Selected Excerpts
ToI
If you have novel ideas, you can promptly receive responses from like-minded partners in the DAO community and establish decentralized credit endorsements through feedback among individuals. Platforms like Taobao, Douyin, and WeChat provide individual sellers with abundant user connections and channel tools, and these platform-level data feedback streams rapidly reflect market changes. The proliferation of open-source tools on GitHub and SaaS enables individual developers to quickly access powerful resources, allowing them to discover feedback signals for niche needs in more decentralized and independent roles, avoiding external dependencies due to insufficient capabilities and protecting the most valuable independent feedback sources. The development of the "Toi" (Toindividual) ecosystem is also accompanied by numerous success stories.
Gary Brewer founded the website BuiltWith, which operates with only one person in the entire company and earns over $14 million annually. This website can accurately identify the specific technologies used by other websites. For example, for a given website, A/B testing uses Optimizely; search uses Algolia; marketing automation uses ConvertKit...
BuiltWith can detect over 61,105 technologies, covering 673 million websites globally. Leveraging the power of big data can create value, such as providing "Internet Technology Development Trends," which allows viewing technology trend data from 2000 to the present. Taking "Popular Web Technologies in Chinese Websites" as an example, this page includes 28 subcategories such as "Marketing Automation," "E-commerce Hosting," and "Enterprise Website Management Systems."
Discoveries in Nobel Prizes
Research on Nobel laureates revealed a simple rule (the Ising model) that can give rise to new order: in a network of adjacent nodes, chaos emerges if each node is self-oriented; if nodes are externally oriented and learn from neighboring nodes, they tend toward order. Moreover, the evolved structures are non-uniform across different scales. For human society, this rule may manifest as a cultural and behavioral tendency, similarly triggering different social orders and structural evolutions. Just as the emergence of a well-ordered society can stem from the widespread adherence of each individual to the simple rule of "respect those around you."
Gradual punitive measures. Initially, if someone violates a rule, we can remind them very kindly. If that doesn't work, gradually escalate the punishment. Conflicts should be resolved fairly and quickly, without allowing them to escalate. Groups should have a certain degree of autonomy. Local matters should be decided locally, and so-called authorities outside the group should not dictate how the group should act. The principles mentioned above can be applied at various scales. We are talking about small groups composed of individuals, but large groups composed of small groups can also operate this way, as long as each small group is treated as an individual, with representatives participating in the affairs of the larger group.
End Feedback: The First Point of Contact for External Survival Pressure and Opportunities
Amazon has a rule personally established by Jeff Bezos: if a product or promotional rule receives more than two similar complaints, frontline customer service personnel have the authority to press the "red light" button, directly removing the product or promotional rule from the shelves. This method forces relevant personnel to solve the problem immediately. This "Andon system" originates from Toyota's lean production model, where frontline operators have the right to stop the entire assembly line upon discovering any abnormality, exposing and resolving the issue.
A former vice president of Amazon China once modified the "Andon" process, shifting the "Andon authority" of frontline personnel upward, requiring approval from a supervisor for removing more than two products and from a manager for more than five. The Senior Vice President of Global Operations at Amazon scolded him severely, considering it outright formalism. He believed that managers sitting in offices with air conditioning and coffee would not understand what was happening better than the employee on the phone chatting with a customer. Making the "Andon" process so complex prevents frontline employees from quickly and effectively helping users solve problems. Even if customer service personnel press the wrong button, they are not held responsible; the responsibility lies with management, who must find ways to solve the problem from the perspectives of recruitment, training, and systems.
Amazon's philosophy is:
First, provide employees with the right tools to solve problems;
Second, rely on the correct judgment of frontline employees. Only by combining these two can the good intention of "customer-centricity" be implemented into concrete actions through mechanisms.
Implicit Characteristics in Long-Cycle Feedback and Indirect Feedback
People perceive long-cycle feedback processes as composed of too many small uncertain factors but overlook that the combination of these small uncertain factors exhibits another kind of deterministic order at a higher level, akin to the short-term market's technical volatility and Warren Buffett's long-term value investment foresight. In reinforcement learning algorithms, such problems are referred to as "sparse reward" problems or long-term credit assignment problems in training environments, which involve how to allocate contributions of system members to outcomes in complex learning systems. Biological intelligence relies on hierarchical cognitive mechanisms to solve such problems, and hierarchical reinforcement learning is also drawing inspiration from this approach to address them.