When LABUBU rejected AI, only the narrow door of "relationships" remained open for the toy industry.

       Recently, a picture of a refrigerator priced at 5,999 yuan has been widely shared on social media. This is not a discussion about appliances, but rather a celebration of desires. A joint refrigerator launched by the IP LABUBU, a trendy toy brand, was once quoted at an astonishing 100,000 yuan on the secondary market. In the face of such an absurd premium, the first reaction of most people was to scoff: it's just a bubble, a hype, or an intellectual tax.

       However， if you only see the price and only see the frenzy of capital， you actually miss the most fatal signal of this wave. At a time when AI toys are being hyped and regarded as the next trend by capital， why did an IP with top traffic not launch "AI + LABUBU"， but instead choose an ice-cold appliance?

       This choice itself is a deep-water bomb aimed at the industry.

I. Clear Retreat： Not because they don't understand， but because they have seen through it

       Many people instinctively interpret it as： Is Poppa Mada not optimistic about AI toys？ Is it because they think the technology is too abstract?

       The truth is exactly the opposite. Companies like Poppa Mada， which have top-level data observation capabilities， are more aware than anyone of the popularity of AI. They did not do it because they couldn't understand， but because they understood it too clearly.

       They have seen the true face of AI toys today： It is not a mature product， but an experimental item full of uncertainties. The experience is unstable (sometimes smart， sometimes stupid)， the cost structure is heavy (the cost of local computing power and cloud invocation is high)， and the user relationship is difficult to sustain (it becomes obsolete after three days of novelty). In terms of business logic， it is far from forming a "definite business".

       While Poppa Mada has completed the original accumulation of IP assets， its current strategic focus is "maximizing monetization efficiency". Therefore， they chose a more stable and proven path： using refrigerators and small appliances to physically occupy users' living space； maintaining scarcity through limited editions and lotteries； and creating emotional consensus through the attribute of social currency.

       This is a perfect battle about "inventory operation". And currently， AI toys still cannot fit into this precisely operating system.

II. Illusion of the Window Period： The Only Opportunity Left for AI Toys

       LABUBU launching a refrigerator does not mean that AI toys have no chance. On the contrary， this leaves an extremely crucial but also extremely short window period for the AI toy industry.

       This window period is not for you to rush (because you can never catch up with the technological iterations of big companies)， but for you to seize something more fundamental： the "first emotional bond" between users and characters.

       We must clearly recognize that once a giant like Poppa Mada solves the technical stability problem of AI， when they enter this field with already formed IP， deep user trust， and clear character cognition， simply "functionally better" or "cheaper" will be completely uncompetitive. What can you use to compete against a "old friend" that users have loved for five years?

III. Misaligned Competition： AI Toys Should Not Compete on Function， But on "Relationship"

       Since direct confrontation is not possible， what should the AI toy team now compete for？ Many teams are still addicted to parameter competitions： who has more accurate voice recognition， who has larger model parameters， who has a cooler hardware form.

These are tactical diligence， but strategic laziness. What really needs to be competed for is only three things：

1. The moment of "the first emotional bond".

       The moment when users have this feeling： "This thing seems to really care about me." This moment is more important than any dazzling function. Once the emotional bond occurs， users will tolerate your imperfections， will continue to pay for your evolution， and will generate high substitution costs. This is the starting point of all long-term value. 

2. Focus on "relationship depth", not on the number of users.

       There is a fundamental difference between AI toys and traditional hardware. Selling is not the end; it's the beginning of a relationship. So, stop focusing on 100,000 downloads. What really matters is: How many people actually have a conversation with it every day? How many people experience emotional fluctuations due to its "joy, anger, sorrow, and happiness"? And how many people regard it as an "existence" in their lives, rather than just an object?

3. Grab "role definition rights".

       In the future AI toy market, it won't be a competition of products, but a competition of roles. Just like in the world of martial arts: Who is the "one who understands loneliness the most"? Who is the "most clingy, most healing existence"? Who is the "even though sharp-tongued, the warmest buddy in critical moments"? These mental highlands, once occupied, will form extremely high barriers. And this is precisely what traditional IP companies are best at, and it is the only breakthrough opportunity for AI-native companies.

IV. The Forgotten Battlefield: Grab "Channels for Relationship Occurrence"

       When discussing product definition, we often overlook the transformation of channels. But here, if the "channels" still remain at product distribution in stores and e-commerce promotion, it would be too outdated.

       The essence of AI toys is not "selling products", but "allowing users to have a relationship with a role". This means that the truly valuable channels are not "places for selling products", but "scenarios where relationships occur".

       Looking back at the explosive path of LABUBU refrigerators, you will find that it did not suddenly explode with a single product. It repeatedly completed one thing through offline stores, limited edition systems, and exhibition spaces over a long period: Let users encounter it again and again. This high-frequency, low-pressure contact greatly reduces users' psychological barriers: From "knowing this role" to "starting to like it" and then "willing to pay a premium for it".

       Looking at the current distribution paths of AI toys, what is it? It's the comparison of parameters on e-commerce pages or the brainwashing of advertising. Essentially, this is forcing users to make purchasing decisions in a "completely unrelated" situation. This is simply anti-human nature. You want users to spend hundreds or thousands of dollars for a "stranger role that hasn't spoken or established an emotional connection"? It's too difficult.

       Therefore, the true meaning of "grabbing channels" is: Who gets the first chance to create the first real interaction between users and the role?

Such scenarios must have three characteristics: Experiential (able to touch and talk), with an emotional atmosphere (lighting, music, environment creation), and can be stayed (not just passing by quickly). That's why offline experience stores will become unprecedentedly important in the AI toy era. It is not a supplementary channel, but the core starting point. Only there can users let down their guard, be attracted by the role, say the first sentence, and establish the initial emotional connection. Once this step occurs, the subsequent online conversion and repeat purchases will be natural.

V. Conclusion: When IP Starts to "Speak"

       Returning to the LABUBU refrigerator incident. Its core lesson is not about how high a product has been priced, but that: The IP is breaking through the boundaries of categories and occupying users' lives in all aspects.

       Today it is a refrigerator, tomorrow it could be clothing, and the day after tomorrow, when the technology is mature, it will be that "silicon-based life" that can speak and accompany you crying and laughing.

At that time, these top-level IPs will no longer sell cold products, but a role that can accompany you for a long time. This is the true ultimate goal.

       So, if you are still thinking: "Should I hurry up and make an AI toy and seize the opportunity?" You may have gone astray. A more thought-provoking question is: When these top-level IPs with millions of fans finally start to speak, why would users still stay with you? The time for AI toy startups is truly running out. Either they must seize the "relationship" opportunity as quickly as possible under the shadow of the giants; or else they will simply be crushed by the downsizing trend and become just a footnote of history.