Apple was one of the first companies to popularize the concept of sampling data without having access to any identifying information about the users whose data was being sampled. This approach has been shown to be effective in collecting useful data while minimizing privacy concerns.

The original Mac file system was created in the early 80s, and it was later updated to HFS Plus in the late '80s or '90s. The new file system is a completely modern design that takes into account the changing way people use computers and storage devices. It's designed to work on solid-state storage, which is more common than traditional hard drives.

One of the key benefits of the new file system is its ability to provide snapshots of files and folders. This allows users to easily recreate previous versions of their data if they need to. The new system also supports duplicate files, making it easier for users to manage large collections of data without having to duplicate each item.

The old Mac file system was also prone to corruption, which could make it difficult for users to access certain files or even render them unusable. The new design aims to address this issue by providing a more robust and fault-tolerant system.

The HFS Plus format is getting a complete overhaul with the Apple file system. It's designed to work seamlessly on all Apple devices, including iOS and macOS, without requiring any significant changes to how users interact with their data.

"WEBVTTKind: captionsLanguage: enyou're listening to the Apple Insider podcast Apple kicked off its worldwide developers conference on Monday and apple Insider was on the scene with iOS 10 Mac OS Sierra watch OS 3 and a new TV OS all due to arrive later this year developers have a lot to look forward to in 2016 as usual Dan Diller was in San Francisco covering the event live and we go to him now in this special edition of the Apple Insider podcast so with me today for a very special episode of the Apple Insider podcast WWDC 2016 Edition is Dan diler live from San Francisco how's it going Dan it's good it's uh really windy outside so trying to find a quiet place to do this indeed here is a little bit of music going on as well well this is the quietest place I found so far so it won't kick us out while I'm doing it here in somebody's Lobby it's a party so uh what did you think of today's announcements there was a lot going over there was some things that were expected there were some things that I had kind of speculated about that um ended up being addressed I think they were if you look at kind of the order they went through things um starting with watch OS what they attacked was the the most obvious problem with it which is speed of apps um right now it's kind of it's not terribly useful to have apps on the watch because you launch an app and the screen shuts off before it loads and um there there are a few apps that I use on my watch but um having you know tackling that as the problem really shows that they are aware that that's that's the weakest link to be patched right so instead of going down uh feature by feature as they listed it what was the what was the number one thing that stuck out in your mind as uh as somewhat surprising that they that they announced on stage um the I guess it isn't surprising because this is something that I wrote about before but um one of the things that I anticipated that they kind of went further along than I thought they would was uh what they're doing with messages on iOS and on all their platforms um and kind of more broadly me the the things they're doing with messages is opening it up to third parties to basically run um essentially an app within messages and it uses the same extension architecture that they debuted a couple years ago to open up things like uh keyboards and last year there was content blockers in Safari so what they're doing is they're they created a mechanism that safely opens up basically a plugin architecture and they're now um creating more extension points or or places you can plug in a specific type of extension throughout the operating system doing it you know on the Mac and they're doing it on iOS and what it's doing is it's allowing developers to extend the features that Apple parts of the platform that have been under kind of controlled um by Apple in the past now they're kind of opening that up and it was interesting that they did it with messages and they're also doing it with maps and then they gave a series of other examples that they've done that with but those are the two things that I kind of had a wish list I I wanted them to do that and what they delivered was kind of more than I was expecting right so and I know uh was it uh Craig Figi who uh kind of teased the the square integration um that was an example of one of the things you could do so it's it's kind of like this idea of messages as a platform is kind of starting to take shape uh I think other companies are also working on this kind of Technology right like um say Snapchat maybe for for example or something like that well yeah basically everyone is is working on on making messages because no one else has a platform I mean Google has a platform Microsoft has is a platform that nobody uses and apart from that I mean apple is the platform for for mobile devices um and Google's not fully control of of Android the individual lenses can do whatever they want and Google has what now like a half dozen different messaging related apps so um what all the other companies in the mobile industry are doing what many of them are doing most of the most successful ones starting with Facebook um and in China the equivalent of Facebook and Twitter WeChat things like that they're all um their platform is messaging they're a messaging app and they're turning it into something more than that they're turning it into basically the equivalent of if you had a a whole OS so they're adding ways to buy things they're adding ways to um kind of talk in an entertaining way that captivates people and keeps them engaged and with messages um Apple's been kind of incrementally that on their own last year they introduced some new features of being able to record video right as right within messages app record audio send your voice things like that now what they're doing is they've added a series of other things that are kind of reminiscent of There features from Facebook there's features I've seen in Chinese apps where you can do things like you know the idea of stickers and um a number of other things that they demonstrated where you're having sort of engaging ways to chat with people right like the new chat Bubbles and the uh the backgrounds and all that s chat and it it you know hits the screen and jiggles everything around in various different ways so adding that kind of personality and then also um the more interesting aspect of that is opening up your third party developers you can basically create messaging apps and the simplest thing is stickers so you don't even have to be a developer you have basically just put images on sale for some you know trivial amount that people buy and you can not only send them like giant Emoji but you can also peel them off and stick them on top of each other or stick them on top of an image so it's you know a clever a clever idea there right also that's kind of the entry level point you can do much more complex things and some of the things that they um it demonstrated was like you mentioned square cash being able to send cash um being able to order things you can um have one of the app ideas that they demonstrated was this cone creation app where one person picks a cone type like an ice cream cone one person picks the cone it sends it to the other person and when they they get it it's like a object that they can manipulate right and also there yeah right and also they're they're integrating um stuff from their own ecosystem right they're bringing in uh digital touch um to messages and they're kind of uh making a more cohesive experience between all the uh all the different Hardware uh systems that they have going on did yeah so on on the watch that's one of the things I wrote about a couple of the articles that I did was you know they had this kind of interesting thing to to message on the watch with but you can only message other watch owners and so if you don't know butn other people that have a watch it's kind of limiting what you can do and um so it's kind of obvious to me I mean it's like it seemed obvious that that's one of the things they'd be working towards um I wasn't certain that that's what they're going to do but they they really went beyond that so not only can you do the things you can do on a watch where you can you know draw a little sketch or something but you can also um sketch characters on the watch as well right right and um of do the heartbeat thing and and other other interesting ways to communicate right so they they just kind of went