Summary of Key Points
This article discusses the organizational changes brought about by the AI era: whether it's Silicon Valley tech giants giving up management roles to become frontline engineers or domestic companies splitting off independent AI departments and promoting small-team collaboration, all point to one trend—the traditional organizational logic centered on "managing the number of people" is being replaced by "intelligence density" (talent + computing power). AI has lowered the barriers to innovation, broken down job boundaries, allowing small teams to create significant value, and redefined the core value of engineers (problems that AI cannot solve, human understanding, and decision-making direction).
1. Silicon Valley Giants "Downgrading": The Standards for Career Success Have Changed
In the past, the career path in Silicon Valley was "work on technology → advance to management → manage more people." However, now a group of leading figures are doing the opposite: Workday's CTO resigned from his executive position to become an ordinary engineer at Anthropic; Tesla's former AI director, Karpathy, left management to get closer to AI models. Why? Because in the AI era, "how much intelligence can be utilized" is more important than "how many people can be managed." Just as being a factory manager was once prestigious, now those who can master AI tools and directly participate in model development are the most sought-after in the industry—the measure of career value has shifted from "management scope" to "technical depth and AI application capabilities."
2. Innovation No Longer Comes from "Top Down": From "Must Do" to "Most Desirable"
In traditional large companies, innovation followed a process of "superiors setting directions → business departments submitting requirements → development teams scheduling projects," which was long and far from user needs, resulting in projects that were merely necessary rather than what users actually wanted. But the AI era is different: Kimi, the intelligent assistant from Moon's Dark Side, was created by a few engineers as an internal project; OpenClaw, a globally popular intelligent agent, was developed by one person alone. A contestant from Ant Group's Hacker松 said, "Previously, fixed teams had narrow perspectives, but now by bringing together people from different roles, we can solve real problems from various angles." AI has reduced the technical barriers to innovation, allowing frontline personnel closer to users to take action directly, turning innovation from top-down design into grassroots-driven initiatives.
3. AI-Native Organizations: Small Teams Can Achieve Great Things
In the AI era, companies don't need to accumulate large numbers of employees; they need to focus on "intelligence density" (talented individuals × powerful computing power). For example, DeepSeek's core team consists of only a hundred people, Moon's Dark Side has always had over a hundred members, and although OpenAI has thousands of employees, its core research teams are still small (referred to as Pods). Large companies are also adopting this approach: ByteDance's Seed and Alibaba's ATH have independent AI departments that report directly to their bosses; Ant Group has launched "AI Builder Pods"—3-5 member cross-functional teams that make their own decisions without the need for dedicated product managers or designers, relying on AI engines and universal architectures to support the entire company's operational needs. Small teams are more flexible and can iterate faster, making them more suitable for AI-driven innovation.
4. Job Boundaries Are Dissolving: Non-Technical Personnel Can Also Innovate
AI has broken down professional barriers: Algorithm engineers used to have no idea how engineering could utilize AI, but after cross-team collaboration, they suddenly understood; non-technical staff can also use AI tools to innovate. For instance, at Ant Group, people from different departments work together to solve problems using AI, eliminating the need for distinctions like "this is a product issue" or "that's a development issue." AI acts as a "universal adhesive," bringing previously separate roles together, allowing everyone to contribute to innovation—innovation is no longer limited to technical teams; anyone who can identify problems can try to solve them with AI.
5. AI Does Not Replace Engineers, but Defines Their Value
Some worry that AI will lead to job losses for engineers. This is a misconception. Ant Group's practice shows that while AI can generate code, it cannot determine "what problems users need solved"; it can optimize processes, but not whether they are user-friendly; it can speed up iteration, but not judge the correctness of technical directions. The value of engineers lies in things that AI cannot do—understanding human behavior, defining problems, and setting directions. AI is a tool, not a replacement; it frees engineers from repetitive tasks to engage in more valuable creative work.
In conclusion, the article emphasizes that in the AI era, the most valuable asset in an organization is when employees' first reaction to a problem is "can I try using AI to solve it?" This atmosphere of "everyone can innovate" is the essence of an AI-native organization.