Apple's Finally Bringing Generative AI To Our iPhones (Their Most Ambitious Update Ever)

Apple will finally be entering the Generative AI arena in iOS 18. Here's me doing my best at explaining the current rumors and behavior behind what Apple claims will be the "most ambitious iPhone update in history"

📕 Corrections 📕

Running Siri or Large Language Models (LLMs) directly on iPhones won't inherently reduce AI hallucinations, but it does offer Apple greater control over the user experience, particularly in terms of consistent performance and availability, independent of internet connection. The reduction of AI hallucinations is primarily achieved through careful model design, the use of curated training data, and the model's ability to learn from personal data. However, it's important to note that Apple's commitment to user privacy means that personal data used for learning would remain on the device, not transmitted to external servers.

I was exaggerating when I said ‘Apple is going to figure out how to run an LLM locally on all of our phones’. It may be partially run on our phones and partially run in the iCloud, with our more sensitive information never leaving our device 🤞🏼. This is all up in the air at this point.

ChatGPT has the API, is the fuel behind a ton of software innovation, and doesn’t JUST live in our browser, of course. But from the everyday consumer experience standpoint, it kind of does.

‘Siri 2.0’ is my term for the upgraded version of Siri iOS 18 is speculated to bring

iOS 18 will probably bring generative AI features to a select few of built-in apps (messages, music, etc), but eventually it could be connected with other Apps.

The exact number of writers who wrote Siri’s responses is unknown, but is estimated to be about 20

🤖 Apple GPT 🤖

There is substantial evidence that Apple is indeed working on its own generative AI technology, which is being referred to internally as "Apple GPT." This technology is being developed under the leadership of Apple's AI chief John Giannandrea and is part of a broader effort within the company to advance its AI capabilities

📝The Research Paper 📝

Apple has made a key breakthrough in deploying large language models (LLMs) on iPhones and other devices with limited memory capacity. This is crucial for running advanced LLMs in resource-limited environments and expanding their applicability. The researchers have developed a technique that uses flash memory to store AI models, bypassing the limitation of insufficient DRAM on mobile devices. This method allows for efficient large language model inference with limited memory, which could enable iPhones to run complex AI assistants and chatbots on-device.

💎 Rumored New Siri Features 💎
• Revamped interaction with Messages app, enabling users to field complex questions and auto-complete sentences more effectively

• Auto-generated Apple Music playlists

• Pages and Keynote offering AI-assisted writing and slide deck creation

• Deeper integration with Shortcuts, allowing users to automate complex tasks more efficiently

Connect with me
TapLink ➔ https://taplink.cc/tylerpoelking
TikTok ➔   / tylerpoelking  
Instagram ➔   / tyteachestech  
Facebook ➔ https://www.facebook.com/profile.php?...