A good comparison is having the coffee pot ready to brew by the time you walk over to it in the morning. Now, this broader idea is called “ambient computing,” in that these devices can handle tasks or monitor without the need for you to make a request. The microphones that power Alexa are also used to listen for events that might be happening in your home. An Echo Dot or a fourth-gen Echo can be used to control connected devices like smart bulbs or even a garage door through another device. So there’s a great benefit for customers in that they’re getting that sort of continuous improvement,” says Reid.Īs we’ve seen in recent hardware launches and with software rollouts, Alexa can do more than tell you what the weather is, tell a story or crack a joke. “The way that Alexa is built with the service largely in the cloud, it allows us to roll out capabilities hourly if we wanted. Amazon also offers a privacy tab within the Alexa app for handling these controls. Your conversations or requests with Alexa are uploaded, and you can see those transcripts, but you can also remove them after they’ve been used for their request. Privacy is also central to the Alexa experience Reid and Koopmans note just that. And so it’s a really important part of when we launch a new locale,” explains Reid. “It’s not just accents or the words they use, but it’s how they actually will ask for things. And Reid notes that rolling out a language, such as Arabic earlier this year, takes time and starts with a lot of data collection. Since Amazon’s Alexa is cloud-based, you’ll need an internet connection.Īnother neat aspect going back to the start of the request is Alexa understanding how we as humans speak across different languages and locales. There can be hiccups every once in a while, but Alexa has gotten the hang of it, especially with these more routine tasks like timers or the weather. If you’re asking for a song from Spotify or Apple Music, it goes through the partners’ technologies to generate the proper result.Īnd this all happens pretty darn fast to get you your results reasonably quickly. If it’s a simple request, that can happen on the device, but more legwork is required when you ask for music or open a third-party skill. And Reid describes the following steps as a domino effect. It takes your voice and transcribes it to figure out what you’re asking. That’s partly thanks to an on-device processor, which works to speed up these requests. More recent Echo devices, such as the fourth-gen model, are faster than older Echos at picking up the wake word and getting a response back. That’s the first step, and it’s a key one that we’ve seen Amazon improve on. that speech processing goes to the cloud to get transcribed,” says Reid. So once the customer states their desire - you know, play CNN, set a timer, etc. “When you say the wake word, it hears it, lights up and streams to the cloud. One of the first questions I get from friends and family members alike is how these smart devices work, and Alexa can be a real head-scratcher for non-techies. Your CNN account Log in to your CNN account
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |