Design and tech agency Uniform recently held a panel discussion at SXSW with the title Every little thing’s gonna be AI , exploring the potential impact of new technology on the music industry. Here, Michael Shorter, senior creative technologist at Uniform, writes on existing and imminent innovations that are changing music, and how designers can control it.
Music’s relationship with technology has always been an uneasy one. Ever since Dylan went electric, the introduction of new and different has caused friction and fissure. Depending on your perspective, the 808, Napster and Garageband have either revolutionised or ravaged the music industry. However, initial upset almost always heralds a new way of thinking, be that the creation of dance culture, globalisation of music or democratisation of audio production.
Recently Uniform built Solo, an Artificial Intelligence radio that can read and reflect the listener’s emotions through song choice. One of many conversations about AI, Solo forced us to think about the relationship between people, music and machines.
It’s a conversation we took to this year’s SXSW in a bid to understand where and how easily this kind of emotive AI might sit in an industry pummelled by change. In conversation with industry veterans, trendsetters and academics, we sought to understand how future tech can, and will impact the industry.
While robots aren’t going to be writing hit songs, they are helping artists plot tours that will sell, grow communities with chatbots and make intuitive recommendations based not only on listening habits but also location and weather.
But is this what we want? Will the ever-improving algorithm simply hone our tastes into the ultimate genre-specific playlist? Discover Weekly is great, but what about gut feeling, tacit instinct and surprise? It’s hard to believe a machine will ever replicate that emotional connection to music.
The kind of muzak generated by Jukedeck and Sony’s Flow Machine software is currently pretty awful but it won’t be that way forever. Much of the charts are already a convergence of sound and style as labels compete for a global pop hit. The ethics of profit sharing with AI are pretty murky but you can guarantee it’ll be cheaper than a human songwriter’s cut. It’s unlikely development of compositional AI will stall at the first (chilling) rendition.
"Admit it, Tupac’s hologram was weird, and the algorithm already exists to compose a brand new AI-Tupac album. Do we really want to go down that path?"Michael Shorter, senior creative technologist at Uniform
We don’t get to pick and choose how and when new technology will be applied to our preferred artform. The same technology that analyses songs for copyright infringement can also be used to predict creative decisions and compose new music by dead artists. Admit it, Tupac’s hologram was weird, and the algorithm already exists to compose a brand new AI-Tupac album. Do we really want to go down that path?
If we learned anything from SXSW’s AI fixation, it was that the gap between AI fact and fiction is wide, but ever diminishing. Robots aren’t likely to write a number one anytime soon, but that isn’t going to stop them from trying. The labels (and Metallica) warned us about Napster, but it didn’t stop us from digitising music. Like it or lump it, new music technology isn’t going away.
A set of drums and a Roland TR- 808 both technically do the same job, yet it takes vastly different skills to get a harmonious sound out them. Turntables, synths and the programmes like Logic and Cuebase mean artists no longer need perfect pitch to create. More than that, as bits of kit, they were central to the creation of hip hop and dance culture.
Music production in 2017 bears little resemblance to the 1970s. Yet both periods stared down a musical revolution brought about by emerging technology. Just as Roxy Music, Parliament and Herbie Hancock got to grips with synths, changing the musical landscape forever, so must contemporary creators get to grips with AI.
As the agencies and developers behind AI we can forge ahead with pop’s equivalent of the Terminator, wiping out future starlets with impunity. Or, we can design for a more emotional connection to machine learning. For Artificial Intelligence to complement artistic production and not replace it, we need to place the talent firmly at the centre of all we build.