There’s something about having an AI personal assistant that still seems to belong more to the realm of science fiction rather than reality. Controlling the various gadgets around your house with simple verbal commands. Surfing the Internet without having to type anything on the keyboard.
Having an assistant that never rests, never sleeps, remembers all your appointments and dates perfectly, and who listens to your comments and responds like a human being. No wonder every new such type of AI on the market is scrutinized with great interest by the general public.
However, voice recognition technology used in such types of AI is still not perfect. Every AI that carries voice recognition capabilities, from Google Home to Siri, Alexa etc, has different levels of efficiency at listening, interpreting and responding to human queries.
Let’s take a look at how well the various AI home devices fare on the basis of voice recognition capabilities.
At birth, Amazon Echo was extremely limited in terms of what orders it could follow. Its voice assistant, Alexa, couldn’t order food, play songs in the order you wanted, or even order things from Amazon. But thanks to the relentless efforts of developers, using Alexa skills kit and partnering with other major companies, Echo can today perform all those activities, in addition to many more.
However, Alexa doesn’t do translations. Another problem is that the technology hasn’t reached the stage where it can answer casual or indirect questions. People rarely use perfect English in everyday life, and this can make it difficult for Alexa to respond to a question or a command until the user become very clear about what is being asked.
Google’s personal assistant, known as Google Now, can be found on the Google search app or chrome web browser. Google Now is the best assistant at voice comprehension for when you’re carrying out basic searches. Integration, on the other hand, is not Google Now’s strong suit.
Telling Google Now to set up a reminder or send an email can result in erratic behavior, either searching out your phrase on its search engine or simply sending you to Google calendar.
Siri faced a lot of criticism in the beginning for being unable to follow simple voice commands and being unable to understand different accents. But the tech behind the app has come a long way since. Siri’s users today, number in the millions around the world. Such widespread usage has allowed Siri to adapt to languages spoken in different dialects in different regions.
In addition, Siri today shows an impressive ability to carry out a conversation with its user while sounding more like a real person than a machine. This is the area where Siri’s voice recognition system has the advantage over the other two AI home devices.
It’s been on the market longer than them, and in that time its compiled plenty of data from daily interactions with users to refine its response capabilities by taking into account modern phrases and slang. This allows Siri to respond to verbal commands more efficiently than either Alexa or Google Now.