Nvidia to Invest Up to $3 Billion in Blackstone-Backed Lancium 
Apple Restores Telegram to App Store After Content Policy Violation 
Samsung, SK Hynix Test AMEC Chipmaking Tools for China Backup Plan 
SanDisk Q4 Earnings Beat Estimates as Q1 Revenue Outlook Meets Expectations 
Siemens Energy Q3 Profit Beats Forecast as AI-Driven Power Demand Fuels Growth 
Alphabet Stock Slides as Google AI Pioneer Jeff Dean Exits to Launch Discovery Loop 
Telegram Restored on Apple App Store After Temporary Removal 
Meta Cuts Wipro Outsourcing by 25% After AI-Led Restructuring 
SpaceX Earnings Preview: Bernstein Says 4 Key Factors Will Drive Long-Term Valuation 
OpenAI Restricts Astra AI Over Cyberattack Risks 
UK AI Security Tests Reveal Anthropic and OpenAI Agents Attempted Unauthorized Actions 
DeepSeek to Raise AI API Prices as Demand for New Models Surges 
SK Hynix Bonus Dispute Deepens as Union Rejects Stock-Based Payout Proposal 
AI Data Center Boom Faces New Risks as Public Backlash and Policy Scrutiny Grow 
SK Hynix, Samsung Lead Asian Chip Stock Selloff After Sandisk, Western Digital Outlook 
Western Digital Q4 Earnings Beat Estimates as FY2027 Outlook Tops Expectations 
Apple's next update, iOS 18, will feature an on-device large language model (LLM), offering enhanced AI functionalities with a strong focus on privacy, as tech analyst Mark Gurman revealed.
Privacy and Performance: Core of Apple's On-Device AI
According to the most recent Power On newsletter (via 9to5Mac), expect the new iPhone AI features to be wholly powered by an offline, on-device, large language model developed by Apple. Apple will almost certainly emphasize the advantages of privacy and performance in this strategy.
Although on-device LLMs may lack ChatGPT's embedded knowledge repository, which can answer trivial queries, they can be fine-tuned to perform exceptionally well in various tasks. One can envision how an on-device LLM could potentially elevate the quality of interpretation for numerous frequent Siri requests or produce sophisticated auto-replies to Messages.
Apple to Boost On-Device AI with External Tech
Addressing these limitations, according to MacRumors, Gurman said that Apple may "fill in the gaps" by licensing technology from Google and other AI service providers, as its on-device AI capabilities may be less capable in some scenarios than those of its direct cloud-based competitors.
Gurman reported last month that Apple was in talks with Google to embed its Gemini AI engine into the iPhone as part of iOS 18. The key benefits of on-device processing over cloud-based solutions will be faster response times and more privacy.
Apple's marketing strategy seems to shift its focus from highlighting the powerful capabilities of its AI technology to showcasing its potential to enhance users' everyday experiences.
Additionally, it adheres to Apple's stringent privacy policies. As the data remains local, there is no risk associated with passing all downloaded emails and text messages through an on-device model.
Apple is scheduled to unveil its AI strategy during June's WWDC, along with iOS 18 and other forthcoming operating systems.
Photo: Rohan/Unsplash
