One of the least accurate counter arguments I heard from VCs for why voice wouldn’t be a huge market was that it was only a small incremental speed improvement to use a voice command vs flipping a light switch. However, that thought failed to integrate the incremental improvement in time over a wide range applications, like music playback.
Reducing the number of steps in any task, especially when shared, adds huge value. An example of this was my discovery of music in my late teens (meaning, when I formed a preference).
Before 1999, you’d go to a music store and try to listen to the songs that they’d allow you to sample (with headsets worn by thousands of people… lice paradise). CDs did change things, especially that libraries would loan them out. You could borrow, burn, return.
Discovering a song that wasn’t Top 40 (in any year or genre) was impossible.
Then there was Napster. Napster was just an eye-opening experience. Any song, including bizarre ones, could be found. I loved the opening to the Drew Carey Show “Five O’Clock World”. I had it in my library within five minutes of searching. Hipster Image’s “Make Her Mind” from a Levi’s commercial… done.
The RIAA and MPAA were extremely slow to catch up on why people went the Napster (and then LimeWire and then KaZaa… and then Torrent) route. People used the services because 1) they worked for search and 2) it was so much more convenient than anything else available.
The result was the cutting of many middlemen.
When it comes to voice, there is likely a market for a more private, more accurate, and more personalized assistant. Maybe one that involves a live concierge? If it supplies even a modicum of additional convenience, it will be worth it.