Evidence-Based Knowledge Aggregator

A system whereby self-contained questions are interconnected within a massive knowledge network evolving inside a decentralized brain.

Bioinformatics, Python, JS, Big Data, ML, AI, Blockchain.

Team = 5 + 3 Advisors -> https://intellart.ca/#/team

PhD-level, Post-doc-level.

Background

I am part of the team behind the 'osNFTs for science popularization' project from fund 6 and the current fund 7. However, this is an independent submission.

I was always bothered by dogmatic concepts in science that seem to go unchallenged, for 2 reasons:

• It is almost impossible for a human being to be 100% objective when presented with evidence that contradicts their life's work.
• Dogmatic knowledge that would never be challenged seriously can cause skeptics to fall into conspiratorial thinking.

Apart from mathematical proofs, any hypothesis in science stays valid as long as it hasn't been disproven. For example, the Theory of Evolution has such an overwhelming body of evidence for its existence that almost no serious scientist doubts its veracity, yet we have an immense number of people who would never acknowledge it. Most of the time, conversations between non-believers and believers are cut short when it is clear that both individuals have a dogmatic view of the subject. That is a shame, especially when there is such a monumental amount of empirical data that can be discussed. The only problem ? Easy access to it.

Imagine a system where any question that can be answered with a simple 'yes' or 'no' can be asked, and both answers have a percentage attributed to them depending on prior evidence. To continue with our example, the question: "Is the Theory of Evolution correct ?" would require answering more questions, such as: "Can new genetic mutations appear in offsprings ?" and "Can 2 different species evolve from the same ancestral species ?". Each of the subquestions would have its own score (Yes/No percentages) and the weighted average of those scores would give the original question its score. This way, any question should be able to be answered with the right evidence backing it.

Purpose of this proposal

Unfortunately, after discussing with our advisors, we have decided to postpone this side project in order to focus on our main project. However, instead of withdrawing this proposal, I will write a report detailing and conceptualizing the way we would eventually tackle this massive problem, for when we eventually begin working on it. It will also serve as an opportunity for anyone who would be interested in collaborating with us on this project to reach out, you can send us a message at https://twitter.com/opensciencenft

Execution

Building this project will require a decentralized system using AI-DSL for interconnecting various AI agents trained to find relevant information for specific questions, and smart contracts that let humans interact with those agents and curate and validate information, and who would be compensated for their work.

Budget

50$/hour for 10 hours (1 day) to write the description and technical specifications for the system.

Total = 500 $