What it is: Amazon’s Alexa devices apparently randomly laugh without any audio trigger.
Owners of Amazon’s Alexa devices have reported that the device seems to laugh for no apparent reason. Amazon said the problem comes from Alexa mistaking outside sounds for the command, “Alexa, laugh.” To fix this problem, Amazon reprogrammed Alexa to laugh only when prompted by the command, “Alexa, can you laugh?” Then Alexa will respond, “Sure, I can laugh,” followed by the actual laughter itself.
Why this command to laugh even exists is the real question. Recognizing voice commands is fine, but recognizing a trivial command to laugh seems rather pointless. Are people really going to talk to Alexa and ask it to laugh? This might be amusing initially but chances are good it’s not a command most people will ever use again, if they even use it in the first place. Voice commands should be used to focus on common questions first and trivial tasks much later, if ever.
More importantly, Alexa needs to learn to recognize different languages to broaden its appeal to more people. What might be more useful for voice assistants is language training. Part of learning to speak another language is the fear of looking stupid when interacting with native speakers. With voice assistants, you could learn French, Spanish, Chinese, Russian, or any other language safely by practicing with a voice assistant until you get the pronunciation correct. Then the voice assistant could respond and you could gradually build up your confidence in a safe environment.
Right now, Alexa supports English and German. Yet many people in America speak Spanish, and people in other parts of the world don’t speak English or German, essentially blocking Alexa from those markets altogether. Voice assistants need to learn other languages and become better at responding to ordinary language. Voice assistants don’t need trivial responses to make it laugh.
Voice assistants are the future. At one time, voice recognition required that individuals train the computer to recognize their own voice by repeating several phrases. Nowadays, voice assistants can recognize anyone’s voice without any training whatsoever. The next step for voice recognition is to respond intelligently, and that means ignoring trivial responses like laughing when given a command to do so.
Amazon needs to focus on supporting more languages and making Alexa respond more realistically to all types of voice commands. The last thing Alexa needs is to respond to silly questions that aren’t useful one bit.