The technological landscape is witnessing a seismic shift as artificial intelligence (AI) continues to penetrate various industries. This shift is especially prominent in consumer electronics, a sector that has long been waiting for an AI hardware renaissance. In the past few years, mobile phones and personal computers (PCs) shone brightly at the center of this field, drawing in most of the technical resources, including AI chips, architectures, and software ecosystems that were predominantly built around these two devices.

However, a new era is dawning. AI is beginning to "escape" from the confines of these central devices, making its way into accessory domains that have long been overlooked by the industry. Devices such as headphones, mice, and keyboards—once considered mere supporting roles—are now evolving into new species of AI hardware. This transformation is fundamentally altering public perception regarding AI hardware.

This change is not merely a coincidence; it largely stems from a decentralization of AI hardware. Various AI firms that have attempted to seize a slice of the hardware market are playing a significant role in propelling this movement forward.

Advertisement

Nevertheless, current iterations of "AI mice" and "AI keyboards" have not yet delivered strong performances; many offerings are more about hype than substance. It's easy to dismiss these as simple marketing gimmicks, yet it is undeniable that the buzz surrounding these concepts may be sketching the outlines of AI accessories. As hardware and AI companies begin to address their shortcomings collectively, a decentralized era of AI hardware is sure to follow.

The trend toward decentralized AI hardware can be seen as a strategic maneuver by AI firms looking to penetrate the hardware space. While these companies may excel in algorithms and software services, they generally lack hardware expertise. Instead of attempting a direct entry into the highly competitive smartphone and PC arenas where large corporations dominate, a more pragmatic approach has emerged—focusing on the accessory market.

The development of hardware is no easy feat; it's a challenge marked by high barriers to entry and significant capital investment, especially concerning PCs and mobile devices. Companies need to navigate patent landscapes successfully, invest heavily in supply chains, brand development, and channel distribution.

Given that the markets for PCs and smartphones are already saturated and experiencing slow growth, jumping into these domains is akin to throwing stones at armored vehicles—a highly risky endeavor with a protracted return on investment. Consequently, many AI firms are opting instead to partner with established manufacturers in the PC and smartphone segments. For example, the much-discussed AI smartphones are backed by firms like Baidu, ByteDance, and MiniMax.

While competition for quality customers is fierce in the smartphone market, the pool of premium clients is limited, leaving AI companies at a disadvantage when it comes to negotiating power. Compared to challenging established markets head-on, entering the accessory domain offers a lower-cost, more operationally feasible route. Components such as mice and keyboards frequently evolve, and the barriers to entry are not as daunting as those associated with smartphones and PCs. Brand loyalty among users is also not as established. Take keyboards as an example: following the expiration of CHERRY's patent protection in 2014, the mechanical keyboard market exploded as new players entered the fray each year.

Another compelling factor driving the current enthusiasm in the AI accessory market is the relatively low cost of experimenting with these products. This dynamic can be perceived as a war of “periphery penetrating the center.” Discussions surrounding AI hardware are commonplace within industry circles, yet they remain unfathomable for average consumers. For an everyday user, the idea of immediately purchasing an AI smartphone or PC feels overwhelming, akin to expecting someone to devour an extravagant Chinese banquet without proper preparation.

Surveys conducted by Photon Planet reveal that although a wave of AI smartphones has already hit the market, in-store visits from potential customers intrigued by AI are sparse. As a result, many brand representatives are hesitant to proactively promote AI features. This indicates that until consumers have established a solid understanding and habit around AI consumption, attempting to sell them costly AI smartphones or PCs could backfire.

Before now, exposure to AI for most consumers was largely through AI applications available in app stores. AI accessories, with their more affordable price points, present an alternate means for users to engage with AI—effectively allowing AI firms to strategize a circumvention approach by making their products accessible. Moreover, while AI is often just one functional module among many in main devices, it takes center stage in accessories, bestowing new value propositions and enhancing user interaction experiences. This is particularly true when AI capabilities are marketed as indispensable selling points.

Observational data from Photon Planet show that in the review section of a specific brand’s AI mouse on Taobao, numerous novice users expressed excitement over features like AI-generated content and PPT production. Although the underlying AI model might not be top-tier, for these “novice users,” this mouse represents an introduction to AI-assisted work. Consequently, AI peripherals can serve as an effective vehicle for promoting AI hardware and validating market demand.

Even before the surge of large models boosting AI hardware's relevance, several AI accessories had already undergone quiet iterations. For instance, iFlytek, a player rooted in the smart voice sector, launched voice-enabled mice and keyboards around 2020—early attempts at the “smartization” of PC accessories. Although the idea of speaking commands to control peripherals might invoke memories of comical scenes from films like Stephen Chow's “The God of Cookery,” these products aimed to cater to slow typists using various input methods that appeal to those not familiar with advanced typing.

