My Website Idea

I have some ideas, and for multiplication effects (both in attracting users and efficiently providing usefulness) I think it's best to combine them into one website.

Yes, the idea is bloated and has everything but the kitchen sink. I think simplified versions, with just a few basics, are possible. Then more features could be added over time...

Basically, it's supposed to be an advanced social media website. Advanced, like comparing Photoshop to MS Paint. But it would be good to make the advanced functionality optional, because lots of people would do fine with a simplified version that was nevertheless more advanced than other websites out there. This page only lists the features. To see why I think these features are useful, see this other page.

Possible names: Explicit Culture, Long Talk.

Proposed Basic Features
The website would contain a number of modules and apps, all built on top of the basic "memes" module.

Trust Aggregation
See Trust Aggregation. Users input evaluations, and each receives their own unique set of estimates computed from this network and data.

Match Computation
OkCupid has a pretty good system for finding matches for dating, based on questions the users answer. This kind of thing needs to be generalized for other kinds of matching and alignment. This could be implemented in simple and/or advanced ways.

Kinds of matching:
 * "We're the same!" kind, like "we both work in the same industry" or whatever
 * "Asymmetrical match", like a market for exchanging

Reason Chains
This module for the creation and storage of "memes", and their associations. Memes are just statements. But they can have associations, connections to other memes. For example, semantic connections (the definition of each word), or reason chains. These features make each meme more valuable, because they contribute to each other.

Types of connections:
 * semantic/logic (including information compression/summarization)
 * empirical factual
 * goal-oriented

Apps
These features don't seem to need much extra functionality. Mostly, they would just be simplified suites for doing tasks that could theoretically be done with all the more basic modules. Some of these could be subsets of each other, I haven't figured it all out yet.
 * facts, values, beliefs
 * biography
 * endeavors
 * evaluation
 * comparisons?

Identifying Duplicates
Allow users to identify when two items are copies of each other, or close, when two pseudonyms refer to the same person, stuff like that.

List Creation and Modification
Creating lists is useful on social media. There are friends lists, block lists, or lists of items that belong to certain other categories. Creating lists should be made easy, not requiring tediously adding or subtracting items one by one. So, tools should be available to modify lists, add one whole list to another, or subtract one whole list from another, or find the intersection of the two. Copying lists others have made. And all kinds of search functions should allow for the creation of a list made out of all the results of the search. Stuff like that.

Visualization
Graphs, charts, and so on. Using data on the website.

Broader Evaluation Metrics
Maybe even allowing people to create their own values and using them to create evaluations.

Logic Verification Templates
Allow user to input parameters, and allow the computer to use them to produce verified logical syllogistic reason chains.

Bayesian Calculations
For a simple version of this, could just have a template for creating one calculation, naming and quantifying the inputs.

More advanced: it might be useful to combine aggregated evaluations with Bayesian calculations for certain truth claims, from one end of a reason chain to the other, and show the results for each "link" in the chain.

Customizable User Interface
"Palettes" of buttons, drop down menus, and so on. Creatable and shareable with lists. Also: simple and advanced versions of the website, perhaps beginner, simple, medium, and advanced.

Comparison
Perhaps an extension or advanced version of "matching and alignment", it would be useful to be able to compare large sets of data, for example the reason chains of different users or groups of users.

Fixes for Issues
Some of the proposed features, while potentially useful, are also potentially "useless". For example, if everyone can create their own metrics, how will we compare them? And what about the creation of duplicates?

Many of these issues arise because of freedom of many individuals not matching each other. So a simple fix would be to impose a solution: limiting freedom.

But I hope a more generalized set of tools could be developed to allow the users to decide how to handle things. For example, allow the creation of a computational rule/tool, then allow that rule to be shareable with other users, or bundled with sets of other rules/tool to share. As for duplicates, since users can add any semantic links between memes that they like, they can say that the two are identical, and the program should be able to then treat them that way.

See also:

 * Division of Information Labor