Imagine you ask your digital assistant or some conversational interface of your choice where you can get a „Berliner“ now. What will happen? How could a conversational interface distuingish between terms with multiple semantic meanings, semantically ambiguous terms or how could conversational interfaces deal with contextual – in the example below: regional language differences and/or natural talking behaviors of people?
For example, how could a conversational interface (like a voice interface) know, what a person who lives in Berlin means when she uses the term „Pfannkuchen“ which is in turn called „Berliner“ in other regions of Germany. Other examples may be „Tempo“ (= the brand vs the actual product which is a handkerchief ) , „Hamburger“ (=People live in Hamburg vs the food)
People have so much cognitive processing/work to do if they just want to achieve something quick. They have to talk and getting used to the language of the machine instead of just talking like they are used to talk. So as usual, the machines dictate how we should talk to them. I hear often the terms „human-like“ or „natural“ interaction but to be honest – instead I sometimes feel more and more like going back to a command line interface. :)