As the AI wave grew more intense, manufacturers’ ambitions in hardware quickly expanded. iFlytek led the charge by introducing several AI learning tools, aiming to inject new life into their older voice-enabled products, hoping to ride the coattails of the burgeoning AI trend.

Take the AI mouse AM50 as an example. Beyond its traditional features like speech-to-text and translation support, it now includes popular AI functions such as smart interaction, automated writing, presentation generation, and image generation, bundled into its drivers to entice users to spend.

This approach primarily targets novice users since seasoned AI practitioners are likely aware of more specialized and mature applications for drawing and writing. Therefore, iFlytek’s AI accessories focus on educating the market while simultaneously fostering a symbiotic relationship between hardware and software to ultimately spur the growth of the AI ecosystem.

Contrast this with ByteDance’s AI headphones strategy, which has taken a “software drives hardware” approach. ByteDance aimed for the XR market and even acquired the Pico headset from Tencent. However, the XR domain remains inadequately validated, leading to Pico's eventual discontinuation due to slow progress.

Learning from past setbacks, ByteDance pivoted back to hardware early this year by acquiring headphone manufacturer Oladance to launch the AI headphones, OlaFriend. The key feature of these headphones lies in their seamless integration with Doubao, enabling hands-free interactions. Doubao, a leading domestic AI application, has attracted considerable heavy users. Therefore, the high compatibility between ByteDance’s AI headphones and Doubao could draw in new consumers, promoting sales of these AI hardware offerings through established AI applications.

At present, devices like mice and keyboards have seen considerable optimization over years, reaching a level of core functionality comparable to the simplicity of chopsticks. The performance metrics for interaction devices like these—including response times, accuracy, and DPI—have already exceeded human limitations. Except for branding, there remains little room for improvement.

As innovation stagnates, the cycle of relentless competition can further reduce profit margins. The remaining large manufacturers leverage established supply chains and brand equity to maintain profitability, while newer entrants often get caught in a cycle of “over-competition,” limiting their profit margins. The narrative has shifted to one where established companies can afford to be complacent, while smaller players feel compelled to test the waters.

Even Finalmouse, which stirred interest through minimalist designs, has relied on collaborations with streetwear brands to maintain relevance; custom keyboards have become a path some manufacturers follow to counter price wars by producing bespoke options to increase their edges. AI peripherals may serve a similar role, enabling companies to push up prices—a disguised “intelligence tax,” if you will.

However, the current surge of AI accessories does not come without pitfalls. Many companies appear ensnared in a common misconception: an overemphasis on AI functionality often overshadows the fundamental quality of the hardware itself.

Industry insiders have noted instances where AI peripherals released by firms like iFlytek lack the basic functionalities found in standard devices. One particular voice mouse displayed a mere addition of a microphone circuit compared to traditional models, while the buttons remained unchanged, not transitioning to more advanced mechanisms seen in contemporary high-end peripheral devices.

Additionally, however insightful the AI features may be, user feedback indicates that the left-click button can sometimes fail—echoing the sentiment that a "voice mouse" must still perform competently as a mouse at its core. Given that iFlytek has updated their products from voice-focused models to AI-enabled peripherals, complaints regarding sensitivity and performance still echo in the comments on their flagship store. Users frequently report that these devices require special surfaces to function optimally and that errant recordings often pick up background voices due to poor microphone directionality.

As one examines an array of AI accessories, a similar trend emerges: a tendency to prioritize software over hardware integrity. Consider, for instance, a certain AI headset whose AI functionalities are aggressively marketed, but lacks the acoustic quality found in its price range among professional-grade options.

On the flip side, while peripheral manufacturers possess considerable technological insights, they often lack proficiency in AI capabilities themselves. For example, Sonic Hardware recently rolled out the ALTO KEYS K98M keyboard featuring integrated AI functionalities, allowing users to activate AI applications with a single press. Reports suggest that this iteration can access tools like ChatGPT, or, domestically, Wenxin Yiyan. This indicates a trend of collaborative evolution in the AI accessory landscape. Instead of working in isolation, AI companies and hardware makers are starting to come together to complete the puzzle.

Ultimately, AI's role in hardware must not be used as a shield to obscure poor quality. Misguided AI peripherals that focus more on feature gimmicks than essential hardware integrity may risk disappearing amid market evolutions. Although the current wave of AI accessories may seem widespread and progressive, it has yet to find true grounding in the consumer mindspace. To narrate the story of AI accessories compellingly, a deeper relationship between software and hardware development must be fostered, rather than simply piling on concepts.