Google’s demonstration in April 2018 of how its Assistant was able to phone a hairdresser and make an appointment without the person on the other end of the phone being at all aware that they were speaking to a machine showed how far voice – and AI – has come in a short time.

“The Google device has started to separate the consumer from the business they’re interacting with,” said Sam Hill, head of engineering at the BBC, and is opening up the prospect of bots talking to bots. “It’s a very interesting future exploding,” he suggested to the Digital Content Summit (London, May 2018).

It was particularly impressive, said Ben Bale, creative innovation director at DRUM, to see how the Assistant managed when calling a Chinese restaurant: “it dealt with accents, it dealt with the unexpected responses it got.”

But rather than the robotic assistants instructing people, most considerations of voice start at the other end, with people instructing the assistants - and they are increasingly comfortable doing so. More and more people are using digital assistants on their phones to carry out searches, for example. And the growing penetration of smart speakers, such as Google Home and Amazon Alexa, means that more people are engaging in potentially lengthier exchanges with devices – a development which may expose the current limitations of voice assistants which can be reduced to repeating “I don’t know the answer to that one” when faced with more complicated or follow-up questions.