The Power of Chat GPT: Unlocking Recipe Solutions and AI Insights

As I watched the keynote presentation, I couldn't help but feel impressed with what was being showcased with Apple Intelligence. However, amidst all the excitement, I found myself pondering about the best version of a consumer AI product that could be created. In this context, I realized that the most effective approach would require access to a vast amount of user data, such as emails and messages. This led me to wonder which companies had the capabilities to pull off such a monumental task.

While Apple has been known for their strong stance on privacy, I began to think about whether they could truly deliver on this promise when it comes to AI solutions like Contextual AI. The potential risks of data loss were too great to ignore, especially when the server hardware is being manufactured and then validated by a third-party before being sealed and deployed to the private cloud compute centers. The sheer number of steps Apple takes to ensure data security is indeed impressive, but it also raises questions about the effectiveness of their approach.

One aspect that caught my attention was Microsoft's lack of involvement in creating AI products for consumers. As they no longer manufacture phones, I'm not convinced that they have the necessary resources or expertise to deliver a product like this. On the other hand, Google has access to a vast array of devices through its various products and services, including Gmail and Google Docs. However, their history with data collection and usage raises significant concerns about entrusting them with sensitive user information.

The notion that Apple's approach to privacy is unique and differentiating was something that resonated with me. Their emphasis on security and validation processes is unparalleled, making it a compelling argument for why they could potentially deliver on this promise. Moreover, the fact that their software is publicly auditable adds an extra layer of transparency and accountability.

As I continued to ponder the possibilities, I found myself wondering whether Google or Microsoft could truly replicate Apple's approach. While both companies have made significant strides in AI research, I believe that their focus on data collection and usage might lead them down a path that prioritizes efficiency over user security. This is particularly concerning when it comes to creating an AI product that has access to sensitive user information.

The key factor in creating an effective consumer AI product lies not only in its technical capabilities but also in how well it integrates with the broader ecosystem of devices and services. Apple's seamless integration of their AI solutions into their existing hardware and software environment is a major strength, allowing them to deliver a cohesive experience that users can trust.

In conclusion, while there are other companies working on AI solutions, I firmly believe that Apple's consumer AI product stands out from the rest due to its unique approach to privacy and security. By combining innovative technical capabilities with a commitment to user-centric design, Apple has created a product that not only works well but also instills confidence in users. Whether or not other companies can replicate this success remains to be seen, but for now, it seems that Apple is the leader in delivering AI-powered solutions that prioritize user security and satisfaction.

