I think this analysis is kind of pointless. Who asks their phone trivia questions? You want directions, reminders, etc. Siri seems a lot smarter about following spoken context in that situation.
I can't speak extensively for Siri (infrequent use on my iPad) or Google Now (no Android), but I've never had to repeat an instruction to Cortana for a reminder or directions. Not repeating myself when giving voice instruction is essentially my holy grail.
I haven't used Siri in ages; what kind of context does it keep? I know that Google Voice Search will do coreference across consecutive queries (i.e. resolving "it" and "him" accurately).
Google Now will do some coreferences, but Siri is almost modal. You can talk to it like an assistant instead of trying to formulate everything as a search query. I had a Nexus 5 almost a year before getting my 6+, and I was always jealous of how much more practically useful Siri was on my wife's iPhone 5.
That is, indeed, fantastic. Does it hold context? I mean, can you say "Find a gas station." and then "Find a McDonald's near that" and have it work? That would be super.
Strongly agree. Isn't this testing the "knowledge box", voice recognition, and NLP at once? It seems like the bulk of this could have been done in a desktop web browser.