full hog and and created all kinds of different ways to message and also integrating Emoji because if you're trying to use emoji sometimes you know you go looking through the the emoji keyboard and there's a lot of stuff to go through so you know unless there's a couple things that you normally send out you know regularly sent out it's hard to you know kind of memorize 300 different emoji icons you could possibly use so the idea of finding words in your text and offering you can change them into Emoji if you want to somebody joked I saw on Twitter that they could maybe do the that in the opposite direction so old people crazy lineup emojis from somebody it's like here put this back into English for me because I'm sure someone will uh create an app that will uh hook into the API somehow but speaking of um uh kind of like the data detectors kind of thing they Apple announced a lot of uh AI driven uh projects that are just now coming to fruition so they announced a lot of stuff with Siri today that was one of the main that was kind of one of the one of the overarching themes I I felt through the entire presentation it's it's now on um it's going to be coming to Mac OS well the the newly renamed Mac OS and they announced some other things with uh with Siri what what did you think about that uh well they built in a whole API for developers to extend it so basically developers come up with a couple of vocabularies of words that um a Siri can use to help sugar to know what to do with the uh what you say so when you talk in this the first thing it does is figure out what you're saying and then it has to parse the language that you said and kind of derive the intent that you have and they have a really smart system that works in a lot of languages and people have kind of compared it to echo which is one of the articles that I wrote Amazon's work Amazon is the people that have used the echo compare it to Siri and they say oh it works better and there's a couple reasons for that you know technically um has more speak you know it's plugged into the wall so it has uh Wi-Fi strong Network signal and it has constant power so it can have multiple mics going and do better voice recognition but one of the other things that Apple's not been as as good at historically is figuring out what you actually meant to say and when it works well it's great but it's hard to get it to do very complex things and so they've been doing a lot of work on that and some of the stuff they showed was really impressive but it's really an engine that's it it not only um works well in English in terms of pulling things apart but it's also designed to work in the 30 something other languages that apple support Syria and what that does is it gives developers the ability to all they have to do is um provide some Clues to thei about what's going to um why users would be using sir for their app and then it basically Apple's doing the work behind the scenes to make it work everywhere so around the world and languages the developer doesn't even need to know how to speak right right so that's a definite Advantage um that Siri has over say Alexa on the echo right right and is is the the article that I wrote that was talking about an analyst who' done some studying was looking at how people use it even just in the United States and he was showing you know clearly the most people using any kind of voice service is overwhelmingly um the traffic is going to sy so Apple knows a lot about what people are asking for and that gives them the ability to um respond to that and then know what to work on MH there's now like a billion devices that are listening uh for Siri commands um and as it becomes better and more useful there's even going to be more right so this is kind of like opening the door to what people have kind of been waiting for with Siri up until now you could do a limited number of tasks and they were you know they were fairly well implemented you could you uh tell your device to work with uh you know certain trusted apps but now that it's open to third-party um developers it kind of uh blows the door wide open to a whole new world of possibilities right yeah I mean it's basically your you're voice enabling a lot of different apps so Siri is I mean Apple's working on the framework of Siri and then other developers can extend it in ways that um do something interesting that users of the app would want to do yeah did you did you see see anything uh that caught your eye maybe maybe uh either in the keynote or uh you know just kind of walking around listening to people Chit Chat did you uh did you did you see anything that was particularly uh uh eye-catching as far as what could be happening in the near future with Siri as far as a app integration goes uh well some you know some of the examples they gave were things that we've already seen competitors doing I mean just being able to tap into things like Pride sharing you know call me an Uber kind of thing um there's right now there's a limited number probably six different categories of things that Siri six different categories of basically app like functions that Siri can handle um and that's going to increase in the future but U it's things like calling a cab or messaging or vo um placing a a voice call so apps like Skype or like um anything else that's that's used for voice calls can respond to uh commands that would normally patch through to messages or making a phone call and can be used the same way so if you have a contact that you always use WeChat to talk to them or if you have something that a particular person who always on Skype you can say that you want to Skype this person and Siri will know how to handle that right and there's also a couple other um AI related announcements that they had today uh including I think they they dropped the uh the buzzword machine learning so what are they doing on that front uh well there's a couple things I mean the the computer learning in terms of uh features for photos was really interesting because they're doing a lot of the same things that um Google has has been doing with photos um it was kind of impressive you know looking at your photos and being able to um I don't know how many years ago it was that they introduced face detection and in photos you could do all this you could categorize all your um you could teach your computer you teach your Mac how to categorize your photos by the people that are in it and it worked pretty good but you couldn't do it on your phone so you're just kind of thinking what your computer had done and now phones are powerful enough to where they're doing that on the on the phone and so in addition to people and and geot taking places there's also it can it does a lot of calculations on the photos and picks out objects and um places like knowing that you're in the forest or knowing that you're on the beach and can categorize your photos into groups and I think a lot of us take just tons of pictures and so you end up with sort of overwhelming bunch of pictures um it's kind of hard to sort through sometimes so being able to to ask Siri by your voice or even just snping on query and saying I want to see pictures of of you know vacation in Lake Tahoe something like that you can put together a lot of um things and have like functional access to your photos or a pictures of a certain person um things like that and it also does it kind of automatically so it creates um albums dynamically from things that are related together so pictures you took at a party or pictures you took on vacation so that was really an interesting way to make photos more useful MH and kind of find signal in the noise um some of the other machine learning they're doing relates to suggesting what you're going to type next and looking intelligently at content instead of just like what word do you normally type after that word but looking at okay you're talking about this subject that means you're going to use this word not that word and kind of offering things that would be useful to be able to do and just kind of underline principle is Apple's doing a lot of that in the Silicon on the device rather than um taking when you ask a question having to send it to a server and do it on the cloud and then send back a result being able to do it on the device would be faster it's also an element of privacy there which is something that Apple was putting a lot of attention on yeah they did mention something about a a what was it differential um uh what do they call it there's a term for it I believe it's a uh differential privacy I think it's called it's a certain type of basically it's like the idea of there's a lot of math behind it but it's taking samples of information from people and being able to pull out Trends and in general um on kind of a simpler level a lot of companies already do an um an anonymizing data that's what Apple did with the genius feature in iTunes be able to understand