Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

While Apple's long-term strategy is on-device AI, it must still use cloud providers like Google for the most powerful models. To reconcile this with its privacy-first brand, Apple is leveraging NVIDIA's confidential compute, which encrypts data and models even during active processing, thus maintaining its privacy guarantee off-device.

Related Insights

Apple's ability to distill Google's large Gemini models into smaller, proprietary versions reveals a strategy to accelerate its own on-device AI development, not just rely on Google's tech. This gives Apple a 'cheat code' to catch up quickly and power its core vision for local AI on iPhones.

As AI companies push for more data collection, Apple can differentiate by leveraging its brand trust. By building AI devices that prioritize user privacy, Apple can capture the premium market segment wary of constant surveillance, turning privacy into its key competitive advantage against rivals like Meta and OpenAI.

Apple's seemingly slow AI progress is likely a strategic bet that today's powerful cloud-based models will become efficient enough to run locally on devices within 12 months. This would allow them to offer powerful AI with superior privacy, potentially leapfrogging competitors.

Apple isn't trying to build the next frontier AI model. Instead, their strategy is to become the primary distribution channel by compressing and running competitors' state-of-the-art models directly on devices. This play leverages their hardware ecosystem to offer superior privacy and performance.

Technologies like Intel TDX and NVIDIA's Confidential Compute encrypt AI workloads directly on hardware. This guarantees that even the physical server owner cannot access the data, allowing anyone to contribute hardware to a decentralized network without needing to be vetted or trusted.

Apple is considering deeper reliance on Google Cloud for its AI services because its own 'private cloud compute' infrastructure is reportedly only 10% utilized. This low usage reflects the lackluster public reception of Apple Intelligence features, making the massive internal investment economically inefficient and pushing the company toward external partners.

By licensing Google's Gemini model, Apple avoids the messy and potentially brand-damaging process of training large AI models on vast datasets. This "privacy washing" allows them to deliver competitive AI features while outsourcing the associated privacy risks and controversies to Google, preserving their carefully crafted image.

Apple is focusing its AI efforts on creating a seamless ecosystem of AI-powered hardware (iPhone, AirPods, glasses) that leverage models from partners like Google. Their competitive advantage lies in device integration and user experience, not competing in the costly model-training race.

Instead of competing in the cloud, Apple's advantage is in hardware. By equipping computers with massive RAM, they can run powerful local AI models. This preserves user privacy by keeping data on-device and sidesteps trust issues with cloud-based AI providers like OpenAI and Google.

The Apple-Google AI deal isn't a simple API call. Apple is incorporating Gemini models directly, allowing it to adapt them for products like Siri while processing data within its own on-device or "private cloud" infrastructure. This structure is key to upholding its stringent user privacy standards.