PANews reported on December 14 that Ethereum co-founder Vitalik Buterin stated on the X platform that Deep Funding combines the two concepts of 'value mapping' and 'refining human judgment', among which:
1. Value mapping: Don't ask 'How much has X contributed to humanity?', but rather ask 'How much of Y's merit belongs to X?'
2. Refining human judgment: The open market of artificial intelligence will fill in all weights, and the human jury will randomly sample.
Regarding the first concept (value mapping), it has gradually become a popular paradigm in recent years for good reasons. Abstractly measuring contribution is very difficult: if you ask people how much they would be willing to spend to save N birds, they would say $80 when N=2000, and also $80 when N=200000. However, more 'localized' questions like 'Which is more valuable for C, A or B?' are easier to handle. The second concept (refining human judgment) is based on some ideas from my information finance article: condensing human judgment in information finance. Anyone can use any method (such as AI) to suggest weights for all graph edges, and a human jury will conduct a detailed analysis of a randomly sampled set of edges. The final submission that is most compatible with the jury's answers will determine the final output.