OK Google, are you listening?
As a communications agency we tend to focus on how people relate to each other, but our technological world is becoming more and more about how machines communicate with us – and us with them. 44’s Tom Ives looks at what brought us to this place, and the challenges that come from living in a sci-fi society…
We talk to machines all the time, whether it’s asking Siri where the nearest post office is or telling a customer service line why you’re calling before you’re put through to an actual human. This was once seen as the stuff of science fiction but is fast becoming a normal part of our daily lives. But how did we get here?
The first human speech recognition technology was invented in the 1950s, and in the 1970s the US Defense Department invested heavily to get it up to the standard of a toddler’s vocabulary.
By the 1980s a doll called Julie could recognise words and respond with phrases in truly creepy Pinocchio fashion. And in the 1990s speech recognition came to home computing and the first voice-controlled telephone menu system soon followed.
Gradually we improved upon these comically clunky tools, and in 2007 the iPhone was launched.
Soon after, Google busted out a voice search app to use on the iPhone. Google had a great advantage when it came to developing voice recognition – for years it had been indexing billions of questions (made through searches) and what the most likely answers were.
Apple’s own speech-recognising personal assistant Siri appeared in 2011 and could actually answer you back, search apps for calendar events and give you funny answers to philosophical questions like ‘What is the meaning of life?’ – the answer’s 42.
Coming in a number of different human voices, Siri was set up as a personality – a part of your phone, tablet, PC or TV that you could have a conversation with.
We should have seen this coming; sci-fi has been preparing us for decades. 2001: A Space Odyssey popularised the idea of the bodiless voice of an artificially intelligent super computer, ‘HAL 9000.’ There was something really creepy about the concept, however, as the film cast the AI as evil rather than benevolent, using its superior intelligence to prey on the weaknesses of humankind.
But Siri (so far) has been less sinister. And equivalent offerings from other companies soon followed: Microsoft’s Cortana in 2013 and Amazon’s smart speaker Echo in 2014 featuring speech recognition and a personal assistant named Alexa. Just last week, Apple entered the market with the HomePod, which answers via the Siri personal assistant and offers high fidelity sound that up until now, only Sonos wireless speakers were specialising in.
Two things gave Alexa a competitive edge:
– Firstly, it recognised the voice of almost everyone around, almost all the time. It’s so good at recognising its wake word ‘Alexa’ that it causes problems for owners – for example by responding when its name is mentioned on the radio or TV.
– Secondly, it is part of the Amazon ecosystem, meaning you can tell it to put toilet roll on the shopping list (or just order it) from the toilet. Pretty handy.
Alexa can access the combined intellect and data of every Amazon computer and customer. And it’s very popular in the home, where people don’t mind talking to themselves.
Like chat messaging, the Echo is another new tool that marketers are starting to treat like any other channel. As with any early product or technology, they’re also sometimes royally screwing it up…
Earlier this year Burger King thought they would try a disruptive new concept, where their TV ad would intentionally trigger the audiences’ Google Home products, by saying “OK Google, what is a WHOPPER®?”
This made the device respond by loading (and reading aloud) the WHOPPER®’s Wikipedia page. In preparation, the page was edited by the company’s marketing department to make it slightly more ‘sales-y’. And the internet responded as you might expect.
Trolls made dozens of edits to the description to claim it was toxic and included ingredients such as children, poison, toenail clippings and rat. Just three hours after the public heard the malicious edits, Google stopped Home responding to the advert. Wikipedia then locked the WHOPPER® entry for edits from the general public.
This turned out to be an epic fail by Burger King. The brazen plug also prompted an epic response from the public. And there’s since been a lot of talk about how easy it was to try to abuse the technology for marketing purposes and the security implications this raises for the future.
These new technologies open up huge new potential, and dangers, for how we interact with each other. But what we can tell is that tech seems to be flying well ahead of rules and regulations, so we’ll have to wait and see what happens next!