Investors are starting to look for AI projects on Bilibili and Xiaohongshu.

null

Author: Investment界

"In the AI era, all hardware deserves to be redone."

This statement may still sound somewhat radical in 2024, but by 2025, its value will sharply increase— as we see, a wave of AI smart hardware startups is rising, funding is pouring in, and the latest valuation of the smart ring Oura Health across the ocean has soared to an astonishing 70 billion yuan.

A fiery scene, sweeping investors with FOMO. This time, besides reviewing over a thousand business plan emails, they surprisingly unanimously focus their deal sourcing on content communities where geeks and young people gather.

The underlying logic is also quite simple: compared to parameter sheets and roadshows, what exposes the real demand for a consumer-grade AI smart hardware earlier is whether this product can be understood, discussed, and questioned in public.

Ultimately, all those tech products aimed at the masses must go through a crowd to test them.

AI hardware explodes, investors scramble for underwater projects

We are experiencing a big year for AI hardware.

Since 2025, extending from the wearable device track, more segmented smart hardware products like AI glasses, AI toys, AI recording cards, AI rings, AI headphones, companion robots, Agent Box, and others have collectively ushered in a boom.

Amidst the fervor, some details reveal the excitement in the investment circle: in the second half of 2025, we repeatedly see top institutions issuing hero posts recruiting AI/smart hardware investors, with job descriptions almost all requiring a base in hardware capital, Shenzhen.

Good projects are almost all fighting to be "visited three times," especially those still underwater and with lower valuations. Investors are even starting to target potential stocks that haven't yet resigned, for example, the coffee shop next to DJI is crowded with VC and FA urging people to start a business.

(Comments in the pre-sale video of the AI sticky note product)

"After this content was posted, many technical discussion directions appeared in the barrage and comment sections, and more users began sharing their similar needs: some mentioned they often participate in cross-border exhibitions, some talked about communication barriers when working remotely with overseas teams, and others shared experiences of using translation tools that are not convenient enough when accompanying foreign friends and relatives for medical treatment or handling foreign-related affairs."

"These feedbacks not only helped us expand the influencer material directions but also directly influenced our thinking about the product line," said Wu Zhen.

(Comments in the video comment section of UP主影视飓风)

Gathering Enthusiasts

Remember January 2024, CES.

A small orange box called Rabbit R1, branded as "the world's first AI hardware device," sparked a frenzy in the tech circle, hailed as the iPhone of the AI era, and sold 100k units upon launch.

But the subsequent results were not successful. On the contrary, doubts about "being less functional than a phone in actual use" kept emerging. At that time, the once-dominant Rabbit R1 failed to clearly explain to users: what exactly can Rabbit R1 do now and in the future? What real problems can its innovations solve?

Even today, user education remains the most important topic for AI smart hardware—when AI enters consumer electronics, the product consumption logic is no longer limited to "improving life efficiency," but also includes higher-level needs like "releasing personal creativity" and "freeing the brain." At this point, whoever can help users understand, discuss, and pay in real scenarios truly crosses the threshold from tech product to consumer product.

But this is not easy. Regarding user education for AI glasses, INMO's CMO Wu Zhen frankly told us: this is not simply transmitting product information; it also needs to dissolve a psychological barrier—that is, users must first complete a "try-on" in their minds, confirming that integrating this into daily life feels natural and unobtrusive, before moving to the next step.

In Wu Zhen's view, the process of psychological building for AI smart hardware users may take longer than most consumer electronics categories. "When one day users naturally wear AI glasses out, without reminders like 'I need to use AI glasses today,' this behavioral change could be more convincing than many data points."

Shuffle begins, attention war heats up

As enthusiasm peaks, reshuffling is quietly underway.

The market has now reached a consensus: 2026 will be the year for commercial validation of AI hardware. This also means that those chasing the trend and frenzied capital will need to become more rational and calm.

A scene of the big wave sifting is unfolding: at the end of 2025, the AI wearable device Friend AI Necklace faced severe market resistance and stalled; after peaking in 2024, Rabbit R1's negative feedback about product capability led to high return rates, serious reputation decline, and ultimately, issues like unpaid wages and cash flow exhaustion.

Meanwhile, smart players are beginning to cautiously adjust and optimize their decisions. In February 2026, market rumors indicated that the Doubao AI glasses project was paused, and at least for the foreseeable cycle, this product line would no longer be considered a direction to pursue; earlier in January, vivo's AI glasses project was halted, citing difficulty in differentiation.

All these reflect industry segmentation.

But undeniably, this trillion-yuan value track still offers entrepreneurs enormous certainty opportunities. According to AICC forecasts, by 2030, the global AI-related hardware market will easily surpass trillions of dollars. In China, the industry predicts that by 2026, the market size of consumer-grade AI hardware (excluding phones and cars) will break through 1.27 trillion yuan, reaching 2.56 trillion yuan by 2030.

Opportunities for withdrawal and entry coexist now. OpenAI just announced it will launch its first AI hardware device in the second half of the year, and Meta plans to double its AI smart glasses annual production capacity to 20 million units by the end of 2026. Domestically, a batch of AI smart terminals like JD's AI desk lamp, AI cooking machine, AI mattress, and AI wheelchair are brewing for new sales.

People still firmly believe that the entrance battle behind AI smart hardware will not stop; the AI era remains a user sovereignty era.

The wave will push everyone forward. But when attention, product strength, and user cognition begin to compete on the same table, the final contest has just begun.