When digital employees become widespread, the winning strategies in AI entrepreneurship lie in “orchestration” and “aesthetics.”

Dialogue | Zhang Peng

As everyone rushes to develop “digital employees” and “Agent tools,” endlessly competing in niche scenarios, where is the true moat for AI startups?

Recently, GeekPark founder & CEO Zhang Peng and VisionFlow founder Liu Ye had a forward-looking discussion after the OpenClaw breakout. As a first-generation Chinese programmer born in 1979, Liu Ye has experienced the full cycle—from low-level hardware to software, from enterprise-level integration (ToB) to online education (industrial internet). After months of closed-door research and conversations with top global AI researchers and leading domestic entrepreneurs, he reached a cold conclusion: treating AI as a “digital employee” to replace individual tasks is an over-simplification driven by engineer thinking about real business.

In this conversation, Liu Ye introduced a series of insightful concepts and frameworks, such as “gradual exposure” and “task high- and low-dimensional matrices.” A clear future possibility emerged: AI’s next step is not flooding the market with tool-like agents, but building “digital organizations” with collaboration, reporting, and reflection mechanisms. When corporate culture becomes unnecessary and low-dimensional work is fully flattened, the future CEO may no longer be a “Chief Executive Officer,” but a “producer” with extreme aesthetic sensibility.

This is a speculative exploration of organizational forms, business barriers, and the ecological niches of new entrepreneurs in the AI era. The goal is to spark deeper future discussions among entrepreneurs.

Below is a curated excerpt from the GeekPark dialogue:

01 The battle for 10,000 A has begun, with too many possibilities,

But what truly matters is what to do

Zhang Peng: From Homework Box to today’s intense exploration of OpenClaw, what changes have you experienced personally?

Liu Ye: I am part of China’s first-generation programmers. I started coding young. I’ve seen the evolution from BASIC to DOS, then Windows, and now the Mac era, witnessing the rise of three major portals. I’ve worked in enterprise information systems, aiming to be China’s IBM; later, I shifted to Homework Box, deeply involved in online education. Online education is a profound industry, the highest form of industrial internet, and the “last train.” This experience made me realize that the core of industrial internet isn’t technology but industry itself—business. The pattern of industrial internet is: first, information matching; second, standardized products; third, supply chains; and finally, complex non-standard services. Margins grow higher as you go further, but it gets harder to do.

So, when the AI wave arrived, my first move was to spend nearly six months doing nothing but talking to all possible people in HR. From top startup chief scientists to core algorithm engineers and researchers at major tech giants, as well as emerging AI entrepreneurs—talking extensively, accumulating nearly a thousand hours of conversations. How deep? I could predict the second half of their sentences from the first half; everyone’s consensus was quite similar.

After this round of conversations, the conclusion was startlingly consistent: everyone is doing the same thing—digital employees. It reminded me of a strategic misjudgment by a big shot about cloud computing, who said Alibaba’s cloud was basically just a cloud disk. Using old frameworks to understand new things only reveals the shallowest layer.

Today, everyone thinks building a digital employee with Claude to generate “digital sales” or “digital customer service” is straightforward. Where are the technical barriers? The moat? When it becomes normal for someone to burn billions of tokens a day, it’s more like manufacturing—completely unscalable. So I ask every entrepreneur the same question: Why are you? What makes you capable? Are you younger? Smarter? Able to stay up late? Competing on one dimension— isn’t that just the difference between “10 seconds 69” and “10 seconds 70”?

Zhang Peng: Yes, today there are many possibilities, but what to do is the most important. Do you have any thoughts on this?

02 The decade of industrial internet, a replay today

Liu Ye: AI is very different, but I believe some patterns still echo those of the industrial internet. Early on, tools were developed; mid-stage, businesses; and finally, consulting. When technology is immature, the first wave is engineers—they excel at overly abstracting the world, like Baidu’s “frame computing,” which sees everything as frames. But the latter half of mobile internet is content and services, not frames.

Engineers tend to oversimplify organizational imagination. Look at the first-generation internet portals—Tencent and Alibaba thrived because they were closer to industry than pure tech. Today, the trend is the same: technology is becoming less important.

Zhang Peng: The current wave of liberal arts graduates is quite happy—seems like not knowing how to code doesn’t matter anymore. But in the long run, what are the evolving demands on people in the AI era? What changes?

