This week advertising titans M & C Saatchi unveiled a new technology which has got the world of marketing talking in a big way. It’s an artificially intelligent screen that uses algorithms and the public’s reactions to continually optimise itself. The technology, currently being trialled on two screens in central London, monitors people’s reactions to a poster and uses their engagement (or lack thereof) to decide what elements of an advert to keep and what to discard.
It seems as though the industry’s having a dual reaction: On one hand, we’re all marvelling at the innovation, and wishing we’d thought of it first. On the other, we’re quaking in our boots at the imminent redundancy of our jobs that this could bring about (especially for me, who only just started last week). I’m going to argue that whilst this technology is undoubtedly brilliant, we’re not going to lose our jobs over it any time soon.
First of all, although this technology is incredibly innovative, it’s not capable of innovating by itself: It starts with a set number of inputs (fonts, layout options, colour schemes, textures, phrases) and works downwards from there, discarding the elements that just don’t work. It’s essentially reductive, with an inbent logic that can only work with what it’s already got. Whilst sifting through ideas and elements to figure out what’s going to work is of course a vital part of the formation of any advert or campaign, it’s only one part of the process, and only one that comes after ideas are generated in the first place. That’s this technology’s crucial flaw: It’s capable of original combinations, but not original ideas and material.
It also lacks a certain degree of critical capacity that it seems A.I. just can’t get right yet. For example, the technology features a program that can actually write its own copy using a predetermined bank of words and phrases. This would be great, if only it worked: One of the poster combinations shown in the above video features the tagline ‘Bahio is the New Steam!’, a line that makes so little sense that any copywriter worth their salt wouldn’t even allow it to get to the ‘Should-I-say-this-out-loud?’ stage of the brainstorm, let alone emblazon it on a very much public Central London poster. The same goes for ‘Sip a Delicious Early Blend Every Circumstance’, which is not so much clunky as just plain rubbish. It seems that we have a long way to go when it comes to programming common sense.
At first, what I really thought was valuable and great about the M & C Saatchi video is how the novel form of producing the advert was itself part of the ad campaign, which would have been very coolly postmodern and self-aware were it not for the fact that the company in question, ‘Bahio’, is completely made up. This doesn’t bode well for any Londoners who had a particularly profound reaction to the poster and are now scouring the city for a delicious cup of Bahio Coffee (It’s the ‘new steam’, or so I here).
The other big problem here is one of feedback: People’s reactions to the poster on the street are ultimately arbitrary and depend on innumerable other outlying factors (the weather that day, the extent to which they’re in a hurry, the overall mood of the city that morning). Arbitrary data only achieves arbitrary results, and this seems to be true of what M & C Saatchi have found so far: The fact that ‘people like hearts’, mentioned as a ground-breaking result in the video, doesn’t seem to be a revelation worth going to all that programming effort for.
Ultimately, we need forward, daring thinkers to come up with forward, daring ideas. No algorithm can replace that, something that Saatchi’s recent endeavour has only served to highlight. One thing’s for sure: in 1989, when the Berlin Wall finally gave way and onlookers were met with a huge ‘Saatchi & Saatchi: First Over The Wall’ poster on the other side, it was a work of advertising genius. And you can be sure that it wasn’t an algorithm that came up with that.