Human Verification Architecture Proposal
Problem: LLM’s make it impossible to know what content is human generated. I do not wish to spend time reading blogs, articles, and tweets that are not human generated. Solution: A trusted domain signs written content that is verified via some heurisitic mechanism to have been written without the aid of technology for content. Constraints: verification service does not store any content or hash; simple attestation service only users can include a “human-generated” stamp on their published content, allowing any user to check that the content was written by a human. High level Architecture Mobile application using Google Play Integrity API or iOS App Attestation. Unfortunately, authenticity of web apps cannot be reliably verified in the way that mobile apps can. ...