English
Follow us
Gamereactor
Videos
HQ

A lecture about AI and virtual worlds with Professor Richard Bartle at Gamelab Tenerife

You always have to learn when you talk with Richard Bartle and here David Caballero had a very interesting chat with the professor and PhD on AI about many of the things defining video games. During the conversation, Bartle tackles aspects such as (in order of discussion): - Employment and talent - Metaverse tech used in virtual worlds - Traditional types of players (killers-achievers-socialisers-explorers) - How a modern MUD would be - Procedural generation - Better NPC and AI being smarter than human - Dynamic difficulty - NFT-linked items - New means for coding

Audio transcription

"We're in sunny Tenerife for Gamelab Video Games Summit, European Video Games Summit, and we're honoured to have you again, Richard, with us.
Thank you so much for joining us."

"We met in Barcelona before.
Some time has passed.
We were talking a bit about virtual worlds and how things were back then, and then we went through a pandemic and many things changed."

"So first of all, what can you share of what you guys have been doing behind the scenes, about the European games industry and how to improve it and how to make it move forward?
Well, the summit that we've just been having is very interesting.
Lots of people, lots of opinions."

"Some of them are opposing, but mainly they're in the same directions.
And the problems we have are to do with employing people with the right range of talents, getting enough money to employ the people with the right range of talents, and making sure that the games that you've made are discoverable so that you make the world's best game."

"But if nobody knows it exists, you may as well not have made it.
So those are the three big issues.
Plus, there was a big discussion about diversity, which can be helpful and can be a hindrance depending on how hard you go at it."

"And of course, back then, we also talked about virtual worlds, which is your thing.
So back then, you told me, you know, we didn't get there yet in terms of what you can offer players to freely express themselves and live in their virtual world.
So how do you feel now?
Well, we still haven't got there yet, but we're getting closer."

"A lot of the technology that's being developed by the Metaverse people, which isn't going to work for a Metaverse, nevertheless, is going to be very useful for virtual worlds.
So I'm seeing that as a positive.
It'll be easier to make a virtual world."

"Smaller teams will be able to do it, and they'll be able to make worlds that are better than the ones that we've been, well, playing for quite a while.
So I am optimistic, but we still haven't really got to the stage where the worlds that we create feel alive enough to be real to the players."

"We'll get there, but again, ask me after the next pandemic.
All right.
That about the worlds.
What about the players themselves?
Of course, one of your famous studies is about classifying players."

"So there's this classification that you did back in the day.
So how did that change with the years?
Well, it hasn't changed in the years.
I mean, I expected it to change."

"I didn't think it would last six months, but people still play for the same reasons that they always played.
But the proportions are different.
So nowadays, people are playing more for community, so they're more social."

"There are more social players in the past when there was more of the players who were playing to try and win.
So we still got people playing for the same reasons, but it's shifting because the longer you play a game, then the less you're playing it to try and win it, the more you're playing it for other reasons, perhaps to explore how the game world works or to make friends or playing with your friends, just hanging out with them."

"So although the player types are still there, the player bases are moving in different directions, but they still all fit within the same original format that was killers, achievers, socialisers, explorers.
You mentioned the metaverse before, but there are also other concepts that are changing gaming as of late."

"So for example, how do you see AI or procedurally generated elements changing the shared worlds in online games?
Well, my PhD is in artificial intelligence, so you asked the right person.
I'm a big fan of procedural generation."

"If the procedural generation is under the control of the game's...
Yeah, yeah.
If it's not, then no, but it can help do a lot of heavy lifting.
And if the designer can then change it, tweak it afterwards, then that's fine."

"It's not quite the same as getting an AI to write a book for you and then just changing a few words.
It's not like that.
We need a canvas upon which to work."

"Create me a canvas, and until eventually you've got the canvas that you like, and then you start saying, well, this would be a good place for a city.
I'm going to put a city there.
This kind of thing, then that's good."

"Also, non-player characters in virtual worlds, we can make them a lot smarter than they are at the moment.
We can give them learning abilities.
Not that that's necessarily a good thing, but we can do all sorts of things to make them behave more realistically."

"And the more realistically they behave, then the more realistic the world itself feels.
So AI is bringing quite a lot to games.
Also, it's taking a lot out of games.
There's a lot of the research done in AI has come out of games research, because games are a perfect testbed for AI."

"I'm very optimistic about it.
There are some limits, though.
There are some limits, though.
One thing I particularly don't like is the use of artificial intelligence to what they call dynamic difficulty adjustment."

"So you're playing in the game, and the AI is thinking, well, what you really want is for this person to be a bit harder or a bit easier.
And you might not, because they don't know how much you like.
At this moment, they don't know whether you want a difficult opponent or an easy opponent or this one's just fine."

"But they still keep doing it.
And I mean, when I was playing World of Warcraft first, I kept getting beaten by this NPC called Hogger, level nine, beat me up every time.
So the first thing I did when I got to level 50, which was a level cap at the time, was I went back and I beat the life out of..."

"And he stood no chance, because this is for what you did to me.
But with dynamic difficulty adjustment, I'd have gone, this is for what you...
And he would have been the same level and have beat me again.
So I didn't want that."

"It also depends on your mood and how you feel that day.
But the challenge should be the same.
You just have to be up to it.
Yeah, yes."

"If the challenge is the same every time, then you've got a bland game.
What you want is for the challenge to be harder when you want it to be harder and easier when you don't.
And so you can back off if you think this is too hard."

"But if it's easy, well, then you think, oh, well, I'll just go somewhere where it's a bit harder.
Or maybe I'll just take on more of them at once and these sort of things.
So these are decisions which players like.
But if you try and balance the game like some kind of director pulling strings, that's the use of AI that I don't particularly like."

