Another point is that all the major providers of these assistants own a platform, and try their hardest to keep that platform for themselves. As such, it would be very difficult to see effective communication between them as they compete tirelessly for the top spot in your mind. This also keeps smaller companies from growing to accommodate more universal access (like how you have the same app on your Android phone in your pocket and iPad on the couch) because they don't have the data to access or mindshare to be accepted.
Hopefully this will change but as it stands, the ones with power to build these systems at level we expect want to control it all but can't.
General purpose robot assistants won’t truly be useful until they can work with each other, in all contexts.
highlights the real problem here: user expectations of chatbots are astronomical.
It's not that our expectations are astronomical, rather that the tools we have already are astronomically simpler and guaranteed to work (using google to get an answer/find restaurants/geolocate). Currently, there's no alternative to manually searching in google/yelp/maps in terms of getting a consistent reasonable answer.
Siri is inconsistent and almost guaranteed to misunderstand you.
here's no alternative to manually searching in google/yelp/maps in terms of getting a consistent reasonable answer.
that's exactly my point. the expectation is that chatbots should be equal to the full power of the biggest and best services. combined. in natural language.