Research project · Misinformation, AI & cryptography
News Currencies
Working data frameworks and speculative demonstrations on misinformation — how AI tools inherit bias from their sources, and what cryptography can and cannot do about it.
News Currencies is a critical study of media ecology and of mis- and disinformation online, carried out through functional data frameworks — open on GitHub to try yourself — and speculative demonstrations.
A focus of the project is the productization of AI, and how tools built on sources such as Wikipedia inherit and reproduce their biases. Wikipedia's "no original research" policy is meant to preserve neutrality, but it also shapes the foundation on which AI's picture of the world is built. The project extends that scrutiny to the infrastructure underneath AI, and uses critical cryptography to examine the ambivalent role of encryption and blockchains — tools that can protect against disinformation and also obscure the truth.
The demonstrations argue that technical fixes alone do not reach the root causes of misinformation, which lie in the structure and governance of digital media itself. A sturdier response is distributed: digital literacy, ethical news consumption, and a critical rethinking of how platforms are run.
Code
- Critical Data Framework — the foundational codebase integrating blockchain with AI, including new ERC-20 contracts paired with native currencies and other digital assets.
- dadasWebClient_Demo — web3 client for interacting with ERC-20 contracts and liquidity-provision protocols.
- dadasSubGraph — a custom subgraph for indexing and analysing on-chain activity.
- ipfs-sqlite-blockchain — decentralized data management combining IPFS, SQLite and a blockchain.