The Remaining Cocktail Party Problem
Despite years of voice-first products being on the market, the cocktail party problem remains.
A few years back, I was sitting in a meeting with two lawyers. One of the lawyers was trying to dictate into a digital recorder a letter while the other was trying to add commentary. The first lawyer burst out “ONLY ONE OF US CAN TALK AT A TIME!”
Yesterday’s experience with my Echo Show reminded me of this. I was trying to ask Alexa to play a song, but there was a gaggle of people speaking. It was frustrating to be rejected all the time by Alexa on trigger. I had to walk up close to the device to get it to work properly. Ugh.
Despite years of the Echo being on the market (and Google Home), it seems, at least in some form factors, the cocktail party problem remains elusive. Maybe it’s not as big an issue in the field as it is a technical challenge for bragging rights among ambitious engineers. However, in some circumstances, it does make for frustrating interactions.