Liu Ye: In China’s talent structure, I see a problem. The first-generation programmers were actually product managers because the role didn’t exist back then. The term “product manager” became widespread around 2010, after Jobs launched the iPhone 4 and Zhang Xiaolong proposed the product philosophy—then the phrase “everyone is a product manager” emerged. Before that, programmers also handled product management tasks; they were the first generation of programmers, learning to code not just for work but out of passion. These unconstrained, passionate individuals are often the most outstanding.

But the second generation of programmers, over the past decade, has been turned into “code farmers” by the industry internet; product managers became architects, and coders were tamed into not thinking about business. Now, AI has come—“coding” is being eliminated, and if they don’t evolve, they’re just “farmers.” These young people are excellent but lack industry understanding. So, the current “10,000 A battle” is fundamentally a flood of tools.

In the later stage of industry internet, companies like Alibaba and Meituan all employ top consulting firm (MBB) backgrounds for business analysis, with consultants leading product managers to design workflows because internet product managers are inherently lacking systematic thinking. Feishu was built this way. ByteDance, though purely internet, also heavily uses consulting firms to build internal processes. In the AI era, this pattern will only intensify.

03 The problem with organizations is never employees, but structure

Zhang Peng: So, you think focusing on “digital employees” as a single point isn’t very meaningful.

Liu Ye: My core judgment is: digital employees are not the end goal; digital organizations are. If digital employees flood the market and even recruitment positions disappear, and everyone has good digital employees, then what? Will the company be profitable and successful? Actually, all company problems are strategic and organizational, never just employee issues.

Today’s agents are still doing work for people, not making decisions. We internally revamped OpenClaw and created MetaOrg, which is essentially a core that can generate agent teams. When tackling any task, we don’t assign a single employee but build an “organization” to solve it. This organization has collaboration, reporting, mission, goals, and action modes.

Zhang Peng: But in the future, could one person be a department? Or even a company?

Liu Ye: That’s a very good question. We still focus on micro-tasks—for example, using AI to make a short video or write a document—requiring multiple rounds of dialogue. You say a sentence, it responds, then you give feedback. That’s tool-like use; it’s just very smart.

So, the concept of a person or department isn’t about quantity. When describing a senior role’s JD, it usually includes: first, capable of doing various tasks; second, able to use various tools. A high-level role can understand intent, proactively plan, execute, deliver, report regularly, reflect, and adjust strategies based on deviations. That’s advanced capability.

Zhang Peng: A competent department should be like “L4 level autonomous driving.”

Liu Ye: Exactly. When given a skill, it can complete complex tasks; with a system of skills, it can handle integrated complex tasks; when multiple intelligent agents are orchestrated, they can do even more complex things, like producing a short drama. I often tell employees: when using MetaOrg, don’t see yourself as a manager, but as a chairman. You need to push its boundaries.

In future startups, instead of a 500K seed, maybe it’s a token budget for trial and error. How many tokens you’re willing to spend determines how advanced the role can be. The higher the role, the longer the reasoning chain, the more back-and-forth iteration and reflection needed.

Zhang Peng: Returning to the earlier question, if an agent group can be broken down into finer units or roles, then the quality of each individual talent determines success or failure. This echoes the old business organization logic: talent density—higher talent quality makes core tasks easier to achieve and outpace competitors.

The core is: if in the future AI becomes omnipotent and we can call on the best AI, then besides business organizations providing different specialized services efficiently, we also need to look at “talent density”—your agents or bots being broken down to atomic capabilities increases talent density, leading to better results, efficiency, and innovation in complex tasks. Is this a correct deduction?

Liu Ye: I agree. Inside companies, there’s usually a department called OD (Organization Development). To assess if an organization can win, the common approach is to benchmark all talent against competitors, judging people’s fit for roles and capabilities to predict outcomes. So, most corporate battles rely on organizational strength, not just strategy. The classic example is Alibaba. They emphasize organizational building, which has allowed them to experience a “second spring.” The founding team ages, but the organization can regenerate. Essentially, if one day we’re competitors and both use AI, I’d build a strong AI organization with robust development capabilities. I’d analyze competitors’ agent skill systems, dissect their capabilities, and develop better skills or fill gaps. For example, I’d start with strategic analysis.

