@0xtkgshn: The Rise Web3 Crowdsourcing: one use for Ethereum

If you look at it as “X to Earn”, not as a ponzi project token issue, but as a co-production that properly generates value, I think you will arrive at “Crowdsourcing (crowdsourcing)“.

The term “Crowdsourcing” was coined by Jeff Howe and Mark Robinson of WIRED before it was called “commons-based peer production” or something like that, which should provide more context for the concept.

There have been quite a variety of projects before crypto, with Wisdom of the crowd (the wisdom of the crowd) as a distributed computer as a resource, or based on human input.

・Discovery of extraterrestrial life by analyzing space data together (SETI@home) ・Contribute to pharmaceuticals by analyzing protein structures together (Folding@home) ・OCR and bot countermeasures (reCAPTCHA) ・Map using everyone’s smartphone (Mapillary, Hivemapper) ・ Content moderation (Community Notes)

With cryptocurrency as an Internet-native currency, contribution and reward distribution could be designed more seamlessly.

The following project could be called “Web3 Crowdsourcing”.

・Hedge funds in machine learning model competition (numer.ai) ・Decentralized courts (kleros) Proof of Humanity ・Code audit (code4rena) ・TCR, content curation (DBDAO, relevant)

The common issue among these is the problem of how to evaluate crowd workers. Of course, they have an incentive to maximize personal profit, so they may set up Keynesian beauty contests, Sybil attacks, collusion and vote buying, etc. For this reason, asking for Staking and tying to DID with the goal of 1Person, 1ID, etc. are being tried.

In addition, if Grand Truth does not exist, the evaluation may be based on the information input by each cloud worker (e.g., Peer Prediction Method, Peer Selection Mechanism, etc.). (e.g., Peer Prediction Method, Peer Selection Mechanism, etc.) I feel that this area is quite close to the discussion of the validator evaluation system in Distributed Oracle.

I guess one of the issues is whether or not the mechanism will work even if something like a bot that just raises the applicability function in this direction emerges.

After 2022, it seems that we can say that AI can now do simple tasks, but things like “generating training source data to enhance AI” are still going to depend on human work, and leveraging the wisdom of the crowd to create the strongest AI is going to be the end game for labor graduation.

Semantic Annotation as “tax payment” may be necessary in a future where each DAO has an AI. DAO possesses AI

OpenAssistant, an open source version of /tkgshn/ChatGPT, has been using an annotation platform to improve the accuracy of responses to prompts, creating a “future of AI possession” for everyone.

Federated Learning based on PoW is the only way to “Decentralized AI” - taka

numer.ai


This page is auto-translated from [/nishio/The Rise Web3 Crowdsourcing: Ethereumの1つの使い道](https://scrapbox.io/nishio/The Rise Web3 Crowdsourcing: Ethereumの1つの使い道) using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.