The unknown unknowns of creative AI
Melting into the keyboard with Caterina Moruzzi
What can a 300-year-old musical instrument tell us about AI? Ask Caterina Moruzzi. An AI researcher at University of Edinburgh, she originally trained as a concert pianist.
“There are quite a lot of interesting parallels,” she told me. When she plays, “sometimes you don’t know where the instrument stops and the body starts.”
Could AI be like that? Questions like this changed Moruzzi’s career. She also studied philosophy, and during her PhD research, she first encountered AI.
“During the time I was visiting McGill University in Montreal, Ian Goodfellow was developing Generative Adversarial Networks. And I thought, ‘This might change my PhD, might change the nature of music.’ I started working on GANs, and how technology is changing our creative processes.”
Now, her insights are indeed changing how we think about creativity.
“The way I think about creativity is not the individual genius, or as only applied to the arts, but the idea of creating something novel, with some kind of agency. The key to this is when someone or something knows when to stop creating.”
Knowing when to stop, she says, is a measure of intentionality – doing something with a purpose, knowing what you want and where you’re going. “And I think that’s something that machines shouldn’t take away from creators,” she adds.
“I would say AI hasn’t reached this stage yet,” she says. “But GANs were a huge step forward, because there is a kind of self-judgment within the system. Generative AI doesn’t work exactly like this, but the idea of internal competition is still present. But for me, until it knows when to stop, it’s still a tool.”
Creating with intent
Intentionality is not the only dimension of creativity that Moruzzi looks at in relation to AI. Others include novelty, problem-solving, surprise, and value. By deconstructing creativity into these different dimensions, she hopes to be able to apply that, and to measure, the extent to which computers can be considered creative.
“In computational creativity,” she says, “it’s important to have a way for people, and also computers, to evaluate their creativity. So it might be impossible to say exactly what creativity is, but at least we can narrow down some key dimensions.”
Considered this way, even aspects like problem-solving can be useful. “Some people say that in art, we don’t always have problems to solve,” she says. “But the history of art was about artists’ development, and the development of tools – like the piano. It’s because they had problems to solve. Even if the problem is, ‘How can I express myself, given that the tools I have available are insufficient?’”
This indeed impacted how the technology of the piano evolved. “It had to keep up with the needs of different pianists,” she says.
The unknown unknowns
In her research, she draws from the work of Margaret Boden, who theorised three different types of creativity.
“There is creativity that combines elements from different areas,” says Moruzzi “Then there is exploratory creativity – to explore something we didn’t think about. And then transformational creativity – completely going outside how we see things.”
“So the first one,” she continues, “is what is already known, but remixing it – this is how generative AI works. The second is about the ‘known unknowns’ – exploring what we think can exist, but we don’t know exactly what it is. And the third is the completely unknown unknowns. What Boden says is that only this third kind of creativity [in addition to the capacity of evaluating the outcome] would classify AI as creative as humans are. For now I believe that AI is at the exploratory creativity stage.”
“Do you think it can get there?” I ask.
“Maybe we can reach transformational creativity through AI,” she replies. “Of course there are ethical issues around data collection, but for the first time we can harness huge amounts of human knowledge in one place, that as humans we wouldn’t be able to process.”
Such “artificial creativity,” as she calls it, is not intended to make machines as creative as humans, but simply to create some measures that could apply to both. “It’s trying to give a less anthropocentric way of understanding creativity,” she says, “trying to level the field.”
Getting emotional
I thought of artist Mona Hedayati, who in a previous interview told me that the way humans operate is completely different from machines, and we only regard it as creative because it doesn’t conform to our understanding. I asked Moruzzi about this, and she referred to Daniel Kahneman’s 2011 bestseller Thinking Fast and Slow.
“People have one system that immediately and intuitively reacts to the environment,” she explained, “and another that is more logical – reasoning, slowing down – when we’re confronted with something unexpected. That second, more analytical, system is closer to how machine learning works. The other, more impulsive, reaction can often be wrong, but that’s the beauty of human nature.”
That intuitive, “fast” thinking is related to emotions – something that machines of course don’t have, but more and more AI researchers are trying to quantify emotions for the benefit of machines. And now, some AI systems can recognise people’s emotional states better than other people can, based on the sound of their voice, or their facial expressions.
“I think there are two very distinct types of emotions,” Moruzzi tells me. “Again, the very impulsive kind that comes from our animal nature – fear, panic, attraction etc. And then there are the more complex emotions. I see these as more easily reproducible in a machine. They’re not exactly rational, but we can better trace what triggered them – our background, our environment.”
Back to the body
We start to pull apart what separates humans from machines, and how much the two can be compared. The big difference, of course, is that we have a body and can move around. For Moruzzi, this doesn’t mean that robots might be any more creative or intelligent than humans. Instead, it means a different kind of intelligence. We talk about how some animals have a sensory-motor intelligence that has nothing to do with their brains, or organisms that, if you cut them into five pieces, each piece still seems to “think” in the same way.
But Moruzzi is interested in the body for other reasons.
“Thinking about embodiment and AI,” she says, “most people think of robots. But there are so many other technologies we deal with on a day to day basis. Look at all the tools for DJs to modulate sound. Or even the iPad and a pen can be interesting – how you use it in an embodied way impacts your creative process.”
This, of course, goes right back to her experience as a pianist. That an instrument which is just a kind of button box can be so expressive suggests some hope for computers. If we can get away from the text prompt as an interface with AI, this might carry its, and our, creativity much further.
“There are interesting developments in AI hardware,” she says. “It’s early days, but it’s exciting to think of new ways of interacting with these technologies. We’ve had the PC and the smartphone for too long now. Lots of exciting developments – it’s an exciting time to be working in this area.”
Want more? Read a complete transcript of our interview here. Find out more about Dr Moruzzi’s research here.