Alexa is Implementing Self-Learning Techniques To Better Understand Users (theverge.com) 27
In a developer blog post published this week, Alexa AI director of applied science Ruhi Sarikaya detailed the advances in machine learning technologies that have allowed Alexa to better understand users through contextual clues. From a report: According to Sarikaya, these improvements have played a role in reducing user friction and making Alexa more conversational. Since this fall, Amazon has been working on self-learning techniques that teach Alexa to automatically recover from its own errors. The system has been in beta until now, and it launched in the US this week. It doesn't require any human annotation, and, according to Sarikaya, it uses customers' "implicit or explicit contextual signals to detect unsatisfactory interactions or failures of understanding."
The contextual signals range from customers' historical activity, preferences, and what Alexa skills they use to where the Alexa device is located in the home and what kind of Alexa device it is. For example, during the beta phase, Alexa learned to understand a customer's mistaken command of "Play 'Good for What'" and correct them by playing Drake's song "Nice for What."
The contextual signals range from customers' historical activity, preferences, and what Alexa skills they use to where the Alexa device is located in the home and what kind of Alexa device it is. For example, during the beta phase, Alexa learned to understand a customer's mistaken command of "Play 'Good for What'" and correct them by playing Drake's song "Nice for What."
so it learned to reboot, cache queries.. (Score:1)
and 'google'.
wow.. that's some 'machine learning'.
Re: (Score:1)
If Alexa can learn like that (Score:3)
tossing in the option of a camera to read sign language for the deaf would be interesting.
Finally, Al for real, not the fake stuff (Score:1)
If this is not artificial life I don't know what is.
Finally, people are starting to realize... (Score:1)
That products like these are not to help people, that is an (un)intended consequence, but
a paid-for-by-the-consumer experiment for data collection and group analysis for control.
This is not a tin-foil hat theory. Think about it, they known everything about you in your home --
SPL, temperature, ambient light, how often you (and your children) have visitors), and
background noise (yes, that background noise as well) to name a few.
Do they operate as a smoke detector as well? Thought so...
Do you really want to
All the better to eat you with (Score:3)
Come closer my dear... so I can better understand you.
First mover advantage (Score:2)
Google Home AI is better by a mile: better speech recognition, language understanding (you mostly can just say things without adhering to a rigid syntax), and speech synthesis, better music service and question answering, too. Yet people treat Alexa as though it’s extraordinary in some way other than being the first on the market.
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
I've tried both. On the low end - Home Mini vs Dot - the Echo device wins by a mile on hearing its wake word in a noisy room while music is playing and still picking up what you want to say. Plus they have audio out to better speakers. Even still, it's useless compared to a Google Home product on what it can actually do.
4 year old using Alexa (Score:2)
To repeat my old mantra. (Score:2)
Thank God! (Score:2)
For example, during the beta phase, Alexa learned to understand a customer's mistaken command of "Play 'Good for What'" and correct them by playing Drake's song "Nice for What."
Thank God! We were all waiting on tenterhooks for that problem to be solved!
Re: (Score:2)
https://www.youtube.com/watch?... [youtube.com]
Better advertise to (Score:2)
learning (Score:1)
Sign it still has a long way to go.... (Score:2)
Alexa's issue (Score:1)