you know, leveraging the whole Dungeons and Dragons experience. When you're dealing with a really great Dungeon Master,

They illuminate every single non-player character experience you encounter. Even the monsters, right, are, you know, they are being well-designed and made appropriate for the gaming experience so that you are engaged and you buy it, right? Like, it's boring to play with a dungeon master who just sits in and rolls the dice and goes, ⁓ you're dead!

Carrie (20:52)
Mm-hmm.

Well, this is why I'm so obsessed with people in the arts getting involved in AI and in tech in general. It's just, adds such an incredible layer of storytelling and depth and humanity.

Owen (21:16)
Yeah, like that's kind of...

Sure.

100%. Well, and that's just it. As an artist, I look at a lot of stuff that comes out of AI. I go, ⁓ that's pretty good. Not great, but pretty good.

Carrie (21:49)
Mm-hmm.

Owen (21:49)
In the hands of artists, some amazing things happen. Right? And there are friends of mine who are using ⁓ image generation and they're art directors, but they're not using it as the end result. They're not trying to get a black box that solves all of their creative problems for them. They're trying to get the ideas out. And then they're working with

Peter Bishop (21:54)
Mm-hmm.

Carrie (21:55)
100%.

Owen (22:18)
⁓ experts who can take that idea and refine it further, whether it's illustration, whether it's ⁓ kit bashing in a ⁓ 3D tool, whether it's ⁓ painting, whether it's copywriting, whatever, right? It's a leapfrog. AI for me is a lot like having a great improv partner.

Carrie (22:36)
Mm-hmm.

Ooh, I like that. And that's like nail on the head. This is a theme we've been hearing throughout this podcast is it's that in-between brainstorm buddy that helps an expert get from idea to execution that much faster because it's just a tool that helps them get their idea from one place to the other, not something that just gives you this thing.

Peter Bishop (22:47)
Mm-hmm.

Owen (22:48)
Right?

Right?

How do we set?

Carrie (23:16)
if you don't know what you're doing.

Owen (23:16)
Well, and,

and, and it's also ⁓ tricky because it says yes to everything. Right. And, and it's like, if I say, okay, chat GPT, for example, ⁓ here's some text where you translate it to Spanish and it goes, sure. Spits out a bunch of Spanish and I don't know Spanish. I'm a non-expert.

Carrie (23:25)
Especially right now.

Peter Bishop (23:26)
Yeah.

Carrie (23:45)
Mm-hmm.

Owen (23:45)
So

I go, cool, awesome. Here's hoping, right? And this is one of the things that as an educator, helping my students learn how to live in a world of AI and how to use tools when they are non-experts and trying to figure things out is, okay, don't trust the AI, trust yourself and

Carrie (23:48)
Looks good to me.

Peter Bishop (23:51)
Mm-hmm.

Carrie (24:13)
Hmm.

Owen (24:14)
If you, when you're first starting out, take the stuff that it generates and verify it. Right. Take the stuff that it creates and go, wow, I really like what it did. Does it make sense? And it's, it's a bit slower. It's a bit, it's, but it's no different than going out and doing the research by hitting the stacks and reading the readings and building up your knowledge base. Right.

So one leapfrog is fine, but the leapfrog gets you a certain distance and then you have work to do beyond that. I'm fortunate that I'm old and I have this history behind me that I can look at a thing and very quickly suss out what it is and what it isn't, right? You were about to say I beg your pardon.

Carrie (24:52)
Mm-hmm. And that's.

Yeah. Well, no, that's okay.

I was just going to say just about the fact checking piece there. It's, think some people are like, it's annoying to go and look up. I'm like, do you have any idea how long this would have taken you before this tool? Like this is still only going to take you 10 minutes, even if you do your due diligence and fact check something and you can fact check by the link that it gave you. It's really easy. It's so easy to fact check it. And

Owen (25:27)
Yeah. Yeah.

Carrie (25:31)
this would have taken you hours beforehand. And it's so funny how as soon as we get our hands on something that makes our lives easier, we're like, let's make this make our lives as easy as possible, as fast as possible. And it's like, let's pump the brakes just a little bit. Yeah.

Owen (25:46)
Right? Absolutely.

⁓ And we've always had tools that try and make our lives easier. And every single time there are limitations to how far that tool can get us.

And so, so AI is no different. And in fact, we're at the point now where the flavors of AI, so moving from Claude to Google, Gemini to GPT to wherever is

There are biases built in and there's styles and come out of that generative process that there, isn't, it will be, it will be what resonates for you. Right. And so one of the things that I've been saying in presentations that I've been doing is saying bias is not a bad word. We treat bias as a bad word.

But we need to recognize that bias is something that we all have. And you have to decide what is the bias that you like? What is the bias that you want? And then seek out the products that resonate for you in terms of what that bias gives you. Where it becomes kind of nefarious is when that bias is hidden from you and it isn't clear what path you're being led down.

Carrie (26:58)
interesting.

Peter Bishop (27:07)
Hmm.

Carrie (27:13)
Hmm.

Peter Bishop (27:19)
You know, I like that it's kind of freeing in a way to just accept the fact that there is bias in the tools that you're using and not try to fight it. Just choose tools that align, right? Because I think the notion that you're going to find these things that are completely logic-based and unbiased is probably a complete myth because they're all made by humans.

Owen (27:43)
100%. Yes. Yes, they're made by humans. And there is no such thing as the super singularity AI that knows everything. For crying out loud, we can barely define human intelligence.

Carrie (28:00)
Yeah.

Peter Bishop (28:02)
Or lack of.

Owen (28:03)
Right?

Right? And that's something that the AI researchers regularly acknowledge is they say, we don't really know what intelligence is. That's why we prefer to use the phrase machine learning. But that's not as fun and cool and sexy in the media.

Peter Bishop (28:18)
Mm-hmm.

Carrie (28:19)
Mm-hmm. it's,

I feel like that's one of the most interesting parts of AI when Peter and I first started talking about this, gosh, over a year ago, ⁓ is, yeah, somebody's built something and usually they can get up on stage and say, I built it by doing this and this and this. And they get up on stage and they're like, we kinda know what's happening, but not really.

And we gave it one little prompt to be ⁓ more cordial, and now we've had this huge wave of it being far too, ⁓ whatever the word, sycophantic, I think they've been using. ⁓

Owen (29:03)
To the tune of several million dollars.

Carrie (29:05)
Yeah, exactly.

Peter Bishop (29:06)
Yeah.

and I got one last question here from me anyways. Because you and I have seen similar... Okay. We've seen a similar set of disruptions in our lifetime, like computers and the internet and all these things. Is this, in your mind, is this the same history repeating itself or is this completely different?

Owen (29:13)
Only one?

Carrie (29:15)
I've got like 20, so that's fine.

Owen (29:17)
Run!

Be both.

⁓

My friend and mentor, ⁓ Patrick, ⁓ said to me repeatedly while we were working through my dissertation, he said, we are in a period of massive epistemological change. And the last time the human society went through this kind of epistemological change was with the printing press, right?

Carrie (30:12)
Hmm.

Owen (30:14)
Meaning is changing. Right? And meaning is changing. Right? 100%. And, and the, so I'm a big Marshall McLuhan fan and, ⁓ and I, I, I believe that the medium is the message. Absolutely.

Carrie (30:15)
Praise be.

Owen (30:41)
⁓ I also, I love his notion of, of rear view mirrors where every new tool that we get, we use in old ways. Right? Right? Right? And so that's why to this day, talk shows on TV still have a mic on the desk and the interviewer sitting at a desk because that's the way they had the radio studio set up. Right? And when

Peter Bishop (30:51)
Mm-hmm.

Carrie (30:52)
interesting.

Right.

Owen (31:07)
TV when TV first came along they just took the camera and put it in the radio studio and fired away ⁓ for the talk shows and yeah that's okay.

Peter Bishop (31:17)
Yeah, I remember this like, sorry to interrupt Owen, but I remember this

notion in design, I can't remember the name of it, but they were saying like, yeah, when websites first came out, that's why they made them look like brochures, right? Was because that was how they could think of it's like a book online and it took a long time to disrupt that too.

Owen (31:30)
Yeah.

Carrie (31:36)
Well, and we still say when we're designing a website above the fold, which is referring to the newspaper, but you're just talking about what you can see on the screen.

