1. Web3 Social
   • Leverage ML models to solve problems of content moderation, targeting, and personalization in Web3 social.
2. Web3 Gaming
   • Build a new type of generative on-chain or Web2.5 games with AI acting as a storyteller, NPC, and a generative environment that is unique and personalized for each player.
3. Web3 Content
   • Help users, builders, and investors navigate the complex world of Web3 by building generative content, such as summaries of podcasts and research articles, or content modality conversions (e.g., convert Vitalik's latest post into a podcast or YouTube video).
4. Security
   • Use AI to help users better understand the security implications of various Web3 tools and protocols. Help developers to navigate the complex Web3 security landscape and automate auditing.
5. On-chain AI Characters
   • Build dynamic and generative characters that live on the blockchain for people to own or interact with for fun and profit.
6. Multi-Agent Systems
   • Connect multiple models using on-chain infrastructure to enable multi-agent interaction, reduce hallucinations, and ensure verifiability of the results.
7. Self-Sovereign AI
   • Enable people, communities, and organizations to build, train, and run inferences on sovereign AI models with private data.
8. Decentralized Marketplaces
   • Open crypto-powered marketplace of models, datasets, evaluations, and compute for AI development.
9. Data Provenance
   • Help users get paid for curating and selling data to train AI models. Ensure copyright compliance via on-chain or cryptographic data provenance.
10. Data Banks
    • Build systems for collaborative and crowdsourced curation of RAG datasets, data labeling, and synthetic data generation, so that the whole world can participate in and earn from training better AI models.
11. Governance of AI Models
    • Build DAO tooling for training and fine-tuning of new and better models.
12. Funding AI
    • Leverage crypto's ability to fund, tokenize, and rally people around strong ideas to fund the next generation of AI. Tokenize AI models.
13. Decentralized Evaluation and Benchmarking
    • Build a decentralized and open version of ML benchmarks. Add economic incentives for people to thoroughly test and evaluate models to produce objective scores.
14. Decentralized AI Research and Development
    • Help ML researchers and engineers to contribute to open-source models and projects and get paid via programmable crypto incentives.
15. Your Own Idea!
    • There’s no limit to what AI and Web3 technologies can do.