Meta CTO explains why the smart glasses demos failed at Meta Connect

timpera 17 hours ago | parent | next [-]

I absolutely loved the failed demos, it felt much more genuine than Apple's conference.

laweijfmvo 17 hours ago | parent | prev | next [-]

Here’s a better response for next time: “I’m embarrassed by our recent failures to demonstrate our latest products. Our customers absolutely deserve better and I’ll be personally making certain that all of these issues are fully resolved before any products ship to paying customers.”

Is that really so hard? If I can write this on the toilet why can’t one of the top tech companies and their massive PR budgets do it?

	▲	freetinker 16 hours ago \| parent [-]
		PR budgets exist to _prevent_ exactly this sort of response.

▲

pbrumm 8 hours ago | parent | prev | next [-]

I wonder how many audio conversations were overheard and uploaded during that moment.

Seems like an attack vector for forcing your devices to start recording the mic and transmitting it.

Even if it just takes down the wifi through maxing out internet connection or cellular network. Play a couple seconds of audio and gigabytes get uploaded.

▲

csande17 15 hours ago | parent | prev | next [-]

> When the chef said, ‘Hey, Meta, start Live AI,’ it started every single Ray-Ban Meta’s Live AI in the building.

Wait, isn't that even worse? The problems with the chef demo looked like run-of-the-mill AI hallucinations (and I'm not really convinced they weren't, unless Meta has a fleet of dumber models and a feature that automatically switches to them during high-load periods without informing the user). But actually, anyone within earshot of someone wearing Meta glasses can yell commands that the AI automatically accepts? There's no wakeword personalization, or microphone beamforming, or anything?

▲

OhMeadhbh 15 hours ago | parent | next [-]

I came here to make that same comment. I'm not really a fan of face mounted cameras, but even less of a fan of privacy eroding, untested, not fully thought-out face mounted cameras. And the sad thing is they have (or maybe had) some pretty decent product, research and engineering people over at reality labs. This thing really shouldn't have been this bad. Seems like a management screw-up.

	▲	fakedang 14 hours ago \| parent [-]
		> but even less of a fan of privacy eroding, untested On par for Facebook/Meta though. Move fast, break shit.

▲

beefnugs 13 hours ago | parent | prev [-]

someone will come up with a sweet subaudible way to trigger every pair of glasses in the wild. automatically dox yourself, release all your private photos, docs, everything the ai has access to

▲

raffraffraff 15 hours ago | parent | prev | next [-]

> When the chef said, ‘Hey, Meta, start Live AI,’ it started every single Ray-Ban Meta’s Live AI in the building. And there were a lot of people in that building,”

You're kidding me. This is comedy fail. They should have put out some rakes to step on.

	▲	karmakaze 7 hours ago \| parent [-]
		And the follow-up, creation of a single point of failure. > That alone wasn’t enough to cause the disruption, though. The second part of the failure had to do with how Meta had chosen to route the Live AI traffic to its development server to isolate it during the demo. But when it did so, it did this for everyone in the building on the access points, which included all the headsets.

▲

sixtram 17 hours ago | parent | prev [-]

For me, it seems the issue with the recipe was due to pre-recorded segments. The guy quickly realized that the "AI" was ahead of the script and cut it short to avoid making a bigger mess during the demo.

As for the Zuck segment, I believe it was a race condition.

Here is the video:

https://www.reddit.com/r/interestingasfuck/comments/1nkbqyk/...