how to put playlists together for you Google does it um there's a lot of companies that figure out what makes sense in a situation based on what a lot of people have already done but what apple is doing and I think they've patented it is looked into how to collect data in a limited way in a way that can't be faced back to you um because there's been attempts to anonymize Data before and one of the examples was n they were putting people's reviews out there and they were stripping their name off it so it was technically Anonymous but because you could cross reference that with other reviews that were already published you could figure out who the people were and so it becomes a privacy problem and there's a lot of examples like that where um you may think that you're protecting privacy but you're actually not really and what Apple's doing is creating a a mechanism where you can actually sample people's data you can take their data without affecting their privacy right I think the way I don't know if I'm right but the way I understand it is that they're taking little Snippets of data and also introducing instead of stripping information they're introducing like little little bits of uh like white noise for example I guess you'd say and then over time they get enough of these data points that they can uh project Trends um based on you know a per user or from tens of thousands of users and just kind of collect all that data together and derive some sort of result or whatever they whatever it is they're looking for right it's it's um because up till now people have been talking about Apple won't be able to do the kinds of things that Google and Facebook and other kind of surveillance ad networks do because they're they're not collecting all your data they're not they're not creating this huge dossier of information about you to know what you like what you're going to click on and what you might buy if we put that ad in front of you and Apple's been saying we don't do that we don't collect a bunch of data on you because we're not trying to sell you we're not trying to sell you to advertisers we're trying to basically sell you our computer um and there's been a lot of criticism it's kind of a Google talking point and a lot of journalists have kind of pick this up and run with it that this idea that Apple can't collect data and so they can't have any of this useful sort of um crowdsourced learning about how things work and you you know that's clearly not the case because Apple was one of the first companies doing that with things like genius but what they're showing now is that you can sample data without having without introducing a privacy problem so you get you're benefiting from um a company that's has some access to a lot of people's data but they don't need to know where any of it actually came from so it's not just your names taken off it's also you know the series of processes that you're talking about result in useful data without uh having a bunch of information on a server about you that if that server gets hacked guess what everybody or anybody can find out and use that information against you in various ways whether it's hurting your credit or stealing your identity or any number of other other things right so it goes along with uh Apple's Credo of uh keeping user user data or private data private right yeah it really changes that conversation yeah yeah touch upon this on stage today but um the HFS plus format the file uh file system format is getting a little bit of a uh little bit of an overhaul with the Apple file system quote unquote so what is that all completely overhauled yeah completely overhauled so um the original Mac file system was from the early 80s and then in the late ' 80s uh or is it late ' 80s maybe late '90s think it's said 98 they turned it into HFS plus and kind of improved a lot of things that were really old and needed to be fixed making it a 64-bit file system I think um but the there's a lot of U assumptions that the file system makes because it was basically designed at a time when you used floppy discs and you know hard drives have a certain profile of how they how they're used and now virtually all computers use I mean all iOS devices don't even have a disc they don't have a physical disc in them they have memory or um flash or uh Solid State Storage anyway and so Apple's designed a new file system that's completely modern so it's kind of a jump similar to what um lvm did for the develop the compiler and what Swift is doing for language so Apple's really uh they know what they need in terms of a technology put a lot of effort into making a core technology for the future and that's what this apple file system is so on theuh customer facing side what what can users expect from this new file system are they going to see anything different any uh it supports some of the things that we talked about with um was it ZFS things like uh um my brain is fading um snapshots is that the right word where you can for example one of the one application is if you were running a fleet of computers in school you can have kids doing all kinds of things with settings and you can uh wipe it all away and go back to a specific State there's also um a lot of cases where things are files are duplicated and it allows you to have like logical copies of files without taking up extra space and um one of the other things that they mentioned was you know optimization for for running on solid state storage and I haven't seen a lot of information about it there's a whole session about it tomorrow that I'm I'm super interested in but uh one of the common problems that affects a lot of people whether you're on a Mac or whatever is file corruption just because the the design of the file system is uh somewhat fragile so if something goes wrong it makes your pictures not readable it makes problems with your dis that are you know sometimes difficult to figure out what's going wrong here and um that's been kind of like a brewing problem for a long time and I think that's one of the things that they're targeting um with this completely new rethought file system right okay well you're going to be uh attending a bunch of sessions this week and um hopefully you can get back with us on Thursday to discuss what you've learned I'm sure your brain will be super full by that point yeah it's a lot of information to go over yeah I mean just today uh two hours of uh I don't know it's pretty pretty action-packed one thing right after the other they didn't they didn't waste much time uh fluffing things up this uh this WWDC all right well uh Dan thanks for taking uh some time to chat with us and we'll talk to you later this week have a have a good uh WWDC all right thanks sounds good all right and that was a special WWDC 2016 edition of Apple insiders podcast we'll be back to our regularly scheduled programming on Friday when Dan dilder returns and we wrap up Apple's big event see you then Iyou're listening to the Apple Insider podcast Apple kicked off its worldwide developers conference on Monday and apple Insider was on the scene with iOS 10 Mac OS Sierra watch OS 3 and a new TV OS all due to arrive later this year developers have a lot to look forward to in 2016 as usual Dan Diller was in San Francisco covering the event live and we go to him now in this special edition of the Apple Insider podcast so with me today for a very special episode of the Apple Insider podcast WWDC 2016 Edition is Dan diler live from San Francisco how's it going Dan it's good it's uh really windy outside so trying to find a quiet place to do this indeed here is a little bit of music going on as well well this is the quietest place I found so far so it won't kick us out while I'm doing it here in somebody's Lobby it's a party so uh what did you think of today's announcements there was a lot going over there was some things that were expected there were some things that I had kind of speculated about that um ended up being addressed I think they were if you look at kind of the order they went through things um starting with watch OS what they attacked was the the most obvious problem with it which is speed of apps um right now it's kind of it's not terribly useful to have apps on the watch because you launch an app and the screen shuts off before it loads and um there there are a few apps that I use on my watch but um having you know tackling that as the problem really shows that they are aware that that's that's the weakest link to be patched right so instead of going down uh feature by feature as they listed it what was the what was the number one thing that stuck out in your mind as uh as somewhat surprising that they that they announced on stage um the I guess it isn't surprising because this is something that I wrote about before but um one of the things that I anticipated that they kind of went further along than I thought they would was uh what they're doing with messages