Owen (31:36)
100 %

Yeah. ⁓

Peter Bishop (31:41)
Right.

Owen (31:44)
Yeah. Unreal Engine, ⁓ the interface that you work in is the stage. ⁓ The things that you drag from the content drawer ⁓ out onto the stage are actors. And so we use things that keep us grounded in a space that we're familiar with and comfortable with.

Carrie (31:53)
Mm-hmm.

Peter Bishop (32:03)
Right. Casts.

Owen (32:13)
But at the same time, when you think the medium is a message, you have to start to wonder, well, am I limiting what I can think about here? Like we're using chat GPT because we know what chat box can do and what chat feels like. But is that, you know, is that the ultimate generative AI interface? Right? Is that, is that.

Carrie (32:39)
Yeah, my dream is still something along the lines of what Tony Stark has. Like where you're just, you're in your garage, you're asking it questions, it's doing stuff for you. You know, that's what I want. I don't want anything more than that. Like that's the ceiling for me. But I agree. It's just like there's, I think what's...

Owen (32:44)
Right?

Peter Bishop (32:45)
Hmm

Owen (32:55)
I love it.

Right.

Carrie (33:06)
both interesting and difficult about AI and another reason why think artists are able to use it almost a bit. Everybody thought it was going to be the engineers and the data scientists and not well the data scientists are crushing it but ⁓ the creatives can go in because there's so much potential with AI and it's literally we cannot fathom how many things it could do. It's almost hard to focus on one thing like can we make it this kind of interface? Can we

Can we change it in this way? we? But creative people are used to living in that head space of like, where can this go? So.

Owen (33:41)
Well, and I...

As an artist, I adore hallucinations. They are not, they're signal enhancers for me, right? When an AI hallucinates, I'm like, ooh, cool, trippy, right? And it's fascinating to me how people have, sometimes I hear the complaint is, it didn't get it right.

Carrie (34:00)
Totally! ⁓

Peter Bishop (34:00)
Hehehehehe

Owen (34:17)
I typed in this prompt and it didn't get it right. didn't, it didn't give me back the thing that I wanted. And, and it's, it's always interesting to me to think about, well, are you perhaps being a bit too rigid in your expectations? And is it a case of perhaps you need to be a little more open to creative interpretation? Right. And, and I,

Carrie (34:43)
Yeah.

Owen (34:45)
I go back again, I go back to my theater roots and I remember directing shows where I would get an actor showing up for an audition, do an incredible thing and then shows up for rehearsal and they're a little bit different. The way that they interpret a scene, the way that they perform, the way that they interact with the other actors around them. As a director, I had to get really clever about drawing

drawing out the different aspects of each one of these wonderful, incredible, creative souls who are interpreting things differently from what I expected. Because if I just simply told them to do exactly what I wanted them to do the way that I wanted them to do it, they would have started becoming very wooden and their performance would have been lacking. No, not exactly.

Carrie (35:25)
Mm-hmm.

Yeah. And neither of you would have been doing your job, right? And

I find that interesting too, because we talk about not personifying AI, but at the same time, you have to treat it in this human way. And I've had this question before, but I'm interested in your answer. When we talk about art and we talk about humans, ⁓ what we love about the human experience and human art is that it's imperfect, because humans are imperfect.

but then AI came along and when they were imperfect everybody was like, oh it's not perfect, this isn't good enough. So it's interesting to me, you know, how do you see the difference between an imperfection from a human being and an imperfection from AI in terms of art?

Owen (36:21)
Mm-hmm.

Great question. ⁓ So I think a lot of the game right now, the cat and mouse game that we're playing with AI is trying to detect it. Right? So we're caught in this ha ha fooled you. Right? As opposed to, you know, when I look at a painting by an amazing artist, I willingly suspend my disbelief.

Peter Bishop (36:33)
Mm-hmm.

Carrie (36:34)
Mm-hmm.

interesting.

