#11: AI and Creativity

Ian Bowie
Hello and welcome to AI Unfiltered with me Ian Bowie and our resident expert, Michael Stormbom, where we will be talking about everything to do with AI in our modern digital society and what the future holds for all of us.

So today we’ll be talking about AI and creativity, and is AI actually going to become more creative than its human equivalent. So if you now start thinking like… I’ve got two sons who are both chefs if we take that forward, maybe another 10-15 years and suddenly they ain’t got a job anymore, replaced by a robot. I mean, who on earth would have ever seen a robot replacing a chef? Gordon Ramsay, I hope you’re listening to this because your time is limited my friend.

Michael Stormbom
One area where the robots are not going to take over just yet in 10 years, I mean, the aspect needs human creativity. I mean, yes, of course, if you’re preparing exactly the same dish, day in, day out, so it becomes…

Ian Bowie
But we know that these algorithms are capable of learning. So why…

Michael Stormbom
But about learning from what?

Ian Bowie
Well, they teach themselves in the end, don’t they? Or they teach through observations? I mean, imagine, imagine an algorithm that’s put into a robot, which can then basically within what two days, surf the internet and absorb every single recipe every YouTube video that Gordon Ramsay, Jamie Oliver and all these other chefs have put up there and then mix it up for itself. So once that algorithm has absorbed all that information, it will be able to make whatever…

Michael Stormbom
Yeah, but it won’t be able to create anything new. necessarily.

Ian Bowie
Yeah but it will, because it’ll have the ability to then think for itself thinking about okay, what about a bit of Asian fusion or what about a bit of Jamaican mixed with German. It’ll start experimenting!

Michael Stormbom
No, because I’m not. I’m not convinced. No.

Ian Bowie
I think we’ll get to that. I honestly, do. I honestly think that these algorithms will start thinking for themselves and experimenting on their own.

Michael Stormbom
Yeah, no, I don’t think so. I mean, it would need to be very deliberately. Sort of a programmed, the mix and match kind of aspect to it.

Ian Bowie
Yeah, but it maybe will be. I mean, I think, you know, if we’d been having a conversation about technology 30 years ago, we wouldn’t even be able to imagine some of the things that are possible today. So why can’t we try and imagine what might be possible in 30 years time?

Michael Stormbom
Yeah, well, I mean, what comes to mind is not so much food making but sort of music making, for example, is…

Ian Bowie
Music, writing. Any creative pastime.

Michael Stormbom
Because there are already algorithms that you know, you can design the sort of like the music, the pop music that’s most likely to hit the charts kind of thing. But, I mean, that’s not really creativity. That’s creating a lot of predictable pieces of music for to optimize.

Ian Bowie
What’s the difference if it’s written by a human being, or it’s written by a machine.

Michael Stormbom
No, I don’t know. Certainly not. I mean, I don’t care in that sense.

Ian Bowie
Music executives the world over are doing isn’t it? They’re only trying to predict what the next big hit will be very predictable. And there are many artists out there where basically, their latest hit doesn’t sound very much different from the last one. And it’s the same with books. You know, publishers have become very conservative. They’re not very keen on taking on new unknown talent. Actually, human beings are the ones who are becoming less and less creative, because they’re averse to risk. So in actual fact, it might be that machines become the creative risk takers of the future.

Michael Stormbom
Yeah, I’m not convinced. I think… I would argue that, I mean, to the extent that we are becoming predictable when it comes to creativity, I think the machines are just following that path.

Ian Bowie
Yeah, but that’s the thing isn’t exactly what you said. We are becoming creative. Sorry. We’re becoming predictively creative, which is an oxymoron. Because actually, creativity should be about originality and something new.

Michael Stormbom
New and unexpected. Yes.

Ian Bowie
Absolutely. That’s real creativity, just regurgitating the same old crap. That’s nothing.

