Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Re-read the article and the above stood out for me, clearly it is recording, as in like "show me that drawing I looked at at store xyz this afternoon" - this is creepy, no other words for it.

Once Visual Intelliigency has identified the book then it could keep a Amazon Link or just a quickly composed reference entry "<title> Foo bar baz</title> <author> Fred Flintstone </author> <isbn> 1234567890 </isbn>. ).
It doesn't have to store the picture at all.

Same with a drawing. Just clone it with a conversion to a marked up text and box file. (notes converts hand written notes into text , math , and/or drawings. ) [ To scan/read the drawing it would have to convert parts of it to text. Similarly could convert to GraphML/svg/etc some open standards drawing file format. If a standard-like flowchart/bubblechart/etc trying to 'remember". ]

Siri has a memory of things previously done ( how long and persistent probably should be a user setting. )
It really doesn't make sense to store the RAW picture if it has already extracted the indexible elements from it (e.g, turned it into a XML file like GraphML and extracted the text data. ) The extraction is quite likely much smaller than the RAW imagine also so cuts down on storage requirments.
 
Lidar is what's used for FaceID. It supplements a camera to provide depth information. As it's about depth, it can't help with flat text.
Face ID does not use LIDAR. It uses an infrared laser dot projector and a camera array to read a 3D map of the face plus whether the subject’s eyes are open or not.

LIDAR is a part of the rear camera array in iPhone Pro and iPad Pro models.
 
Once Visual Intelliigency has identified the book then it could keep a Amazon Link or just a quickly composed reference entry "<title> Foo bar baz</title> <author> Fred Flintstone </author> <isbn> 1234567890 </isbn>. ).
It doesn't have to store the picture at all.

Same with a drawing. Just clone it with a conversion to a marked up text and box file. (notes converts hand written notes into text , math , and/or drawings. ) [ To scan/read the drawing it would have to convert parts of it to text. Similarly could convert to GraphML/svg/etc some open standards drawing file format. If a standard-like flowchart/bubblechart/etc trying to 'remember". ]

Siri has a memory of things previously done ( how long and persistent probably should be a user setting. )
It really doesn't make sense to store the RAW picture if it has already extracted the indexible elements from it (e.g, turned it into a XML file like GraphML and extracted the text data. ) The extraction is quite likely much smaller than the RAW imagine also so cuts down on storage requirments.
So you agree that it will record "something", really unbeknownst to us at this point.
This in itself is creepy ...
 
Yeah I can see some use cases around navigation or directions. Or looking at a restaurant you’re passing and adding it to a list. But none of these outweigh the privacy violations/exploits.

If the photos are just held long enough to extract the textual metadata about them and the dumped what is the expliot. If the camera feed go shoveled into random applications then there is an exploit. But if treated similar to TouchID/FaceID and goes into a Apple control controlled stack then the exploits are much narrower. Limited the types and contexts that the text/metadata can be extracted from an image and it is even more narrow.
 
Once Visual Intelliigency has identified the book then it could keep a Amazon Link or just a quickly composed reference entry "<title> Foo bar baz</title> <author> Fred Flintstone </author> <isbn> 1234567890 </isbn>. ).
It doesn't have to store the picture at all.

Same with a drawing. Just clone it with a conversion to a marked up text and box file. (notes converts hand written notes into text , math , and/or drawings. ) [ To scan/read the drawing it would have to convert parts of it to text. Similarly could convert to GraphML/svg/etc some open standards drawing file format. If a standard-like flowchart/bubblechart/etc trying to 'remember". ]

Siri has a memory of things previously done ( how long and persistent probably should be a user setting. )
It really doesn't make sense to store the RAW picture if it has already extracted the indexible elements from it (e.g, turned it into a XML file like GraphML and extracted the text data. ) The extraction is quite likely much smaller than the RAW imagine also so cuts down on storage requirments.

It's refreshing (and very rare) seeing a well-reasoned response. Seems many automatically go for the worst possible sky-is-falling use cases.

There's a ton of potential in how this tech can be used without going for the creepy possibilities (that Apple wouldn't use in any case).
 
Re-read the article and the above stood out for me, clearly it is recording, as in like "show me that drawing I looked at at store xyz this afternoon" - this is creepy, no other words for it.

Or, is Apple leaking this intentionally to see user reactions? Before a release?
And presumably all these new products will require the 27 version of the OS, so why all these reference in 26.7??
For one thing, streaming usable video over the bandwidth of something like Bluetooth LE is not possible; secondly, any other way to stream video to be able to record consumes too much energy to be sustainable with an AirPod’s tiny battery. Even Meta glasses had to have those thick legs to be outfitted with a bigger battery to enable recording. Come on people, do some basic analysis and you’ll see recording for more than a few seconds is not happening.
 
