At first glance, Voice Inbox may appear to be a better version of Siri in certain aspects: it has a human-level robustness and accuracy on speech recognition, and a similar level of accuracy in understanding the user’s intention when adding calendar events.

But being an voice assistant like Siri is never the goal or intention.

Siri functions as an agent that users interact with, where they speak to it and wait for its response. On the other hand, Voice Inbox does not provide any spoken responses, users don’t “talk” to anyone other than themselves while using it. Instead, Voice Inbox transcribes the user’s voice messages into text and saves them to the user’s journal, with the added capability of adding calendar events for the user if mentioned.

To illustrate the difference, let’s consider an example of a hospital visit1:

With Siri:

  • User: “Add an appointment with the doctor on next Friday 3pm”
  • Siri: “‘Doctor’ is scheduled for …”

With Voice Inbox:

  • User: “I have just finished my physiotherapy for today, this time I feel way better than before, the doctor said once I feel better I can visit less frequently, so I scheduled my next session on next Friday 3pm”
  • Voice Inbox will
    1. add the above to the user’s journal
    2. add an event to the user’s calendar

The shift here is that with Voice Inbox, users are not instructing anyone to do something (add the event), they are actually writing part of their daily journal, Voice Inbox works quietly in the background, requiring no user participation. This way, after a dump of thoughts, users can quickly return back to their tasks at hand.

In a word, if Siri is say-and-wait, Voice Inbox is say-and-go.

  1. The example comes from another post The Birth of Voice Inbox ↩