"When you said that you have to set a limit, I thought that you were not going to talk about the difficulty, but about the old cliche of AI being too much and characters in this case, or perhaps droids becoming like humans and Westworld being something real."

"I have actually written a book about how to be a god, which talks about you as a designer of a virtual world, creating non-player characters who are as smart or maybe smarter than you are.
What kind of moral responsibilities do you have towards those non-player characters?
Because if they're as smart as you are, I mean, they can't do anything to you in the real world because they're trapped in a virtual world."

"So killer robots, it's not a problem.
But are you allowed to switch their world off?
Could you switch off the server, kill a billion non-player characters who are as smart as you, who don't know that they're in a game world, they think it's a real world, just like we could be living at the moment."

"The movie Free Guy is more or less about that.
Yeah, it is.
That's one of the areas I'm looking at at the moment.
We don't have this AI at the moment, but we'll have it in, well, how many years do you want?
50, 100, 1,000, a million?
We've got the rest of eternity."

"And eventually, we will have AI characters that are as smart as us or smarter than us.
And if you want a computer the size of a planet to do it, you can have a computer the size of a planet because that far down the line, we'll have those things."

"So we need to think now about how to handle these non-player characters while we've got the chance before we develop them.
And if they are smarter than us, well, I mean, it's not actually hard to be smarter than humans in these games."

"Back in the old text days, I wrote a non-player character and I gave him all these abilities that he could figure out what you were doing.
And I had to dumb him down because he kept beating the players.
He knew, aha, so you're carrying the potions."

"If I try and steal your potions, I'll be able to drink those potions.
So I'm going to do that.
And the players go, what?
What happened?
What happened?
You just took them."

"I mean, so, yeah, he was, I had to dumb him down.
And that wasn't using particularly sophisticated AI.
Perhaps today, nowadays, would you use those characters in a modern version of Mad?
How would you picture that game nowadays?
Well, nowadays, I wouldn't do it as text because nobody will play text."

"Well, some people do, but people are so into the idea that video games has a video component.
You can imagine why they would.
That, so I'd have to make it a real world."

"And I wouldn't set it in a fantasy environment.
There's too many of those around.
Similarly, science fiction, similarly horror.
Probably set it in a contemporary world because that way you can sell items of clothing which have got no gameplay effect, but does a heck of a lot of them."

"And in a modern world, you can do things with...
Link to NFTs?
No, no, no, no, no, no, no, no, no, no, no, no, no, no, no.
And not NFTs, no."

"In a game world, artificial scarcity isn't really fair.
In a virtual world, if you're limiting how many of something there are through artificial means, well, you're screwing people over, aren't you?
And if it's got a gameplay effect, that's even worse."

"But even if it doesn't, you're still forcing people who are interested in having something to pay for something, which, well, if you've already paid for the game, or if your presence in the game is helping other people to pay for things in the game."

"So I'm not a fan of NFTs in games at all.
In the rest of the world, well, yeah.
I mean, okay, well, there might be Ponzi schemes, but then so is modern art."

"That's just buying something that everyone's agreed is worth a lot.
But whether it is or not, well, that's just a case of whether other people are going to buy it.
So I'm not a fan of NFTs in games."

"And all the things about moving things from one game to another game and back again, you know.
So I've got a bow and arrow here, so that's a ranged weapon.
And I'm going to go into this game here."

"It's science fiction, so what's a ranged weapon?
So my bow and arrow now turns into a laser gun.
Well, and then when I come back, I want the bow and arrow.
Well, that sort of thing, that doesn't work."

"Because bows and arrows in a survival game might be very rare.
And everybody in this game here might have a ranged weapon.
So if you move into there, then what was very rare isn't very rare anymore.
And so what you want is to say, well, how can I assign a value to this object, which will then have an equivalent value in another one?
And it is, but it's easy if you say, well, what happens is I sell this object and I get money and then I buy an object in there and that's worth the same."

"So you just, you're going through money and that's a lot easier than trying to move things using some kind of protocol between games.
All right, closing one.
We've talked about game design, characters, artificial intelligence, but what about coding and programming?
Of course, you are into that as well."

"So how do you see that evolved in the past few years?
And how do you see developers using new tools and new means for better coding?
Yes, well, when I started out, coding was a creative thing that you had to do.
I had to write my own packages to use an array."

"We didn't even have arrays.
We only had vectors, one-dimensional arrays.
From then on, it changed from creative like that to be more problem solving.
How do I do this?
And you're trying to figure things out."

"And then after that, it went on to, okay, so I've got these different units that I'm taking that have already been written and I have to stitch them together to make them work as a bigger tapestry.
And then nowadays, in games, a lot of it is drag-and-drop programming using Blueprints and things in Unreal, Unity, where you're drawing images and you are coding, but you're doing it visually."

"And that limits what you can do because it doesn't have the fidelity.
You can't go very detailed, but what it does do, you can do really quickly.
So when we started out, it was like embroidery, stitching away.
And then it became more like the tapestry making, where you're getting pieces of cloth and sticking them all together."

"And nowadays, it's more like building an outfit.
Well, I'll just go and buy that jacket and that shirt and that tie.
And then these cufflinks, yes, I should show you my cufflinks.
They're Tetris pieces."

"And yeah, it has changed and it'll probably keep changing.
But ultimately, someone at the bottom has to code in a programming language.
And those are the people who the industry probably needs the most.
And there's some art to it as well."

"All right.
So the sun is setting.
You cannot see it, but the sun is setting in the background.
And that's a beautiful moment to call it a day."

"So thank you so much, Professor.
Looking forward to seeing you again.
Thank you.
Looking forward to seeing you again."

Gamelab

More

Videos

More

Movie Trailers

More

Trailers

More

Events

More