MONAI

The MONAI token is the native utility asset of the Monai ecosystem. It plays a central role in facilitating the economic and computational interactions that support the protocol:

• Access to Inference: Users pay MONAI to query the AI models. These queries can be one-off or through subscription-based models.
• Incentives for Compute Providers: GPU node operators earn MONAI by providing computing resources needed to process model inferences.
• Staking and Delegation: Token holders can stake or delegate MONAI to node operators, influencing which nodes receive more queries and rewards. This staking also subjects node operators to slashing if they perform poorly or submit wasteful governance proposals.
• Governance Participation: Governance proposals can include adopting new models, hardware upgrades, or identifying issues like censorship. Proposers are incentivised through rewards from a governance treasury initially funded with 2.5% of network rewards.

The token was launched on Ethereum to enable broader accessibility. A 1:1 migration to Monad is planned once the Monad mainnet is operational.

The Monai project is led by a multidisciplinary team with expertise in AI, blockchain, and systems development:

• Jarvis (Business Development Lead) – Previously worked in AI since 2017 and has a background in venture capital.
• Dr. Henry (Chief Technology Officer) – Holds a Ph.D. in Computer Science with a specialisation in neural networks and over 15 years of experience in AI R&D.
• Josh (Blockchain Lead) – Focused on decentralised system integration and the Monai-Blockchain interface.
• Jordan (Systems Developer) – Ensures non-AI infrastructure runs efficiently.

The machine learning team includes engineers, data scientists, and researchers with prior experience in major AI labs and projects. Their collective responsibilities include refining Monai's training datasets, enhancing model architecture, and ensuring continuous performance improvements.

The team credits foundational research by OpenAI, Google, and others as vital for making advanced models possible, and draws on academic literature for moderation, safety, and bias-related challenges.