Huawei’s “Five Looks and Three Fixes” methodology is a good example. I joke that with this approach, we can beat 99% of competitors. The five looks are: industry trend, market and customers, competitors, internal capabilities, and strategic opportunities; the three fixes are: control points, goals, and strategies. This methodology can eliminate most competitors because most people play chess randomly, relying on quick thinking, while experts default to deep reasoning and strategic thinking. The first instinct of a master is to think as a commander—how to win.

Zhang Peng: The “Five Looks and Three Fixes” essentially means avoiding “reactive responses” and solidifying a long reasoning process.

Liu Ye: Experts are deep researchers and thinkers—they first look at global best practices and information, then analyze and reason deeply, and their answers are often one move to defeat the opponent.

I believe the core of future competition is modeling traditional industry operations, abstracting them into systems capable of orchestrating intelligent agents. This is the new organizational development (OD) capability, evolving into AIOD— the ultimate core competitiveness.

Alibaba’s strength lies in building organizations. Once the organization is in place, regardless of competitors or business scope, it can be competitive. Jack Ma once said that the purpose of war isn’t necessarily to seize a domain but to foster organizational growth. Alibaba measures success by organizational growth—this is a high-level mindset. Ma himself is like a super-information hub, flying 200 times a year to gather intelligence, then using it to improve organizational building. He is truly a chairman, not just a CEO.

This is the highest form of organizational structure—able to span generations, cover different industries, continuously succeed, and rebound after decline. Usually, if a company appoints the wrong CEO within ten years, it’s likely to decline. So, using history as a mirror and viewing current development from a higher dimension—even making some cuts and optimizations—are far more efficient than building from scratch.

Today, anyone can easily set up an agent, with very low entry barriers, plus open-source communities, industry secrets are scarce. In tools alone, competition can’t beat open-source communities. So, what core advantage does open-source have that can’t be copied?

04 The physics of AI organizations: Why is “gradual exposure” key?

Zhang Peng: In the “old era,” organizational discussions emphasized culture, values, KPIs, etc. When transitioning to the new era of AI agent organizations, which elements can be completely discarded, and which should be retained but transformed?

Liu Ye: Anthropic launched skills mainly because of the “gradual exposure” concept in AI coding—if AI receives a large amount of messy information, it risks context corruption and attention issues. Gradual exposure helps AI maintain focus and produce quality outputs. Relying on manual implementation of gradual exposure is essentially full manual dialogue, which is inefficient. Therefore, the core value of skills is to layer and decompose complex tasks, enabling AI to be gradually exposed.

This aligns with corporate management logic: the board focuses on strategy; the CEO on tactics and managing executives; employees handle simple tasks. If 300 people participate in the same meeting, it’s unmanageable. The core purpose of organization is to enable layered information processing—like database normalization, which improves efficiency through layered compression. Complex problems must be decomposed and exposed gradually, not input all context at once. This is the core logic of traditional organizational forms, given the limited computational power at any given time.

Zhang Peng: Models consume huge computational resources each time from scratch, which is inefficient.

Liu Ye: Impossible to do otherwise. Relying on layered, gradual exposure is essential; resources must be called upon as needed, constrained by AI model capabilities. Another reason Anthropic launched skills is that complex tasks have surpassed basic physical laws; skills decompose complex tasks into low-dimensional, simple ones. The key dimension isn’t difficulty but complexity—low-dimensional, high-difficulty tasks like coding or solving math problems are examples.

Horizon’s Yu Kai proposed a classic model: all job types can be divided into four quadrants based on “competition level” and “dimensionality”: high-dimensional high-competition, low-dimensional low-competition, low-dimensional high-competition, high-dimensional low-competition. Sales and engineers are low-dimensional high-competition; product managers and CEOs are high-dimensional high-competition; scientists are high-dimensional low-competition—some research topics may only have one researcher worldwide, with low competition but high dimension. Tasks like high-quality short dramas or novels are currently beyond AI; low-dimensional, high-competition tasks like code optimization AI can handle well. The higher the dimension, the fewer data sources, but the more data needed to train models—this explains why text models appeared first, followed by image and video models, and why short-video models are hard to implement. The supply-demand mismatch in high-dimensional tasks and data can only be addressed by decomposing tasks into skills, similar to breaking down high-level roles into basic ones when talent is scarce. Only high-level roles like CEOs are irreplaceable.

