All products featured on WIRED are independently selected by our editors. However, we may receive compensation from retailers and/or from purchases of products through these links.
Artificial intelligence is spreading through our lives. As it moves from new feeds to productivity tools, the boundary between what is human-made and machine-made is becoming almost invisible, resulting in vast shifts in how we perceive and interact with the world.
AI has surprised us with how it can move, see and hear, but also, with what it can create. Once thought of as a solely human trait, creativity is now a debated topic. Can AI fully master the creative process? Or is a thought process constrained by what us humans define it as?
In 2016, Google DeepMind's AlphaGo system trounced the world’s reigning champion, Lee Sedol at Go – an ancient Chinese game – by using its creative intelligence to devise a winning strategy. Although AlphaGo has been described as years ahead of its time, AI is rapidly sweeping into our everyday lives and its advancement has begun to transform and aid creative processes, as well as challenge what we believe to be ‘creative thinking’.
In one game against Sedol, the Google system's 37th move was one no human would think to make. "It's not a human move. I've never seen a human play this move," one fellow Go champion said, before adding that it was "beautiful".
Using its ability to learn human creativity from vast datasets and quickly produce solutions to tough problems, AI is creating a new landscape of creative opportunity. “As humans we are very good at making sophisticated decisions to select the best fit by taking into account convoluted considerations,” says Daghan Cam and Michail Desyllas, Founders of Ai Build, who use AI and robotic technologies for additive manufacturing. “However, we are not very good at producing large arrays of possible solutions to choose from. We lose our ability to take these decisions efficiently and consistently as the number of possible solutions increases.”
AI is a solution to this issue and is enabling designers such as Cam and Desyllas to construct better designs by accelerating their thought processes and increasing their creative options. By using AI to automate the production process, Ai Build reduces material usage, labour and costs whilst creating complex designs that were once impossible to devise and build. In one instance it used concrete in irregular shaped molds to reduce the amount of material needed.
Creative AI is being used as more than a workmanlike tool, however. It is also influencing the music industry, giving musicians the ability to better understand their creative process and even themselves. Beatboxer Reeps One, also known as Harry Yeff, is the first voice artist to perform a ‘battle’ with an AI opponent. Via a deep-learning programme, a machine was programmed to categorizes Yeff’s sounds, but also create its own music.
After taking part in chess tournaments and using AI to improve his game, Yeff was initially fascinated with developing what he calls a ‘pet self’ – one which challenges you and leads you towards a new innovation. “I loved the idea of how an opponent can interrupt you and force you to become more,” Yeff explains. “AI is not an ‘other’ – a hijacked sci-fi representation; it's a tool, and it's something that we’re already engaging with every day.” AI can allow musicians such as Yeff to better understand and develop their creative abilities and, similar to Sedol’s competition with AlphaGo, raise their game.
“It blew my mind when they told me that in a ten minute clip of me performing, I technically made 140 fixed variations of kick drum and that's because of the subtle human changes, which are far beyond human observation.” Although AI has enabled Yeff to further unlock his creative potential, he stresses how science fiction has influenced people to view AI as a separate entity and the importance of artists in breaking down this stigma. “It’s a responsibility of artists like myself to go into these spaces with no agenda and try to raise awareness and engage with it to make sure that people are seeing how these things can be used,” Yeff explains.
Although AI is giving artists increasing insight into their creative process and abilities, it also beginning to inform practitioners of the kind of creative output society favours. As far back as 2015, people were confusing computer-generated music with classic pieces from the composer J.S. Bach.
Similarly to Ai Build and Yeff, visual artist Anna Ridler is using AI as a tool to shape her creative strategy. In Ridler’s latest project ‘Tulips,’ she took photographs of flowers and produced an algorithm that created thousands of notional tulip varieties with unique characteristics. Although Ridler’s project illustrates the creative possibilities of AI used in visual art, Ridler says recommendation algorithms are changing the industry by influencing artists and curators to create work which society is more likely to engage with.
“A lot of curation of art and design is now being driven by data and what people seem to like,” she says. “Netflix has probably had more of an impact on film making than any deep fake technology.” In an auction at the end of October 2018, an AI-generated portrait of a fictitious man sold for £341,000.
In addition to giving creatives more options, insight into their work, and the awareness of audience engagement, AI is influencing creativity by being programmed to complete mundane tasks so that practitioners can focus on more creative ones. This is widely illustrated in the gaming industry, where AI helps automate game development tasks such as testing programs.
Yet, AI used in creative practice today is self-controlled, meaning that it works within a controlled environment and is programmed to do what its creators tell it. “Creative people still want a lot of control over what AI is allowed to do,” says Mike Cook, an AI researcher at Queen Mary University of London and programmer of ‘Games by Angelina’ — an AI that builds its own games.
“That's not a bad thing, and it's still letting AI enhance our work in incredible ways, but I think the next step – accepting AI into our work as a contributor – it will be a huge shift in the way people create.” In the future, Cook thinks that AI might take on a fully creative and personal role in playing and designing games. For example, in future versions of Minecraft, an AI might be programmed to have an active relationship with the player and help them achieve their goals. DeepMind has already proved it can play the infinitely complex StarCraft better than humans.
Cook also hopes to see AI making it easier for people to make games as well as play them. “I hope that AI game designers will be able to act as teachers, mentors and collaborators for people, helping them do things they struggle with, and giving feedback on the games they make,” says Cook.
As AI provides an additional and vastly diverse ‘creative arm’ in a number of industries, it is challenging the possibilities of creation and giving us more understanding of creativity. However, as its initial conditions are created by human intelligence, AI is still a human endeavour, and although it is starting to develop creative characteristics, its lacks the complex link between imagination, abstract thinking and episodic memory which scientists believe makes up creativity.
Yet as human creatives are praised for their originality, many of the most famous artistry has been influenced and mimicked – used to drive new movements and trains of thought. As historian Yuval Noah Harari once wrote, “humans are essentially a collection of biological algorithms shaped by millions of years of evolution,” which, when thought of like this, brings us closer to AI than we might have thought.
Reeps One is speaking at WIRED Pulse: AI at the Barbican on June 15; get more information and book tickets on wired.uk/ai-event.
This article was originally published by WIRED UK
