Apple's 2017 Decision: The Foundation for Apple Intelligence on M1 Chips

Apple's Visionary Move in 2017

In a revealing interview on The Circuit podcast, Apple executives Tim Millet and Tom Boger shed light on a pivotal decision made by the company in 2017 that set the stage for Apple Intelligence on M1 chips and beyond 1

1

. This strategic move demonstrates Apple's long-term vision and commitment to integrating artificial intelligence into its devices while prioritizing privacy and on-device processing.

The Birth of the Neural Engine

The journey began in 2017 with the introduction of the first version of Bionic chips featuring the Neural Engine. Initially conceived to enhance computational photography in iPhones, the Neural Engine soon became the focus of a much more ambitious plan 2

2

.

Transformer Networks: A Game-Changer

That same year, Apple's engineering team came across a scientific paper titled "Attention is All You Need," which introduced transformer neural network models. Recognizing the potential of this technology to revolutionize AI, Apple made a crucial decision to completely rethink the architecture of the Neural Engine 1

1

2

2

.

Long-Term Vision and Development

Tim Millet, Apple's Vice President of Hardware Engineering, explained that the team understood the development of technologies like transformer networks would take years. However, they also realized that adapting their chips to these capabilities would require a similar timeframe. This foresight led to the development of a more advanced version of the Neural Engine, which debuted with the M1 chip in 2020 2

2

3

3

.

M1 Chip: The Turning Point

The introduction of the M1 chip in 2020 marked a significant milestone in Apple's transition to its own silicon. The enhanced Neural Engine in M1 chips allowed for efficient running of advanced neural networks directly on the device, offering greater power, speed, and an unparalleled level of privacy 2

2

.

Apple Intelligence: Privacy-Focused AI

Apple's approach to AI stands out due to its focus on on-device processing. This strategy not only enhances performance but also ensures user privacy by keeping data processing local rather than relying on cloud servers 2

2

.

Current and Future Applications

Today, users of M1 chip devices and later models can enjoy features such as object removal in Photos, intelligent notification sorting, and real-time personalized emoji generation. Apple Intelligence is set to expand further with upcoming Image Tools and a new version of Siri with personal context awareness 2

2

.

Industry Implications

While some may argue that Apple has lagged in the AI race, this revelation demonstrates the company's deliberate and patient approach to integrating AI technologies. By focusing on hardware capabilities first, Apple has positioned itself to offer unique AI features that align with its privacy-centric philosophy 1

1

2

2

3

3

.

As the tech industry continues to evolve rapidly in the AI space, Apple's strategy of long-term planning and hardware-software integration sets it apart from competitors relying heavily on cloud-based AI solutions.