/biz/ - Business & Finance

A reminder that no one needs oracles:


Any project that tells you they provide oracle services isn't actually providing them. They're just saying they think the answer is correct based on popular voting.

Let's say you're trying to collect rainfall data for a city. Normally you'd rely on a central, trusted source for rainfall data. A weather station or science lab with a few collection points set up through the city.

It's an easily manipulated scenario, but it's otherwise accurate and auditable.

Then some autists come to town and they say they can collect rainfall data better. They set up twice as many rainfall collection points through the city and decide that they're going to popular vote on the rainfall total. The votes are weighted by seniority to protect against manipulation.

The most senior rainfall collection node has a quirk in its reporting. A tree in a neighbor's yard prevents it from collecting much rainwater. It starts reporting half of what it normally does. Lower ranked nodes start seeing that they're not being rewarded for their "incorrect" voting and decide to modify their collection to better match the senior rainfall node.

Soon, a high majority of all nodes are reporting this universal "fact". Meanwhile, the false data is being reported to anyone buying.

This is a simple demonstration of why popular voting with trustless data collection is worthless.

Multiple trusted independent sources aren't perfect, but it's still a better solution.