This article from WeChat public account: Reflections (ID: zuomoshi) presents a comparison between AI-related conferences and Zhao Benshan’s sketch “Selling Crutches.” Reflecting on the development process of the metaverse, one can draw parallels to the process of “selling crutches.” It involves combining a sci-fi concept with advancements in new technology, packaging it, amplifying the output power to promote various self-justified stories, and then seeing Mark Zuckerberg playing the role of Fan Wei by selling crutches (Facebook changing to Meta), but where is the metaverse in all this?
AI conferences also seem to have a similar meaning: lots of imagination but minimal application. If AI is indeed as powerful as the internet, essential for daily life, where are its super applications?
The fatal mindset lies in viewing applications from a technical and strategic perspective, as these perspectives often fail to identify super applications, making it challenging to recognize useful applications. The common-sense issue here is that no matter how advanced AI becomes, it is not a mirror image of the entire world; super applications are destined to emerge at the intersection of the world and AI.
To create new super applications, one needs an outward-looking perspective that spans two domains. However, the technical perspective, focusing on AI’s possibilities, often leads to a top-down strategic approach, typical of research institutes. While the first principle is a starting point, it can mislead individuals into thinking they understand the essence of the world, which can hinder progress.
Reflecting on the PC and internet era, one can see how Xerox’s research institute invented the graphical user interface, later transformed into the foundation of two trillion-dollar businesses by Jobs and Gates. Alibaba and Tencent then successfully implemented the internet infrastructure in China. Conversely, Facebook’s early success was due to a low-tech college social product, leading to strategic missteps in the app selection and acquisitions that ultimately caused setbacks.
To find AI’s super applications, one must focus on experiencing new technologies firsthand, considering what they can truly accomplish, and creating a feedback loop of experiences.
Returning to the scene and establishing a personal testing ground reveals numerous new possibilities. By envisioning what was previously impossible but now feasible, like replacing human teachers with AI in products such as VIPKID, one can explore the potential of AI applications further.
Innovating in real-world scenarios, such as developing an AI-generated storytelling platform for children, requires attention to detail and seamless interaction, rather than dwelling on past product successes. By reshaping roles and leveraging imagination, new super applications can emerge, offering a more practical approach to AI development.
In conclusion, AI’s shift from a device-centric to an intelligence-centric model necessitates a focus on precise role allocation and intelligent-centered applications to navigate the evolving landscape effectively. By emphasizing the concept of “intelligent value density,” one can measure the richness of AI’s capabilities in different scenarios, ensuring a more effective and impactful AI implementation.
By adopting a methodology that transitions from specific to general and vice versa, AI product development can strike a balance between hands-on experimentation and strategic planning, ensuring a comprehensive approach to product innovation and deployment. Ultimately, returning to the essence of roles and experiences is crucial for successful AI applications, steering clear of blind faith in technology and focusing on tangible outcomes.