"WEBVTTKind: captionsLanguage: enso WWDC happened and Apple has blessed us with yet another one of their brand new inventions Ai and by AI we mean Apple intelligence because in classic apple fashion they just will not use a pre-existing Tech term one that's been out for decades they take artificial intelligence tweak it and repackage it to us as Apple intelligence but I would say that this time around they actually have something that's pretty unique I think that this product is arguably the best consumer AI we've seen on the market and I would also go as far as to say that I don't think other companies can easily pull off something quite like this so the first question you might have is what is Apple intelligence and I guess the best way to describe it is it's an AI system that lives on any kind of computing product from Apple a new Computing product from Apple so an iPhone a tablet a Mac and the purpose of this AI system is just there to elevate the whole user experience when it comes to Communications and work and there are three main areas that they're integrating this apple intelligence the first is with text so the generation and understanding of text the second is with image generation and then the third is with Siri it's been a completely overhauled system badly needed uh and Siri is now at the point where I used to think it was like the worst digital assistant on the planet and now it's probably going to be the best okay let's start off with that first one text so with apple intelligence your devices can look at text and it can proofread it rewrite it or summarize it and it can obviously work with first-party apps like mail or notes but also works with third party apps and this can be super useful like you can imagine like proof reading something for spelling or grammar and they have examples of rewriting paragraphs in a different tone or like as a poem and they also had examples where they summarized big walls of text into something more concise but none of this is like groundbreaking Innovation right we've seen most of this kind of tech before but you'll see in a second all of this ties into a much larger picture uh now the second thing that they showcased was image generation now they've broken it into like two components the first is Gen emojis the second is image playground so first gen emojis it's their generated emojis they basically use these text prompts and profile photos if you want to be able to create these personalized emojis and you can use them in messages or stickers and reactions and they look fun like they're a little cheesy but I think my kids my family they're just going to love these things and it's very on brand for Apple like the idea of using generative AI to create these custom emojis it's perfect for the brand uh now the second thing is a little bit more interesting to me so it's image generation that they call Image playground and the idea here is that users are able to create images from scratch and you only get a few art styles to choose from uh you can use existing photos to help with the text prompt but it's all done on device all of that generative AI is just happening right on the chip of whatever phone tablet or computer that you're using and you can use these generated images in messages to send to people or you can use it in your workflow so if you have a project that you're working on or some kind of document that you're working with you can pull up the the image playground tool and it'll contextually identify the words around the area that you're working on and create an image for you based on that context now some of the image playground demos they did didn't look great to me like they had an example of someone sending a birthday wish with one of these generated images and it just felt really gross to me like the image itself seemed dated as if it came from a really old iteration of this kind of tech uh it just looks not great but also the idea of getting one of these low effort images as a birthday wish I don't know it just feels like this would get stale very quickly if people use it too much uh but if you'll notice the images that were produced with this image playground none of them were really photo realistic like if you compared to something that mid Journey would generate these Apple intelligence photos seem much more playful but also just lower quality I think part of it is that Apple doesn't want photo realism here I think they wanted something that's meant to be more playful and fun almost like a toy but I also don't think that the hardware if they're doing this all on device I don't think the hardware could even handle that right now uh now the Third Way that they've integrated Apple intelligence into this whole Apple ecosystem is with Siri so this new Siri runs an llm that is local on device and it can now perform actions based on context from emails and messages like it can add an event into a calendar or it can make reservations or it can give you information about an appointment or flight that you had booked like you could take a conversation that you had previously and then reference it with some data from the maps app or like the photos app and then it can draw a conclusion and then act on that conclusion it can perform an action based on all of that context it is so much better than anything we've seen before from Siri and this is I think the way that we've imagined AI assistance to be like this weirdly intelligent tool now the way that it works is that right now Siri will try to perform all of these requests on device like if you ask it to do something it'll do its best to handle it locally on device but you're going to run into limitations these are llms you're dealing with right large language models keyword being large and on a device with 8 gigs of RAM it can only be so large before it doesn't even run on these systems that is the cuto by the way these systems need 8 gigs of RAM to to pull off Apple intelligence right now so if the system thinks that a particular request is too complex to handle on device it will send that request out to the cloud where there are servers with much stronger compute power and then it'll send the result back to the phone but the third variant is that if Siri thinks that the question is better suited to chat GPT like if you're asking it a more generative question or a more creative question like the example they gave was like if you have a bunch of ingredients you know you got this fish and this vegetable and you want a recipe for it chat gbt is really good at that type of stuff so it will then prompt you and say hey chat GPT is really good for this type of question do you mind if we send it over to them and if you want to you can get chat GPT to handle that request and it'll send you that answer now as I was watching the keynote I was obviously impressed with what they're showcasing with apple intelligence but there's another part of me that realized the best version of this consumer AI type product is going to come from a company that has full access to your digital life like your emails and your messages like and so I think the only companies I could pull off a product like this are going to be maybe Microsoft but Google and apple and I would say that Microsoft doesn't have a real shot here to to be that like all-encompassing product because they don't make phones anymore and and I think if you're going to create a product like this it has to be on a phone you know what I'm saying like it has to be such a