Claude Code founder Boris Cherny’s sharing at the Sequoia Conference was incredibly informative, and I heard many viewpoints completely for the first time. This guy really has a good understanding of AI. I’ll share my summary.

01 Code Is No Longer Scarce

For a large number of mainstream development scenarios, the act of writing code by hand is already becoming an inefficient task. Previously, when a function needed to be delivered, engineers would sit down, first figure out how to implement it, and then write the code line by line. In this process, the engineer’s greatest value was: whether they could write, how well they could write, and how quickly they could write.

Now, the working method is different. For the same function, what the engineer does is more like: first clearly explaining the requirements, breaking the task into several parts and handing them over to an Agent, setting an acceptance standard, and then checking whether the results run by the Agent are correct. If not, adjust the prompts and let it run again. AI can already handle most of the Coding tasks.

Overall, the value of engineers has shifted from whether they can write code to whether they can break down tasks, whether they can clearly explain the goals, whether they can accept the results, and whether they can manage Agents. This change is actually very similar to the industrial revolution. Code itself is no longer a scarce commodity. The ability to write code is becoming a basic skill, like being able to use PPT.

02 Like the Gutenberg Printing Press

Coding is changing from a professional skill to a basic ability. This can be compared to the printing press in 15th-century Europe. Boris believes that AI’s impact on software is an accelerated version of the printing press revolution. Software will be completely democratized within a few decades, becoming something that anyone can master. Eventually, being able to create software will be as natural as sending a text message.

03 What Abilities Are Most Important?

When the barrier to writing code is lowered to an extremely low level by AI, what truly distinguishes a person’s ability is their product sense and a genuine understanding of a specific field. After AI flattens the execution threshold, the gap in judgment is magnified.

This directly rewrites the meaning of the word generalist. The future generalist is a cross-disciplinary full-stack. Some people understand product, design, and engineering at the same time. The Claude Code team is like this: engineering managers, PMs, designers, data scientists, finance, user research, everyone is writing code. With AI assistance, writing code has become a common language for everyone.

04 SaaS Moats Are Eroding

In the past decade or so, there have been several consensuses in the SaaS industry that have almost been taken as axioms, namely switching costs and workflow lock-in. But with a sufficiently strong model, the logic of things begins to change. Now, you can directly throw the interfaces and data structures of both sides to the model, let it figure out the mapping relationship itself, and what originally took several months may take only a few days to run a usable version.

Models like Opus 4.7 excel at understanding a complex process, breaking it down, and reassembling it in a new environment. Therefore, the moats built on data accumulation and process accumulation in the past are eroding. For those who are doing SaaS, this may be bad news. But for all customers using SaaS, and teams preparing to build the next generation of SaaS, this is a real window of opportunity.

05 The Best Era for Entrepreneurs

The number of startups that truly disrupt industries in the next 10 years may be 10 times more than in the past 10 years. Small teams can use AI to create products that are at the same level or even better than those of large companies. Conversely, it is a liability for large companies to truly use AI. Because the muscles that made money for large companies in the past are now stuck on the road to AI truly realizing its value.

06 MCP Will Not Die

MCP will not die. It allows the model to be directly connected and used, and the model can understand and adjust itself without a programmer translating for it. Boris calls API the Human Developer Interface and MCP the Model Interface Protocol. In the AI era, all services should be MCP-ized by default.

07 Computer Use Is Still Important

Many people now talking about Computer Use feel that this direction may not work. But Boris sees a completely different level. What he really values is that Computer Use solves the biggest pain point in AI implementation: in the real world, there are a large number of systems that have neither API nor MCP. These systems will never wait for a perfect API to save themselves. In the short term, major models should still be improving their Computer Use capabilities.

[AI Product A Ying]

RichSilo Exclusive Analysis:

AI’s Disruptive Impact on the Crypto Landscape: Code, Moats, and New Opportunities

Boris Cherny’s insights at the Sequoia Conference aren’t just about the future of software development—they signal profound shifts that will reshape the blockchain and crypto markets. As a crypto market analyst who has witnessed multiple paradigm shifts, I believe Cherny’s seven judgments provide a roadmap for understanding the next wave of innovation and investment opportunities in the intersection of AI and blockchain.

Code Abundance and the Democratization of Development

Cherny’s observation that “code is no longer scarce” resonates powerfully with blockchain’s trajectory. If AI can automate traditional coding, smart contract development faces similar disruption. This democratization threatens the competitive advantages enjoyed by established dApp development teams and could accelerate the emergence of no-code/low-code blockchain platforms.

