Artificial Intelligence and voice assistants

Comments (0) Technology, Trends, TRIBE Social Magazine

smart_home
Pinterest

Thanks to progress in Artificial Intelligence, in the last years the development of voice assistants made great strides: in fact, carrying out researches throughout the smartphone just by pronouncing: “Hey Siri” or “Ok Google” is by now a diffused habit, but this is only the beginning. The next future of this technology will orientate more and more towards a use that results natural and totally integrated to people’s daily life, starting from the house where they live.

Connection of physical and virtual objects via communication tools (the so-called “Internet of Things”) sees one of its main application in domotics – or Home automation –, which makes the direction where some of the biggest companies are already stepping into. Amazon, Apple and Google, all along the most forefront bearer over the time.

Kitchen at the HGTV Smart Home 2013 located in Jacksonville, FL

Kitchen at the HGTV Smart Home 2013 located in Jacksonville, FL

We are talking about home automation as one of the greater technology trends of 2017: so at which point of the evolution are we?

The point of development of Vocal Assistants

Although the developments in the range of vocal recognition are easily visible, before this technology will become part of anyone’s common daily life, there are still some improvements to make, especially in two directions:

  • The total understanding of all natural conversation, which means that the voice assistant should be able to understand any real conversation in any language or dialect, spoken with any accent. To reach this goal and make it happen that a given voice command is always recognized by the assistant, the virtual assistants have to learn at doing it: how? What basically these companies are missing are the data: in this case, recordings of real human conversations to provide to their computer so that they can analyse and memorize them in order to then interact.
  • The understanding of a conversation even with a background noise. This means, allowing the voice assistants to develop the skill of splitting a single human voice from an eventual chatter and from external interferences with the ability to figure out the command.

Voice assistants around the house are real by now

As already stated, some big ever-present companies are investing a lot in technologies related to development of voice breaks with increasing precision and functionalities. Amazon and Google, especially, have already launched their devices that arise the goal to make the smart home an accessible reality.

amazonECHO

Echo, by Amazon, is already present in the market from 2014 (not yet in the Italian one) with its voice assistant Alexa. Alexa has been really appreciated by users for its skills in carrying out simple tasks like answering questions, order delivery food or check the state of a flight. However, some limits are clear at the same time: above all, the misunderstanding of the vocal commands recognition occurs when these are not well specified talking slowly.

If Echo is configured as a simple Bluetooth speaker able to solve efficiently some tasks assigned by the user, the recently born Google device, Home, goes further. First, it is equipped with the strongest research engine ever. This allows ‘Home’ to achieve better results in the researches; moreover, it’s able to contextualize a conversation: if in a first research something was mentioned, in the next one it’s possible to refer to that same thing just using a pronoun, which was not possible with Echo. Google Home is surely better even for what concerns the audio functionalities. Throughout Chromecast it’s indeed possible to connect more devices to obtain a stereo reproduction or to benefit from one’s own music in multi-room mode, that is spreading the same track in all the rooms of the house simultaneously.

google-home-orange

Pros for Echo are (for the moment) the compatibility with a higher number of devices produced by other studios committed to achieve the home automation and the simplicity of its use: all the skills are triggered by simple commands talking to Alexa, without any need of other apps or password, something that unpleasantly occurs with Google Home.

Finally, having to do with smart home and voice assistants means that it must be mentioned Jarvis, Facebook creation (to all extents, since it’s installed at Zuckerberg’s). Presented by the CEO of the most famous social network as a personal challenge to undertake by the 2016, Jarvis is the prototype of voice assistant able to elaborate the natural language, facial recognition and backup learning. This last characteristic permits the virtual assistant to learn habits and tastes of the user, which is something that distinguishes this software from Google Home and Echoes.

Despite the progress, the same Zuckerberg states that there are still many step forward to go before controlling the whole house through Jarvis or any voice assistant. It’s necessary to teach to the system how to learn efficiently and autonomously, more than reaching a status in which all the devices are lined up on the same standard, such a way that they can connect and interact with each other.

Comments

comments

Comments are closed.