on iOS and on all their platforms um and kind of more broadly me the the things they're doing with messages is opening it up to third parties to basically run um essentially an app within messages and it uses the same extension architecture that they debuted a couple years ago to open up things like uh keyboards and last year there was content blockers in Safari so what they're doing is they're they created a mechanism that safely opens up basically a plugin architecture and they're now um creating more extension points or or places you can plug in a specific type of extension throughout the operating system doing it you know on the Mac and they're doing it on iOS and what it's doing is it's allowing developers to extend the features that Apple parts of the platform that have been under kind of controlled um by Apple in the past now they're kind of opening that up and it was interesting that they did it with messages and they're also doing it with maps and then they gave a series of other examples that they've done that with but those are the two things that I kind of had a wish list I I wanted them to do that and what they delivered was kind of more than I was expecting right so and I know uh was it uh Craig Figi who uh kind of teased the the square integration um that was an example of one of the things you could do so it's it's kind of like this idea of messages as a platform is kind of starting to take shape uh I think other companies are also working on this kind of Technology right like um say Snapchat maybe for for example or something like that well yeah basically everyone is is working on on making messages because no one else has a platform I mean Google has a platform Microsoft has is a platform that nobody uses and apart from that I mean apple is the platform for for mobile devices um and Google's not fully control of of Android the individual lenses can do whatever they want and Google has what now like a half dozen different messaging related apps so um what all the other companies in the mobile industry are doing what many of them are doing most of the most successful ones starting with Facebook um and in China the equivalent of Facebook and Twitter WeChat things like that they're all um their platform is messaging they're a messaging app and they're turning it into something more than that they're turning it into basically the equivalent of if you had a a whole OS so they're adding ways to buy things they're adding ways to um kind of talk in an entertaining way that captivates people and keeps them engaged and with messages um Apple's been kind of incrementally that on their own last year they introduced some new features of being able to record video right as right within messages app record audio send your voice things like that now what they're doing is they've added a series of other things that are kind of reminiscent of There features from Facebook there's features I've seen in Chinese apps where you can do things like you know the idea of stickers and um a number of other things that they demonstrated where you're having sort of engaging ways to chat with people right like the new chat Bubbles and the uh the backgrounds and all that s chat and it it you know hits the screen and jiggles everything around in various different ways so adding that kind of personality and then also um the more interesting aspect of that is opening up your third party developers you can basically create messaging apps and the simplest thing is stickers so you don't even have to be a developer you have basically just put images on sale for some you know trivial amount that people buy and you can not only send them like giant Emoji but you can also peel them off and stick them on top of each other or stick them on top of an image so it's you know a clever a clever idea there right also that's kind of the entry level point you can do much more complex things and some of the things that they um it demonstrated was like you mentioned square cash being able to send cash um being able to order things you can um have one of the app ideas that they demonstrated was this cone creation app where one person picks a cone type like an ice cream cone one person picks the cone it sends it to the other person and when they they get it it's like a object that they can manipulate right and also there yeah right and also they're they're integrating um stuff from their own ecosystem right they're bringing in uh digital touch um to messages and they're kind of uh making a more cohesive experience between all the uh all the different Hardware uh systems that they have going on did yeah so on on the watch that's one of the things I wrote about a couple of the articles that I did was you know they had this kind of interesting thing to to message on the watch with but you can only message other watch owners and so if you don't know butn other people that have a watch it's kind of limiting what you can do and um so it's kind of obvious to me I mean it's like it seemed obvious that that's one of the things they'd be working towards um I wasn't certain that that's what they're going to do but they they really went beyond that so not only can you do the things you can do on a watch where you can you know draw a little sketch or something but you can also um sketch characters on the watch as well right right and um of do the heartbeat thing and and other other interesting ways to communicate right so they they just kind of went full hog and and created all kinds of different ways to message and also integrating Emoji because if you're trying to use emoji sometimes you know you go looking through the the emoji keyboard and there's a lot of stuff to go through so you know unless there's a couple things that you normally send out you know regularly sent out it's hard to you know kind of memorize 300 different emoji icons you could possibly use so the idea of finding words in your text and offering you can change them into Emoji if you want to somebody joked I saw on Twitter that they could maybe do the that in the opposite direction so old people crazy lineup emojis from somebody it's like here put this back into English for me because I'm sure someone will uh create an app that will uh hook into the API somehow but speaking of um uh kind of like the data detectors kind of thing they Apple announced a lot of uh AI driven uh projects that are just now coming to fruition so they announced a lot of stuff with Siri today that was one of the main that was kind of one of the one of the overarching themes I I felt through the entire presentation it's it's now on um it's going to be coming to Mac OS well the the newly renamed Mac OS and they announced some other things with uh with Siri what what did you think about that uh well they built in a whole API for developers to extend it so basically developers come up with a couple of vocabularies of words that um a Siri can use to help sugar to know what to do with the uh what you say so when you talk in this the first thing it does is figure out what you're saying and then it has to parse the language that you said and kind of derive the intent that you have and they have a really smart system that works in a lot of languages and people have kind of compared it to echo which is one of the articles that I wrote Amazon's work Amazon is the people that have used the echo compare it to Siri and they say oh it works better and there's a couple reasons for that you know technically um has more speak you know it's plugged into the wall so it has uh Wi-Fi strong Network signal and it has constant power so it can have multiple mics going and do better voice recognition but one of the other things that Apple's not been as as good at historically is figuring out what you actually meant to say and when it works well it's great but it's hard to get it to do very complex things and so they've been doing a lot of work on that and some of the stuff they showed was really impressive but it's really an engine that's it it not only um works well in English in terms of pulling things apart but it's also designed to work in the 30 something other languages that apple support Syria and what that does is it gives developers the ability to all they have to do is um provide some Clues to thei about what's going to um why users would be using sir for their app and then it basically Apple's doing the work behind the scenes to make it work everywhere so around the world and languages the developer doesn't even need to know how to speak right right so that's a definite Advantage um that Siri has over say Alexa on the echo right right and is is the the article that I wrote that was talking about an analyst who' done some studying was looking at how people use it even just in the United States and he was showing you know clearly the most people using any kind of voice service is overwhelmingly um the traffic is going to sy so Apple knows a lot about what