Problems

Today, AI models breakdown and have problems

  1. Biased content from the start

Feeding the LLM biased content for creator's selfish purposes.

  1. The model produces biased, fake content

AI tries to guess information to bridge knowledge gaps and produces false content as a hallucination.

  1. New biased content published

Users do not have any guidelines to act on the biased content. 4. Fake, biased information is then fed back to the AI

AI is continually trained on generated false data resulting in model collapse

The dFusion Protocol governs AI through a Knowledge Mining

  • Community contributes knowledge

Nodes nominate new knowledge in accordance with the protocol constitution to enter knowledge pool.

  • Nodes Validate Governed By Protocol Constitution

AI powered nodes filter, validate and vote on whether new information can enter the knowledge pool.

  • Added to Knowledge Pool

If information is validated then it succesfully added to the knowledge pool.

  • Distribution of Rewards

Successful contributions reward nodes for participation

Last updated