Rui CarmoI have been pondering the Audio Intelligence Live Rewind feature that Apple announced in the last keynote, and even though I know I am very late to this entire discussion, I still want to go on the record about it.
In short:
- The feature is stupid, and was demoed poorly (you can usually just ask people to repeat what they said, and the idea that anyone would prefer using the watch is bonkers)
- Rather than adding value to the watch, it makes it a target (companies and specific locations are likely to start banning its use)
- In the current AI/privacy climate, Apple will likely not launch it at all, regardless of what they stated (and, hopefully, not in the EU, either)
I would very much prefer to map double-tapping the crown to a favourite app rather than such a dubious “feature”, but the key thing for me, though, in a world that is now awash with always-on AI recording of everything (or, as we old folk used to call it, “surveillance”), is that Apple is so late to this that it’s ridiculous.
The Voice Memo Conundrum
However, the ambient listening privacy controversy is not really what I’m concerned with, although I am dead certain that however well Apple has tried to explain that they don’t really store any raw recordings, nobody in the real world will understand the difference.
What I’ve been annoyed about for years are voice memos.
The experience of using Voice Memos on the watch has always been terrible, because even if you add it to a complication like I have, you still have to tap to record and tap to stop, which adds a lot more friction than, say, the Pebble Index 01, and it doesn’t let you do anything with the audio.
It doesn’t even auto-rename the clips themselves to anything meaningful, which is par for the course in any transcription app. I have a dozen voice memos on my iPhone that are just named after my address, because Apple clearly assumes I am recording voice memos on a mobile phone, and that location is paramount–which is just a bad assumption.
In fact, as I discovered the other day while trying to build a simple Mac app that would take voice memos and auto-transcribe them better (because the default transcription model is simply useless), the entire voice memos “ecosystem” on Apple’s platforms is completely broken:
- Syncing either doesn’t work at all or takes forever (by the time a Watch recording syncs through from the iPhone to the Mac, it would have been simpler and faster to send a carrier pigeon)
- You can’t automate any of it–the best you can do is poll the recording index on your Mac and hack your way towards retrieving the audio.
What Apple should have done, years ago, is to just let me run shortcuts from the watch (remember, that worked in Workflow before they acquired the team and nerfed it into what is Shortcuts today).
That way, I could, in these modern times, build a shortcut that would take audio or a transcribed prompt from the watch and do my own thing, either transcription and summarisation (like the Index 01 does), or, who knows, actual automation of other things, like I’ve been complaining about for ages.
But at least my Series 9 can add To-Dos and take notes using Siri.
I do have to tap the display to confirm the notes (even if I can’t really access any of that from anywhere else, since they broke all the iCloud integration APIs), but hey, it was enough to ramble three short, garbled sentences and remind me to write this out–three weeks ago.
I finally found the note today, and decided to write it up before tomorrow’s event, where I expect them to parade a potpourri of iPads on a stick to try to make HomeKit relevant–years after Amazon and Google already shipped their own kitchen-top displays.
Gotta take the small wins, I guess.