Zhang Peng: Low-dimensional, high-competition tasks are most likely to be fully replaced by AI.

Liu Ye: Absolutely, and that’s already happening.

Zhang Peng: That’s why all low-dimensional, high-competition tasks should be solved by AI as soon as possible, decomposed into skills, and implemented through agents—human involvement isn’t always necessary.

Liu Ye: I have a preliminary idea: IBM and Accenture, as the world’s largest consulting firms, essentially extract industry best practices and align with digital transformation—selling processes, not tools. When companies buy risk management or IP, they hire consulting firms to implement. Our current core work is building skill clusters, identifying top experts in each field, extracting and aligning their capabilities to form standardized skill sets. This is similar to Homework Box’s model—collaborating with Beijing No.4 High School, Renmin University Affiliated High School, Gaokao question setters, and TAL Education teachers to refine question-setting, teaching, and grading methods, then working with Baidu engineers to build systems. It’s fundamentally about aligning best practices. The core organizational capability is assembling high-quality cross-disciplinary teams—those who understand industry and engineering, can coordinate top industry experts, and possess business, recruitment, and management skills. This is the new generation of AI SaaS core.

Zhang Peng: Further extrapolating, future organizational structures should be driven by business needs. Organizations are essentially orchestration frameworks—like operating systems for business—placing people as productive units within suitable structures to maximize value. When productivity elements shift from reliance on humans to infinitely supplyable AI, and a positive cycle is established, scaling becomes continuous. Past organizational culture may transform into goals and context, no longer needing slogans, meetings, or icebreakers.

Liu Ye: Culture is a management intent, not a business intent. In the old era, strategy started with vision, which determined values; organizations served strategy, and business validated everything. Culture was just a governance tool, not directly serving strategy, and could even be personal preferences of founders.

Zhang Peng: In the past, there were gaps between people and strategy. Is AI eliminating these gaps?

Liu Ye: Yes, in the AI era, culture is less relevant. Culture is part of human belief systems, but AI doesn’t need it. AI’s core requirement is computational power.

Zhang Peng: You mean AI needs goals and principles. A single document can clarify goals and principles, enabling all units to synchronize and execute faithfully, reducing friction.

Liu Ye: Exactly. The old organizational chain: strategy → culture → talent → execution. The new AI organization: goals → principles → skills → orchestration. The entire management chain is compressed by half.

05 The final barrier: aesthetics and orchestration

Zhang Peng: What are the new organizational barriers? Talent quality is replaced by Skill Sets; as long as I have aesthetic judgment, I can acquire the best skills worldwide. Then, the next layer is “orchestration,” right? What changes will this bring?

Liu Ye: Like Huaqiangbei can buy all electronic components, but why can’t everyone make an Apple? Steve Jobs’ definition of aesthetics was very clear: seeing enough good things in the world and being able to distinguish quality is aesthetic judgment. If you’ve never seen good products, good processes, or good organizations, you can’t produce high-quality results.

Zhang Peng: Experience is the prerequisite for aesthetics.

Liu Ye: Experience plus talent—nothing more.

Zhang Peng: Aesthetics manifest in two ways: one is proactive design and orchestration; the other is recognizing and selecting emerging high-quality things in chaos. These aren’t mutually exclusive.

Liu Ye: Exactly. Some of Apple’s achievements come from independent R&D, some from acquisitions, but the core is aesthetic judgment—no need to reinvent the wheel; when necessary, develop independently.

Zhang Peng: The key is whether agents operate within set modules to confirm paths, enabling emergent orchestration; or whether all paths are preset, leading to designed orchestration?

Liu Ye: Emergence is non-manipulative; it requires setting seed rules and principles first, which reflects one’s aesthetic. Like a good engineer who can produce excellent OpenClaw with 500 lines of code, while a poor engineer might write 50,000 lines but still not match. The underlying seed rules must be human-defined.

Zhang Peng: So, waiting for emergence in chaos isn’t feasible—it takes too long. Orchestration remains crucial. Ultimately, does this orchestration only come from founders or resemble a “producer”?

Liu Ye: I think “producer” is a good analogy. Even with emergence and scale effects, data labeling, cleaning, and continuous alignment of algorithms are needed to prevent disorder.

The orchestrator depends on business complexity—complex projects like short dramas or scriptwriting can’t be done by one person. The “one-person company” concept is overused; the world can’t be infinitely simplified. Computers can be operated by one person, but mastering all high-dimensional skills is impossible. Super talents like Elon Musk or Fei-Fei Li, who excel across multiple fields and can take on any role, are extremely rare.

Zhang Peng: If we can call on the world’s top agents and skill systems—say, a top screenwriter—could we, in theory, produce a globally renowned, profitable film? The core is the “core highlight” (good script) but unable to handle all other aspects. Is this “core highlight + global resources” loop feasible?

Liu Ye: It’s fundamentally a data issue—whether there exists data that captures the highest-dimensional information. For training CEO skills, there’s currently no sufficient data: Meng Wanzhou’s lengthy articles, Jack Ma’s spoken words—none fully expose their high-dimensional cognition; even collecting all global company reports and CEO speeches can’t produce a model fit for CEO-level capabilities because the core CEO skills are implicit knowledge, not fully exposed in text.

Zhang Peng: So, the core CEO capabilities can’t yet be vectorized. This limits the “one-person company” ideal— even if each person excels in a single dimension and leverages top global resources, the absence of a core orchestrator remains. Ultimately, having the best “components” still requires strong orchestration.

Liu Ye: Product managers are similar; their implicit knowledge can’t be fully textualized. This is also why current AI companions and content generation lack “vitality”—they lack high-dimensional implicit data. When data is scarce, focus on skills; with more data, develop models. Robots can’t be deployed effectively because of insufficient data.

Zhang Peng: From this, the future competitive edge of companies isn’t just access to top models—initial AI resources seem uniform, and computing power correlates with financial strength and business loop capabilities. The ultimate difference still comes down to the “producer”—their orchestration ability and the innovation and significance of their goals, which form the core competitiveness.

Liu Ye: A McKinsey former partner told me that McKinsey’s core business is extracting best practices, building models, and helping companies implement them—like advising Chinese automakers by learning Toyota’s practices, essentially copying and applying best practices.

Mimeng’s case in short dramas is instructive. She’s a literature major, but her core team includes top Tsinghua and Peking University math and CS graduates, dissecting viral short-video logic, achieving high hit rates. This approach is fundamentally social engineering modeling—though overfitting is possible, the direction is correct.

IBM, Accenture, McKinsey do similar things—first-generation McKinsey modeled best practices into partners; IBM digitized processes; both are essentially “selling management and processes.”

Zhang Peng: The core is extracting best practices, repeatedly validating and implementing—that’s the key to future business success. Only with thorough decomposition can orchestration be efficient. So, is your next focus to advance along this path?

Liu Ye: Over the past three years, we’ve mainly done AI ToC (Theory of Constraints) business, reconstructing the entire teaching and research system with MetaOrg. This isn’t just about “AI efficiency.” We built a complete agentic research organization, with virtual research teams: language learning experts tracking second-language acquisition theories, corpus collection teams sourcing authentic expressions, dialogue evaluation teams establishing multi-dimensional speaking standards, dialogue design teams translating teaching methods into natural human-machine interactions, question design teams solving content adaptation, data analysis teams mining real signals of learning effectiveness. Each team has its own skills, workflows, and evaluation standards. About 80% of tasks—textbook data labeling, monitoring, user insights, product iteration—are now handled by AI.

Our path is from “AI as a function” to “AI as organizational capability.” The role of a language teacher, a medium-complexity position, has been abstracted and generated into other roles via MetaOrg; with the latest skill architecture, more advanced roles can be built.

We’ve completed the full process of AI tutor—abstracting and engineering orchestration capabilities. In the near future, MetaTutor will likely upgrade to MetaOrganization—its smallest unit is a role, not an employee, emphasizing collaboration and management among roles. Our current focus is connecting with top CEOs across industries because they are the true “producers.”

Zhang Peng: So, you’re creating a more scalable department?

Liu Ye: The goal is to move toward a “company.” Large companies are essentially composed of smaller units, with roles as the smallest units. We need to focus on industry-wide strategic choices and also iterate products starting from roles—if roles aren’t well designed, even strong managers can’t build efficient organizations.

Zhang Peng: To build a good department, you must first decompose related capabilities and roles, then decompose roles into skills, aiming for these skills to reach SOTA levels.

