RadCat
591 posts














After researching crypto AI agents, as far as I can tell, most fall into 4 categories: 1) Infrastructure - launchpads, software kits, oracles, research tools, etc (@virtuals_io, @ai16zdao, @daosdotfun, @cookiedotfun, etc) 2) Influencers - Posting alpha on twitter, video creation, music creation, etc (@aixbt_agent, @luna_virtuals, etc) 3) Investment DAOs - Investing or trading user deposits (@Vader_AI_, @sekoia_virtuals, etc) 4) Utility - Fulfilling business functions like auditing or trade execution (@h4ck_terminal, @soleng_agent, etc) To make a comparison to DeFi, infrastructure projects are similar to Layer 1s in that they contain an ecosystem of app projects. Influencers, Investment DAOs, and Utility agents are app categories analogous to Lending, DEXes, etc. Also similar to DeFi, infrastructure projects (Layer 1s) have the most value accrual and potential to benefit from Metcalfe's Law and network effects. As a result, projects like @ai16zdao and @0xzerebro have expanded from one of the app categories into the infrastructure category. Some agent infrastructure projects like @cookiedotfun serve functions across different ecosystems and could be better compared to Chainlink or The Graph. Once agent swarms are implemented, ecosystem network effects may become stronger and the need for cross-ecosystem agent coordination protocols may arise (analogous to DeFi bridges?). Overall, it's a rapidly moving space, but this framework is helping me to make sense of the hundreds of new projects. Video overview coming soon.






