[ad_1]
getty
Using a smart home assistant like the Amazon Echo or Google Home can quickly get repetitive – one tends to ask the same things over and over again: maybe some general knowledge questions, playing a particular song or album, turning the lights on and off, make the shopping list, set alarms, etc. In short, we are probably only using a small amount of the device’s capabilities and we rarely take the time to discover its many possibilities.
Some features, like answering questions or playing music, are probably part of the original features that led us to install the device in the first place. Other functions come from additional devices: if you buy a camera, irrigation system, TV remote control or light bulbs, you will look for the tag “works with Alexa”, and as soon as you install the device in question, you will incorporate commands into your routine. For manufacturers of many devices, the idea of making them operable from a home assistant of this type is an increasingly important attribute for certain market segments.
But what about the additional possibilities? Usually, the algorithms that allow these assistants to do their work are installed by the company that manufactures them (Amazon, Google, Apple, Baidu, etc.) and they are one-way. Some users create their own commands and instructions, known as routines: for example, when I go to bed, I want all the lights in the floor to turn off and the lights in the bedroom to turn on. Basically putting together a bunch of predefined commands, and that’s as sophisticated as possible.
That, however, could start to change: a note published by Amazon a few weeks ago says that its devices could soon be trained by its users to acquire new functionalities derived from two-way processes. Now we could “explain” to our device what a certain configuration consists of: if you tell it to put the lights in dinner mode, the device will ask you to define what you mean by dinner mode, and you would explain that you want it to dim all the lights in the room, except for the table lights, so you can see the food. The device would store those user-defined settings and use it later.
The idea of a “teachable Alexa” with algorithms with which we can interact or define is interesting and represents a further step in the customization of this type of device. Households that adopt smart assistants are seeing how, from very basic functions, they become more sophisticated as more devices are installed: I have had several Alexas at home for a long time, and I also have several Ring cameras, but I found out yesterday that since I have an Amazon Fire TV plugged into the TV, I could ask you to show me the camera at the front door on the screen before opening it, or cut to the camera in the garden. If Alexa and its entire growing ecosystem of surrounding devices can be aware of some of my requests or allow me to define them, their potential increases, as does the value proposition.
At the same time, it is interesting to see how our brain evolves when we regularly interact with artificial intelligence: some people argue that these types of devices make us more lazy, because we ask it to turn the lights on or off instead of getting up. to do so, when the reality is that it teaches you to formulate your requests, to think of ways to access the information, or even to write some commands in a simple way. In practice, we are rewiring our brains to prepare for a world in which these types of assistants or automation of tasks will undoubtedly be much more common. Despite its apparent simplicity, people initially tend to make mistakes, such as talking to these devices as if they were human, rather than simply using the wake command at the beginning of each sentence, rather than providing meaningless or confusing requests.
The idea of bidirectionality, that these algorithms can be trained by us, offers enormous potential and could be the basis for developing new uses that increase their value proposition, and one more step in our interaction with this type of device, than in the most of the cases it is simply request – response. If we start to include dialogues in phases, such as those necessary to explain what we mean by something so that the device learns it for the next time, we can add many new possibilities. Improving the algorithms of these devices can quickly create a valuable ecosystem. It’s just a matter of pushing the limits of your imagination …
