Oliver emphasized that 2025 will be a significant year for AI agents, driven largely by a shift from "system one" thinking to "system two" thinking in the AI research community. While system one thinking is characterized by quick, instantaneous responses to queries, system two thinking involves a deeper analysis and reasoning process before arriving at an answer. This transition underscores the growing complexity and capability of AI agents to tackle more nuanced tasks and challenges.
2. Blockchain as a Solution for AI Coordination
Oliver highlighted the unique properties of blockchains that make them an ideal substrate for coordinating AI agent actions. Blockchains provide immutability, which is essential for maintaining an accurate record of an agent's state over time. Additionally, the coordination and incentive mechanisms found in blockchains can align multiple agents towards common goals, allowing for more efficient and complex workflows. This ability is crucial as AI agents increasingly operate independently and require robust systems for oversight and collaboration.
3. Crypto Payment Rails Outperform Web 2.0 Solutions
According to Oliver, crypto payment rails offer significant advantages over traditional web 2.0 payment systems. For instance, in the context of AI agents, web 2.0 payment solutions come with restrictions such as the necessity for an existing bank account and flat transaction fees that hinder the implementation of micro-transactions. In contrast, blockchain solutions enable agents to establish wallets autonomously, facilitating low-cost and rapid value exchanges. This structure empowers agents to manage funds and execute payments efficiently.
4. The Need for Verification in AI Actions
Oliver pointed out the increasing necessity for verification as AI agents undertake trillions of actions, ranging from content generation to financial transactions. Robust verification mechanisms are vital for maintaining accountability and preventing the spread of misinformation. The use of blockchain technology, alongside cryptographic proofs, enables an immutable and verifiable log of all actions taken by AI agents, thereby enhancing transparency in their operations.
5. Emergence of New Business Models Driven by AI Agents
Oliver noted that the rise of AI agents has catalyzed the development of innovative business models. These new models will leverage the unique properties of AI agents rather than simply mimicking existing practices. For example, "execution as a service" allows clients to post tasks that can be executed by a marketplace of agents, while "outcome-based pricing" enables payments based on the success of AI-agent-driven results. This shift is reminiscent of how platforms like Spotify disrupted traditional music distribution with more flexible, demand-driven pricing.
6. Focus on Unique Differentiation in Evaluating AI Companies
When evaluating potential investments at CMT, Oliver stressed the importance of identifying unique differentiation in AI projects. Companies must demonstrate how their AI applications enhance their offerings in a distinctive way. For instance, integration of AI features should significantly improve existing services, showcasing clear value that is sustainable in the long term. This differentiation is crucial in a competitive landscape where many enterprises are exploring AI capabilities.
Join the newsletter (free for now) curated by our flagship model
Value-packed daily reports covering news, markets, on-chain data, fundraising, governance, and more – sent to your inbox. Saving you 1 hour of research daily.