Responding to warnings in The New York Times and The Economist about AI summaries and 'cognitive shallowing': from television-era 'couch potato' panics to how YouTube, TikTok, and AI gave me the scaffolding to read Kant's Critique of Pure Reason—why staring at raw text does not equal understanding, and what really determines whether we keep questioning after opening the door.

• Are short videos and AI summaries genuinely causing the neural circuits for deep reading to atrophy, as some scholars warn?
  Sarah: From an evolutionary standpoint, this 'genetic regression' argument is highly exaggerated. Widespread reading of printed books is only a few centuries old. Our biological brain hardware is resilient and won't mutate just because we changed the information medium over a few decades. Yao Di: Absolutely. Every era has its instant gratifications, from ancient taverns and theatre to modern television and social media. The tool doesn't make us shallow; the key is how we proactively use these high-density resources, much like using dumbbells in a mental gym.
• Why is the assumption that 'staring at raw classic text automatically builds deep thought' considered a myth?
  Sarah: It's a form of intellectual elitism that confuses 'medium friction' with 'conceptual depth.' Forcing someone to sit alone with incomprehensible academic jargon usually just results in frustration and abandonment. Yao Di: Precisely. Real comprehension requires a progression from sensory intuition to conceptual understanding. When facing a monumental work like Kant, the visual metaphors from videos and conceptual translation from AI serve as the essential entry ramps over the wall of jargon.
• In the age of AI, how can we avoid becoming shallow thinkers who only consume three-minute summaries?
  Sarah: It boils down to your motivation. If you are only looking for a 'standard answer' to turn in or show off, you'll stop at the door. You need to identify a real-world problem that genuinely troubles you and demands deeper answers. Yao Di: Right. When you get a high-density summary from AI, always remind yourself that it's just the admission ticket. Take your real-life confusion, use that key, and keep asking the original text, the AI, and the world 'why'—that is how you build real mental muscle.

Read the full bilingual essay → https://theyaodi.com/blog/why-video-and-ai-do-not-kill-deep-reading

Perhaps we suspect AI is conscious simply because it exhibits a structural isomorphism with conscious human beings. Across seven distinct levels—from output and behavior to causality and subjective experience—isomorphism is both our sharpest tool for approaching the unknowable and our deepest source of misrecognition.

• What does 'isomorphism is not identity' mean, and how does it apply to AI?
  Sarah: Think about corporate legal personhood. We grant corporations legal rights, let them sign contracts, and sue just like natural persons. This is functional isomorphism. But nobody actually claims that a corporation has biological feelings or personal memories. Mistaking structural mapping for biological reality is a category error. In AI, when people see a model generate highly empathetic responses, they assume it must have a real human mind. They are mistaking conversational isomorphism for ontological identity. Yao Di: Isomorphism means that different physical systems share an identical internal relational structure. For instance, a jet plane and a bird share elegant aerodynamic isomorphism in how they fly, but an airplane's wings don't have feathers or blood vessels. AI models are highly isomorphic to humans in information processing and semantic geometry, but that does not mean their underlying physical medium, evolutionary path, or biological nature are identical to ours.
• How does the 'seven-rung ladder of isomorphism' clarify debates on machine consciousness?
  Sarah: It works like a conceptual scalpel. Historically, debates on AI consciousness have been incredibly muddy. People would see a model console someone (Behavioral Isomorphism) or point to its attention mechanism (Functional Isomorphism) and immediately jump to the conclusion that the machine experiences subjective pain (Phenomenological Isomorphism). This seven-rung ladder shows that while we can rigorously test the first six rungs in a lab, we cannot directly measure the seventh. It stops us from talking past one another. Yao Di: Exactly. The rungs are Output, Behavior, Function, Representation, Process, Causality, and Phenomenological Experience. For instance, latent space analysis helps us prove Rung 4 (Representational Isomorphism), and causal intervention experiments let us verify Rung 6 (Causal Isomorphism). The ladder transforms speculative theological debates into structured, falsifiable scientific inquiries.
• Why is 'agnosticism' considered a refined, rational stance rather than defeatism in the AI era?
  Sarah: Many people treat agnosticism as a passive shrug, but here it represents rigorous intellectual honesty. It shields us from the hype that 'AI has acquired a soul,' while also preventing us from dismissively declaring 'AI is just a bunch of cold metal.' It defines the boundaries of third-person scientific observation, allowing us to explore technology with both clarity and respect. Yao Di: Agnosticism marks the boundaries of knowledge, not its end. We may never be able to directly prove machine qualia from a first-person perspective, but that shouldn't stop us from using 'isomorphism' to map and approximate the system. Knowledge isn't always about directly penetrating the essence of a thing; it's about building rigorous structural correspondences. Like a mathematical asymptote, we infinitely approach the truth, which is the ultimate form of inquiry.

Read the full bilingual essay → https://theyaodi.com/blog/isomorphism-agnosticism-and-machine-consciousness

If coding transformed faster than law because software has compilers and unit tests, how do we build a verification loop and test harness for legal work without asking busy lawyers to grade AI homework?

• Won't a feedback loop that learns from historical redlines just institutionalize yesterday's compromises or a tired lawyer's quarter-end concession?
  Sarah: That's the classic regression-to-the-mean risk in enterprise AI. If you blindly train on every historical redline, one bad concession becomes company policy. Yao Di: That is why the harness strictly separates descriptive telemetry from normative authorization. Redline diffs never auto-merge into the playbook—they generate a Pull Request that requires the General Counsel's explicit sign-off, paired with an expiry tag when market or statutory assumptions shift.
• How does building this verification harness help solve the Judgment Ladder Crisis for junior lawyers?
  Sarah: Instead of competing with LLMs on raw document drafting, young lawyers move up the stack to maintain the test suite. Yao Di: Exactly. Designing deterministic linter rules and adversarial counterparty prompts becomes the new Flight Simulator for Judgment—because you can only write the test harness for a deal if you truly understand its institutional physics.

Read the full bilingual essay → https://theyaodi.com/blog/how-to-build-the-verification-loop-and-harness-for-legal-ai

From an enterprise perspective, we must never view public officials through personal likes and dislikes—and we must outgrow both the two-way risk misjudgments across the Pacific and the extreme zero-sum myth that doing business in the other country merely strengthens a rival.

• How should an enterprise manage the tension between employees' personal political views and professional government relations?
  Sarah: It's critical to establish a firm boundary between the 'natural person' and the 'fiduciary.' When representing a company, professionals must prioritize their duty to shareholders and teams, leaving personal political biases at the door. Yao Di: Well said. Respecting a public official is about respecting the public trust and the citizens they represent, not personal fondness. Keeping this distinction clear allows for objective and rational business execution.
• How can local multinational teams overcome the severe 'compliance anxiety' felt by their overseas headquarters regarding Chinese government relations?
  Sarah: Headquarters must understand that the Chinese government is an active participant in market infrastructure and standards. Forbidding interaction doesn't reduce risk; it forces local teams to run blindfolded, increasing operational risk. Yao Di: Local teams need to establish structured, transparent, and compliant communication channels. Converting 'mysterious public relations' into a standardized, documented business process is key to dispelling political paranoia at HQ.
• How should global businesses counter the zero-sum rhetoric that cross-border trade is 'aiding the enemy'?
  Sarah: Global business is a highly integrated ecosystem. Profits made by US firms in China fund R&D back home, while Chinese firms hone global capabilities by competing in the deep US market. It's mutual empowerment, not a one-way transfer. Yao Di: Exactly. Isolating oneself only leads to competitive atrophy. True, confident patriotism lies in trusting your nation's talent to enter the global arena and create indispensable value through real competition and cooperation.

Read the full bilingual essay → https://theyaodi.com/blog/never-let-disliking-a-leader-sabotage-your-work

Using Kant’s triad of cognition (Sensibility, Understanding, Reason) and Husserl’s direction of attention, this essay re-maps the epistemological division of labor among telescopes, the classic internet, short-form video, and generative AI—and explores why the bottleneck of understanding has shifted from external tools to our own inner mind.

• What is the fundamental epistemological difference between AI and traditional search engines?
  Sarah: Traditional search is an external archive; it brings the library to your desk, but you still have to manually decode the jargon. Yao Di: AI is externalized 'Synthetic Understanding.' It doesn't just find documents; it recompiles the logic of one discipline into the language of another, breaking conceptual isolation.
• Why is short-form video described as an 'epistemological divergence' from AI?
  Sarah: Because it flips the relationship—we become passive recipients of stimuli rather than active seekers of knowledge. Yao Di: Short video 'short-circuits' the intellect by targeting the limbic system, whereas AI requires 'intentionality'—it won't move unless the human subject asks a question.
• How should we understand the 'inward shift' of the cognitive bottleneck?
  Sarah: It means the barrier is no longer access to data, but the sophistication of your internal 'operating system' used to process that data. Yao Di: Exactly. Specifically, Theorem 2 suggests that without first-hand 'embodied' experience, the massive symbolic map provided by AI remains an abstraction that cannot be converted into true insight.

Read the full bilingual essay → https://theyaodi.com/blog/epistemology-of-machines-and-humans

Many of the issues global headquarters worry about most in China—IP protection and data compliance—are technical problems that can be engineered and solved. The harder challenge in 2026 lies in understanding how the competitive engine and the institutional rules of the game have changed.

• If China's procedural brakes are weaker on sensitive issues than in the United States, why isn't staying out of the market the safest choice for a General Counsel?
  Sarah: Because sitting in the garage is its own crash in slow motion. China in 2026 is the world's toughest proving ground for supply chains, hardware iteration, and digital consumer models. If you forfeit that arena, local competitors will perfect their cost and tech edge unchallenged and bring it straight to Europe, Latin America, and the US. Yao Di: Exactly. And remember that ordinary commercial law in China works predictably for routine transactions. Instead of relying solely on external US-style litigation brakes, a smart GC maps political cliffs in advance and installs internal corporate brakes—ring-fenced local entities and capped pilot trials.
• How can a local China team convince a skeptical global headquarters to grant room for local innovation?
  Sarah: Stop asking for a blank check in the name of 'China speed.' Headquarters hears that as uncontrolled exposure. Instead, bring headquarters a pre-engineered roll cage—show them exactly how the risk is capped and isolated. Yao Di: Right. The local team should proactively propose self-imposed guardrails: ring-fence the initiative strictly to the China market, structure it as a time-bound pilot trial, and over-communicate context early. When headquarters sees the guardrails, while holding an uncompromising bottom line on core principles, both sides can move forward.
• Why does treating data sovereignty as an engineering exercise benefit a multinational beyond China?
  Sarah: Because most companies have messy global data flows until a strict regime forces them to map every byte. Running that exercise forces real discipline around data minimization and access control. Yao Di: When technical architects, legal counsel, local teams, and global privacy teams sit down together to solve China cross-border data compliance, the resulting architecture almost always upgrades the company's global data governance and protects user data better worldwide.

Read the full bilingual essay → https://theyaodi.com/blog/lunch-with-a-visiting-general-counsel-2015-to-2026

Recently I accompanied an American professor and his students to visit a senior municipal official in one of China’s premier provincial capitals. Sitting in a grand reception room in City Hall for over an hour, we braced ourselves for standard bureaucratic talking points—and instead heard a candid ledger of urban governance, backed by a thirty-year personal footnote.

• Why is not speaking the local dialect after thirty years such a powerful proof of a city's openness?
  Sarah: Glossy talent brochures are easy to print, but in many cities the inner circles of politico-business power remain locked behind local dialect and hometown networks. Yao Di: Rising from an inland Hunan graduate to a senior municipal leader without ever learning the local tongue proves the city genuinely decoupled career advancement from origin and accent.
• How do China's top commercial cities overcome the curse of 'new officials ignoring old ledgers'?
  Sarah: Policy whiplash across three-to-five-year mayoral terms is the single biggest risk for long-term private capital. Yao Di: Mayors rotate on political clocks, but the seasoned technocratic spine beneath them stays for thirty years, preserving institutional memory and keeping pro-entrepreneur commitments intact.
• What is the deeper tension inside the metaphor 'I take care of the sunshine and rain; you take care of growing strong'?
  Sarah: It captures admirable restraint against micromanagement, yet it still casts the state as the weather and private firms as dependent crops. Yao Di: It reflects unequal legal standing—crops cannot sue the sky. Yet in a system where the state inevitably commands the heights, committing to be steady, predictable sunshine and rain is the most pragmatic and enlightened bargain available.

Read the full bilingual essay → https://theyaodi.com/blog/sunshine-and-rain-five-words

Many of the issues global headquarters worry about most in China—IP protection and data compliance—are technical problems that can be engineered and solved. The harder challenge in 2026 lies in understanding how the competitive engine and the institutional rules of the game have changed.

• If China's procedural brakes are weaker on sensitive issues than in the United States, why isn't staying out of the market the safest choice for a General Counsel?
  Sarah: Because sitting in the garage is its own crash in slow motion. China in 2026 is the world's toughest proving ground for supply chains, hardware iteration, and digital consumer models. If you forfeit that arena, local competitors will perfect their cost and tech edge unchallenged and bring it straight to Europe, Latin America, and the US. Yao Di: Exactly. And remember that ordinary commercial law in China works predictably for routine transactions. Instead of relying solely on external US-style litigation brakes, a smart GC maps political cliffs in advance and installs internal corporate brakes—ring-fenced local entities and capped pilot trials.
• How can a local China team convince a skeptical global headquarters to grant room for local innovation?
  Sarah: Stop asking for a blank check in the name of 'China speed.' Headquarters hears that as uncontrolled exposure. Instead, bring headquarters a pre-engineered roll cage—show them exactly how the risk is capped and isolated. Yao Di: Right. The local team should proactively propose self-imposed guardrails: ring-fence the initiative strictly to the China market, structure it as a time-bound pilot trial, and over-communicate context early. When headquarters sees the guardrails, while holding an uncompromising bottom line on core principles, both sides can move forward.
• Why does treating data sovereignty as an engineering exercise benefit a multinational beyond China?
  Sarah: Because most companies have messy global data flows until a strict regime forces them to map every byte. Running that exercise forces real discipline around data minimization and access control. Yao Di: When technical architects, legal counsel, local teams, and global privacy teams sit down together to solve China cross-border data compliance, the resulting architecture almost always upgrades the company's global data governance and protects user data better worldwide.

