BY ROB MCGOWAN, STRATEGY DIRECTOR.
I received further evidence that artificial intelligence has well and truly crossed over to the mainstream this weekend while listening to BBC Radio 6 Music on Sunday morning. On the early show they ran a short piece about ‘Solo’ the ‘emotional radio that plays music to fit your mood’. Solo is a wall-hung device with a circular screen at its center that takes a photo of the user as they stand in front of it. This image is sent to Microsoft who run it through their box of tricks, analysing facial features and calculating an ‘emotional score’ with values for happiness, sadness, anger and so forth. This is then joined with data from Spotify who hold music valence ratings for their entire catalogue. Spotify then chooses the song that, in theory at least, most closely matches your mood. It’s a neat idea, and while it may not be perfect for all situations (I may be sad but maybe I want that uplifting house tune to snap me out of it for example), it is a nice illustration of machines really trying to respond to emotion rather than just behavior.
For the last couple of years or so the marketing community has been fairly aligned in the sense that we have tended to try and add some balance to the hype surrounding AI by reminding the market at large that successful marketing is a blend of many different skills, centered around humans. While technology and artificial intelligence algorithms are an important and increasingly well-known part of the mix, we marketers are typically keen to remind people that without the creativity, abstract thought and ideas of humans, marketing activity can seem a bit dry, a bit obvious, or dare I say it, a bit robotic. While it may be fairly self-serving to espouse that point of view (marketing agencies don’t, on the whole, tend to own or develop technology – hence it is in our interest to remind people that ‘people buy from people’) I also happen to think that for the moment at least it is true.
What excites me about Solo is that it is attempting to read emotions – distinctly ‘natural’ rather than technical or solely behavioral traits – and combining this analysis with other human generated behavioral datasets. And it is doing so in a fun and commercially attractive way. It’s fairly easy to see how adaptations of this idea could spill over in to all sorts of things we like to do. Perfect movie for a hangover? Best drink to have on the day you fall in love? Ideal book for a dreary Sunday afternoon? It’s more nuanced than the “you just did this so I am going to send you that” marketing that is commonplace.
Our job as marketers, as well as having the ideas about how to engage humans with products, is to think about how we can create and join up different sets of data and deploy them in fairly abstract ways. This is what excites me about AI – augmenting the ideas we have with the processing power to create approachable, useful, playful things that didn’t seem possible just five years ago. As Mike Shorter, one of the lead technologists who worked on Solo puts it, “When you try Solo, you think not only about how technology is changing, but more importantly how its relationship with us is changing.”