Yazılar

Apple to Integrate Google’s Gemini AI Model to Power New Siri, Bloomberg Reports

/in /tarafından

Apple plans to use Google’s massive 1.2 trillion-parameter Gemini AI model to power the next generation of its Siri voice assistant, according to a Bloomberg report published Wednesday. The deal, expected to be finalized soon, would see Apple pay roughly $1 billion annually for access to Google’s advanced AI technology.

The partnership marks a major step in Apple’s artificial intelligence ambitions, as it seeks to close the gap with rivals like Microsoft and OpenAI in generative AI. The Gemini model — one of the world’s most powerful — is set to serve as a temporary solution while Apple continues developing its own large-scale AI systems.

Bloomberg reported that the Gemini integration would enable Siri to handle more complex queries, offer improved conversational capabilities, and process natural language more fluidly than before. Apple’s current in-house models are far smaller in scale and capability compared with Google’s.

Neither Apple nor Alphabet has commented on the report. The collaboration underscores the growing trend among major tech firms to partner on AI development, even as they compete fiercely in the space.

Apple’s use of a Google-built AI model mirrors its long-standing partnership with the search giant, whose engine already powers Safari search functions on iPhones — another deal worth billions annually.

Amazon’s AI-Enhanced Alexa Set for Major Upgrade in February 2025

/in /tarafından

Amazon is preparing to unveil a significant overhaul of its Alexa voice assistant with the introduction of a generative AI-powered service, marking the most substantial update since Alexa’s original launch over a decade ago. The event, scheduled for February 26 in New York, will feature Panos Panay, head of Amazon’s devices and services team. While the company has remained tight-lipped about specifics, it is clear that the event will focus heavily on Alexa’s transformation.

This upgrade promises to take Alexa beyond its current capabilities, allowing the AI to engage in more complex interactions and respond to multiple requests in a single session. The new Alexa will act as an “agent,” capable of performing tasks on behalf of users without their direct input, making it a more integrated tool for daily activities like scheduling and shopping. Despite the excitement, Amazon faces significant challenges in ensuring the new AI system delivers accurate responses without the “hallucinations” common to generative models.

Although initially launching with limited access and no fees, Amazon is exploring potential subscription charges of $5 to $10 per month. Classic Alexa, the version currently in use, will continue to be available for free but will no longer receive new features. The decision to proceed with the generative AI version will be finalized during a “Go/No-go” meeting set for February 14, with Amazon executives aiming to resolve remaining concerns about performance and speed.

Alexa was originally envisioned by Amazon founder Jeff Bezos to resemble the voice-activated computers from Star Trek, capable of handling a wide range of tasks from controlling home devices to managing communications. However, after several years of stagnant innovation, Alexa’s functionality has largely remained limited to basic tasks like setting timers or checking the weather.

This new generative AI-driven version, internally referred to as “Banyan” or “Remarkable Alexa,” is expected to help Amazon recapture the interest of users by making Alexa smarter and more versatile. The company has also invested $8 billion into AI startup Anthropic to support the AI’s development. According to analysts, if 10% of Alexa’s 100 million active users were to pay for the new service, Amazon could generate an estimated $600 million annually.

Apple Clarifies Siri Privacy After $95 Million Settlement

/in /tarafından

Apple has clarified its stance on Siri’s privacy practices following a $95 million settlement in a class action lawsuit that accused the company of recording private conversations after unintentional activations of its voice assistant. The lawsuit alleged that these conversations were then shared with third parties, including advertisers.

Apple denied the claims, stating that it has never sold or used Siri data to build marketing profiles. The company emphasized that no data was shared for advertising purposes, and no audio recordings were retained unless users explicitly consented to improve Siri’s performance. As part of the settlement, Apple agreed to pay up to $20 per device to affected users of Siri-enabled devices like iPhones and Apple Watches.

The company clarified that certain Siri features do require real-time data input from Apple servers to function correctly, but it uses the minimum amount of data necessary. Apple also reiterated its commitment to enhancing privacy features for Siri in the future.

This settlement comes amid ongoing legal scrutiny, including a similar lawsuit involving Google’s Voice Assistant, which is currently pending in federal court in California.