Sounds dystopian to me, I'd want to reconcile it by not allowing "one-party consent" for people to record me.
Not sure if the state laws you're referencing are in reality limited to phone calls, but I strongly dislike unregulated public camera use.
Your vision (no pun intended) is the story of the Black Mirror episode "The entire history of you", IMO from the show's golden age.
edit; I know that surveillance cameras pass this line already, but here they have to be announced with signs. And even when they aren't, to me state or police surveillance is different from potentially everyone stealthily recording me in private or public spaces.
It's possible the state laws in question (Tennessee) only apply to audio recordings, which would suit my desire. I also don't believe that the idea of a rolling buffer that normally discards its contents to be morally against the idea of notification of recording, or of seeking someone's consent.
I'd be fine with glasses that only record audio in such a way, that illuminate an LED once the "record" button has been pressed. If audio is being recorded into a buffer at all times, but then discarded unless triggered to start "recording", then maybe that should not count as "recording" under the law.
As a practical matter, if one is in a situation where such recording is warranted, by the time you press the record button, you've already missed important information that's relevant to the context of the recording. Allowing a 60-second rolling buffer that then gets dumped to storage when "actual" recording starts should be allowed.
Cameras everywhere just keeps everyone honest, right? Nothing to hide, nothing to fear, right? What's acceptable now will always be acceptable in the future, right? My mind never changes, whose does?
The point of this idea is that it would be under control of the individual wearing the glasses. I would most definitely not want it to be syncing to the cloud or some stupid shit like that. The buffer, and the storage, would need to be entirely contained within the glasses (or other device, if it turns out audio is a legally safer way to implement something like this).
As I mentioned in a sibling comment, I'm not against a visual notification of such recording once the "start saving to storage" button has been pressed. At the same time, I realize that the 60 seconds or so leading up to pressing that button is also often vital (otherwise dashcams wouldn't use a rolling buffer). And in such a situation where audio (or video, in applicable jurisdictions) is being recorded only in volatile memory and overwritten when the buffer is exhausted, I don't think a recording notification should be necessary unless the user has actively engaged non-volatile recording. In that sense, it's similar to the difference between streaming and downloading media. Both are technically the same, but the intention of "streaming" is to download the media and decode it without storing it in a non-volatile fashion.
I think you're thinking about this a bit naively, concentrating on the utility without considering the detriments.
Look at social media. WE are the ones who surveil ourselves. Yes, the big social media companies process all that data and use it against us, but we are the ones who give the pictures, videos, and words to them. There's really no good way around this either. I put those same things on my blog and they still get scrapped.
So what ends up being the difference? It's not synced to the cloud, but we put it there anyways. Do you really think most people are just going to take the videos and not share them? Do you think most people are just going to run a NAS at home? In an ideal world, yes. But I don't think we're anywhere near that happening. So a good portion of those videos just get put online somewhere and bad actors have access.
Non-volatile recording doesn't really exist. We're on HN and I'd expect most people here to be familiar with how easy it is to download a streamed video. yt-dlp will do that for a lot more than youtube.