Session #1

Mar 18, 2025

Objective and Roadmap

This session was the first official meeting of the AI Agent Lab. The goal was to establish foundational knowledge and set up tasks for the upcoming weeks. The immediate objective was to explore use cases, define criteria for excellence, and create a preliminary list of promising crypto investment prospects within the AI agent space.

Next session (on Thursday):

  • Refinement of the initial prospect list (creating a "golden list").

Defining AI Agents

Oskar provided a straightforward definition:

“An AI agent is an autonomous digital entity that interacts with its environment, processes information, makes decisions, and takes actions.”

Use Cases Discussed

The following key use cases were explored in-depth:

  1. DeFi Analysis Assistance
    • AI agents handle information overload, analyze on-chain/off-chain data, and assist in investment decision-making.
       
  2. Trading Assistance
    • Simplifying complex processes, such as decentralized exchange operations. Example: BankerBot.

  3. Bridging Fragmentation
    • Overcoming the lack of compatibility between blockchain ecosystems (Ethereum, Solana, Base, Arbitrum, etc.) to simplify user interaction.

  4. Social and Marketing ("More eyes, more buys")
    • Automation of social media personas (e.g., Terminals of Truth and meme coin marketing).
    • Built-in virality using bots (e.g., BankerBot on Twitter/X).

  5. Social and Companionship
    • AI companions addressing social isolation. Example: Elisa—AI girlfriend companion.

  6. New Economic Models ("X-to-Earn")
    • Play-to-earn (e.g., Axie Infinity), move-to-earn, and the novel "simp-to-earn" model.

  7. Agents as a Service
    • Automation of tasks such as crypto portfolio tracking, tax reporting, and copy trading from successful wallets.

Group Input on Use Cases

Participants expanded on use cases:

  • John highlighted market analysis, DeFi automation, and governance/DAOs.
  • Dan discussed Real-World Asset (RWA) tokenization and environmental modeling predictions.
  • Ewald emphasized defining clear criteria for "good" projects, including utility, sustainable business models, short- vs. long-term profitability, entertainment factors, and simplicity for users.

Tools and Methods for Discovery

Oskar demonstrated two key platforms for discovering and analyzing AI agents:

  • CoinGecko: Filtering AI agent coins and analyzing market data.
  • Cookie.fun: Advanced platform tracking AI agents by category and ecosystem, providing sentiment analysis from Twitter/X.

Participants were instructed to use these platforms, and also utilize a custom-built AI assistant (GPT-based) to rapidly analyze whitepapers and documentation.

Criteria for Excellence

The group discussed initial ideas on what makes AI agent projects succeed. Suggested factors include:

  • Developer community activity
  • Narrative and hype
  • Social media mindshare (measured through tools like Cookie.fun and Grok)
  • Ecosystem health (growth in developer community, project integrations)
  • Ease of use and onboarding for newcomers
  • Clear value proposition and use-case utility

Ewald and John suggested leveraging AI tools (like Grok or Perplexity) to analyze why past AI agent projects succeeded, using this historical data to build predictive success criteria for future investments.

Assignments and Next Steps

Participants received the following tasks:

  • Research (due Thursday):
    • Identify 2–10 promising AI agent crypto projects.
    • Use CoinGecko, Cookie.fun, and the provided AI assistant to analyze them.
    • Submit these via the provided agent analysis form.
  • Communication:
    • A WhatsApp group was created for real-time communication, questions, and interim scheduling.
  • Long-Term Focus:
    • Aiming towards creating a refined shortlist ("golden list") of high-potential AI agent crypto projects next week.

Technical Tools

Oskar also introduced participants to an AI workflow automation tool (N8N.io), suggesting it might be useful for building AI agents or automation flows in the future.

Conclusion

The session set the stage for deep analysis and collaboration, with concrete tasks assigned for the coming days. Participants were encouraged to collaborate openly through the WhatsApp group and the community platform.

Categories