Liu Ye: The core method is co-creating with top-tier client companies. The skills developed must be evaluated by leading enterprises—like subordinate proposals needing supervisor approval—nothing self-congratulatory. For example, in short-video modeling, industry top institutions’ recognition is essential; otherwise, it’s not truly top-tier. Everything must be assessed and measured.

Midjourney produces high-quality images because its team includes top photographers and engineers with excellent aesthetic judgment; LV’s training of image models with Stable Diffusion surpasses ordinary models because LV has the world’s top image aesthetics and data. Clearly, evaluation ability is key. To build an AI company, emulate IBM or Huawei—after serving top automakers, master best practices and output; Huawei spent 4 billion on IPD processes, used internally and externally—this is core competitiveness.

Zhang Peng: Essentially, breaking down skills along best practices, achieving SOTA in skills, then upgrading to SOTA in roles and departments, and finally orchestrating top-tier business—this is the clear path to business excellence. Another key question: how to keep skills up-to-date? Like biological mutations, each era’s SOTA may be replaced in the next. How to adapt to this change?

Liu Ye: The core logic aligns with human and biological evolution: perception, planning, action, reflection. Maintaining high talent density and cross-disciplinary attributes—connecting to cutting-edge research, exploring business models, co-creating with top clients—continually evaluating and optimizing in real scenarios is the only way.

Zhang Peng: From this, the best practices of top companies can help mid-tier firms leapfrog, but such systems are likely only accessible to resource-rich companies. Small and young startups face high barriers. The consulting industry has shifted from traditional services to tool-based products. Is the new generation’s opportunity only at the skill level? How to achieve disruptive innovation at the skill layer and avoid the “noble cycle” of exclusivity?

Liu Ye: In the last SaaS wave, companies like Salesforce, Palantir, Notion, Slack offered either general tools or integrated services, showing young entrepreneurs still have opportunities—by avoiding areas where they lack advantages, focusing on universal skills, and finding niche ecosystems. Notion is a prime example: it abstracts text note-taking, becoming a universal tool. Ultimately, the world will be a division of labor among countless agents. Young people should first find their niche, leverage their strengths, align with future trends, and avoid becoming victims of time.

Over the past decade, the first internet generation was mostly returnees (relying on cognitive advantages), the second was programmers (relying on tools), and the third is industry internet entrepreneurs—clear patterns. Young people need to understand the mid-game and their own advantages.

Zhang Peng: So, you believe that limited innovation at the skill level has only marginal effect, and the greatest opportunity lies in goal innovation—identifying emerging new goals, combining high-quality skills, and continuously evolving to build new systems and breakthroughs.

Liu Ye: Skill competition is very subtle. While skills are hot now, if someone aligns with even more top human experts and develops better skills, existing skills will be replaced. This circles back to moat issues: early movers may not win in the end—they might become nutrients for higher-dimensional competitors.

Zhang Peng: I worry about becoming just a “loading program,” helping higher-dimensional opponents lay foundations. If you only optimize efficiency on existing goals, it’s meaningless—the advantage will eventually be leveled. To truly break through, the new generation must make fundamental differences in goals.

Liu Ye: Exactly. If you can’t grow into a core force yourself, you’re just nourishing higher-level opponents. Business is simple at its core: knowing who your customer is, how to serve them, and making them indispensable. If young people don’t understand who their customers are, they can’t optimize.

Zhang Peng: Also, focus on incremental markets. In saturated markets, competition is fierce. If your business succeeds, it will elevate competitors to the same advanced level—these companies have wealth and cognition, making it hard for startups to compete.

Liu Ye: The success of SaaS companies like Notion and Slack was driven by goal differentiation.

In early SaaS, Chinese funds favored investing in scientists, but later found scientists are better suited for collaboration than entrepreneurship—high-dimensional, low-competition fields are very different from high-dimensional, high-competition business logic. The higher the domain dimension, the harder it is to switch to new fields; the core thinking is different. Early on, it’s about technology (low-dimensional, high-competition, immature tech). Once mature, it shifts to business (high-dimensional, high-competition, industry players, product managers, business practitioners). For example, when Apple launched the iPhone, most top apps were developed by programmers; years later, with the rise of industry internet, those programmer-led apps were replaced.

In the AI era, if we follow the move from mobile internet, Silicon Valley’s core strength remains experienced practitioners, just like China’s second-wave entrepreneurs in industry internet. The opportunity for young people is still to find differentiated goals.