With the first beta of watchOS 27.3, Apple introduces the two main Audio Intelligence features on the new Apple Watch Series 12 and Ultra 4. We’re talking about Live Rewind and Siri Recap, which thanks to the S11 chip let us revisit what we just heard and access summaries of the most important conversations. Let’s see how they work.
Revisiting a conversation and remembering its details
Live Rewind shows us a transcript of the last 15 seconds of a conversation whenever we press the Digital Crown twice. To prevent this feature from being activated covertly, the watch plays a sound and shows an animation when you activate it, alerting people nearby even if we have the Apple Watch on silent.
Siri Recap, meanwhile, creates summaries with a title and the key points of our conversations. We can schedule this feature by time and location, check the summaries in the Siri app, and save them for later. The summaries are automatically deleted after seven days, unless we decide to keep them.
Privacy and reliability, the two keys to their success
As we already saw when talking about Audio Intelligence’s privacy, the Apple Watch processes audio in a protected environment that no app, nor even the operating system itself, can access. It then uses Private Cloud Compute to generate the summaries. Both features require our explicit activation.
In both the summaries and the rewinds, the text doesn’t identify the participants, so it’s useful only for reference, not as proof of a conversation, since it’s impossible to know who said what.
For now both features are only available in English and join sound recognition and automatic song recognition via Shazam, which is also included in this version of the operating system. All in all, we have a very natural way to enjoy artificial intelligence from our wrist.
