Why doesn’t Siri speak in a whisper?

It is customary to say nasty things about Siri. There are enough grounds for her criticism. Apple’s voice assistant (by the way, the first such assistant for smartphones in the world) still does not support the most valuable and useful functions of assistants that appeared on the market much later than it. From 2012 to 2017, Apple forgot about this direction (or maybe just scored). Siri expanded its scope and even changed for the better, but slowly and somehow without a twinkle. Amid the incredible breakthroughs of competitors and their achievements, it seemed that nothing was happening to Siri . To say that Google Assistant and Amazon Alexa are ahead of Siri in their development is to say nothing. But things are about to change and Siri will have better times.

Finally, Apple remembered the existence of Siri

Perhaps Apple is working on an all-new voice assistant that no one else can match. The best assistants today have a lot of flaws, and Siri 2.0 would be great. Moreover, in the field of artificial intelligence and machine learning , Apple is doing very well these days. But alas, even if Apple is working on this, no one outside of the narrow circle of those in the know knows about it. It is difficult to explain the lack of real breakthroughs in the last four years (Google and Amazon have gone far ahead, and their gap from Siri has been continuously increasing all these years).

Last week, April 1, the Siri development unit went public. And this is not a leak, not a rumor, or an invention of a couch analyst. This is an official document published by the US Patent and Trademark Office. April Fools’ draw is excluded. And this patent application is just about ways and methods of implementing functions in Siri that have long been used in competing voice assistants.

Why doesn’t Siri speak in a whisper?

Until now, despite its venerable age, Siri does not even try to take into account the peculiarities of the environment. Siri does not pay any attention to either the noise level or the volume with which Apple assistant is asked a question. The volume of the answer does not change. When asked in a whisper, Siri will respond as usual. Probably too loud. Now this will be over. Siri-enabled devices have both “hearing” organs (microphones) and processing power to adequately assess the characteristics of sound and the noise it generates. A patent application filed by five Apple employees describes how and how to intelligently respond to these and many other circumstances.

If Siri can teach this, then HomePod will also be able to speak quieter.

And these are not separate ways and methods that work on their own – they are combined into a system of interaction with the outside world. Formally, Apple is catching up with competitors that have gone far ahead. In fact, it seems that no one has yet dealt with these issues so thoroughly and deeply. Just catching up is boring, and it’s not customary at Apple.

Alas, Apple’s patent applications do not oblige it to do anything. Applications are published for a variety of reasons. For example, when you need to mislead imitators and make them spend energy, time and money on something completely useless. Why hasn’t Apple implemented these features until now? Competitors’ implementation of these functions is protected by their patents. Coming up with something of your own, not only different, but also better than that of competitors is not an easy task. Judging by the published application, Apple engineers and scientists succeeded. It seems to be