I second you Mohammed, especially for AIKit. However I think they can even go further than just on-device ML, we should be able to use the Private Cloud Compute environment as well.
AIKit is a great step forward, but unlocking Private Cloud Compute would be a game-changer. It bridges the gap between on-device performance and cloud-scale capabilities, all while preserving user privacy.
If Apple opens that door (even in a limited capacity via entitlements or API constraints), it could massively boost real-world ML use cases on iOS. 🔐🚀
Thanks for bringing that up — curious how you envision developers could integrate PCC in practice. Would love to hear more!
Ideally I think Apple should provide a suite of AI models usable either on device or PCC: LLMs, translation, transcription, vision, image generation etc.
Performance would depend on which version you’re using.
And they could give us incentives to use those APIs as opposed to any third party AI vendor by making it free with limits, similar to CloudKit.
I second you Mohammed, especially for AIKit. However I think they can even go further than just on-device ML, we should be able to use the Private Cloud Compute environment as well.
Absolutely agree! 💯
AIKit is a great step forward, but unlocking Private Cloud Compute would be a game-changer. It bridges the gap between on-device performance and cloud-scale capabilities, all while preserving user privacy.
If Apple opens that door (even in a limited capacity via entitlements or API constraints), it could massively boost real-world ML use cases on iOS. 🔐🚀
Thanks for bringing that up — curious how you envision developers could integrate PCC in practice. Would love to hear more!
Ideally I think Apple should provide a suite of AI models usable either on device or PCC: LLMs, translation, transcription, vision, image generation etc.
Performance would depend on which version you’re using.
And they could give us incentives to use those APIs as opposed to any third party AI vendor by making it free with limits, similar to CloudKit.
Exactly 🤞🏻🤞🏻