Original Source: Shenmou Finance (chutou0325)

On September 16, the Nubia NaviX Ultra officially hit the market. Developed in collaboration between ZTE and ByteDance, this device represents the second generation of the "Doubao Phone," explicitly positioned as an "AI agent smartphone."

The first-generation model, released as an engineering prototype last December, was priced at 3,499 yuan and limited to 30,000 units, which quickly sold out. On the secondary market, prices surged well above the original retail price. The new Doubao Phone starts at 5,999 yuan, or 5,499 yuan after national subsidies. Prior to its launch, cumulative reservations on JD.com alone exceeded 370,000 units.

Nine months have passed since the release of the limited engineering model. Now, the Doubao Phone faces a critical question: not whether the technology is feasible, but whether this new category can survive and thrive in the competitive market.

01 From Engineering Model to Commercial Product: Addressing Three Key Shortcomings

Based on official specifications, the Doubao Phone's hardware is not particularly groundbreaking among this year's flagship models. It features the fifth-generation Snapdragon 8 Supreme platform, a 7,100mAh battery, a BOE 6.78-inch screen, and a triple-camera setup that includes a 200MP main sensor, a 50MP ultra-wide lens, and a 64MP periscope telephoto lens. Notably, it uses Changxin Storage's LPDDR5X with a maximum speed of 10,667Mbps—marking the first time a high-speed domestic memory chip has been used in a mass-produced smartphone.

What sets this phone apart from ordinary flagships is the independent AI button with fingerprint recognition on the right side. Pressing this button activates an intelligent agent embedded in the system, rather than launching an app. This agent connects to the Seed full-duplex voice large model, supporting continuous conversations that can be interrupted at any time and recognizing over a dozen dialects, including Minnan, Hakka, Shanghai, and Cantonese.

The manufacturer highlights its capabilities in four key areas: understanding, execution, memory, and security.

In practical scenarios, the phone functions as follows: saying, "Book me a high-speed train ticket to Shanghai next Friday and a hotel near Hongqiao Station" can autonomously complete the entire process of searching, comparing prices, and placing orders. Saying, "Summarize what was discussed in this Feishu group today" can read messages, generate summaries, and forward them to specified individuals. Pointing at a product and saying, "Help me find the best price" can identify the screen content, compare prices, and provide purchasing recommendations.

Compared to last year's engineering model, the biggest changes in this generation are not about being "smarter" but about becoming "more like a commercial product," with three main improvements:

The first is positioning. The previous Nubia M153 was a technical preview version aimed at industry insiders and early adopters, with a limited release. This generation is the consumer version, starting with 512GB storage, four color options, and following a mass-production retail strategy.

The second is rules. The previous generation relied on "screen reading" to simulate operations directly, but it faced significant backlash from app developers just one day after launch.

This generation introduces a comprehensive set of rules, launching the Screen Automation Operation Declaration Protocol (SAEP) and initiating a 30-day rule announcement period. Third-party apps can decide whether to allow AI to perform screen operations. During the announcement period, the Doubao Assistant will only operate system apps, ByteDance-owned apps, and explicitly consenting third-party apps. For sensitive actions like payments, fingerprint button confirmation is required, and agent permissions are tiered.

The third is commercialization. The first-generation Doubao Phone was priced at 3,499 yuan with only 30,000 units available. The new generation starts at 5,499 yuan for mass production, with inventory scaled up to around 200,000 units.

Overall, this generation essentially repackages what worked technically but faced ecological barriers in the previous generation into a commercially viable product.

02 The First Generation's Setbacks Set the Stage for the Second Generation's Actions

The second-generation Doubao Phone focuses on "rule-setting" largely due to the setbacks faced by the first generation.

The first-generation Doubao Phone gained overnight attention thanks to its aggressive screen-reading technology, allowing the agent to complete cross-app tasks by "seeing" the screen, clicking, and inputting like a human, without needing open APIs from apps.

The advantage of this approach was immediate results without relying on ecosystem cooperation. However, the drawback was equally clear: it bypassed app developers' authorization, making conflicts almost inevitable as it scaled up.

In December last year, just one day after the first-generation engineering model went on sale, users reported that their WeChat accounts were flagged for "abnormal login environments" when using Doubao to operate them, with some accounts forcibly logged out or temporarily frozen. Frequent issues like human verification, crashes, and forced logouts occurred in Alibaba-owned apps like Taobao, Xianyu, and Damai. Apps from Agricultural Bank of China and China Construction Bank suspended logins and payments due to "risky environments."

These countermeasures had multiple reasons. When Doubao operated on behalf of users, users no longer opened these apps themselves, leading to lost ad impressions, in-app traffic, and user data. More seriously, there were risks of privacy breaches, and platform risk controls aimed to uphold the principle that "users must operate apps themselves."

The second generation responds with the SAEP protocol and a 30-day announcement period, allowing apps to declare whether they allow AI operations. "Whether operations are possible" is no longer solely determined by Doubao but requires confirmation from both user and app authorization.

In fact, this restraint is not entirely Doubao's choice. On July 15 this year, the Cyberspace Administration of China publicly released, for the first time, Record Filing Information (filing information) for generative AI services on mobile devices from seven manufacturers: Apple, Huawei, vivo, OPPO, Xiaomi, Samsung, and Nubia. What AI can and cannot do on phones is now being brought under regulatory frameworks. The Doubao Phone's "rule-setting" reflects both passive corrections after hitting walls and proactive adaptation to tightening industry regulations.

However, despite improvements in user privacy and security, several issues remain unresolved:

First, ecological boundaries persist. Currently, the Doubao Phone's voice operations primarily cover ByteDance's own products, with only a few third-party apps like Caocao Chuxing integrated. High-frequency apps like WeChat, Taobao, and Alipay remain inaccessible. An agent that "gets things done" faces an obvious ceiling if it cannot handle the most common tasks.

Second, success rates are a weakness. While manufacturers claim an end-to-end task success rate exceeding 80%, this figure comes from scenarios they themselves define. Looking at the broader intelligent agent industry, the China Academy of Information and Communications Technology, citing OSWorld evaluation data at the World Artificial Intelligence Conference, noted that the general success rate for agents performing long-chain tasks exceeding 500 steps is only 18%.

Third, whether "one-sentence errands" are a rigid demand remains unanswered. The first-generation engineering model's hype in the secondary market quickly faded after launch, and how many of the 370,000 reservations for this generation will convert to actual sales remains to be seen after the initial launch.

03 Four Approaches, Each Betting Differently: Consensus Is Wavering

If we view the Doubao Phone merely as a new product, its sales performance would only affect ZTE's quarterly financials. However, this phone represents more: it is the first time the "AI agent smartphone" category faces market scrutiny in a mass-produced form. Beyond it, several other manufacturers are betting on different directions at the same time.

Honor pursues an "embodied" approach. Its Robot Phone, released on August 12, features a four-degree-of-freedom titanium alloy gimbal inside the body, allowing the camera to rotate, track, and even nod or shake its head. Priced starting at 9,999 yuan, Honor's answer is to give the phone a "body," letting AI move from the digital realm on the screen into the physical world.

Jieyue Xingchen takes a "native" approach. Its STEPX Neo runs on a self-developed intelligent agent native operating system, Step AOS, which reconstructs the system from the ground up. It includes a personal intelligent agent, Amoo, and an end-side model, StepEdge, along with an interactive secondary screen on the back.

The STEPX Neo is currently the only intelligent agent phone to pass the Intelligent grading L3 Test (Intelligence Level 3 testing), with ecological partners including Meituan, Alipay, Didi, and Ctrip. However, the phone has not yet provided a mass-production timeline, and officials admit they have not pre-determined shipment volumes.

Huawei continues its self-developed closed loop. Leveraging the synergy of its Kirin chips, Pangu large model, and HarmonyOS, the HarmonyOS 7, released in September, upgrades Xiaoyi into a system-level intelligent agent capable of perceiving over 200 system-level user data points, invoking over 2,100 system capabilities, 2,000+ HarmonyOS intelligent agents, and 500+ partner-selected skills. Huawei's confidence in this approach stems from controlling everything from chips to systems in-house.

These four approaches offer four different answers to "What should an AI phone be?" Disagreements are not necessarily bad; they indicate that no standard answer exists yet and that no one has truly succeeded. Who is right or wrong remains unjudgeable at this moment.

What these four approaches reveal is a fork in the evolution of smartphones. Integrating intelligent agent capabilities into phones is a settled direction; the real suspense lies in whether this will upgrade "smartphones" to a new generation or spawn an independent new category.

This suspense has no immediate answer because the industry has not yet figured out "what an AI phone should be."

For over a decade, smartphones have had a nearly uniform form factor: a touchscreen and a grid of app icons. Now, intelligent agents have become a new entry point beyond icons, but manufacturers are betting on vastly different forms for this new entry point.

The same is true at the standards level. In May this year, the Ministry of Industry and Information Technology implemented the national standard 《Intelligent grading of artificial intelligence terminals》 (Artificial Intelligence Terminal Intelligence Levels), dividing terminal intelligence into four levels from L1 to L4. In July's first batch of tests, only 66 products from 17 companies reached L3 (assisted level), with none yet touching the highest L4 (collaborative level).

The market is also evolving. According to Counterpoint Research's June 2024 forecast, generative AI-supported phones will account for 45% of global shipments by 2026 and 52% by 2027. The market is growing, but how it will be divided remains undetermined.

Ultimately, the answer must come from users. Previously, what drove users to upgrade phones was better screens, faster chips, and improved cameras. As these three dimensions reach their limits, AI intelligent agents are hoped to become the new reason to upgrade.

Whether this hope will be fulfilled depends on two things: first, whether intelligent agents can transition from novelty features to daily necessities, and second, whether they can disrupt the ecosystems of super apps. The former determines if users are willing to upgrade for it, and the latter determines how much it can "get done."

Currently, the Doubao engineering model proved "technically feasible." Now, it is the market's turn to prove "it's worth buying." The Doubao Phone has submitted its answer sheet, and the ones grading it will be every real user in the coming months.

* Image sourced from the internet. Please contact us for removal if infringement occurs.