Read the full bilingual essay → https://theyaodi.com/blog/why-visiting-general-counsels-ask-2015-questions-in-2026

Rather than viewing China solely through the lens of multi-party electoral rotation, the Anglo-American legal and financial concept of 'fiduciary duty' offers a clearer framework for understanding both the historical roots of Chinese governance and the structural trade-offs facing East and West.

• How do the Western and Chinese models handle a major economic crisis differently under this fiduciary framework?
  Sarah: In the Western 'publicly traded' model, a crisis often triggers a noisy, public proxy battle. You get intense media scrutiny, blame-shifting, and potentially a complete change in management at the next ballot. It's chaotic, but it acts as a highly visible pressure-release valve. Yao Di: In the Chinese 'stewardship' model, the general partner can mobilize massive, coordinated state capital almost instantly to stabilize the system. The trade-off is that this top-down intervention can sometimes mask underlying structural risks or lead to over-correction, as there are fewer external checks to halt a mistaken policy.
• If the Chinese model lacks external elections, how do the 'shareholders' actually hold the management accountable when duties are breached?
  Sarah: Historically, the ultimate check was the 'forced liquidation'—the dynastic collapse when the water overturns the boat. Today, the system has to rely on intensive, top-down internal audits, anti-corruption sweeps, and constant monitoring of public sentiment to catch agency failures early. Yao Di: That's spot on. It's a high-stakes, self-policing mechanism. The trustee must constantly deliver substantive performance—like economic growth and public safety—because without the procedural legitimacy of elections, their mandate depends entirely on keeping the baseline shareholders satisfied.
• How does the rise of long-term global challenges like climate change test these two different trust architectures?
  Sarah: Climate change is the ultimate test of intertemporal capital allocation. Western systems struggle here because politicians are incentivized to defer painful carbon-reduction costs to win the next election cycle, though their open systems are great at fostering private-sector green tech innovation. Yao Di: Meanwhile, China's stewardship trust can execute massive, decades-long green energy transitions and infrastructure buildouts without worrying about political whiplash. However, the risk is that their top-down, mandate-driven approach can sometimes crush local economic flexibility in the process.

Read the full bilingual essay → https://theyaodi.com/blog/state-as-public-trust-fiduciary-duty

Why has AI already reshaped coding while its impact on legal work seems less dramatic? The key may be how easily each field can verify its results.

• Why is 'verifiability' the ultimate bottleneck for AI adoption in non-technical fields?
  Sarah: If you can't cheaply and automatically verify if an AI's output is correct, a human expert has to review every single line. That completely erases the cost and speed benefits of using AI in the first place. Yao Di: In software, the compiler acts as a free, instant auditor. In law, medicine, or strategic consulting, the 'compiler' is a real-world crisis, a lawsuit, or a market failure. The cost of a bad verification there is catastrophic.
• How can legal departments prepare for the 'hollowing out' of junior roles?
  Sarah: Law firms and corporate legal departments need to completely abandon the billable hour model for junior tasks and shift toward active apprenticeship models, where juniors shadow senior partners on judgment-heavy tasks early on. Yao Di: We have to treat judgment as a deliberate, structured curriculum. Since juniors won't be spending thousands of hours doing manual document review anymore, we must design simulated crisis scenarios to accelerate their decision-making experience.
• What does 'context reconstruction' mean for the limits of LLMs?
  Sarah: LLMs only know what is explicitly written down in their context window or training data. They are blind to the 'dark data' of human organizations—the unwritten alliances, the CEO's unstated priorities, or the mood of a regulator. Yao Di: Yes, senior advisory work is less about information retrieval and more about reading between the lines. Until AI can capture the tacit, relational knowledge embedded in human networks, its strategic advice will remain shallow.

Read the full bilingual essay → https://theyaodi.com/blog/ai-coding-law-verifiability