For crypto investors, this means reevaluating development-focused projects. Platforms that facilitate AI-powered smart contract creation—such as thirdweb or Alchemy’s suite of tools—may gain significant traction. The value proposition shifts from “who can build better” to “who can conceptualize and manage AI agents more effectively.”

The Gutenberg Moment for Blockchain

Comparing AI’s impact to the Gutenberg printing press is particularly apt for blockchain. Just as printing press democratized knowledge, AI is democratizing software creation—including blockchain applications. This suggests we’re on the verge of an explosion of user-generated dApps and protocols, potentially bypassing traditional venture capital gatekeepers.

For investors, this means focusing on infrastructure that supports this democratization: decentralized compute networks, AI-friendly blockchain frameworks, and user-friendly development tooling. Projects like Bittensor and Render, which provide decentralized compute resources, may benefit significantly.

Product Sense as the New Moat

When execution barriers fall, judgment becomes paramount. In blockchain, this implies that tokenomics, user experience, and real-world utility will increasingly differentiate successful projects from mere technological achievements.

This creates opportunities for projects that combine deep industry expertise with strong token design. Data DAOs that aggregate specialized, verifiable data—particularly in niche verticals—could become valuable as AI systems seek high-quality training data. Investors should scrutinize projects not just for their technical merits but for their understanding of market needs and their ability to translate that into effective token incentives.

Eroding Moats and the Rise of Composable Finance

Cherny’s point about SaaS moats eroding has profound implications for DeFi and blockchain-based services. The ability of AI to rapidly map and integrate between different systems threatens the competitive advantages enjoyed by established protocols. This plays directly into blockchain’s strength of composability.

Investors should favor protocols designed for maximum interoperability and those that can leverage AI to enhance their services. Modular blockchain solutions, which provide specific functionality that can be easily integrated, may outperform monolithic approaches. Additionally, oracles that facilitate seamless data transfer between AI systems and blockchains will become increasingly critical.

The Entrepreneur Renaissance

The assertion that we’re entering “the best era for entrepreneurs” aligns with blockchain’s promise of democratizing access to capital and tools. Small, agile teams with domain expertise could outcompete larger, slower-moving incumbencies by leveraging AI for development and blockchain for tokenization.

This creates opportunities for niche protocols that solve specific problems rather than attempting to build all-encompassing platforms. Investors should seek teams with deep domain expertise combined with technical prowess, as these “cross-disciplinary full-stack” teams will be best positioned to leverage the AI-blockchain convergence.

MCP: The Hidden Infrastructure Opportunity

Cherny’s endorsement of MCP (Model Context Protocol) as the “Model Interface Protocol” suggests a critical infrastructure layer for AI-blockchain integration. While most focus is on LLMs, the protocols that enable seamless communication between AI systems and blockchain services may represent undervalued opportunities.

Investors should monitor developments in MCP-compatible blockchain infrastructure. Projects that position themselves as bridges between AI models and blockchain protocols—particularly those that can translate AI decisions into on-chain actions—could capture significant value as the AI-blockchain ecosystem matures.

Computer Use: The Practical Bridge to Legacy Systems

The persistence of Computer Use as important reminds us that AI systems must interact with existing infrastructure—much of which lacks clean APIs or blockchain integration. For blockchain, this means the most practical near-term applications may involve AI agents that can interact with traditional financial systems and Web2 platforms.

Investors should prioritize projects that demonstrate practical utility in bridging AI capabilities with existing systems, particularly in areas like legacy financial infrastructure or supply chain management. Real-world use cases that solve immediate problems will likely outperform purely theoretical approaches.

Strategic Investment Implications

Infrastructure First: Focus on providing the foundational layers for AI-blockchain integration rather than application-layer solutions.
Data DAOs: Support projects that aggregate verifiable, specialized data that AI systems can leverage for decision-making.
AI Agent Marketplaces: Platform that enable deployment and monetization of specialized AI agents for blockchain interactions.
Modular Solutions: Favor protocols designed for interoperability and composability over monolithic approaches.
Domain Expertise: Prioritize teams with deep understanding of specific industries combined with technical capabilities.

The convergence of AI and blockchain represents not just a technological shift but a fundamental rethinking of value creation, distribution, and governance. Investors who understand these dynamics will be positioned to capture the next wave of innovation in this rapidly evolving ecosystem.