people are asking for and that gives them the ability to um respond to that and then know what to work on MH there's now like a billion devices that are listening uh for Siri commands um and as it becomes better and more useful there's even going to be more right so this is kind of like opening the door to what people have kind of been waiting for with Siri up until now you could do a limited number of tasks and they were you know they were fairly well implemented you could you uh tell your device to work with uh you know certain trusted apps but now that it's open to third-party um developers it kind of uh blows the door wide open to a whole new world of possibilities right yeah I mean it's basically your you're voice enabling a lot of different apps so Siri is I mean Apple's working on the framework of Siri and then other developers can extend it in ways that um do something interesting that users of the app would want to do yeah did you did you see see anything uh that caught your eye maybe maybe uh either in the keynote or uh you know just kind of walking around listening to people Chit Chat did you uh did you did you see anything that was particularly uh uh eye-catching as far as what could be happening in the near future with Siri as far as a app integration goes uh well some you know some of the examples they gave were things that we've already seen competitors doing I mean just being able to tap into things like Pride sharing you know call me an Uber kind of thing um there's right now there's a limited number probably six different categories of things that Siri six different categories of basically app like functions that Siri can handle um and that's going to increase in the future but U it's things like calling a cab or messaging or vo um placing a a voice call so apps like Skype or like um anything else that's that's used for voice calls can respond to uh commands that would normally patch through to messages or making a phone call and can be used the same way so if you have a contact that you always use WeChat to talk to them or if you have something that a particular person who always on Skype you can say that you want to Skype this person and Siri will know how to handle that right and there's also a couple other um AI related announcements that they had today uh including I think they they dropped the uh the buzzword machine learning so what are they doing on that front uh well there's a couple things I mean the the computer learning in terms of uh features for photos was really interesting because they're doing a lot of the same things that um Google has has been doing with photos um it was kind of impressive you know looking at your photos and being able to um I don't know how many years ago it was that they introduced face detection and in photos you could do all this you could categorize all your um you could teach your computer you teach your Mac how to categorize your photos by the people that are in it and it worked pretty good but you couldn't do it on your phone so you're just kind of thinking what your computer had done and now phones are powerful enough to where they're doing that on the on the phone and so in addition to people and and geot taking places there's also it can it does a lot of calculations on the photos and picks out objects and um places like knowing that you're in the forest or knowing that you're on the beach and can categorize your photos into groups and I think a lot of us take just tons of pictures and so you end up with sort of overwhelming bunch of pictures um it's kind of hard to sort through sometimes so being able to to ask Siri by your voice or even just snping on query and saying I want to see pictures of of you know vacation in Lake Tahoe something like that you can put together a lot of um things and have like functional access to your photos or a pictures of a certain person um things like that and it also does it kind of automatically so it creates um albums dynamically from things that are related together so pictures you took at a party or pictures you took on vacation so that was really an interesting way to make photos more useful MH and kind of find signal in the noise um some of the other machine learning they're doing relates to suggesting what you're going to type next and looking intelligently at content instead of just like what word do you normally type after that word but looking at okay you're talking about this subject that means you're going to use this word not that word and kind of offering things that would be useful to be able to do and just kind of underline principle is Apple's doing a lot of that in the Silicon on the device rather than um taking when you ask a question having to send it to a server and do it on the cloud and then send back a result being able to do it on the device would be faster it's also an element of privacy there which is something that Apple was putting a lot of attention on yeah they did mention something about a a what was it differential um uh what do they call it there's a term for it I believe it's a uh differential privacy I think it's called it's a certain type of basically it's like the idea of there's a lot of math behind it but it's taking samples of information from people and being able to pull out Trends and in general um on kind of a simpler level a lot of companies already do an um an anonymizing data that's what Apple did with the genius feature in iTunes be able to understand how to put playlists together for you Google does it um there's a lot of companies that figure out what makes sense in a situation based on what a lot of people have already done but what apple is doing and I think they've patented it is looked into how to collect data in a limited way in a way that can't be faced back to you um because there's been attempts to anonymize Data before and one of the examples was n they were putting people's reviews out there and they were stripping their name off it so it was technically Anonymous but because you could cross reference that with other reviews that were already published you could figure out who the people were and so it becomes a privacy problem and there's a lot of examples like that where um you may think that you're protecting privacy but you're actually not really and what Apple's doing is creating a a mechanism where you can actually sample people's data you can take their data without affecting their privacy right I think the way I don't know if I'm right but the way I understand it is that they're taking little Snippets of data and also introducing instead of stripping information they're introducing like little little bits of uh like white noise for example I guess you'd say and then over time they get enough of these data points that they can uh project Trends um based on you know a per user or from tens of thousands of users and just kind of collect all that data together and derive some sort of result or whatever they whatever it is they're looking for right it's it's um because up till now people have been talking about Apple won't be able to do the kinds of things that Google and Facebook and other kind of surveillance ad networks do because they're they're not collecting all your data they're not they're not creating this huge dossier of information about you to know what you like what you're going to click on and what you might buy if we put that ad in front of you and Apple's been saying we don't do that we don't collect a bunch of data on you because we're not trying to sell you we're not trying to sell you to advertisers we're trying to basically sell you our computer um and there's been a lot of criticism it's kind of a Google talking point and a lot of journalists have kind of pick this up and run with it that this idea that Apple can't collect data and so they can't have any of this useful sort of um crowdsourced learning about how things work and you you know that's clearly not the case because Apple was one of the first companies doing that with things like genius but what they're showing now is that you can sample data without having without introducing a privacy problem so you get you're benefiting from um a company that's has some access to a lot of people's data but they don't need to know where any of it actually came from so it's not just your names taken off it's also you know the series of processes that you're talking about result in useful data without uh having a bunch of information on a server about you that if that server gets hacked guess what everybody or anybody can find out and use that information against you in various ways whether it's hurting your credit or stealing your identity or any number of other other things right so it goes along with uh Apple's Credo of uh keeping user user data or private data private right yeah it really changes that conversation yeah yeah touch upon this on stage today but um the HFS plus format the file uh file system format is getting a little bit of a uh little bit of an overhaul with the Apple file system quote unquote so what is that all completely overhauled yeah completely overhauled so um the original Mac file system was from the early 80s and then in the late ' 80s uh or is it late ' 80s maybe late '90s think it's said 98 they turned it into HFS plus and kind of improved a lot of things that were really old and needed to be fixed making it a 64-bit file system I think um but the there's a lot of U assumptions that the file system makes because it was basically designed at a time when you used floppy discs and you know hard drives have a certain profile of how they how they're used and now virtually all computers use I mean all iOS devices don't even have a disc they don't have a physical disc in them they have memory or um flash or uh Solid State Storage anyway and so Apple's designed a new file system that's completely modern so it's kind of a jump similar to what um lvm did for the develop the compiler and what Swift is doing for language so Apple's really uh they know what they need in terms of a technology put a lot of effort into making a core technology for the future and that's what this apple file system is so on theuh customer facing side what what can users expect from this new file system are they going to see anything different any uh it supports some of the things that we talked about with um was it ZFS things like uh um my brain is fading um snapshots is that the right word where you can for example one of the one application is if you were running a fleet of computers in school you can have kids doing all kinds of things with settings and you can uh wipe it all away and go back to a specific State there's also um a lot of cases where things are files are duplicated and it allows you to have like logical copies of files without taking up extra space and um one of the other things that they mentioned was you know optimization for for running on solid state storage and I haven't seen a lot of information about it there's a whole session about it tomorrow that I'm I'm super interested in but uh one of the common problems that affects a lot of people whether you're on a Mac or whatever is file corruption just because the the design of the file system is uh somewhat fragile so if something goes wrong it makes your pictures not readable it makes problems with your dis that are you know sometimes difficult to figure out what's going wrong here and um that's been kind of like a brewing problem for a long time and I think that's one of the things that they're targeting um with this completely new rethought file system right okay well you're going to be uh attending a bunch of sessions this week and um hopefully you can get back with us on Thursday to discuss what you've learned I'm sure your brain will be super full by that point yeah it's a lot of information to go over yeah I mean just today uh two hours of uh I don't know it's pretty pretty action-packed one thing right after the other they didn't they didn't waste much time uh fluffing things up this uh this WWDC all right well uh Dan thanks for taking uh some time to chat with us and we'll talk to you later this week have a have a good uh WWDC all right thanks sounds good all right and that was a special WWDC 2016 edition of Apple insiders podcast we'll be back to our regularly scheduled programming on Friday when Dan dilder returns and we wrap up Apple's big event see you then Iyou're listening to the Apple Insider podcast Apple kicked off its worldwide developers conference on Monday and apple Insider was on the scene with iOS 10 Mac OS Sierra watch OS 3 and a new TV OS all due to arrive later this year developers have a lot to look forward to in 2016 as usual Dan Diller was in San Francisco covering the event live and we go to him now in this special edition of the Apple Insider podcast so with me today for a very special episode of the Apple Insider podcast WWDC 2016 Edition is Dan diler live from San Francisco how's it going Dan it's good it's uh really windy outside so trying to find a quiet place to do this indeed here is a little bit of music going on as well well this is the quietest place I found so far so it won't kick us out while I'm doing it here in somebody's Lobby it's a party so uh what did you think of today's announcements there was a lot going over there was some things that were expected there were some things that I had kind of speculated about that um ended up being addressed I think they were if you look at kind of the order they went through things um starting with watch OS what they attacked was the the most obvious problem with it which is speed of apps um right now it's kind of it's not terribly useful to have apps on the watch because you launch an app and the screen shuts off before it loads and um there there are a few apps that I use on my watch but um having you know tackling that as the problem really shows that they are aware that that's that's the weakest link to be patched right so instead of going down uh feature by feature as they listed it what was the what was the number one thing that stuck out in your mind as uh as somewhat surprising that they that they announced on stage um the I guess it isn't surprising because this is something that I wrote about before but um one of the things that I anticipated that they kind of went further along than I thought they would was uh what they're doing with messages on iOS and on all their platforms um and kind of more broadly me the the things they're doing with messages is opening it up to third parties to basically run um essentially an app within messages and it uses the same extension architecture that they debuted a couple years ago to open up things like uh keyboards and last year there was content blockers in Safari so what they're doing is they're they created a mechanism that safely opens up basically a plugin architecture and they're now um creating more extension points or or places you can plug in a specific type of extension throughout the operating system doing it you know on the Mac and they're doing it on iOS and what it's doing is it's allowing developers to extend the features that Apple parts of the platform that have been under kind of controlled um by Apple in the past now they're kind of opening that up and it was interesting that they did it with messages and they're also doing it with maps and then they gave a series of other examples that they've done that with but those are the two things that I kind of had a wish list I I wanted them to do that and what they delivered was kind of more than I was expecting right so and I know uh was it uh Craig Figi who uh kind of teased the the square integration um that was an example of one of the things you could do so it's it's kind of like this idea of messages as a platform is kind of starting to take shape uh I think other companies are also working on this kind of Technology right like um say Snapchat maybe for for example or something like that well yeah basically everyone is is working on on making messages because no one else has a platform I mean Google has a platform Microsoft has is a platform that nobody uses and apart from that I mean apple is the platform for for mobile devices um and Google's not fully control of of Android the individual lenses can do whatever they want and Google has what now like a half dozen different messaging related apps so um what all the other companies in the mobile industry are doing what many of them are doing most of the most successful ones starting with Facebook um and in China the equivalent of Facebook and Twitter WeChat things like that they're all um their platform is messaging they're a messaging app and they're turning it into something more than that they're turning it into basically the equivalent of if you had a a whole OS so they're adding ways to buy things they're adding ways to um kind of talk in an entertaining way that captivates people and keeps them engaged and with messages um Apple's been kind of incrementally that on their own last year they introduced some new features of being able to record video right as right within messages app record audio send your voice things like that now what they're doing is they've added a series of other things that are kind of reminiscent of There features from Facebook there's features I've seen in Chinese apps where you can do things like you know the idea of stickers and um a number of other things that they demonstrated where you're having sort of engaging ways to chat with people right like the new chat Bubbles and the uh the backgrounds and all that s chat and it it you know hits the screen and jiggles everything around in various different ways so adding that kind of personality and then also um the more interesting aspect of that is opening up your third party developers you can basically create messaging apps and the simplest