Identity

Browsing as: Guest_36

Login

Register New User

Canonizer

Canonizer algorithm:

Algorithm Information

Filter <

As Of

include review

default

as of (yy/mm/dd):

Navigation

Help

Canonizer Main

What Is Canonizer.com

Browse

Create New Topic

Upload File

Canonizer Information

Canonizer.com is a consensus building system enabling people to build consensus where none has been possible before. It is a wiki system that solves the critical problems suffered by Wikipedia. It solves edit wars by providing contributors the ability to create and join camps and it provides a measure of information reliability by providing relative measures of expert consensus. Unlike other information sources, such as peer reviewed documents, where there is far too much information for any individual to fully comprehend (We just blew past 20K documents in the field of consciousness) this open survey system provides real time concise and quantitative descriptions of the current and emerging leading theories. Theories that have been falsified by new scientific evidence are being instantly measured to the degree experts are abandoning those theories for newer better ones. The non repetitive, continually ratcheting up process significantly accelerates and amplifies the education and wisdom of the entire crowd.

Many people jump to the false conclusion that the Goal of Canonizer.com is to measure 'truth' via popular consensus. In fact, the goal is just the opposite. Crowds tend to behave in ignorant herding behavior, not unlike sheep. Various 'camps' and religions have a strong desire and interest in anything that promotes what they believe. They are highly motivated to dismiss or ignore anything that goes against their beliefs. The goal of Canonizer.com is to enable the crowd more rapidly recognize when this is happening, making it easier for them to measure for the quality of a good new theory they may want to pay attention to, even if it is counter to their currently preferred beliefs. The bottom line being our goal is not to measure truth via popularity, but to enable emerging minority theories to be more rapidly herd above any such biased bleating of any herd.

Leading minority experts can first find each other, and then build a concisely defined consensus in a consistent language by pushing lesser important disagreeable issues that inevitably emerge and are traditionally focused on, out of the way into lower sub camps. This allows a consensus to be negotiated and built around and the focus to stay on the most important actionable issues. Since anyone can contribute (even high school students have become educated by making significant contributions) and everyone can choose (selecting a canonizer algorithm) who they think are the experts, it enables the emerging experts to have a concise and quantitative measure of what the primitive popular crowd still believes, comparable to what the emerging experts believe, so they can know, concisely and quantitatively, what kind of evidence is working and what isn't, with a focus on discovering what evidence would be required to convert more.

When an expert publishes an important paper, possibly containing arguments that convince only the author, they need to know who and how many people agree, how many don't, and why. In most all sources today, there tends to be infinitely repetitive and painful yes / no assertions on all sides, on lesser important issues, all making everyone afraid of such topics, preventing good communication. This situation with no clear concise descriptions of the best scientific theories, leads everyone to believe there is or may not be any expert consensus. In this environment the still mistaken crowd tends to easily drown out any emerging minority experts.

With Canonizer.com diversity is valued and drives the system forward, rather than destroying it. Supporters of camps can object to any proposed changes they don't agree with, filtering can then be done by the reader who can select the kind of experts and canonization algorithm they choose to prioritize things. In this way, knowing concisely and quantitatively what everyone currently believes, how the expert opinion still differs from the popular opinion, can eliminate all these communication prevention issues and fears - even making still 'religious' issues quite enjoyable for all. Such can finally enable emerging expert minorities to make progress at being heard above the bleating noise of the crowd at a measurably accelerating rate.

As a demonstration, the currently leading "Theories of Mind and Consciousness" survey topic is now a collaboratively negotiated, concise, quantitative state of the art representation of the working hypotheses of now approaching 50 experts and hobbyists. Building consensus in such fields is impossible. Even attempts to agree on definitions, such as consciousness, can't be done. This survey project is building more consensus than has ever been possible. It already includes, at various levels of participation, diverse experts such as, Steven Lehar, David Chalmers, Daniel Dennett, and a growing number of others. It is already indicating a surprising amount (34 out of 46) of consensus. Nothing like this has ever been possible from a crowed this diverse. Of course in order to make this map more comprehensive, requires the survey participation of all people including you, even if that is to say you are in a "we don't know yet" camp. Our goal is to track all this in real time as ever more scientific evidence eventually falsifies and forces most all experts into the one camp representing the one best theory.

That which you can measure, improves and converts.

Knowing, concisely and quantitatively, everyone else who wants what you do, and what is still standing in your way, is the hardest part. Once you know enough people that want the same thing you do, it will just happen.

James Carroll is the author of this Tech Report:

Canonized list for namespace: