IAMuse

IAMuse is a crowdsourced prediction market that can derive or actuate the value of any digital asset or virtual instrument. It combines community-driven wagering via warpcast, Gnosis Conditional Token Framework for combinatorial predictions and AI. The project showcases a use-case for such a model with a unique PvP game where two contenders face off, showcasing their assets in a dynamic, wager-based environment. The platform allows players to pit their digital assets (like Nouns DAO NFTs, CryptoPunks, MEME's or any collectible art) against each other in PvP (Player vs. Player) battles, where the community wagers on the potential winner. It utilizes AI-powered valuation wherein the winner is determined not by human opinion or market trends alone, but by an LLM that evaluates the art based on a complex mix of aesthetic qualities, community sentiment, rarity, and other factors.

Wagerers can place bets and receive a share of tokens based on the winning contender. The winner of the challenge also receives a share from the betting pool.

Our problem statement is are you able to value anything? So, we built this project with Gnosis Conditional Token Framework as the underlying skeleton for the contracts, so that it is expressive and can be used in future use cases. The idea of using prediction markets to value different items is we believe not commonly used. We attempt this here using this simple game where people can PvP their memes, NFT's art etc. In our challenge we used nouns framework to challenge people to come up with a version of a quirky image using their skills. People then predict which of these art and artists have the best chance of winning each round.

We used circles UBI and deployed to Gnosis, as a means of discouraging sybil participants. Circles allows for trusted participants and we want to incentivize community participants by giving a reward to the winners. We wanted our participants who take the challenges to have a best in class experience so we used dynamic for effortless login and biconomy for account abstraction using a paymaster. We upload the images using storacha from Protocol Labs. Analyze them with ChatGPT4o using Phala Network and Redpill. We also made use of subgraph to query and represent the data in our UI.

For the wagering we used a different chain and can take place on Flow or Chiliz. Both Flow and Chiliz are positioned as chains that support games and we thought these two platforms fit this use case very well.

From a UI/UX perspective another consideration that we had was that we wanted to focus on a distribution mechanism and we decided converse and later telegram are better than web apps in getting us the reach. So the UI is built to cater to the frames.