pervasive and and ubiquitous part of society that if you don't have a phone you can't be the company that just Nails it okay so I would say Microsoft is out but then if you look at Google I mean they have access to the most phones right they have a huge product stack with all the Gmail and Google Docs everybody uses that stuff it's awesome but could they pull it off could they create an llm that ties in their whole product stack I think they could but the biggest thing is privacy because we're not dealing with like regular your browsing habits here you're dealing with like the again the best version of companion AI like this is something that has access to the your most personal stuff and to entrust Google with it that's that's something I think a lot of people would be uncomfortable with because it's Google and they are absolutely in the business of using your data data now some people just don't care about this stuff that's fine but I think a lot of people would be concerned if they knew that Apple or Google or Microsoft was just slurping up your user data while you were using their AI tools now the way that Apple has approached this whole thing about privacy is pretty unique it's I think the biggest differentiating factor from what Apple intelligence is compar to anything else we've seen on the market before okay so Apple has usually has had a pretty good stance about privacy they do their best to try to keep uh user data and personal information private they they seem to at least now when it comes to AI of this caliber like contextual AI you need some pretty sensitive information or it could be sensitive right you're dealing with emails you're dealing with messages that are like those are your messages and to send it out to the cloud there's a concern because remember like I showed before when Apple intelligence thinks a request is too complex to handle on device it sends it to the cloud and that's when you have the risk of privacy loss now the way that Apple has done this is really interesting they are so hardcore about this whole privacy thing when it comes to this type of data they built their own secure data centers for this Cloud element of Apple intelligence and they call it private Cloud compute uh their process seems super hardcore like even when the server Hardware is being made they take these highes photos and then at the data centers it gets validated by a third party before they actually put the hardware in and then seal it up and there's like a crazy number of steps they're doing to ensure that the data that gets sent to the cloud stays secure and also once that request has been fulfilled they say that they they nuke the data completely and nothing gets retained now I don't know how much of this is you know for fact these are very big claims that they're making but it's also seemingly auditable like the softw that they using is publicly auditable it's impressive stuff and then the other thoughts that come into my mind are like I cannot imagine Microsoft and or Google doing something like this like I feel like the the whole context of privacy and and securing user data and personal information and like keeping it like there's no way I feel like they just they just don't care enough about that stuff as companies to put in the resources that this required from Apple intelligence uh okay there you have it I think that this product from Apple their uh consumer AI product is by far the best one on the market not because of its like topend capabilities it's it is somewhat limited but it's just how they integrated all this stuff into their Hardware into this bizarrely tight uncomfortably tight ecosystem it just works so well for this type of product okay there you have it those are my thoughts on Apple intelligenceso WWDC happened and Apple has blessed us with yet another one of their brand new inventions Ai and by AI we mean Apple intelligence because in classic apple fashion they just will not use a pre-existing Tech term one that's been out for decades they take artificial intelligence tweak it and repackage it to us as Apple intelligence but I would say that this time around they actually have something that's pretty unique I think that this product is arguably the best consumer AI we've seen on the market and I would also go as far as to say that I don't think other companies can easily pull off something quite like this so the first question you might have is what is Apple intelligence and I guess the best way to describe it is it's an AI system that lives on any kind of computing product from Apple a new Computing product from Apple so an iPhone a tablet a Mac and the purpose of this AI system is just there to elevate the whole user experience when it comes to Communications and work and there are three main areas that they're integrating this apple intelligence the first is with text so the generation and understanding of text the second is with image generation and then the third is with Siri it's been a completely overhauled system badly needed uh and Siri is now at the point where I used to think it was like the worst digital assistant on the planet and now it's probably going to be the best okay let's start off with that first one text so with apple intelligence your devices can look at text and it can proofread it rewrite it or summarize it and it can obviously work with first-party apps like mail or notes but also works with third party apps and this can be super useful like you can imagine like proof reading something for spelling or grammar and they have examples of rewriting paragraphs in a different tone or like as a poem and they also had examples where they summarized big walls of text into something more concise but none of this is like groundbreaking Innovation right we've seen most of this kind of tech before but you'll see in a second all of this ties into a much larger picture uh now the second thing that they showcased was image generation now they've broken it into like two components the first is Gen emojis the second is image playground so first gen emojis it's their generated emojis they basically use these text prompts and profile photos if you want to be able to create these personalized emojis and you can use them in messages or stickers and reactions and they look fun like they're a little cheesy but I think my kids my family they're just going to love these things and it's very on brand for Apple like the idea of using generative AI to create these custom emojis it's perfect for the brand uh now the second thing is a little bit more interesting to me so it's image generation that they call Image playground and the idea here is that users are able to create images from scratch and you only get a few art styles to choose from uh you can use existing photos to help with the text prompt but it's all done on device all of that generative AI is just happening right on the chip of whatever phone tablet or computer that you're using and you can use these generated images in messages to send to people or you can use it in your workflow so if you have a project that you're working on or some kind of document that you're working with you can pull up the the image playground tool and it'll contextually identify the words around the area that you're working on and create an image for you based on that context now some of the image playground demos they did didn't look great to me like they had an example of someone sending a birthday wish with one of these generated images and it