Owen (36:51)
When I go to watch a piece of theater, I know that I'm not actually in the 1500s, you know, if I'm watching a period piece, no matter how excellent they are with the set and everything, right? When I go and watch a film, I know that I'm sitting in a theater watching something that was recorded. There are moments where I am fooled into believing that what I'm

Carrie (37:00)
You

Owen (37:20)
watching is real and I have to check myself and go, wait a second, nope, that character isn't actually a real robot, that's just an actor, right? ⁓ And right now what we're doing with AI is we're so caught up in the detection and the... ⁓

And because in some respects we're also getting sucker punched as audiences. We're getting, I remember reading about an art competition where someone submitted an AI generated image that was good. They won the art competition and then later revealed that they, ha ha ha, fooled you. I used AI to create this thing. And so there's sort of this.

Carrie (38:11)
Right.

Owen (38:17)
inherent distrust, right? We invent these incredible technologies and then immediately distrust them. ⁓ And rightly so because, you know, we don't want to be made a fool. We don't want to be, you know, we don't want to put ourselves in danger. And so we kind of fall into this protective space, right? ⁓ And it's only when we, when the audience kind of

and gets beyond that space where they can start to say, oh yeah, okay, it's AI, I'm good with that, I'm okay. know, oh yeah, that's not really a train that's about to drive through the movie screen at me. Because those were the early days of film, right? Was the train engine chugging away at the screen, everyone going, ah! Right?

Carrie (38:54)
Mm-hmm. Mm-hmm.

Right, right, everybody's ducking behind their seats.

Yeah.

Owen (39:15)
And we have all sorts of examples of that kind of. ⁓

moment in audience, shift audience awareness, right? Orson Welles, War of the Worlds, the radio play that was performed as a news segment, despite repeated announcements that this was a show by Mercury Theater. It was, these news reports were just so compelling and no one had ever thought that you could do a piece of entertainment in the form of a news report.

Peter Bishop (39:53)
Mm-hmm.

Owen (39:53)
and not

have it be news. And so, so, you know, you, kind of collectively, you know, as society went, guess we better make sure that this isn't a theater play or a, you know, a radio show that we're listening to. Right. And yeah.

Carrie (39:55)
Right.

Yeah, great answer.

Peter Bishop (40:18)
So

Owen (40:20)
Absolutely.

Peter Bishop (40:22)
Okay, on that note, sorry, I really have to wrap. I'm gonna be so late. and this is definitely a part two coming up.

Carrie (40:25)
Yeah, we'll have to have you on again, Owen.

Owen (40:26)
No worries. Well, this was great. Thank you so much. Yeah, I would love to

Carrie (40:30)
This is awesome.

Owen (40:31)
have a part two. That would be a blast.

Peter Bishop (40:34)
⁓ I didn't realize how much material there is to unpack here, but it's been fascinating. think there's so much to what you're talking about. as we, maybe that is because we're about to wrap the season. maybe we use you to kick off the next season. That would be awesome. Right. ⁓

Carrie (40:34)
Absolutely.

Ooh, bookends. I love that.

We'll see you season two, episode one.

Owen (40:54)
Thank you.

Peter Bishop (40:58)
But

All right, on, we really do got to catch up properly, but thanks for joining us for this episode.

Owen (40:59)
Come on.

⁓

Carrie (41:01)
Yeah, thank you so much.

Owen (41:02)
absolute

pleasure. Thank you again for inviting me.

Peter Bishop (41:05)
Okay, we'll talk soon.

Owen (41:07)
Take it easy.

Carrie (41:14)
Okay, you finally, this is your guest.

Peter Bishop (41:20)
Right, yeah, I score one for me, I think. No, it's always weird to catch up in real time, ⁓ well on a podcast, but it does give you a bit of a chance to hear what he's been up to. And that guy has been busy since the last time I heard it.

Carrie (41:25)
Hahaha!

I'd say that's a win.

Yeah, I'm

obsessed with the fact that you thought he was in Edmonton and he's in London. Like, no, I live here.

Peter Bishop (41:47)
Right? I thought he was on holiday.

And I remember that he had a performing arts background, ⁓ but I'd forgotten it. So as soon as he jigged that part of my brain, I'm like, right. Yeah, that's right. He's got, and I liked that idea of, you know, AI and all of this stuff is very technical, usually run by

you know, computer engineers and all that type of stuff. But when you start to add in that human element, which it sorely needs, you know, the performing arts and all those kind of interesting backgrounds really make sense to kind of bring some balance to everything.