My sentiments exactly!

No matter if it's cameras in AirPods, or Apple Glasses...

It makes one wonder, just what is this company up to?

Oh sure there's the warm and fuzzy narrative and marketing speak they promote.

Then there's the potential risks that cameras everywhere present.

Better get used to it. You’re not gonna like how connected the world is within the next 5 years.
 
We may not like the idea of it, but it's becoming an inevitable part of our world.

Remember that it will enable even life-changing things like giving vision to the blind.

Rather than banning a technology that's nearly impossible to do so, we need to figure out how to regulate it in order to protect people the most.

Keeping in mind that in public any assumptions of privacy are nearly moot and have been for a long time. There are hundreds of cameras around you in people's pockets ready to use at any given moment (and many being used) plus surveillance cameras, car cameras and more.
 
For one thing, streaming usable video over the bandwidth of something like Bluetooth LE is not possible; secondly, any other way to stream video to be able to record consumes too much energy to be sustainable with an AirPod’s tiny battery. Even Meta glasses had to have those thick legs to be outfitted with a bigger battery to enable recording. Come on people, do some basic analysis and you’ll see recording for more than a few seconds is not happening.
Sure, whatever you define as "usable", back in the days 320x200 at 15fps was "usable"...
 
So, if that camera can read, as this video suggests the title of a book, what is stopping it from recording? The SW? Recording could happen on the iPhone it is paired to ...

Don't know what to make out of this... other than, not for me
If it focuses on and only resolves to the length of a human arm, it wouldn’t be better than Meta’s cameras for pervity. 🙂
 
So you agree that it will record "something", really unbeknownst to us at this point.
This in itself is creepy ...

It isn't 'creepy' it is how normal converstation works.

Person : Hey SirI what time does my Mom's flight get in today?
Sirei : 6:30 pm.
Person : what baggage claim terminal is for her flight?
Siri : Terminal C .

If Siri doesn't remember the first sentence then there is no way to resolve the 'her' pronoun. People use pronouns and previous references all the time ( he , it , them , this , etc.). Can't find the corresponding noun then really don't know what is being talked about.

What would be creepy is holding a conversation when what you are talking to can't remember the noun that occured 4-6 sentences back.

Pretty sure during one of the WWDC presentations Apple mentioned that Siri Intelligence has a iCloud history ( that is encrypted to only your account/devices). If the sevice has any kind of longer conversational abilities then it has to remember to do the resolutions of the ambiguous terms used. How much goes int othe cloud history ( or local history if have cloud turned off) probably should be a parameter the user sets. But Siri will be precieved as dumb as a rock if it is absolutely zero.


Remember the metadata about a book it just a 'short form' of what can already do now.

Person : Hey Siri create a Note about a book with the title "Foo bar Baz" authored by. Fred Flinstone.
Sire: oK created note "Book wit hthe title Foo bar baz authored by Fred Flinstone.


versus

Person : Hey Siri remember this book I'm holding up so. that you can see.
Siri: OK record the information about that book 'Foo bar baz' by Fred Flintsone.


End result is basically the exact same data. One is more verbose, but not any new material data recorded. Maybe there is a difference of whether the data is stuffed into the Notes app's file versus the Siri Intelligence 'memory' file. If want to keep it very long term it is probably safer in Notes. There is likley some 'rolloff' to Siri's 'memory' file.
 
Last edited:
We may not like the idea of it, but it's becoming an inevitable part of our world.

Remember that it will enable even life-changing things like giving vision to the blind.

Rather than banning a technology that's nearly impossible to do so, we need to figure out how to regulate it in order to protect people the most.

Keeping in mind that in public any assumptions of privacy are nearly moot and have been for a long time. There are hundreds of cameras around you in people's pockets ready to use at any given moment (and many being used) plus surveillance cameras, car cameras and more.

I think it’ll be a proper incentive for the general public to start acting right. People are not gonna get away with the things previous generations did. But the laws that will address the problems currently happening in public will be laid out accordingly.
 
If it can be used by the blind to navigate their local world while not being able to record images of people across a room (or even across the table), I would imagine that most would tend towards whatever’s Apple doing. Then, of course, once the competition gets their hands on it, it becomes the hot new form factor that sucks the wind out of the sails of smart glasses.
 
I highly doubt it’s a camera. I would think it’s something like Lidar - a new tech that can use light to “read” but not to take photos etc and consumes less power.
Lidar can't see things in that way.
Lidar times the bounce back of a laser light to measure distance. The book surface would be seen as one flat surface.
This has to be a camera.

EDIT: Sorry I should have checked that 100 other people replied with this same thing.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.