Siri’s Architect Questions Apple’s AI Strategy, Cites Focus on Aesthetics Over Interaction
Table of Contents
Apple’s Siri once propelled the company to the forefront of the burgeoning chatbot race. Now, as competitors like OpenAI’s ChatGPT redefine the landscape of artificial intelligence, questions linger about why Apple hasn’t delivered a comparable breakthrough. The answer, according to one of Siri’s original architects, may lie in a long-held prioritization of visual appeal over the core quality of user interaction.
The Roots of Siri: From Research Project to Apple Acquisition
Babak Hodjat, a key developer of the natural language processing technology that powered Siri, believes Apple missed an opportunity to capitalize on its early advantage. Speaking on the sidelines of the Web Summit conference in Lisbon, Hodjat asserted that Apple was “picky about how Siri looked and felt,” prioritizing aesthetics over substantive improvements to the assistant’s conversational abilities. Apple declined to comment on these claims.
The origins of the technology stretch back to the late 1990s, when Hodjat served as chief technology officer of Dejima, a startup founded with colleagues from Kyushu University in Japan. Dejima pioneered agent-oriented technology, which ultimately formed the foundation for CALO (Cognitive Assistant that Learns and Organizes), a large-scale US research project funded by the federal government.
In 2007, engineers from the CALO project – excluding Hodjat – established Siri as a separate entity to commercialize the technology. Apple acquired Siri in 2010, integrating the assistant into its iPhones in 2011. Hodjat was not involved in the project following the acquisition. He currently serves as Chief AI Officer at the consulting firm Cognizant, leading a team of AI specialists.
A Shift in Focus: Text Over Voice
Hodjat contends that a crucial misstep occurred when Apple initially positioned Siri primarily as a voice assistant. He notes that Apple emphasized the ability to text Siri, rather than speak to it, even before the introduction of its Apple Intelligence system last year.
“When we started at Dejima, for us the type of interaction was secondary,” Hodjat explained. “You can talk, you can type – it depends on where you are and how you interact.” He further observed that many people feel uncomfortable speaking aloud to inanimate objects, suggesting that the perceived “aliveness” of a moving object, like a car, makes vocal interaction more natural than addressing a stationary appliance.
The Growing Market for Voice Assistants
Despite these potential inhibitions, the market for voice assistants continues to expand. Market research firm Emarketer, a sister company of Business Insider, forecasts that 148.7 million people will use voice assistants this year. Siri is projected to reach approximately 87.3 million users by 2025, according to an Emarketer forecast from May.
.
Hodjat’s insights offer a compelling perspective on Apple’s AI journey, suggesting that a renewed focus on the quality of interaction, rather than solely on visual presentation, may be critical for the company to regain its leadership position in the rapidly evolving world of chatbots and natural language processing.
