Wake The Town
Stuck in Motion
Contemplative Reptile
  • International Edition
  • USA Edition
  • UK Edition
  • Australian Edition
  • Canadian Edition
  • Irish Edition
  • German Edition
  • French Edition
  • Singapore Edition
  • Spanish edition
  • Polish edition
  • Indian Edition
  • Middle East edition
  • South African Edition

OK Google, What’s the Big Deal with Voice and AI?


The brains behind J. Walter Thompson and Mindshare’s Speak Easy research into artificial intelligence chat to LBB’s Laura Swinton

OK Google, What’s the Big Deal with Voice and AI?
Judging by the hype at this year’s CES in Las Vegas, 2017 is all about the marriage of voice technology and artificial intelligence. It’s a situation that J. Walter Thompson London’s Elizabeth Cherian and MindShare’s Jeremy Pounder suspected would develop when, back in December, they decided to embark on some in depth research into the adoption of voice tech, and its potential implication for people and brands.

The initial findings of their Speak Easy research project were published in April – a qualitative and quantitative dive into UK attitudes to voice and AI, with an enlightening slice of neuroscientific insight. In just over a week’s time, they will be revealing their global findings that will drill into the cultural differences and universal similarities across the world.

According to Elizabeth, the research project was driven by curiosity rather than specific client requests (though following the report’s publication many intrigued but wary clients came forward). She and Jeremy shared the feeling that we might be on the verge of a technological and societal tipping point. What’s more, brands like Amazon have been making the technology accessible and affordable.

“It’s been around for ages but the big change is advances in natural voice processing. AI is getting tonnes of investment and as a result it’s developing very quickly. When you interact with these devices now, they’re actually understanding you,” she says. “There’s voice recognition but the real complicated stuff is the natural language processing in the background; understanding intent, getting context right. It’s not there yet but we’ve reached a tipping point.”

That hunch proved to be spot on. The day after the Speak Easy report published, Google Home was launched in the UK. 

And it seems the speed of technological development and the rush of products coming to market is mirrored by user adoption. One of the most fascinating findings was that it doesn’t take long for people to abandon their reticence and awkwardness. Participants were wired up with EEG caps (which measure brain activity) and were rigged up to measure various other bodily responses at Neuro-Insight, which helped with this section of the research.

“A lot of people hadn’t used Echo or Dot before and they had to take 20 minutes to do the tasks,” explains Jeremy. “You could see from the data, their first interaction was characterised by withdrawal, but as they had four or five interactions, their warmth grew steadily with each one. So as people get used to it, you could see how they would take it a bit further.”

Elizabeth explains that, because we instinctively anthropomorphise inanimate objects anyway, particularly those that look human or talk back to us, warmth is understandable. And it seems that the ‘warmth’ has the potential to heat up considerably…

“We thought: ‘what would people say if we asked if they’d ever had a sexual fantasy about a virtual system?’ Will anyone say yes… And 26% came back and said ‘Yes I have’,” laughs Elizabeth. “We were all like, ‘are you serious?’ We’ve asked a lot of countries about this and the UK isn’t even the most perverted!”

So, it seems like adoption may not be as big an issue as you might think when it comes to voice tech. But while that’s great for the technology giants and the brands on the front foot, there is an unexpectedly negative consequence of this ease of use.

“In the qualitative work we did, people expect it to be better than it is,” reflects Jeremy. “That’s where they get frustrated. They say, ‘well it knows I’ve bought all this stuff, why is it not now putting that all together and starting to suggest something else?’. As people get comfortable with it they become quite demanding.”

The technology is developing rapidly and there’s an appetite for AI-driven voice tech… but what are the implications for brands?

One of the big predicted trends is that voice will be ‘liberated’ from the screen. A world where we can access information simply by asking – and do our shopping and keep up with the news and select our playlists – is going to have a huge impact on how advertisers think about their brand.

In terms of search or shopping, being top of mind is going to be more important than ever. For example, with the weekly food shop, If there are no shelves or screens to browse, people are either going to go for the brand that immediately comes to mind or may even find it easier to go for the retailers’ own brands.

And what’s more, marketers, strategists and creatives are going to have to think a lot more carefully about the aural aspects of their brand and communication – a big shift in an industry which has traditionally been more visually-orientated. 

“One of the trends in the report is ‘find your voice’. Brands are going to have to think much more about how they sound. We think that’s particularly the case for brands that you have a relationship with – your bank, your car – you want to be able to talk directly to them and you don’t want to have to go via Alexa. Which is how it works at the moment – you set up a skill through Alexa,” says Jeremy. “Brands are used to talking about tone of voice, but it’s been more about the type of language they use, rather than literally ‘is it a male or female voice?’ or ‘is it a regional accent?’ ‘Is it jolly and funny?’ All these sorts of things.”

And filtering down to production, that could make for lots of new opportunities for voice actors and the like, says Elizabeth. “It’s thinking about what will consumers want to know about us and how will that sound experience be. It doesn’t have to be just someone’s voice, it could be music, all sorts of things. Sound designers are going to have an uplift with sound coming through.”

Of course, regardless of the speed of technological evolution and adoption, it is possible that this screen-liberated future doesn’t manifest quite as we imagine it. For example, some critics, such as tech analyst Benedict Evans at venture capital firm Andreesen Horowitz, suggest that voice may not be the best user interface for all tasks. He also suggests that the fact that a ‘general AI’ (or AI that could answer any question, without specific AIs or ‘skills’ and which functions like a human intelligence) is decades away – but most people when they think of AI imagine this general AI, therefore the experience is ultimately dissatisfying.

But despite these words of caution, there is value in at least beginning to investigate the potential of AI. However, some brands are hesitant. “It’s a classic innovation problem,” says Jeremy. “Do I build something now or do I sit back and let someone else make the mistakes and do my version later?”

But Elizabeth says that it’s a ‘worthwhile experiment, no matter what’. She details a conversation with one man that she had in the process of doing her research, who worked for Radio Player, who had developed a ‘skill’ (voice experience) for Alexa. And the experience had been surprisingly illuminating. He’d had to take everything his business did and distil it down to what was most important to the user, to really think about what sort of things users might need or ask for.

“It’s figuring it out how to be useful, and it really made him think about what he does in a way that he had never done before. It was like a paradigm shift for him,” says Elizabeth. “I’m sure brands are always trying to put themselves in consumers shoes, but by trying to figure out what the conversation would look like, it took it to the next level.”

That curiosity is bound to be reflected at this year’s Cannes Lions. Is the technology ‘there’ yet? Not 100%. I used an AI-driven transcription platform to transcribe my interview (Trint) and, err, I’m sorry to report to my fellow journos that I did need to resort to the ‘old fashioned’ method of typing everything out myself. Liberation from the torture of interview transcription is still some way off. But there are so many innovative and exciting uses of voice-driven AI experiences out there already. IBM Watson’s magical AI art gallery guide or Audioburst, which searches, curates and distils online radio and audio content and delivers them in digestible chunks. Ford is pushing AI research for its cars.

Hotels are figuring out how voice AI can help guests with everything from check out to turning on the lights in an unfamiliar room.Ultimately, you need to try it out for yourself to figure out what the applications and implications for your brand or business might be.
view more - Trends and Insight
Sign up to our newsletters and stay up to date with the best work and breaking ad news from around the world.

Genres: People

Categories: Consumer Electronics, Computers

LBB Editorial, Thu, 08 Jun 2017 16:26:28 GMT