We’re learning to talk to our machines – and they’re speaking back, thanks to technology that is giving our gadgets the human touch.
Amazon sold 11m of its voice-activated Echo devices in 2016, and it’s thought that number doubled last year as consumers got to grips with the idea and Alexa’s abilities grew.
It’s hugely exciting, the technology is getting better… and they will become more and more common.Polly Purves
Rival Google Home is said to have sold in the region of 6m in 2017, while the search engine’s voice assistant features on scores of devices, from phones to speakers.
Microsoft has Cortana, and most recently Apple threw Siri into the mix when it launched Homepod in the UK. Its Siri enabled speaker reacts to voice commands to search for information online, find and play music and set alarms.
Giving machines a voice that is rich in personality, emotion and humanistic character is something a team of around a dozen engineers are working on at Edinburgh-based CereProc, the world's leading developers of speech synthesis systems.
One of their best known examples is Hanson Robotics’ Sophia, the world’s first robot to gain citizenship. She also hit headlines when, during an exhibition of her abilities, she declared “Okay, I will destroy humans”, a statement all the more disturbing given that it was delivered in a youthful, friendly American female voice and not a comic book Dalek.
According to CereProc’s Chief Scientific Officer Dr Matthew Aylett, giving machines believable human voices is a key element in how users will interact with them.
“Speech is a very special type of communication. Speech is part of what we are as human beings, we don’t just communicate information, we communicate who we are, our background, emotion, it’s all wrapped up with our identity,” he says.
“At the moment a lot of voice recognition is clunky, it’s slow, the voices are very neutral and don’t give you that sense of presence.
“Sometimes we get a very posh English accent, usually because these voices are built by companies which aren’t British and when they think of an English accent they immediately think of someone that sounds like Hugh Grant. But of course we don’t all talk like Hugh Grant.”
To prove the point, CereProc has a nifty feature on its website showing its range of British synthesised accents, including a menacing Glaswegian called Dodo and a Black Country girl called Sue.
“When you start to move into more complex applications, it’s important to convey emotion, whether it’s home devices, personal assistants or in games,” adds Dr Aylett.
As well as creating artificial voices for machines, CereProc works on cloning voices – fun if you fancy putting words in President George Bush’s mouth but also potentially life changing for someone who has lost the use of their voice through illness.
According to Dr Aylett, Scotland is an ideal location for voice recognition and synthesis technology. “There’s a wonderful skillset, Edinburgh is the best place for us to be located because of the quality of engineers and access to cosmopolitan language groups.”
Major players, from Starbucks to Ryanair and insurance giant Aviva to most banks, have already launched apps to work with voice enabled devices allowing customers to order their coffee before they leave home, or book their flight through their Amazon Echo or Google Home device.
Voice recognition is creeping into how we use our smartphones too - in recent weeks Santander became the first bank to launch voice activated payments in the UK. Its phone app also enables its customers to check transactions and report a lost card by talking to their phone.
Other banks use voice recognition technology which eliminates the need for customers’ to remember countless passwords and security questions, and instead access services using their voice.
And while that is likely to make it easier for us to get what we want, it will also have massive implications for the businesses that we interact with, while those specialist tech companies at the cutting edge of voice recognition and synthesis will gain ever greater status.
According to Polly Purves, Chief Executive of ScotlandIS, the trade body for the digital technologies industries, voice technology is about to make its presence felt in many areas of our lives.
“We are just at the start of this technology – everyone will get into this in some way or another,” she says.
“Where we are just now is a bit like the move to mobile apps a few years ago. Like apps back then, now everyone whether they are banks, retailers or other businesses, will be looking to provide this.”
While voice recognition technology enables us to interact with machines, eventually the computer at the other end will not just answer back far more intelligently than ever before but will no longer have that robotic tone that reminds us we’re dealing with a machine.
Glasgow has internationally acclaimed Multimodal Interaction Group, and St. Andrews and Dundee universities also offer degree courses in Human Computer Interaction.
The pool of talent is already attracting a new generation of tech companies specialising in machine voice technology. Last September Dublin-based specialist Voysis opened its first UK office in Edinburgh. Within months it had launched its AI voice assistant which it aims to license to retailers which can then integrate it into their own app or mobile website enabling users to shop using their voices.
It means stores can bypass the might of Amazon, Google and Apple’s voice assistants to interact directly with customers using voice technology – enabling them to retain vital data that can be used to improve their customers’ experience and help mould future elements of the business.
Ms Purves agrees that there is potential for Scotland to lead the way.
“There’s no doubt that within the Scottish community there are lots of capable people who moved to produce mobile apps very quickly and who will see the opportunities this presents.
“We saw a gold rush when mobile apps were being developed and that could happen again.
“Scotland is well placed and definitely not lacking expertise. It’s hugely exciting, the technology is getting better… and they will become more and more common.”
The 3rd Digital Transformation conference was held on February 27 at Our Dynamic Earth.
To find out more about interesting technology developments and companies in Scotland visit SDI at https://www.sdi.co.uk/invest/sectors/technology