- Meta is apparently working on glasses with always-on AI
- It can recognize people’s faces and remind you of things you forget
- Existing glasses apparently have all the same sensors, but too small a battery for the feature to be worthwhile
Meta’s next smart glasses might always be watching with a new AI that can track everything you do.
That’s according to a leak from The Information (behind a paywall), which details two pairs of specs the company is working on, according to insiders.
Internally codenamed Aperol and Bellini, the two pairs of specs would feature a new live AI. You start it with a command phrase, like “Hey Meta, start live AI,” and the glasses would start recording your every move.
This includes being able to recognize people with facial recognition, and to remind you of things like your house keys if the glasses saw you forgot them.
The current Ray-Ban Meta smart glasses models are apparently already packed with the sensors they’d need to pull off this same feat of AI.
Unfortunately, those familiar with Meta’s tests with the existing glasses have said the feature cuts their battery life to only 30 minutes, which isn’t ideal.
Meta’s current focus is reportedly to have its in-development glasses – and a pair of Meta earbuds that will supposedly also boast cameras and super sensing software – run for hours on a single charge with this feature switched on.
Always watching
In many ways, I can see the advantage of this always-live AI assistant. Anyone who has locked themselves out wishes they’d had a little software assistant reminding them to pick up their keys before they leave home.
At the same time, it sounds like a privacy nightmare.
People who point somewhat nervously at my Meta Ray-Ban glasses and ask, “Are those recording us right now?” always seem to relax when I show them they’re switched off or explain they only record when instructed, like if I say, “Hey Meta, take a picture.”
Now, imagine I instead say, “Yeah, they’ve been recording you the whole time.” I imagine I’d lose a few teeth.
The Information’s report also suggests Meta has reworked its privacy and safety policy to speed up innovation, which, in conjunction with the huge amounts of data this live AI will collect, isn’t something that fills me with joy. Personally, this seems like the time to tighten your privacy policies to ensure sensitive data isn’t shared in a way it shouldn’t be.
I expect this AI would come with some of Meta’s existing safety features, such as the light on the glasses being on while you’re recording, and it is at least a feature you have to manually activate, but (call me old fashioned) I can’t imagine ever wanting a camera to record my every move.
Others will probably feel similarly about these potential new always-on AI glasses. Just look at the backlash against Microsoft’s always-on AI feature, Recall – and that was an app that just took computer screenshots.
As with all rumors, we’ll have to wait and see what Meta announces officially, but I’m hoping the final product isn’t anything as scary as it sounds right now.