AI agents are reshaping human interactions with software. With a few simple prompts, they can independently plan, execute, and critically assess a broad range of tasks. Agents can be defined as software systems that observe and act upon their environment.  With recent advancements, agent capabilities have grown by quite a lot. This leap forward in technology is reshaping AI as a force for collective benefit. With this, the infrastructure that should be interoperable, and agents should be viewed as a sort of intelligence protocol. This is where crypto’s ethos of distribution and coordination become important. Open source, co-owned governance structures are key to ensuring AI serves the public good. A future where technology isn’t a gatekept tool but a shared resource, democratically managed and transparently operated.

While early AI agents like Auto-GPT hinted at the potential, they were often caught in repetitive loops. The landscape, however, is quickly shifting. Cognition AI's Devin, a coding agent, signals these leaps. Projects in the crypto space, such as Autonolas and Giza, are also pioneering this shift. They merge these innovations with co-ownership. Building alongside, rather than for. These governance models reflect a growing trend in AI towards more advanced but also democratically owned systems.

Soon, agents will not only execute tasks but anticipate our needs, learning from habits to enhance experiences. Integrated with other smart technology like watches, vehicles and phones, these agents will be able to build out a comprehensive model of their target, enabling them to take action and make accurate predictions. As AI agents begin to handle sensitive data, privacy and security become foundational necessities for these autonomous systems. AI agents already can be found completing much of the following familiar roles:

• DeFi Strategy Management

• Autonomous World NPC Behavior

• Smart Contract Analysis and Security

As multi-agent AI systems are becoming more commonplace, governance remains a bottleneck. Crypto's experiments with governance models offer an interesting reference point, and DAOs are the testing grounds for these new paradigms. They exemplify democratic decision-making and transparent operations, and could use these characteristics to govern AI systems. Integrating AI, the “A” in DAO is put to work, enabling these novel structures to become as agile and effective as any other entity. For example, with DAO governance, we ingest tons of information but often run into an oracle problem. Continuously reading proposals and setting transactions. In an agentic system, no one should have to deal with the low level stuff like signing transactions and bridging; it’s abstracted. In this example, we see the benefits going both ways.

Past this, many agree that each country should cultivate its own AI infrastructure as you just can’t outsource your intelligence infrastructure to outside countries. This stance acknowledges a world increasingly reliant on AI, and underscores a desire for sovereignty of agents. While both AI and crypto promise greater global connectivity, governments are wary of entrusting their sensitive information to agents produced by centralized AI corporations. This tension spotlights the advantages of decentralized, open-source models. More importantly, if more of the digital landscape transitions to crypto-adjacent networks, AI agents incapable of interacting with and synthesizing onchain data at the architecture level will fall behind in terms of technical capabilities and security.

This promise of a landscape filled with open-source AI agents is transformative: economic benefits and trust shared by many. But, achieving this vision is no small feat. The complexities of open governance require meticulous design and robust legal frameworks. Crypto has been at the forefront of this challenge for a while now and still doesn’t have it all figured out. But, it’s an ideal testing ground for governance models that could underpin future agentic systems. Soon, there will be more agents on the internet than humans.

AI agents are reshaping human interactions with software. With a few simple prompts, they can independently plan, execute, and critically assess a broad range of tasks. Agents can be defined as software systems that observe and act upon their environment.  With recent advancements, agent capabilities have grown by quite a lot. This leap forward in technology is reshaping AI as a force for collective benefit. With this, the infrastructure that should be interoperable, and agents should be viewed as a sort of intelligence protocol. This is where crypto’s ethos of distribution and coordination become important. Open source, co-owned governance structures are key to ensuring AI serves the public good. A future where technology isn’t a gatekept tool but a shared resource, democratically managed and transparently operated.

While early AI agents like Auto-GPT hinted at the potential, they were often caught in repetitive loops. The landscape, however, is quickly shifting. Cognition AI's Devin, a coding agent, signals these leaps. Projects in the crypto space, such as Autonolas and Giza, are also pioneering this shift. They merge these innovations with co-ownership. Building alongside, rather than for. These governance models reflect a growing trend in AI towards more advanced but also democratically owned systems.

Soon, agents will not only execute tasks but anticipate our needs, learning from habits to enhance experiences. Integrated with other smart technology like watches, vehicles and phones, these agents will be able to build out a comprehensive model of their target, enabling them to take action and make accurate predictions. As AI agents begin to handle sensitive data, privacy and security become foundational necessities for these autonomous systems. AI agents already can be found completing much of the following familiar roles:

• DeFi Strategy Management

• Autonomous World NPC Behavior

• Smart Contract Analysis and Security

As multi-agent AI systems are becoming more commonplace, governance remains a bottleneck. Crypto's experiments with governance models offer an interesting reference point, and DAOs are the testing grounds for these new paradigms. They exemplify democratic decision-making and transparent operations, and could use these characteristics to govern AI systems. Integrating AI, the “A” in DAO is put to work, enabling these novel structures to become as agile and effective as any other entity. For example, with DAO governance, we ingest tons of information but often run into an oracle problem. Continuously reading proposals and setting transactions. In an agentic system, no one should have to deal with the low level stuff like signing transactions and bridging; it’s abstracted. In this example, we see the benefits going both ways.

Past this, many agree that each country should cultivate its own AI infrastructure as you just can’t outsource your intelligence infrastructure to outside countries. This stance acknowledges a world increasingly reliant on AI, and underscores a desire for sovereignty of agents. While both AI and crypto promise greater global connectivity, governments are wary of entrusting their sensitive information to agents produced by centralized AI corporations. This tension spotlights the advantages of decentralized, open-source models. More importantly, if more of the digital landscape transitions to crypto-adjacent networks, AI agents incapable of interacting with and synthesizing onchain data at the architecture level will fall behind in terms of technical capabilities and security.

This promise of a landscape filled with open-source AI agents is transformative: economic benefits and trust shared by many. But, achieving this vision is no small feat. The complexities of open governance require meticulous design and robust legal frameworks. Crypto has been at the forefront of this challenge for a while now and still doesn’t have it all figured out. But, it’s an ideal testing ground for governance models that could underpin future agentic systems. Soon, there will be more agents on the internet than humans.