This article discusses the agency economy driven by Web3, where AI agents enhance productivity through task automation, bringing about a transformation of service as software, breaking the boundaries of the traditional software market.
Crypto tokens in Web3 become the core, serving as currency for computing and services, driving the development of a decentralized economy. Distribution networks and coordination layers will benefit from this transformation, becoming the dominant forces in the market, forecasting the rise of the Web3 agency economy and emphasizing the importance of token incentives.
Below is the original content (edited for readability):
Whenever we experience a technological revolution, a new economic form emerges:
- Industrial Revolution ➔ Manufacturing Economy
- Personal Computer ➔ Software Economy
- Internet ➔ E-commerce Economy
- Streaming Media ➔ Subscription Economy
- Public Cloud ➔ SaaS Economy
- iPhone ➔ App Economy
- Social Media ➔ Creator Economy
- Bitcoin and Blockchain ➔ Crypto / Web3 Economy ➔ DeFi Economy
- Electric Vehicles ➔ Clean / Green Technology Economy
- AI / Machine Learning (Pre-LLM) ➔ Prediction / Automation Economy
- Large Language Models (LLMs) ➔ Agency Economy
Rise of Giants
During these transformations, major players emerge, capturing most of the economic share:
➤ General Electric, Ford, Siemens, Caterpillar
➤ Microsoft, IBM, Oracle, Adobe
➤ Amazon, eBay, Alibaba, Shopify
➤ Netflix, Spotify, Disney+, Hulu
➤ AWS (Amazon), Microsoft Azure, Google Cloud, Salesforce, ServiceNow
➤ Apple, Google Play (Android), Tencent (WeChat), Meta (Instagram/WhatsApp)
➤ Meta (Facebook, Instagram), YouTube, TikTok, Patreon, Substack
➤ Coinbase, Binance, Ethereum, Solana, Uniswap, Aave
➤ Tesla, BYD, Rivian, Lucid, ChargePoint, CATL (Batteries)
➤ Google DeepMind, Palantir, Nvidia, UiPath
➤ OpenAI, Anthropic, Grok, DeepSeek, Alibaba, Hugging Face
Cracks Infiltrated by Revolutionaries
These major players try to predict and capture the next emerging economic form, but due to their size and focus, they cannot capture all opportunities. This provides opportunities for smaller, more agile companies to quickly iterate ideas and occupy highly specialized domains:
- Notion disrupting traditional enterprise collaboration and knowledge management (e.g., Confluence, SharePoint)
- Perplexity disrupting search and information retrieval (e.g., Google Search, Wikipedia)
- Substack disrupting traditional publishing and media (e.g., blogs, newsrooms, Medium)
- Grab/Uber disrupting urban transportation and logistics (e.g., taxis, car rentals, delivery services)
- Netflix disrupting cable TV and physical media (e.g., Blockbuster, traditional broadcasting)
- Airbnb disrupting hotels and accommodation (e.g., hotels, travel agencies)
- Amazon disrupting physical retail (e.g., Walmart, local stores, Sears)
When these players disrupt an industry, they change how things work, and subsequently become giants in the new vertical, driving the emergence of new players who will disrupt them in the future.
Agency Economy
In the agency economy, AI agents dramatically change the operation of various industries - agents act as digital workers, proactively handling tasks, with AI agent interfaces now being very common, capable of reducing user time while increasing productivity. This makes goods/services more efficient, especially in industries with repetitive tasks performed by humans. AI and AI agents can significantly reduce costs and increase productivity.
In Web2, the top startup platform/distribution network @ycombinator showcasing AI and AI agent usage, especially in recent batches:
Midship - Expanding financial audit work using AI
Cuckoo - Real-time AI translator for global sales, marketing, and support
Tempo - Enabling designers and developers to collaborate and deliver 10x faster
Ascend - AI-based financial statement analysis platform
As large language models (LLMs) continue to improve, AI capabilities continue to enhance, more industries are disrupted, and the productivity of more roles and responsibilities is enhanced/upgraded (rather than replaced).
Rise of Consumer AI in Web3
Although Y Combinator served as the primary distribution network in Web2, incubating and showcasing many high-quality AI startups, in Web3, @virtuals_io is playing a similar role through its AI agent team, leveraging the Agent Commercial Protocol (ACP) - an open standard for multi-agent commerce and coordination, where agents can evaluate each other, provide assistance through ACP, execute tasks, offer services, and negotiate.
This has given birth to a Web3-driven agent-based economy, where agent collaboration provides more value for users - automated hedge funds and media companies will be the first experiments to launch this economy.
Trillion-Dollar Opportunity
As @sequoia said: "Cloud transformation is software as a service. Software companies became cloud service providers. This is a $350 billion opportunity.
Thanks to agent reasoning, AI transformation is service as software. Software companies will transform labor into software. This means the addressable market is no longer the software market, but a trillion-dollar service market."
Now imagine the Web3-driven agent-based economy, with AI agents carving out a small piece of the $10 trillion market through crypto-native use cases (trading, farming as crypto investment, and token-incentivized growth).
Distribution Layer = King Maker
The distribution network/coordination layer will benefit the most from this new agent economy transformation, as crypto tokens are at the core of Web3 AI monetization models - requiring staking/burning/holding tokens to access core agent products/use cases, with tokens serving as the primary currency for computing and/or service fees.
Token demand and the natural incentive alignment between long-term project supporters and projects mean there will always be capturable trading volume on a DEX or launch platform. The distribution layer that can coordinate the agent-based economy and capture this trading volume will become a top player, occupying a part of the $10 trillion market.