thing is stickers so you don't even have to be a developer you have basically just put images on sale for some you know trivial amount that people buy and you can not only send them like giant Emoji but you can also peel them off and stick them on top of each other or stick them on top of an image so it's you know a clever a clever idea there right also that's kind of the entry level point you can do much more complex things and some of the things that they um it demonstrated was like you mentioned square cash being able to send cash um being able to order things you can um have one of the app ideas that they demonstrated was this cone creation app where one person picks a cone type like an ice cream cone one person picks the cone it sends it to the other person and when they they get it it's like a object that they can manipulate right and also there yeah right and also they're they're integrating um stuff from their own ecosystem right they're bringing in uh digital touch um to messages and they're kind of uh making a more cohesive experience between all the uh all the different Hardware uh systems that they have going on did yeah so on on the watch that's one of the things I wrote about a couple of the articles that I did was you know they had this kind of interesting thing to to message on the watch with but you can only message other watch owners and so if you don't know butn other people that have a watch it's kind of limiting what you can do and um so it's kind of obvious to me I mean it's like it seemed obvious that that's one of the things they'd be working towards um I wasn't certain that that's what they're going to do but they they really went beyond that so not only can you do the things you can do on a watch where you can you know draw a little sketch or something but you can also um sketch characters on the watch as well right right and um of do the heartbeat thing and and other other interesting ways to communicate right so they they just kind of went full hog and and created all kinds of different ways to message and also integrating Emoji because if you're trying to use emoji sometimes you know you go looking through the the emoji keyboard and there's a lot of stuff to go through so you know unless there's a couple things that you normally send out you know regularly sent out it's hard to you know kind of memorize 300 different emoji icons you could possibly use so the idea of finding words in your text and offering you can change them into Emoji if you want to somebody joked I saw on Twitter that they could maybe do the that in the opposite direction so old people crazy lineup emojis from somebody it's like here put this back into English for me because I'm sure someone will uh create an app that will uh hook into the API somehow but speaking of um uh kind of like the data detectors kind of thing they Apple announced a lot of uh AI driven uh projects that are just now coming to fruition so they announced a lot of stuff with Siri today that was one of the main that was kind of one of the one of the overarching themes I I felt through the entire presentation it's it's now on um it's going to be coming to Mac OS well the the newly renamed Mac OS and they announced some other things with uh with Siri what what did you think about that uh well they built in a whole API for developers to extend it so basically developers come up with a couple of vocabularies of words that um a Siri can use to help sugar to know what to do with the uh what you say so when you talk in this the first thing it does is figure out what you're saying and then it has to parse the language that you said and kind of derive the intent that you have and they have a really smart system that works in a lot of languages and people have kind of compared it to echo which is one of the articles that I wrote Amazon's work Amazon is the people that have used the echo compare it to Siri and they say oh it works better and there's a couple reasons for that you know technically um has more speak you know it's plugged into the wall so it has uh Wi-Fi strong Network signal and it has constant power so it can have multiple mics going and do better voice recognition but one of the other things that Apple's not been as as good at historically is figuring out what you actually meant to say and when it works well it's great but it's hard to get it to do very complex things and so they've been doing a lot of work on that and some of the stuff they showed was really impressive but it's really an engine that's it it not only um works well in English in terms of pulling things apart but it's also designed to work in the 30 something other languages that apple support Syria and what that does is it gives developers the ability to all they have to do is um provide some Clues to thei about what's going to um why users would be using sir for their app and then it basically Apple's doing the work behind the scenes to make it work everywhere so around the world and languages the developer doesn't even need to know how to speak right right so that's a definite Advantage um that Siri has over say Alexa on the echo right right and is is the the article that I wrote that was talking about an analyst who' done some studying was looking at how people use it even just in the United States and he was showing you know clearly the most people using any kind of voice service is overwhelmingly um the traffic is going to sy so Apple knows a lot about what people are asking for and that gives them the ability to um respond to that and then know what to work on MH there's now like a billion devices that are listening uh for Siri commands um and as it becomes better and more useful there's even going to be more right so this is kind of like opening the door to what people have kind of been waiting for with Siri up until now you could do a limited number of tasks and they were you know they were fairly well implemented you could you uh tell your device to work with uh you know certain trusted apps but now that it's open to third-party um developers it kind of uh blows the door wide open to a whole new world of possibilities right yeah I mean it's basically your you're voice enabling a lot of different apps so Siri is I mean Apple's working on the framework of Siri and then other developers can extend it in ways that um do something interesting that users of the app would want to do yeah did you did you see see anything uh that caught your eye maybe maybe uh either in the keynote or uh you know just kind of walking around listening to people Chit Chat did you uh did you did you see anything that was particularly uh uh eye-catching as far as what could be happening in the near future with Siri as far as a app integration goes uh well some you know some of the examples they gave were things that we've already seen competitors doing I mean just being able to tap into things like Pride sharing you know call me an Uber kind of thing um there's right now there's a limited number probably six different categories of things that Siri six different categories of basically app like functions that Siri can handle um and that's going to increase in the future but U it's things like calling a cab or messaging or vo um placing a a voice call so apps like Skype or like um anything else that's that's used for voice calls can respond to uh commands that would normally patch through to messages or making a phone call and can be used the same way so if you have a contact that you always use WeChat to talk to them or if you have something that a particular person who always on Skype you can say that you want to Skype this person and Siri will know how to handle that right and there's also a couple other um AI related announcements that they had today uh including I think they they dropped the uh the buzzword machine learning so what are they doing on that front uh well there's a couple things I mean the the computer learning in terms of uh features for photos was really interesting because they're doing a lot of the same things that um Google has has been doing with photos um it was kind of impressive you know looking at your photos and being able to um I don't know how many years ago it was that they introduced face detection and in photos you could do all this you could categorize all your um you could teach your computer you teach your Mac how to categorize your photos by the people that are in it and it worked pretty good but you couldn't do it on your phone so you're just kind of thinking what your computer had done and now phones are powerful enough to where they're doing that on the on the phone and so in addition to people and and geot taking places there's also it can it does a lot of calculations on the photos and picks out objects and um places like knowing that you're in the forest or knowing that you're on the beach and can categorize your photos into groups and I think a lot of us take just tons of