The Wall Street Journal highlights a key moment in Apple’s changing attitude towards AI. In late 2022, Apple’s SVP Craig Federighi tested GitHub’s Copilot code completion features and was impressed.
This led to a directive for the software engineering group to explore and integrate as much AI functionality as possible.
The article also recounts the historical tension between Federighi’s OS software groups and John Giannandrea’s AI and machine learning division.
Due to cultural conflicts and indecision, these groups struggled to collaborate effectively.
Federighi’s team often developed its own tools independently, while Giannandrea’s work was hampered by limited resources. However, these conflicts appear to have been resolved by 2022, with both leaders now closely aligned.
Apple’s executive team is now more open to adopting cutting-edge large language model (LLM) technology, despite potential issues like AI hallucinations.
This shift is expected to be evident in Apple’s upcoming operating system releases, which will be announced at the Worldwide Developers Conference (WWDC).
New features will include a more conversational Siri, automatic messages and notification summaries, AI-enhanced quick replies, deeper integration of audio transcription services, upgraded Spotlight search, generative AI photo editing, and AI-assisted document generation in Pages.
Additionally, Apple plans to introduce rich AI-powered autocomplete for Xcode, a nod to what initially inspired Federighi’s focus on AI.
A key challenge will be balancing these new AI features with Apple’s commitment to privacy. The company plans a multi-faceted approach, combining on-device technology, AI in a virtual “black box” in the cloud, and a partnership with OpenAI for a ChatGPT-like chatbot.
Details on device compatibility and potential subscription fees for these features remain uncertain.
