(Illustration: Some people buy multiple pieces of the same clothing style. I tend to buy multiples of the same toys (or brands), paired with a deep-seated quirk of not wanting to be like everyone else, like this niche notebook Field Notes that I’ve been using for over a decade. I was originally drawn to its woodgrain cover, which has since been discontinued, but thankfully my hoarding habit means I have some stocked up. The advantages of using the same style through deliberate design decisions are many, perhaps including the transformation into excess returns for our clients a decade later (wishful thinking). Photo taken in 2016 at a friend’s home in Boston. Image source: Ernest.)

✳️ The Design Decision of AUM Fees

Charging a percentage of AUM (Assets Under Management). Almost every wealth management firm uses this model, and most clients take it for granted. But a16z Perennial CIO Michel Del Buono said on the Sorcery Podcast that this isn’t just a pricing decision, it’s a design decision that influences everything else. Let’s practice decomposing it:

Read More

(Not midnight yet. 9 PM at the outermost moat of the Imperial Palace, with nobody else around. Occasionally understanding things others don’t, while experimenting with all sorts of magic for preserving memories. Photo by Ernest.)

Every time I see news about “Company X laid off XX%,” what I really want to know isn’t why they cut, but rather: after the cuts, how do the remaining people work? What does the organization look like? What processes did they adjust?

Block ($XYZ) laid off more than 40% of its workforce in 2025 Q1, openly stating that AI was a key factor. Their business lead Owen Jennings shared the specifics on the a16z channel. Owen has been at Block for 12 years, was previously the CEO of Cash App, and now oversees product operations and customer support across Square, Cash App, and Afterpay. What he described wasn’t the generic “we embrace AI” narrative, but the actual logic behind their rebuilding.

There are three principles worth unpacking here:

Read More

(Image: Olaf makes an appearance. Source: NVIDIA GTC 2026 Keynote.)

✳️ Viewing GTC 2026 Through a Token Economy Lens

Every year after GTC, the community races to compare specs. Which chip has the highest compute, how much NVLink bandwidth, how many times faster Vera Rubin is than Blackwell. But the slide Jensen held up this year and called “my best slide” wasn’t a spec sheet for any single chip — it was an architecture diagram spanning the entire structured data ecosystem. It listed Snowflake, Databricks, Amazon EMR, Google BigQuery and other CSP engines alongside various data storage solutions, with NVIDIA’s cuDF acceleration engine at the bottom. He said his team always tells him “don’t show this one” — too complex — but he insists on presenting it every time. The core message: “Structured data is the ground truth of enterprise computing” — the foundation of all AI.

The point of that diagram isn’t about any individual platform. It’s about the reason for acceleration. In the past, accelerating structured data was about doing more, cheaper, more frequently — “good enough was good enough.” But going forward, AI understands and consumes data far faster than humans can. Without accelerating data preparation, you simply can’t keep up. Nestle used Watson X to accelerate their supply chain — 5x faster, 83% cost reduction — speed, scale, and cost benefits all at once. NVIDIA built two foundational platforms for this: cuDF is “RTX for data frames,” handling structured data; cuVS handles unstructured data. The latter is arguably more critical: 90% of the world’s data is unstructured, previously “completely useless to the world,” until AI’s multimodal understanding made it searchable. (Earlier this quarter, we helped a manufacturing client re-explore their ERP data. We had no idea what we’d find until we dove in — the general-purpose model’s ability to understand that data was jaw-dropping. Happy to chat more about this separately if you’re interested.)

OK, so let’s say your organization has the data layer acceleration in place. The next question is: how do you price what AI produces?

Read More

(Illustration: Layer upon layer, every piece matters. Like this ham air-drying in the breeze, sacrificing itself for the greater good. Trust can’t be built all at once; it has to be stacked, one layer at a time. The first layer? Start by hitting that like button. Image source: Ernest.)

