This article examines the reasons behind the “smart failure of Siri” and the internal disputes at Apple that delayed the development of the voice assistant and left the company lagging in the artificial intelligence race.
Article contents:
- Roots of the problem
- Internal disputes
- Privacy concerns
- Employee frustration
- John Giannandrea’s vision
- Stalled ambitious projects
- Deceptive demo
- Horizon of change
- Frequently asked questions
Roots of the problem
The Information reported that Apple faced numerous obstacles in developing Siri’s AI-powered capabilities. Although Apple was one of the first companies to introduce a voice assistant on smartphones, Siri suffered from sluggish development and remained limited compared to its competitors.
Internal disputes
Between 2018 and 2024, disagreements arose between Apple’s AI teams and senior leadership. Some demanded the development of large models running in the cloud, while others believed it was necessary to focus on smaller models running on-device to preserve privacy. These divisions led to wasted effort and poor coordination between teams.
Privacy concerns
Apple has long been known for its strict commitment to user privacy. This dedication acted as a barrier to collecting massive amounts of data that could be used to train language models, making Siri’s development less seamless compared to competitors like ChatGPT, which relies on vast quantities of data.
Employee frustration
Some employees nicknamed Apple’s AI department “AIMLess” to highlight the lack of clear direction. Issues such as favoritism, salary discrepancies, variations in leave days, and bonuses emerged, resulting in the departure of several experts to competing companies.
John Giannandrea’s vision
John Giannandrea took charge of Apple’s AI department after joining from Google, believing that improving data quality and web analysis would be sufficient to advance Siri. However, the emergence of ChatGPT in 2022 did not prompt Apple to accelerate its pace of work or fundamentally change its model. Giannandrea stressed that chat technologies might not offer substantial core value to the user.
Stalled ambitious projects
Alongside developing the voice assistant, there was a plan to bring Siri’s capabilities to the “Apple Vision Pro” mixed-reality headset, while giving it the ability to receive multiple commands from more than one person simultaneously, but these proposals were rejected or delayed. Even the feature of changing the wake phrase from “Hey Siri” to “Siri” took nearly two years to implement.
Deceptive demo
At the WWDC 2024 developer conference, Apple presented an “exciting” demonstration of new Siri capabilities, such as tracking flight schedules directly from email. However, it later turned out that many of these features were “cosmetic” and unavailable on iPhone beta devices except for the animated colored border feature.
Horizon of change
Reports indicated that Craig Federighi and Mike Rockwell have taken over responsibility for Siri’s development, with a loosening of previous strict restrictions and permission to use open-source models. This development could open the door to redesigning Siri and reviving it as a competitive tool in the artificial intelligence market.
Frequently Asked Questions (FAQ)
1. Why is Siri lagging behind competitors in artificial intelligence?
The most prominent reasons include internal disputes, an intense focus on privacy, scarce available training data, and poor coordination between teams.
2. Does this mean a total failure for Siri?
Apple has not ended the Siri project, but it has acknowledged a delay in its development compared to services like Google Assistant and ChatGPT. New efforts are underway to develop it.
3. What role does privacy play in this delay?
Apple’s adoption of strict privacy policies restricts it from broadly collecting user data, which large models need to learn efficiently.
4. Was there a fake demo at WWDC 2024?
According to reports, some of the featured features were more like simulations rather than existing in actual models, which sparked internal and external criticism.
5. What are the main disagreements between development teams?
They center around one team’s desire to use massive cloud models versus another team focusing on lightweight on-device models, in addition to administrative issues.
6. Does this affect Apple’s reputation?
This may raise questions about Apple’s ability to innovate in the field of artificial intelligence, but the strength of Apple’s brand remains firm in the market.
7. Will Apple use external models?
Reports indicated that Craig Federighi recently allowed teams to test open-source models, which may reflect a shift in the company’s strategy.
8. When will the improvements be available?
There are no confirmed dates, as it depends on the success of the adjustments and the depth of changes in Apple’s administrative and technical structure.
