Artificial intelligence and automated voices are progressing more rapidly now than ever. Virtual assistants such as Amazon's Alexa have introduced this technology into many of our homes in more recent years. But where did it all start, and what does the future look like for AI and automated voice innovations? The reality may surprise you.
A Brief History of AI and Automated Voices
Few would believe that the history of artificial intelligence dates back to antiquity. Yet that's absolutely the case. In Greek mythology, there's talk of intelligent robots and artificial beings. Meanwhile, mechanical statues built in ancient Egypt were believed to be capable of wisdom and emotion.
In a modern-day sense, contemporary artificial intelligence essentially has its roots in the Turing Test of the 1950s. The following decade, inventor Ray Solomonoff would then lay the foundations for a mathematical theory of AI.
However, strictly in terms of automated voices, innovation dates back as far as the 18th century. In 1770, synthetic speech was produced for the first time in mechanical form. What followed was a device tailored toward assisting those who couldn't speak. The same technology was also employed to study human voice function.
Despite first coming to fruition more than a quarter of a millennium ago, such devices were perfectly capable of producing clear speech. Though basic in their construction, they did so by replicating complicated human vocal structures. Even by today's standards, it was an impressive feat of engineering. Over the next 200 years or so, various other advances took place, culminating in the work of the likes of Alan Turing and Ray Solomonoff.
A particular highlight from this period is the work of Charles Babbage and Ada Lovelace to develop programmable calculating machines between 1822 and 1859. Another is Leonardo Torres Quevedo's building of a chess automaton in 1915. Shortly after, the Spanish engineer and mathematician would offer a groundbreaking insight into so-called automata theory.
The earliest electronic speech technology came around 150 years after the first mechanized solutions. Unveiled in 1930, the first machine ever built for electronic speech could perform without human input. A revelation within artificial intelligence and voice automation, it was the work of pioneering electronic and acoustic engineer Homer Dudley. Dudley would later prove central to the development of World War II technology designed to send secure voice transmissions.
Present-Day AI and Automated Voices
By and large, present-day AI and automated voices date back to the synthetic voice systems pioneered in the 1960s. Still, it wasn't until 1985 that artificial intelligence and, in particular, voice automation exploded into the mainstream. With that, and over 35 years later, Stephen Hawking's synthetic voice system remains among the most iconic pieces of AI and automated voice technology ever created.
Advances that had previously taken centuries to occur were now happening from one decade to the next. Hawking himself said at the time that success in creating artificial intelligence would prove to be the biggest event in human history. The theoretical physicist, cosmologist, and author also famously suggested that it may be the last. Thankfully, his fears would prove to be unfounded, at least in the course of his own lifetime.
Since then, improvements in AI and automated voices have only accelerated further. We're now in a position where such technology is employed in aspects of day-to-day life across the board. Once largely the preserve of science and medicine, many of us now have AI-powered personal assistants in our pockets.
A modern-day pioneer of AI and automated voices, multinational technology behemoth Apple first introduced the masses to intelligent voice automation with Siri on the iPhone in 2011. Three years later, Alexa from Amazon would take the same technology out of our pockets and set it down squarely in our living rooms, kitchens, and bedrooms.
Even though it's been hundreds of years in the making, 99 percent of the advances made so far within intelligently replicated human speech have been delivered in just the last few years. Siri, Alexa, and—subsequently—Google Assistant now grow in their capabilities with each passing year.
The leading AI and automated voice products can now mimic any individual's speech accurately after as little as 60 seconds of analysis. More recently, the same systems have developed to the point where they can also detect and replicate emotions, including anger, happiness, and sympathy.
We also recommend checking out our blog on Text to Speech (TTS) Technology and the Uncanny Valley Effect.
The Future of AI and Automated Voices
We're now at a point where advances in artificial intelligence and automated voice technology are measured in years rather than decades or centuries. Already, new developments are coming on a monthly and even weekly basis. In the future, it's not unreasonable to expect highly significant progress within AI and voice automation to become a daily occurrence.
That said, now is, perhaps, a good time to heed Stephen Hawking's warning. Just three years after his passing, a combination of AI and automated voice technology is now being used for increasingly questionable purposes. Central to the rise in so-called deepfakes, synthesized media is being created to mimic the likenesses of predominantly famous individuals.
A primary worry here is that deepfakes are perfectly positioned to manipulate the public using an amalgamation of machine learning, artificial intelligence, and voice automation. Yet, it's crucial not to allow the reputation of AI and automated voice technology, in particular, to become tarnished.
As it stands, the positives far outweigh the negatives. So, what can we expect from the future of AI and automated voice? While artificial intelligence is already successfully mimicking emotions, personality is another matter. Though given the rapid speed at which the field continues to progress, it's unlikely that this will be the case for too much longer.
On a practical level, also expect to see increasingly streamlined conversations, personalized experiences, more widespread integration, and a sustained focus on security. Much beyond this, only time will tell. What must be stressed, however, is that any risk of AI voices taking over completely from their human counterparts remains entirely unlikely.
With that, there should be less to fear and more to be excited about. This is especially true as we continue to charge headfirst toward a wealth of prospects made possible exclusively by ever-growing investment into AI and automated voices.