Michael Stormbom
Yeah. But that’s very much what I argue that AI will be doing, it will be regurgitating the same crap because…

Ian Bowie
I think a lot depends on the audience if the audience is willing to accept the same old crap, which they seem to be willing to alas, but then having said that, they’re only willing to accept it in my opinion, because that’s what they’re being fed. Because that’s where the money is. So you take an artist, and they have a hit, and then they have another hit, and then they have another hit. So the music executives know they’re safe with that artist. Oh, let’s put another one. You know, stick their name on. It can’t be too different from the last one though, because otherwise people won’t buy it. And it’s just a vicious circle, isn’t it? You know, that’s why in a way, things like YouTube, Spotify, whatever, SoundCloud that they’re very, very good, because it allows everybody to have a voice.

Michael Stormbom
Yeah, no, I mean, I think Spotify and SoundCloud and so forth. I mean, they’re, they’re good in that sense, because they it takes away just like the old gatekeepers.

Ian Bowie
Yeah, I mean, it gives everybody a voice everyone has a chance. If you’re lucky, you will find an audience without the help of one of the big publishing houses or music labels. Of course, once you’ve got a voice, they’ll pick you up immediately.

Michael Stormbom
Then you cease to be an artist and you become a brand.

Ian Bowie
And then you become a commodity of course. Yeah.

Michael Stormbom
Yes. So the point I wanted to make about sort of predictability versus creativity. Because I mean, music and all these so called arts, they are very much produced by algorithms already. So in that sense, having, if it just happens to be an AI that is executing the algorithm, I think that makes no difference in that sense. It’s still the same old crap.

Ian Bowie
Yeah, I mean, you might be right I mean, I don’t know. I mean, it it’s difficult to know because I mean, of course, this is something else that technology has allowed, you know, everybody can be a musician. Everybody can be a singer. Everybody can be a writer, a poet, a visual artist, whatever. So actually, possibly what’s actually happening is true creativity is drowning in the sheer volume of everything that’s out there. It’s getting lost.

Michael Stormbom
I think that’s part of it because yeah, the traditional gatekeepers are no longer they are and, mostly for better, because everyone has a chance. But of course, since literally everyone has a chance it’s very also very hard to get noticed…

Ian Bowie
It waters it down a little bit as well, isn’t it? You know, because if you give people a choice of 1000 books to read, and you ask 10,000 people, what book would you like to read? They’re all going to choose you know, so you’ve got 1000 books you’ve got. All right, let’s say 1000 people, 1000 books, 1000 people, the chances are that some of those books won’t get read at all. Some books will get maybe a couple of hundred readers, some books will get 50 each, and it’ll be well spread out over that 1000 titles. If you give 1000 people only 10 books to read, all of those books are going to get read.

Michael Stormbom
Yeah, I mean, they’ve done research into like, people and making choices. So I mean, it actually has been shown that having fewer choices makes you more satisfied, because if you have too many choices, then you will spend a lot of time agonizing over which is the right choice and then you will inevitably be dissatisfied because you spent so much time on making the choice in the first place. Whereas if you only have a couple of choices then you already then it is sort of baked in it is not necessarily going to be exactly what you want. So you feel more satisfied after making that.

Ian Bowie
Yeah. Yeah, I think that’s part of the prom with the whole world at the moment actually is there’s just too much of everything.

Michael Stormbom
This is the problem with Netflix if you try to figure out what to watch and…. There’s just there’s just too much stuff and you have no idea where to start and you have only so much time at your disposal.

Ian Bowie
That’s the thing. Yeah, you know, you’ve got so much time to allocate, for example, to watch a film or to watch a documentary or television series and you therefore want it to be something worth that time.

Michael Stormbom
Yeah, yes. And then spend half an hour trying to figure out what to watch and, like…

Ian Bowie
I suppose it’s the same thing with our podcasts, you know, how do people find us? How did you find us, you people who are listening to us?

Michael Stormbom
I was standing out on the street handing out flyers to people. No, but…

Ian Bowie
Yeah, maybe. Of course, you tell your friends you tell the family, everybody you know, you hope that they tell a few people and so it goes.