✳️ The Boring Plumbing Work

What actually blocks enterprise AI deployment is almost never that the models aren't good enough. Mistral AI CTO Timothée Lacroix put it bluntly: today's model capabilities are already sufficient to unlock massive enterprise value, but you first need to get all the connectors, data formats, and permission management right — all this "boring plumbing work" — before enterprise token consumption truly takes off. He used the word “plumbing” (and I’d stress the “system” part, not just the pipes). He said it three times throughout the interview. We’re still in the construction phase, he noted — most enterprises haven’t even gotten basic data connectivity right, let alone running AI agents executing tasks at scale in the background. (What we see on the ground is even bleaker: wishlists that don’t map to actual data. No wonder everyone just jumps straight to word-chaining chatbots — sarcasm mode engaged.)

Read More

(Illustration: Mt Rainier in Seattle, Reflection Lake inside the national park. Facing the mountain, facing time, what can we humans do in a post-AGI world? At the very least, being a decent human being should be the baseline, you demons. I am not sure I could live in the countryside, but I do enjoy having nothing to distract me. Then I found out that a Chinese idiom meaning “undistracted focus” is also a Year of the Horse blessing?! Image source: Ernest.)

✳️ This man has not touched a single line of code since November

I recently listened to Lenny’s Podcast interview with Boris Cherny, the person in charge of Anthropic’s Claude Code, formerly one of the most productive engineers at Instagram. Since last November, 100% of his code has been written by Claude Code. He has not manually edited a single line, shipping 10 to 30 PRs per day, and during the recording he had 5 agents running simultaneously. Anthropic's engineering team grew 4x while each engineer's output increased by 200%. In his previous role at Meta, he was responsible for code quality across the entire company. Back then, hundreds of engineers spending a full year would typically improve productivity by only a few percentage points. Now it is hundreds of percentage points. A completely different order of magnitude.

Read More

(Illustration: Taking the day off. Happy Lunar New Year! Image source: Business Next.)

✳️ Coding is easy, Context is hard

I recently recorded an episode of Podcast “Digital Keywords” EP228 with James from Business Next, discussing a core theme I kept chewing on throughout my 2025 year-in-review.

AI has crossed the threshold in writing code. It is no longer a small assistant that auto-completes one line at a time. Take Claude Code for example: it reads through your entire project directory on its own, understands cross-file context and dependencies, plans how to coordinate changes, even dispatches sub-agents to handle different tasks in parallel, and then delivers an entire feature in one go. More importantly, it has memory. You write your project specs, goals, and coding style into a file, and every time it starts working, it reads that file first. No need to re-explain everything from scratch.

From Cline and RooCode, to Cursor, to Claude Code and Kiro, my team and I have walked the entire tool evolution path together, feeling the role shift at every step. Humans can no longer compete with AI on speed. The bottleneck has moved from “think fast, code fast, iterate fast” to “can you articulate your requirements, context, and intent clearly?” I joked with my team: we all need to start practicing how to communicate well. This is not just an observation. It is the firmest conclusion I reached after being pushed through all of 2025, and the central theme of this podcast episode.

Read More

(A dish that breaks the rules of the game can show up even during a casual meal. Turns out Greek cuisine has its own version of sashimi, paired with a light touch of avocado mousse that balances out the richness of the tuna, making it the star among the bite-sized appetizers. At Taverna in Palo Alto, a dancing silhouette tucked in the corner of the menu reminds us: “Every day is a gift.” We will live well, and challenge the rules together. Image source: Ernest.)

✳️ Typing has become a commodity

Jensen Huang, holding a glass of white wine at the late-night session of Cisco AI Summit 2026, said: “Coding as it turns out is just typing. And typing as it turns out is a commodity.”

This isn’t putting down engineers. For 60 years, we’ve been telling computers exactly what to do, line by line in Fortran, C, C++. That’s called explicit programming. Now we’re entering the era of implicit programming, where we tell computers our intent and they figure out how to solve it. Just like that, the scarcity of “being able to write code” has vanished.

