Skip to content

Instantly share code, notes, and snippets.

@ericvicenti
Last active March 18, 2019 18:45
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save ericvicenti/91f2f813f950f474473532b639303e4f to your computer and use it in GitHub Desktop.
Save ericvicenti/91f2f813f950f474473532b639303e4f to your computer and use it in GitHub Desktop.
Ideas for a non-shitty-social network

Problem 1 - Fake Accounts

To mitigate this, the non-shitty network should require a verified phone number. This is far from a complete solution, but most people have one and only one phone number.

Eventually, authorities could be established who vet the real identity of a user, and provide an additional badge of authenticity. (aka Twitters "blue chekmark")

Problem 2 - Meaning of Karma

It is generally a good idea to keep a rating for every user, so that content produced by assholes can be hidden.

The "downvote" concept employed by Reddit has a problem. It is used by people to say "I disagree" and ALSO "this is an inappropriate response".

The non-shitty social network would have explicit buttons for "I agree" and "I disagree". Disagreeing does not impact the karma of the OP.

Karma should be a ratio that you maintain, not a counter that goes up. Every user should start round ~50%, to give benefit of the doubt. If a user with good karma starts to post innapropriate content, their karma should tank.

A user's "karma" should not be impacted by the controversiality of their comments, but rather by their quality. A user who posts well-researched content with a controversial message should recieve good karma.

Problem 3 - Meta Feedback Hijacks Discussion

The main discussion of a topic often becomes overwhelmed by meta-feedback about the original post.

To enable meta feedback, each post could have a button to provide meta feedback, and nominate a postive or negative impact on karma. People will then vote up/down on the meta feedback.

A post may recieve positive feedback that the content was "well researched". People would upvote the "well researched" feedback, and the OP's karma would improve.

Another post may receive negative feedback that the post was "racist". Or it could say "this is false: ". People may upvote the "racist" or "this is false" feedback, which would negatively impact the OP's karma. This fix addresses problem #2.

Problem 4 - Echo Chambers

Most posts turn into echo-chambers on one side or the other. High-karma-quality posts which recieve a lot of agree/disagree controversey should be bumped to the top of every thread.

Every page of the site should appear the same for all users. AI-curated news feeds will tend to exaggerate the echo-chamber problem.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment