The Zhitong Finance App learned that the US e-commerce and cloud computing supergiant Amazon (AMZN.US) is further extending the AI layout from an enterprise-level AI computing power resource platform to end-side AI systems, consumers' daily memory, operation task management, and online shopping portals. The tech giant launched the Bee Frontier wearable bracelet device and the Bee Eclipse mobile phone magnetic accessory on October 15. The official price is $74.99 and $99.99, respectively. It generates summaries, reminders, and to-do items by understanding everyday conversations, and can be connected to Alexa+ and Amazon.com. The $11.99 monthly Bee Pro subscription links hardware sales to ongoing service revenue.
Amazon's stock price performance since 2026 clearly shows the characteristics of “rising during the year and recent fluctuations”: Amazon closed up 0.27% to $254.76 on October 15, with a cumulative increase of about 2.25% in October and an increase of about 10.37% during the year; however, as of Thursday, it fell by about 2.92%, retreating about 11.3% from the historical intraday high of $287.20 on August 3. By the close of the US stock market on Thursday, the new product launch can be described as adding important new highlights to Amazon's AI commercialization, and is expected to become the latest positive catalyst for Amazon's stock price and basic outlook.
The business foundation that supports Amazon's AI investment narrative still mainly comes from its largest cloud platform in the world, Amazon AWS. AWS revenue in the second quarter reached US$42.2 billion, up 36.7% year on year, the fastest growth rate in 18 quarters; operating profit was US$16.6 billion, compared to US$10.2 billion in the same period last year. The enterprise AI cloud computing business has contributed considerable profits, and Amazon's C-side consumer AI platform is trying to transform the cutting-edge technology and capabilities of the AI model into more frequent consumer-side AI interactions, AI application subscriptions, and AI-driven online transaction expansion.
Let the big AI model remember everyday promises! Amazon launches two portable AI listening devices
Amazon on Thursday unveiled two new devices to listen to and record users' daily information throughout the day, provide schedule reminders and other AI-driven key consumer insights. These two products belong to the Bee brand that Amazon bought last year.
The Frontier is a bracelet, and the Eclipse is a block device that looks like a portable magnetic charger. Users can attach Eclipse to their smartphone. Amazon says these devices aggregate information from users' calendars, apps, and other channels throughout the day.
According to Amazon, the Frontier bracelet costs $75 and the Eclipse costs $100.
Panos Panay, head of Amazon's device business, said at the TechCrunch Disrupt conference in San Francisco: “They'll listen to your conversations throughout the day. It'll remind you what conversations you've had, remind you what to-do you might have, and tell you what things you've promised others to accomplish.”
He said that since the device continues to listen once turned on, users no longer need to set their own reminders, so they can “focus more on the present” throughout the day. The company said that privacy is entirely controlled by users. Consumer users can control privacy by deciding when to turn on the device and deleting insight information or previous conversations as needed. Amazon doesn't save audio.
In 2019, there were media reports that thousands of workers would listen to users' conversations with Alexa's voice assistant service, and Amazon was criticized for this. The company is currently facing a class action lawsuit over this practice. Amazon denies any wrongdoing.
Bee will keep the data for 14 days, but users can extend the retention period by paying a monthly subscription fee of $12. These devices can also connect to Alexa+, a subscription-based generative AI voice service provided by Amazon.
Amazon Bee will face competition in AI wearables, particularly Meta's Muse Charm — a pendant device equipped with the company's latest smart AI software, and Meta's glasses equipped with cameras and other sensors. Numerous startups have also developed similar devices to record conversations and act on behalf of users.
In recent weeks, Amazon has been intensively releasing a variety of new devices. This is one of Panay's moves to attract consumers willing to pay more for a higher quality AI smart lifestyle and consumer electronics durability. Despite the slowdown in global tablet sales growth, the most sold price of the series of tablets announced by Amazon last week is still up to three times that of previous versions.
Earlier this year, some media reported that Amazon is considering returning to the smartphone terminal market about ten years after the launch of the failed Fire Phone.
From the cloud to the wrist, the cloud collaborates to connect Amazon's AI empire
Recently, Amazon can be described as continuing to focus on accelerating the development of enterprise AI cloud computing superplatforms, consumer AI applications, and end-side AI computing. On the B-side, AWS provides multi-model access through BedRock, supports agent deployment, operation, and tool call through AgentCore, and uses Trainium self-developed accelerators to handle training and inference requirements; Anthropic and OpenAI have made multi-gigawatt scale computing power deployment commitments related to Amazon's AI cloud computing business for many years. Its competitiveness comes from combining models, complex and huge computing resources, and enterprise operating environments, so that customers can use AI services continuously and efficiently.
The latest C-side layout and development trend is that AI is beginning to directly accept consumer intentions... Rufus and Alexa+ have been integrated into Alexa for Shopping, covering product comparisons, price alerts, and automated purchases; Bee further complements the personal context of offline conversations. When users actively connect to relevant services, the products, agreements, and needs mentioned in the conversation can be converted into recommendations, schedules, or to-do items. It is clear from this that Bee's potential value is to increase Amazon's access to user needs and guide them to the subscription and transaction process.
The wearable AI portal and large-scale local AI model inference for C-side users are also Amazon's key layout. Amazon has deployed AZ3 and AZ3 Pro self-developed high-performance AI chips on Echo devices to enhance voice perception and end-side AI model processing and end-side AI agent workload capabilities; Bee uses a device-cloud collaborative architecture, and its Private Compute relies on AWS Nitro to securely isolate data. Officials have made it clear that audio is processed in real time and deleted, while transcribed text, summaries, and to-do items are protected by encryption, so “not saving audio” does not mean that all calculations are done within the bracelet.
Judging from the underlying technical mechanism, individual consumer AI application systems require serial speech recognition, information extraction, memory retrieval, and task execution for a long time, and commercialization therefore also depends on usage frequency and unit service cost; Amazon's recent increasing focus on end-side AI computation and processing of large end-side AI models can reduce ineffective uploads and improve response speed, and the cloud undertakes more complex understanding and task orchestration. The key to whether Bee can become an additional source of profit is whether continuous use can be converted into paid retention and shopping conversion, and cover corresponding self-research and inference costs. This is also the strategic value of Amazon's ability to simultaneously grasp cloud infrastructure, consumer portals, and trading platforms.