From AUDREY to Alexa

A brief history of voice technology

“Siri, what’s the weather forecast for today?”. “Alexa, add cereals to the shopping list”. “Hey Google, tell me a joke”.

For many people using voice activated technology is part of their every day lives. But how did we get here? Here’s Swotly’s brief history of voice technology.

AUDREY leads the way

Back in 1952, Bell Laboratories designed the AUDREY system. An automatic digit recognition machine, AUDREY is the first documented speech recogniser. The machine could recognise strings of digits from zero to nine.

Work to increase the number of words that were recognised continued around the world. In 1962, IBM launched the Shoebox machine at the World Fair. Shoebox understood 16 English words.

From words to sentences

In the 1970s, the aim was to move from machines understanding single words to complete sentences. Harpy – a collaboration between businesses including IBM and academia in the US – was born. Harpy could recognise 1 011 words. That’s similar to the vocabulary of an average 3-year-old.

Not only did Harpy have a much larger vocabulary, but the language model used reduced errors. The model knew what words made sense together, resulting in more accurate speech recognition.

Throughout the 1980s, the vocabularies across all types of voice technologies rose from a few hundred words to several thousand words.

Voice technology reaches the public … but at a cost

In 1990, Dragon Systems released Dragon Dictate, the first consumer speech recognition product. Despite costing a whopping $9 000, like all the products available before it, the user still had to pause in between each spoken word!

It wasn’t until 1997 when Dragon Naturally Speaking was launched that there was a continuous speech recognition product.

Voice technology for the masses

After a quieter decade or so, voice activated technology as we now know it burst onto the scene in the late 2000s. The launch of the Google Voice Search app reached millions of people. Impressive accuracy and predictions were possible as Google collected data from searches.

Apple launched Siri in 2011. Amazon’s Alexa and Google Home followed over the next few years.

Now in the UK 19.7m adults own a smart speaker whilst 20% of UK consumers feel that their smartphones do the job just as well.

So what’s next?

Now’s the time to enable users to ask for more than a joke or the weather forecast. Follow us here at Swotly to see how we’re using voice activated technology to improve the experience of parents with children in education.