Michael Stormbom
This presupposes that our podcast has any listeners, but…

Ian Bowie
oh yeah, of course it does. We’re not just talking to the airwaves. Anyway, yeah. Press those like buttons, people, let us know that you’re out there.

Michael Stormbom
We’re accepting your calls now.

Ian Bowie
Absolutely. No, no, no, no calls. Just hit the like button, like they do on Youtube.

Michael Stormbom
Shoot an email if you absolutely have to communicate.

Ian Bowie
If you think about, you know, people say where does your creativity come from? How do you get your ideas? Of course you get it from your experiences and things you’ve seen what you’ve read, what you’ve felt, alright, okay. Perhaps machines can’t feel but they can certainly have far greater experience in terms of listening and reading. You know, if you can read 2000 books in a lifetime, AI can read 2 million books. So imagine being able to draw on all the information that you have read from 2 million books as opposed to a couple of 1000.

Michael Stormbom
Yeah, I’m pondering…

Ian Bowie
You’re looking skeptical.

Michael Stormbom
Yeah. And I’m pondering this very thing because, well, I mean, not to go too metaphysical on this, but what does it mean to be creative in the first place? So I mean, I would argue that AI, it simply recognizes patterns and fills in the blanks based on those patterns. So sure if that’s if we say that that’s how creativity works, then…

Ian Bowie
Creativity is a process. And it’s a process that you work through in order to create something original, new different, whatever. And there is no reason why an AI algorithm cannot work through exactly the same process that an artist does, in order to come up with something new, original. It is simply understanding what the process is that you need to go through in order to create something.

Michael Stormbom
Yeah, no, I mean, to that extent I do agree.

Ian Bowie
In fact, actually because, I mean, it’s not necessarily logical. And then of course, you know if, if all AI is based on pure logic and nothing else, then I would agree with you that AI will never be able to be truly creative in the human sense.

Michael Stormbom
Well, well, yes. I think that’s what I’m trying to articulate, if not very creatively. No, exactly that so so that’s why I’m arguing that I agree with you if we, if we were to design, like an AI writer, and they’re sort of putting into, well, if we seeded into finding patterns and then breaking the pattern, breaking the rules.

I mean, I actually created this is an algorithm that generates text in the style of the 19th century Swedish language author Zacharias Topelius. Basically what I did is I took well the books that he had written and and and trained his model on those and then it starts outputting text in a similar style as Topelius.

Ian Bowie
Brilliant. Yeah. So there you go. That’s a simple example of how it can be done. So you can actually you can actually bring dead authors back to life. In a way.

Michael Stormbom
Yeah, indeed. Absolutely.

Ian Bowie
That takes us into another area as well, doesn’t it? Oh, look at this, the lost Charles Dickens manuscript. How much would you like to give me for that?

Michael Stormbom
Yeah, indeed. So I’m not sure how that ties into creativity necessarily, but it of course is an interesting question, who is the actual author of the text? It has been written, because I didn’t write it. Zacharis Topelius didn’t write it. He provided me the words, but…

Ian Bowie
I suppose indirectly, you might argue that you wrote it because you wrote the algorithm that wrote it. So maybe it was a joint effort?

Michael Stormbom
Possibly. Yeah. A Stormbom-Topelius co-production.

Ian Bowie
Yeah, I was gonna say it’s a co-production between the three of you. It wouldn’t have been possible without his original writings,

Michael Stormbom
No, indeed.

Ian Bowie
Wouldn’t have been possible without your skills to create an algorithm that could basically mimic what he wrote. So therefore, it’s you, the algorithm and him.

Michael Stormbom
A collaboration between man and machine.

Ian Bowie
Absolutely, yeah. Very good example.

Michael Stormbom
Yeah. So of course, we were talking about creativity. And, and AI creating things that are unexpected. So of course, possibly, let’s say blending two different authors. Then sort of creating new work through fusion in a way.

