Will Apple’s ‘Answer Engine’ Rival Leading AI Chatbots? – AI Magazine
Apple has finally thrown its hat into the chatbot ring.
After months of hesitation, the tech giant has assembled a dedicated team to build an AI system designed to take on OpenAI’s ChatGPT directly.
The company has dubbed its project an “answer engine” and placed it under a newly formed division called “Answers, Knowledge and Information,” Bloomberg reports.
This is a big change for Apple, which has long preferred to acquire or partner rather than build complex AI systems from scratch.
The decision puts Apple on a collision course with OpenAI, whose ChatGPT sparked the current AI boom and Anthropic. Both firms have carved out substantial market share whilst Apple has watched from the sidelines.
Tim Cook has come under increasing pressure from shareholders who have grown frustrated watching competitors race ahead in the AI arms race.
Meta and Google’s parent Alphabet have both integrated Gen AI into their core products whilst Apple’s efforts have appeared slower by comparison.
The iPhone maker had explored several alternatives before settling on the in-house route.
Talks with both OpenAI and Anthropic about enhancing Siri, Apple’s voice assistant, foundered on cost.
The company also considered acquiring Perplexity.
The project has landed in the hands of Robby Walker, a Senior Director who previously ran Siri before engineering setbacks saw him moved sideways.
Walker is now leading the AKI team alongside several former Siri engineers who understand the frustrations of Apple’s current AI limitations better than most.
Their brief is ambitious: build a system that can trawl the web and deliver intelligent responses to almost any query.
The technology could eventually surface as a standalone app, though Apple is also exploring ways to beef up Siri and integrate the capability into Spotlight search and Safari.
Recent job adverts offer clues about Apple’s intentions.
The company is hunting for people with “experience with search algorithms and engine development” to help shape “the future of how the world connects with information” – language that suggests Apple sees this as more than just another product feature.
The timing is hardly coincidental. Apple pockets roughly US$20bn each year from Google for the privilege of being the default search engine across iPhones, iPads and Macs.
But US antitrust regulators are circling and that lucrative arrangement could be torn up, leaving a hole in Apple’s finances.
Building its own search capability would provide a cushion against such a scenario.
More importantly, it would give Apple control over one of the most valuable digital real estate assets: the gateway through which users access information.
“AI-based search represents the future” Eddy Cue, Apple’s Head of Services says during recent court testimony, confirming the company has been exploring partnerships with companies including Perplexity.
Apple’s AI push has hit turbulence, with key researchers jumping ship to Meta’s new artificial general intelligence (AGI) laboratory.
Four members of Apple’s Foundation Models team – the group responsible for building the large language models (LLMs) that power AI systems – have decamped to Meta in recent weeks alone.
The defections include some heavy hitters. Ruoming Pang, who built the Foundation Models team from scratch, left in July.
Bowen Zhang, who led work on multi-modal AI systems that can process text, images and audio simultaneously, followed soon after.
Meta reportedly offered salaries several times what Apple was paying, along with the promise of working on cutting-edge technology.
The talent raid highlights internal frustrations at Apple, where the Foundation Models team has reportedly shouldered blame for the limitations of Apple Intelligence, the company’s consumer AI platform.
Those limitations are real and obvious to anyone who has tried using Siri lately.
Apple’s assistant still struggles with basic queries, often punting users to Google searches or offering a stripped-down interface to ChatGPT.
The problem becomes particularly acute with devices like the HomePod, where users expect seamless voice interactions but often encounter dead ends.
Apple’s answer engine project remains in early stages, with industry observers suggesting a finished product is still “far off.”
The company’s methodical development process, involving extensive testing across hundreds of devices before any public launch, means users shouldn’t expect to see results soon.
But in an industry where being late to market can prove fatal, Apple may need to move faster than its instincts suggest.
“In the competitive AI landscape, Apple is signalling it’s ready to reclaim control over its AI destiny,” says Sarath Nair, Data & AI Director at Cognizant.
AI Magazine connects the leading AI executives of the world's largest brands. Our platform serves as a digital hub for connecting industry leaders, covering a wide range of services including media and advertising, events, research reports, demand generation, information, and data services. With our comprehensive approach, we strive to provide timely and valuable insights into best practices, fostering innovation and collaboration within the AI community. Join us today to shape the future for generations to come.