pictures and so you end up with sort of overwhelming bunch of pictures um it's kind of hard to sort through sometimes so being able to to ask Siri by your voice or even just snping on query and saying I want to see pictures of of you know vacation in Lake Tahoe something like that you can put together a lot of um things and have like functional access to your photos or a pictures of a certain person um things like that and it also does it kind of automatically so it creates um albums dynamically from things that are related together so pictures you took at a party or pictures you took on vacation so that was really an interesting way to make photos more useful MH and kind of find signal in the noise um some of the other machine learning they're doing relates to suggesting what you're going to type next and looking intelligently at content instead of just like what word do you normally type after that word but looking at okay you're talking about this subject that means you're going to use this word not that word and kind of offering things that would be useful to be able to do and just kind of underline principle is Apple's doing a lot of that in the Silicon on the device rather than um taking when you ask a question having to send it to a server and do it on the cloud and then send back a result being able to do it on the device would be faster it's also an element of privacy there which is something that Apple was putting a lot of attention on yeah they did mention something about a a what was it differential um uh what do they call it there's a term for it I believe it's a uh differential privacy I think it's called it's a certain type of basically it's like the idea of there's a lot of math behind it but it's taking samples of information from people and being able to pull out Trends and in general um on kind of a simpler level a lot of companies already do an um an anonymizing data that's what Apple did with the genius feature in iTunes be able to understand how to put playlists together for you Google does it um there's a lot of companies that figure out what makes sense in a situation based on what a lot of people have already done but what apple is doing and I think they've patented it is looked into how to collect data in a limited way in a way that can't be faced back to you um because there's been attempts to anonymize Data before and one of the examples was n they were putting people's reviews out there and they were stripping their name off it so it was technically Anonymous but because you could cross reference that with other reviews that were already published you could figure out who the people were and so it becomes a privacy problem and there's a lot of examples like that where um you may think that you're protecting privacy but you're actually not really and what Apple's doing is creating a a mechanism where you can actually sample people's data you can take their data without affecting their privacy right I think the way I don't know if I'm right but the way I understand it is that they're taking little Snippets of data and also introducing instead of stripping information they're introducing like little little bits of uh like white noise for example I guess you'd say and then over time they get enough of these data points that they can uh project Trends um based on you know a per user or from tens of thousands of users and just kind of collect all that data together and derive some sort of result or whatever they whatever it is they're looking for right it's it's um because up till now people have been talking about Apple won't be able to do the kinds of things that Google and Facebook and other kind of surveillance ad networks do because they're they're not collecting all your data they're not they're not creating this huge dossier of information about you to know what you like what you're going to click on and what you might buy if we put that ad in front of you and Apple's been saying we don't do that we don't collect a bunch of data on you because we're not trying to sell you we're not trying to sell you to advertisers we're trying to basically sell you our computer um and there's been a lot of criticism it's kind of a Google talking point and a lot of journalists have kind of pick this up and run with it that this idea that Apple can't collect data and so they can't have any of this useful sort of um crowdsourced learning about how things work and you you know that's clearly not the case because Apple was one of the first companies doing that with things like genius but what they're showing now is that you can sample data without having without introducing a privacy problem so you get you're benefiting from um a company that's has some access to a lot of people's data but they don't need to know where any of it actually came from so it's not just your names taken off it's also you know the series of processes that you're talking about result in useful data without uh having a bunch of information on a server about you that if that server gets hacked guess what everybody or anybody can find out and use that information against you in various ways whether it's hurting your credit or stealing your identity or any number of other other things right so it goes along with uh Apple's Credo of uh keeping user user data or private data private right yeah it really changes that conversation yeah yeah touch upon this on stage today but um the HFS plus format the file uh file system format is getting a little bit of a uh little bit of an overhaul with the Apple file system quote unquote so what is that all completely overhauled yeah completely overhauled so um the original Mac file system was from the early 80s and then in the late ' 80s uh or is it late ' 80s maybe late '90s think it's said 98 they turned it into HFS plus and kind of improved a lot of things that were really old and needed to be fixed making it a 64-bit file system I think um but the there's a lot of U assumptions that the file system makes because it was basically designed at a time when you used floppy discs and you know hard drives have a certain profile of how they how they're used and now virtually all computers use I mean all iOS devices don't even have a disc they don't have a physical disc in them they have memory or um flash or uh Solid State Storage anyway and so Apple's designed a new file system that's completely modern so it's kind of a jump similar to what um lvm did for the develop the compiler and what Swift is doing for language so Apple's really uh they know what they need in terms of a technology put a lot of effort into making a core technology for the future and that's what this apple file system is so on theuh customer facing side what what can users expect from this new file system are they going to see anything different any uh it supports some of the things that we talked about with um was it ZFS things like uh um my brain is fading um snapshots is that the right word where you can for example one of the one application is if you were running a fleet of computers in school you can have kids doing all kinds of things with settings and you can uh wipe it all away and go back to a specific State there's also um a lot of cases where things are files are duplicated and it allows you to have like logical copies of files without taking up extra space and um one of the other things that they mentioned was you know optimization for for running on solid state storage and I haven't seen a lot of information about it there's a whole session about it tomorrow that I'm I'm super interested in but uh one of the common problems that affects a lot of people whether you're on a Mac or whatever is file corruption just because the the design of the file system is uh somewhat fragile so if something goes wrong it makes your pictures not readable it makes problems with your dis that are you know sometimes difficult to figure out what's going wrong here and um that's been kind of like a brewing problem for a long time and I think that's one of the things that they're targeting um with this completely new rethought file system right okay well you're going to be uh attending a bunch of sessions this week and um hopefully you can get back with us on Thursday to discuss what you've learned I'm sure your brain will be super full by that point yeah it's a lot of information to go over yeah I mean just today uh two hours of uh I don't know it's pretty pretty action-packed one thing right after the other they didn't they didn't waste much time uh fluffing things up this uh this WWDC all right well uh Dan thanks for taking uh some time to chat with us and we'll talk to you later this week have a have a good uh WWDC all right thanks sounds good all right and that was a special WWDC 2016 edition of Apple insiders podcast we'll be back to our regularly scheduled programming on Friday when Dan dilder returns and we wrap up Apple's big event see you then I\n"