just felt really gross to me like the image itself seemed dated as if it came from a really old iteration of this kind of tech uh it just looks not great but also the idea of getting one of these low effort images as a birthday wish I don't know it just feels like this would get stale very quickly if people use it too much uh but if you'll notice the images that were produced with this image playground none of them were really photo realistic like if you compared to something that mid Journey would generate these Apple intelligence photos seem much more playful but also just lower quality I think part of it is that Apple doesn't want photo realism here I think they wanted something that's meant to be more playful and fun almost like a toy but I also don't think that the hardware if they're doing this all on device I don't think the hardware could even handle that right now uh now the Third Way that they've integrated Apple intelligence into this whole Apple ecosystem is with Siri so this new Siri runs an llm that is local on device and it can now perform actions based on context from emails and messages like it can add an event into a calendar or it can make reservations or it can give you information about an appointment or flight that you had booked like you could take a conversation that you had previously and then reference it with some data from the maps app or like the photos app and then it can draw a conclusion and then act on that conclusion it can perform an action based on all of that context it is so much better than anything we've seen before from Siri and this is I think the way that we've imagined AI assistance to be like this weirdly intelligent tool now the way that it works is that right now Siri will try to perform all of these requests on device like if you ask it to do something it'll do its best to handle it locally on device but you're going to run into limitations these are llms you're dealing with right large language models keyword being large and on a device with 8 gigs of RAM it can only be so large before it doesn't even run on these systems that is the cuto by the way these systems need 8 gigs of RAM to to pull off Apple intelligence right now so if the system thinks that a particular request is too complex to handle on device it will send that request out to the cloud where there are servers with much stronger compute power and then it'll send the result back to the phone but the third variant is that if Siri thinks that the question is better suited to chat GPT like if you're asking it a more generative question or a more creative question like the example they gave was like if you have a bunch of ingredients you know you got this fish and this vegetable and you want a recipe for it chat gbt is really good at that type of stuff so it will then prompt you and say hey chat GPT is really good for this type of question do you mind if we send it over to them and if you want to you can get chat GPT to handle that request and it'll send you that answer now as I was watching the keynote I was obviously impressed with what they're showcasing with apple intelligence but there's another part of me that realized the best version of this consumer AI type product is going to come from a company that has full access to your digital life like your emails and your messages like and so I think the only companies I could pull off a product like this are going to be maybe Microsoft but Google and apple and I would say that Microsoft doesn't have a real shot here to to be that like all-encompassing product because they don't make phones anymore and and I think if you're going to create a product like this it has to be on a phone you know what I'm saying like it has to be such a pervasive and and ubiquitous part of society that if you don't have a phone you can't be the company that just Nails it okay so I would say Microsoft is out but then if you look at Google I mean they have access to the most phones right they have a huge product stack with all the Gmail and Google Docs everybody uses that stuff it's awesome but could they pull it off could they create an llm that ties in their whole product stack I think they could but the biggest thing is privacy because we're not dealing with like regular your browsing habits here you're dealing with like the again the best version of companion AI like this is something that has access to the your most personal stuff and to entrust Google with it that's that's something I think a lot of people would be uncomfortable with because it's Google and they are absolutely in the business of using your data data now some people just don't care about this stuff that's fine but I think a lot of people would be concerned if they knew that Apple or Google or Microsoft was just slurping up your user data while you were using their AI tools now the way that Apple has approached this whole thing about privacy is pretty unique it's I think the biggest differentiating factor from what Apple intelligence is compar to anything else we've seen on the market before okay so Apple has usually has had a pretty good stance about privacy they do their best to try to keep uh user data and personal information private they they seem to at least now when it comes to AI of this caliber like contextual AI you need some pretty sensitive information or it could be sensitive right you're dealing with emails you're dealing with messages that are like those are your messages and to send it out to the cloud there's a concern because remember like I showed before when Apple intelligence thinks a request is too complex to handle on device it sends it to the cloud and that's when you have the risk of privacy loss now the way that Apple has done this is really interesting they are so hardcore about this whole privacy thing when it comes to this type of data they built their own secure data centers for this Cloud element of Apple intelligence and they call it private Cloud compute uh their process seems super hardcore like even when the server Hardware is being made they take these highes photos and then at the data centers it gets validated by a third party before they actually put the hardware in and then seal it up and there's like a crazy number of steps they're doing to ensure that the data that gets sent to the cloud stays secure and also once that request has been fulfilled they say that they they nuke the data completely and nothing gets retained now I don't know how much of this is you know for fact these are very big claims that they're making but it's also seemingly auditable like the softw that they using is publicly auditable it's impressive stuff and then the other thoughts that come into my mind are like I cannot imagine Microsoft and or Google doing something like this like I feel like the the whole context of privacy and and securing user data and personal information and like keeping it like there's no way I feel like they just they just don't care enough about that stuff as companies to put in the resources that this required from Apple intelligence uh okay there you have it I think that this product from Apple their uh consumer AI product is by far the best one on the market not because of its like topend capabilities it's it is somewhat limited but it's just how they integrated all this stuff into their Hardware into this bizarrely tight uncomfortably tight ecosystem it just works so well for this type of product okay there you have it those are my thoughts on Apple intelligence\n"