New Paradigm and Investment Logic in the AI + Web3 Era

Author: Dàshān, Founding Partner of Waterdrip Capital

Preface

In recent years, AI concepts have been everywhere, with large models and decentralized technologies intertwined, causing many to be dazzled. As a venture capital firm, we've reviewed hundreds of cross-sector projects and want to share some honest truths—including the real cycle of the AI industry, why AI cannot do without Crypto, and what we prioritize and remain cautious about when evaluating projects.

Note: The content of this article is excerpted from the roundtable discussion "AI+Web3—The Real Revolution" at the "AI + Bitcoin, the Next-Gen Revolution - BTC Vegas Side Event" held on April 28.

1. The "Seven-Layer Matrix" of the AI Industry and Investment Cycles

Although giants like OpenAI, Anthropic, and Google are currently in the spotlight and seem poised to dominate everything, our research shows that the entire AI industry is still in a very early stage.

Being early doesn’t mean you can blindly invest now, especially in overheated fields like AI hardware manufacturing, which require more careful thought. In our view, the AI industry chain can be broken down into seven distinct layers, each requiring a completely different approach to investment:

• Seventh Layer: AI Intelligent Agents

• Sixth Layer: Token Optimization

• Fifth Layer: Token Distribution

• Fourth Layer: Large Language Models

• Third Layer: GPUs

• Second Layer: AI Data Centers

• First Layer: Power Infrastructure

Take the second layer—AI data centers—as an example. This is a typical cyclical industry. Over the past two years, global capital has been aggressively buying land and building data centers, expanding assets heavily. However, these capacities take time to complete, power up, and reach the market. Based on our observations, this massive capacity is likely to come online around 2028.

This means that if everyone continues to pour money into traditional data centers next year or even now, by the time they all open in 2028, there’s a high chance of oversupply. Rents and returns will plummet, and investors who buy in at that point will suffer. Therefore, aligning with the cycle and timing is crucial.

In 2024, we see significant opportunities in the AI computing layer, especially in cross-sector tracks that leverage Web3 for computing power scheduling. We have invested considerable resources and capital into this area. Currently, one of our major AI infrastructure companies, in which we are a large shareholder, is in the final stages of preparing for Nasdaq listing.

2. Why AI and Crypto Are Inseparable

Many ask: since large models in Web2 are already so powerful, why does AI still need Crypto? In my view, this isn’t just about forcing a concept; it’s because, as AI advances, there are two issues that traditional Web2 cannot solve:

1. Recent focus on payments: AI Agents need their own "financial sovereignty"

Traditional bank cards, PayPal, or gateway payments are all designed for humans, with complex identity verification and centralized clearing processes. But future AI Agents (smart proxies) will collaborate and transact in a completely different way.

Their transactions are often extremely micro (e.g., paying $0.0001 to buy a few seconds of computing power), high-frequency, and continuous. Using traditional Visa channels, fees could exceed the transaction amount. More importantly, the issue of "payment sovereignty": AI Agents have autonomous execution capabilities. As they evolve, they will inevitably need to own and control their own wallets, rather than always relying on human bank cards or fearing API bans by centralized giants. Blockchain-based cryptocurrency payments are currently the only technology enabling machines to autonomously settle with each other.

2. Long-term trust: Preventing hallucinations and enabling trustworthy auditing

Everyone knows that large models sometimes hallucinate, producing false or biased answers. In traditional Web2 black boxes, it’s difficult for humans to verify whether an incorrect or biased response is due to underlying randomness or deliberate manipulation by large centralized companies.

This creates a real demand for blockchain. Recording key AI data, prompts, and call paths on-chain in real-time can produce an immutable, traceable, and auditable evidence system. This not only defines data ownership but also provides a reliable way for humans to supervise and verify whether AI behavior complies with regulations.

3. Investment Projects: Honesty as the Metric

Every year, thousands of AI + Web3 business plans flood the market. After reviewing so many stories, our core screening principle boils down to two words: honesty. It may sound obvious, but in today’s industry, honesty is perhaps the most scarce resource. We see honesty on two levels:

1. The team must be honest—reject packaging and embellishment

Founders and core team members’ backgrounds must be completely truthful, with no falsification or exaggeration. The industry often sees teams assembled hastily, trying to chase hot trends by slapping together big-name scientists or prestigious backgrounds, but with no real internal cohesion. Such projects tend to fall apart when facing technical bottlenecks or market shifts. A truly capable team must have complementary skills, and founders must be genuine and committed to their work.

2. The product must be honest—reject PPT-driven hype and false metrics

The actual capabilities of the product and user data must be supported by solid code, node counts, and other tangible indicators. We’ve seen many projects that simply connect an OpenAI API to a frontend, then boast on PPT that they have a "native autonomous large model," or present fake demos to deceive investors. Product development must be realistic—truly solving real problems.

In capital markets, dishonest projects driven by hype and falsehoods might temporarily inflate valuations through flashy financial maneuvers on secondary markets or exchanges. But without real underlying business, bubbles burst, and the final outcome is inevitably zero.

Conversely, honest projects that focus on solid fundamentals may appear slow initially, even somewhat naive in the eyes of speculative investors. But because they are built on a firm foundation, they often endure and grow long-term. In investing, slow can be fast; projects that can survive long-term are the ones we are willing to invest in and support strategically.

4. The Most Undervalued Opportunity by 2026: The Ultimate Fusion of AI + Blockchain + Entertainment

Regarding opportunities that are still undiscovered or severely underestimated, our research points to the deep integration of AI, blockchain, and entertainment as the most promising.

Currently, most capital is focused on AI payments and decentralized compute (DePIN)—hardcore, somewhat dull B2B infrastructure. While these sectors are important, competition is fierce, and they tend to be overlooked are the consumer-facing, entertainment-driven segments that attract mass users and deposit capital.

Our reasoning is as follows: as large models and agents mature, AI will become an efficient workforce capable of replacing most repetitive white-collar and blue-collar jobs. When productivity is vastly increased and material costs are extremely low, society will undergo a fundamental transformation—most people won’t need to work just to earn a living. When that happens, the demand for entertainment and leisure will explode. Where will people’s time and energy go? High-quality entertainment will become an essential outlet.

Future entertainment must fully integrate AI.

Take gaming as an example: every NPC will have autonomous consciousness, with memories, personalities, and social relationships. Players entering the game won’t face simple repeatable responses but will interact with "living" characters capable of emotional responses and even spontaneous on-chain transactions. AI will dramatically enhance the freedom and fun of gaming.

Blockchain’s role will be to define "property rights": it will establish land ownership, the uniqueness of rare items, and build economic order through tokens.

When people no longer need to work in real life, and devote their time, energy, and assets to such a fun, AI-driven virtual world where assets belong to players, the commercial value generated will be enormous. We are actively seeking pioneers in this direction, as it could become the next industry-wide narrative explosion.