Carrie (42:23)
Totally, yeah, I agree. Classic artists coming in to save the world.

Peter Bishop (42:25)
Especially game.

Yeah, gaming, especially because I do feel like it's gotten. And I'm not disparaging game developers, but hey, let's face it, they're probably all nerds, right? So once once you can see that shift over the last few years where everyone started to realize how much of entertainment this is and the budgets got up there and everything. And now they're they are they're like they're like playing a movie. They're just so immersive and.

Carrie (42:39)
Yeah.

Mm-hmm.

Peter Bishop (42:55)
and voiced by real actors and real, the, the, uncanny Valley that they talk about where it's like, it's like just real, but not real enough that it puts you on edge because it's like plastic faces. There's some famous movie that was like right in the bottom of this. can't remember what it was. ⁓ but we feel like we climbed out the other side where the, rendering and everything's getting so good that you can really suspend belief. Like Owen was talking about.

Carrie (43:22)
Mm-hmm.

Well, and it's interesting too, because I think games like The Last of Us really do that, and also ⁓ my most recent obsession, which I know I'm late to the party, but Breath of the Wild. ⁓ But all other games are so ruined now, because it's like that open concept, you can do whatever you want. I can't go back and play a game that's like...

Peter Bishop (43:37)
I'm Carrie.

Carrie (43:50)
just go from point A to point B and do this thing and then go to point C and do this thing. And so, yeah, so they've got to step it up. I feel like this is the way to do it.

Peter Bishop (44:00)
Man, I'm replaying Red Dead 2 right now. I'm not gonna lie, I'm shedding a tear at the end. It's emotional. It's emotional journey, and I keep trying to tell Jill that, but she doesn't really believe me. Yeah, totally. Yeah. My horse died.

Carrie (44:03)
Hmm.

Wow.

She doesn't get it, She's like, why are you crying? It's so beautiful.

shouldn't laugh, that's terrible.

Peter Bishop (44:22)
Don't understand. can't make dinner. My horse is dying.

yeah, it's, it's, they've definitely come a turn to corner, but this is interesting. What he's talking about is just, there's always been that element of like, yeah, you've got some characters that you can interact with even a fully, ⁓ open world concept like breath of the wild and stuff. can do so many things, but there is those limitations when you meet the store owner or when you talk to the person.

little quest giver and stuff. They're all pretty wooden still and how could they not be? There's a million things to do but that future of hey these could all be just like people. Like it could be you know as rich as interacting with anyone.

Carrie (44:56)
Mm-hmm.

Yeah.

Yeah, well-

And I like the idea too of, you know, they have a bit of a memory. It's like if I hit somebody with a sword, maybe next time I see them, they're not just like, hi, how are you? And saying the same thing they always say. Maybe they're like, hey, you asshole, you hit me with a fucking sword last time, but do you still want to buy some Beatles? I don't know.

Peter Bishop (45:19)
Yeah.

You sound like an aggressive gamer.

Carrie (45:27)
I might be.

Peter Bishop (45:34)
But yeah, I don't know. It's interesting because Owen lives kind of in that academia world too, where they're spending time doing research and getting paid for it and coming up with concepts. And like I said, it's his dissertation and trying to find something that no one's ever done before and get people to buy into it and try and change the world with stuff. just think it's just fascinating kind of lifestyle that he runs.

Carrie (46:01)
Yeah, he's at quite a cross section of arts, academia, tech, AI, and all coming to this place where he's creating something pretty incredible by the sounds of it.

Yeah, it's kind of fun.

So this might be our last episode.

Peter Bishop (46:19)
Right, before adventures.

Carrie (46:22)
right before our Inventors sessions, which will come out.

Peter Bishop (46:26)
Yeah, our live podcast

and there's like a 50-50 % chance that it'll actually work.

Carrie (46:31)
We might do it. We might make it happen.

Peter Bishop (46:33)
Well,

I haven't done live podcast before, but I think I'm really excited to do it. And then we've got, I saw the lineup that you sent through and it looks phenomenal. So a lot of great guests.

Carrie (46:44)
Yeah,

I think we're gonna have some amazing guests

Peter Bishop (46:48)
nice work once again and yeah, we'll see you soon.

Carrie (46:50)
And to you.

Yeah.