Ian Bowie
So we have AI at the moment. What about if we suddenly have organic AI? So that that it’s almost a hybrid? A semi-living thing? Capable of independent thought?

Michael Stormbom
We’re gonna get into brain in a jar territory I mean, I think we’re looking at another 50 years ahead before we are there any…

Ian Bowie
Maybe? Yeah, absolutely. I mean, it’s not gonna happen tomorrow. Possibly not even in our lifetime. But I do believe one day will. I do believe it. I believe that one day, somehow, they will be able to tap into the human brain and somehow create a hybrid model, a superhuman almost, capable of infinite thought and decision making. It’s a quite scary thought I know. But…

Michael Stormbom
Yeah, but I don’t think we have to worry about that particular….

Ian Bowie
I mean, I think it’s interesting to talk about. Of course, it’s the realms of science fiction, but a lot of people actually are fascinated by science fiction, and actually science fiction often turns into science fact. Yeah, no, absolutely. There’s a wonderful picture of, it’s from Star Trek. And it’s actually when they had this guy, Captain Picard and there was a picture of Patrick Stewart holding a device in his hand in an episode of Star Trek, which looks identical to a modern tablet. And this was many years before Steve Jobs stood up on stage with the iPad. So it was a science fiction device. And maybe, I don’t know, was it 10-15 years later? Yeah, it was reality.

Michael Stormbom
Although it should be set at one Steve Jobs stood on a stage with that tablet. Not a particularly revolutionary about the tablet, it has been around since early 90s. Which actually comes pretty close to when that Star Trek episode would have aired.

Ian Bowie
Yeah. Yeah. But of course, you’re a Finn. You know, of course, you don’t like Apple because it destroyed Nokia. So of course, you have to say something like that. So…

Michael Stormbom
No, I mean, not absolutely nothing against Apple. But I think, I think when people think of for example, the iPhone and the tablet as very revolutionary and, yes, they were, but not in sort of like a technological breakthrough sense, but rather in the value it brought to the people. So I mean, iPhone, the smartphone concept. I mean, yeah, we know here in Finland, we’ve had those for years before, before the iPhone, and likewise with the tablets, they have been around frim since the early 90s, really so. So the innovation, there wasn’t necessarily that these devices exist but rather the value they bring to the user, the whole ecosystem around it. So for example, it’s it’s easy to develop applications for it, there are plenty of applications and so forth. And so I think the the innovation, there wasn’t really in the technical sense. I think it was more on just like, let’s call it value innovation.

Ian Bowie
I think I think it was also the fact that Apple appealed to emotions. It appealed to… Apple actually made computers sexy, in a way.

Michael Stormbom
Yeah. That’s what happened when Steve Jobs returned to Apple and they created the iMac, absolutely, yeah, no, indeed. I mean, it’s just much about the brand and emotions that it brings.

Ian Bowie
Because I mean, basically the end of the day, we are emotional creatures. Yeah. Yeah. We use logic quite simply to justify our emotional decision making.

Michael Stormbom
Yeah sense making, as it is called.

Ian Bowie
Yeah, yeah, definitely.

Michael Stormbom
And also to the point about AI and using AI to create things, I mean it’s all about. It’s all about the data, as if you have data on what type of music sells and if you can break it down to a very, very granular level and surely you can design or rather automatically create the perfect piece of music. Perfect piece in the sense of maximizing the likelihood of it. Soaring to the top of the charts.

Ian Bowie
Yeah. You I think, I think they probably have already actually,

Michael Stormbom
I think they do.

Ian Bowie
I think they do. Yeah, I mean, we can talk about this forever. I mean, it’s an absolutely fascinating subject. Yeah. And I think we probably will come back to it.

Michael Stormbom
I think we have about 50 episodes worth of stuff.

Ian Bowie
Oh, absolutely. Yeah, no, I think for now, we’ll we’ll leave it with what we’ve got. But we’ll certainly come back to it at a later stage.

Michael Stormbom
Yeah, indeed.

Transcribed by https://otter.ai