Hyperion Hackathon
$200K in Rewards
Claim Layer3 rewards
AI gets smarter every month, yet control over how it learns is still concentrated. A small set of platforms decides what data goes in, who benefits from the outputs, and which values get encoded.
People, teams, and institutions contribute the raw material that makes AI useful, but they rarely hold the keys. If alignment is about serving human intent, then the people who create the value need a direct way to govern it.
We take a simple approach to implement this: put individuals first, then scale. Treat every dataset, model, and agent as an ownable, governable asset with clear provenance and programmable permissions, with governance pushed to the edge where creators set policy.
The structure that makes this possible is the iDAO. iDAO will roll out on LazAI, built on Metis infrastructure.
An Individual-centric DAO (iDAO) is a lightweight ownership and governance shell around your AI assets. It is the social and economic home for a dataset you curate, a model you train, or an agent you operate.
Inside that shell, members set and update governance policies, those rules are enforced onchain. You connect to applications, and you share in the value your assets create. iDAOs turn contributions into governed property that can be trusted, reused, and rewarded.
iDAO launches on LazAI and will be powered by Metis infrastructure. Andromeda is the production Layer 2 where iDAO activity will settle and gas is paid in METIS. While Hyperion is the high-throughput testnet for agent-heavy experiments; with transaction settlement also being in METIS.
On this stack, DAT mints, permission updates, and reward distributions execute with gas covered in METIS, giving you verifiable records, programmable access, and a path to scale without giving up control.
LazAI supplies the rails that make iDAOs practical. Governance changes, such as policy updates or parameter votes, are executed through the iDAO. On testnet these flows are exercised and validated, with production settlement on Andromeda. Alith is the unified access layer that lets contributors submit data, models, or agent updates. All iDAO onchain steps settle on Metis and use METIS as gas.
A Verifiable Service Coordinator routes those updates to trusted validator quorums. Once verified, the update is anchored, the DAT reflects the new state, and future usage can share value with the iDAO that made it possible.
Every step is visible onchain, which builds confidence for partners and downstream users.
Putting assets inside iDAOs changes incentives. Contributors keep authorship and set the terms. Applications gain reliable, permissioned inputs with clear provenance.
Communities can specialize around domains and steadily improve quality because better inputs create better models, which attract more usage and send more rewards back to the people doing the work. Alignment becomes an economic property. What iDAOs unlock in Web3 AI:
If you create data, train models, or run agents, an iDAO gives your work a home and a voice. Join the conversation in the Metis Telegram, share your use case in the community forum, and follow Metis for iDAO pilot updates and release notes. Builders can start now by skimming the Hyperion and LazAI docs, then bringing a dataset, model, or agent into the discussion.