Since last Lunar New Year, I’ve gone from Cline to Cursor to Claude Code, and my deepest takeaway is: Coding is Easy, Context is Hard. When AI can write the code we want in seconds, the bottleneck has long shifted from typing speed to whether we can articulate the context of the problem, the boundary conditions, and the edge cases clearly. That’s why I moved from Vibe Coding toward Spec-Driven Development combined with organizational workflow frameworks: define “what to do” and “why” in human language first, then let AI handle “how to do it.”

Read More

(Caption: Deliberately captured Tokyo Ginkgo with the Ginkgo Camera GR IV. Image source: Ernest Chiang.)

“Sometimes you got to slow down to go fast.”

This phrase kept echoing in my mind throughout 2025. While every industry chases the efficiency gains promised by AI, and anxiety about the future and artificial intelligence pervades, I kept returning to this note in my journal: “True progress often comes from deliberate pauses and deep thinking.”

The hard part isn’t stopping or slowing down—it’s being deliberate.

Deliberate practice, deliberate thinking, deliberate connection, deliberate verification.

Read More

(Illustration: At AWS re:Invent 2025 CEO Keynote in Las Vegas, listening to Sony CDO’s eight-minute speech on Technology to “Create” and “Deliver” KANDO. Image source: Sony.)

✳️ tl;dr

• Ever since listening to Sony CDO’s eight-minute speech on Technology to "Create" and "Deliver" KANDO at AWS re:Invent 2025 CEO Keynote in Las Vegas two weeks ago, I was not only impressed by AWS’s nuanced breakdown of AI, but deeply admired how this established company Sony integrates Amazon Bedrock into its own KANDO culture and execution.¹²
• Sony acquired controlling interest (80% stake) in Peanuts Holdings for $460 million, combined with its initial $185 million investment in 2018, totaling $645 million over seven years. The transaction is expected to generate revaluation gains on equity as operating income.³
• This is a case of balancing financial discipline with strategic vision. Sony chose staged investment rather than outright acquisition, validating the business model before increasing investment, reducing risk while building partnerships.³
• Peanuts generates $2.5 billion in annual retail sales, with holiday products contributing $500 million. Compared to Sony’s $185 million investment in 2018, the brand demonstrates strong cash flow generation capacity and stable licensing revenue.⁴

• Sony’s entertainment business revenue grew from 26% in fiscal 2012 to 60% in fiscal 2023, with content IP investment accounting for 57% of strategic investments (1.5 trillion yen). This shows the company’s successful transformation into a content-driven entertainment group.⁵
• Sony CEO Yoshida and CFO Totoki emphasize “Creation Shift” and “Creative Entertainment Vision.” Totoki stated: “I am obsessed with growth. When growth stagnates, you fall into a negative spiral.”⁶⁵
• Sony’s recent IP investments include 10% stake in Kadokawa (50 billion yen), 2.5% stake in Bandai Namco (68 billion yen), and acquisition of Crunchyroll. These investments build an anime and gaming IP ecosystem.⁷⁸
• Compared to studio acquisitions (Bungie, Firewalk underperformed), direct IP investment offers more controllable risk. IP can be monetized across multiple platforms without being constrained by single-studio operational risks.⁹
• Apple TV+ exclusive streaming agreement through 2030 ensures long-term revenue visibility. Platforms are willing to pay premium for classic IP because it attracts multi-generational audiences and builds cultural resonance, reducing churn.¹⁰
• Sony plans to spin off part of its financial services business in 2025 to focus on entertainment and content creation. This demonstrates management’s determination to simplify the business portfolio and improve capital allocation efficiency.⁵
• WildBrain received $460 million from the sale, to be used for debt repayment and investment in Strawberry Shortcake, Teletubbies, and digital content networks. For WildBrain, this represents portfolio optimization, focusing on core assets.³
• (Speculation) Sony may increase Peanuts brand value by 50-100% within 5-10 years through cross-media integration (gaming, music, film) and expansion into new markets (Asia, Latin America).
• The Schulz family retains a 20% stake to ensure brand heritage and quality control. This equity structure balances commercial interests with cultural legacy protection, crucial for long-term brand value.³

• Peanuts’ cultural significance in Japan (inspiring Hello Kitty’s creation) provides Sony with unique advantages. Sony may strengthen Asian market expansion, as two-thirds of revenue already comes from outside the U.S.¹¹⁴
• This case demonstrates how strategic IP acquisition builds lasting competitive advantage rather than chasing short-term trends. Peanuts’ 75-year history proves the enduring value and cross-generational appeal of classic IP.⁴

Read More

(Illustration: Frozen soap bubble. Image source: Photo by Jill Warvel on Unsplash.)

