MIT Technology Review is reporting that blockchain startup Chainlink has engineered a potential solution to blockchain's oracle problem using a trusted hardware system, called Town Crier, developed by Cornell researchers.
If you're not familiar with the oracle problem, it's essentially this: EDCCs (smart contracts) are deterministic and decentralized, making them hypothetically trustless. You can know that the contract will respond to information input exactly how it is coded to, cutting out the need for a trusted intermediary to act as a guarantor and thereby allowing peer-to-peer transactions. However, these contracts often rely on off-chain data to execute. The relayer of this information is called an oracle.
For example, if there is an EDCC that will automatically transfer ownership of a given asset under the condition that the current owner dies, then the EDCC needs to receive proof of death for that asset to change hands. That proof would come by way of an oracle. Some worry that this defeats the point of blockchain, as it centralizes power to the oracle by requiring user trust.
Chainlink is proposing combining its software with the Cornell University's Initiative for Cryptocurrencies and Contracts-developed hardware system Town Crier to overcome the trust issue in certain circumstances. Chainlink would create a decentralized network of oracles to form a multi-sourced stream of online data to inform an EDCC. This way, power isn't centralized to just one information source or relayer, so it shouldn't matter as much if one oracle is dishonest, or if one website provides an oracle with inaccurate information. The Town Crier hardware would act as a secured bridge between online data sources and the Ethereum blockchain called an enclave, to ensure that the information collected by Chainlink's software isn't tampered with.
Of course, there are many types of information an oracle might provide for an EDCC, and this solution might not apply to all of them, but this certainly promises to be a solid hack at the problem.