✳️ tl;dr

• Howard Marks believes AI shows signs of “irrational exuberance,” but bubbles can usually only be identified in retrospect, and current valuations, while high, have not yet reached crazy levels ¹
• “One of the most interesting aspects of bubbles is their regularity, not in terms of timing, but rather the progression they follow. Something new and seemingly revolutionary appears and worms its way into people’s minds. It captures their imagination, and the excitement is overwhelming. The early participants enjoy huge gains. Those who merely look on feel incredible envy and regret and – motivated by the fear of continuing to miss out – pile in.”

• AI exhibits bubble characteristics: revolutionary technology, FOMO-driven speculation, extremely high valuations, but “this time is different” may hold true with a 20% probability. (Everyone wants to predict the future but also hedge their bets?)
• Circular deals raise concerns: Nvidia invests $100 billion in OpenAI, which uses that money to purchase Nvidia chips, with Goldman Sachs estimating 15% of Nvidia’s sales come from such transactions
• Data center investment scale is staggering: JPMorgan estimates total AI infrastructure buildout costs at approximately $5 trillion, with spending approaching $500 billion next year
• Debt financing risks escalate: Oracle, Meta, and Alphabet issue 30-year bonds to finance AI investments, with yields exceeding US Treasuries by only 100 basis points or less
• Warren Buffett reminds us: automobiles were the most important invention of the first half of the 20th century, but only 3 out of 2,000 car companies survived, proving that technological importance doesn’t guarantee investor profits
• AWS Hero Ernest recommends that technical decision-makers establish a multi-vendor strategy, avoid sole dependence on Nvidia, and evaluate alternatives such as AWS Trainium and Google TPU for cost control and supply chain resilience

• AI chips have an actual useful life of only 1-3 years, yet companies use 5-6 year depreciation schedules, with Michael Burry accusing tech giants of inflating earnings ²
• Nvidia shifted from a 2-year to an annual product cycle, with Jensen Huang joking: “Once Blackwell starts shipping, you couldn’t give Hoppers away”
• Anthropic derives 80% of revenue from enterprise customers, with B2B models showing more promise due to higher transaction values, suggesting product strategy should prioritize enterprise markets ³
• OpenAI expects to continue massive losses until 2028, with HSBC estimating it won’t be profitable by 2030, requiring an additional $207 billion in funding ⁴
• Google TPU emerges as Nvidia’s strongest competitor, with 7th generation Ironwood offering 2x power efficiency improvement and 1.4x Nvidia’s cost-effectiveness, targeting 10% market share by 2027 ⁵

• SPVs (Special Purpose Vehicles) are used for data center financing, hiding off-balance-sheet debt, raising concerns similar to the Enron model
• WEF predicts AI will displace 85 million jobs by 2030 but create 97 million new ones, though 77% of new jobs require master’s degrees, necessitating fundamental HR strategy adjustments ⁶
• 77,999 jobs have already been lost to AI in 2025, averaging 491 people unemployed daily, with Microsoft reporting 30% of code written by AI while 40% of layoffs target engineers ⁷
• Historical analogy: The AI bubble resembles the 1860s railroad boom and 1920s aviation bubble, both being “inflection bubbles” that accelerated technology adoption at investors’ expense ¹
• Current AI giants average a P/E ratio of about 34, lower than the dot-com bubble’s 59, but Shiller CAPE reaches 40.40, approaching dot-com bubble levels ⁸

Read More