THE QUESTION
Why do we stay inside echo chambers?
The root causes, deepest first. Disagreements stay on the map.
closed · back to Actions
Viewing as guest. Enter a handle to join and take part.
This room’s map
Root causes on the left, outlined in green. Tap a claim to see its chain.
a link the room certified contested: the room could not settle it
Every claim on this map, with its own address (18)
Each page says what the room certified around the claim, with the counts, and can be cited, embedded, or read as JSON.
The merged map
Folded from 3 rooms (all demos).
why
Every room on this question built its own map; these are folded into one. The same claim raised in two rooms is one item, and each link shows how many rooms found it. Where rooms found opposite things, each finding stands, with how many of the rooms that judged the link found it.
The rooms certified 51 links between them, but no link was found by more than one independent source (rooms that share a source, or vote as one, count once), so there is no shared map yet. Each room’s own map stands alone: Why do we stay inside echo chambers? (22), Why do people like filter bubbles? (21), Why do people prefer their own information bubble? (8). A shared map appears when two rooms certify the same link; a card that travels between rooms during Generate is what makes that possible.
THE ROOMS IN THIS MAP
Why do we stay inside echo chambers?
⚠ demo · up to 5 cards word for word the same as in 2 other rooms here
counts its verdicts weigh on this map — steward trusted by this instance
Why do people like filter bubbles?
demo · stands 1 vouch from its source, trusted steward _9c10b0b91f
⚠ demo · up to 7 cards word for word the same as in 2 other rooms here
counts its verdicts weigh on this map — vouched for by the steward of “Why do people prefer their own information bubble?”
Why do people prefer their own information bubble?
⚠ demo · up to 7 cards word for word the same as in 2 other rooms here
counts its verdicts weigh on this map — steward trusted by this instance
A room counts once someone accountable stands behind it. ● settled before any map was public; ○ after.
RAISED IN MORE THAN ONE ROOM · 14
| People choose the sources that already agree with them | 3 rooms | on the map |
| Ranking optimises for time spent, not for being informed | 3 rooms | on the map |
| Ad money rewards volume of attention, not quality of it | 3 rooms | on the map |
| The people most online are the least representative | 2 rooms | not on the map: no link to it was certified |
| The rules are written for one country and applied to forty | 2 rooms | not on the map: no link to it was certified |
| Recommendation systems are not inspectable from outside | 2 rooms | on the map |
| Arguments online end in blocking, not in agreement | 2 rooms | on the map |
| Appeals go to the same system that made the decision | 2 rooms | not on the map: no link to it was certified |
| There is no shared set of facts to argue from any more | 2 rooms | on the map |
| Nobody agrees on who should decide what is true | 2 rooms | on the map |
| Local newspapers closed and nothing replaced them | 2 rooms | on the map |
| Schools do not teach how to read a source | 2 rooms | on the map |
WHAT THE ROOMS WOULD DO
Each room judged its own actions against its own root causes, at three quarters of its own votes. Nothing here is a vote across rooms: it counts, cause by cause, what each room found. A room appears once its Actions stage has closed. Each number on the map marks a root cause that action reduces.
Reduced in 1 of 2 rooms:
Ranking optimises for time spent, not for being informed
① Fund local reporting from a levy on ad revenue · Why do people like filter bubbles?
Reduced in 1 of 1 room:
Children learn what a source is from their parents' feed
② Teach source-reading in every year, not one · Why do people prefer their own information bubble?
Recommendation systems are not inspectable from outside
① Fund local reporting from a levy on ad revenue · Why do people like filter bubbles?
③ Teach source-checking in the last two years of school · Why do people like filter bubbles?
The metric was chosen because it was easy to log
④ Independent audit access under a legal safe harbour · Why do we stay inside echo chambers?
No room found an action that reduces these:
- A report takes three days and the post has already done its work · judged in 1 room
- Being wrong in front of people you know is worse than being wrong · judged in 1 room
- Regulation asks for transparency reports nobody can verify · judged in 1 room
- Researchers ask for data and legal says no · judged in 1 room
- The problem is not what people believe but that believing has become an identity · judged in 1 room
FOUND BY 1 ROOM · 51
| People no longer meet anyone who disagrees with them in person | → | Disagreeing in public costs you socially, so most people stop |
| → | Local newspapers closed and nothing replaced them | |
| → | Everyone assumes the other side is more extreme than it is | |
| Children learn what a source is from their parents' feed | → | A report takes three days and the post has already done its work |
| → | The problem is not the algorithm, it is that everyone is watching | |
| Nobody outside the company can check whether moderation is even applie | → | Distrust of one institution spreads to all of them |
| → | Ranking optimises for time spent, not for being informed | |
| Regulation asks for transparency reports nobody can verify | → | People choose the sources that already agree with them |
| → | The people affected worst are in markets we do not staff | |
| Correcting someone publicly makes them dig in harder | → | Local newspapers closed and nothing replaced them |
| Media literacy is taught once, by whoever has a free period | → | Schools do not teach how to read a source |
| Nobody has time to be a citizen after two jobs | → | People are exhausted and take the version that costs least to hold |
| Nobody outside can reproduce a single claim we make | → | Nobody agrees on who should decide what is true |
| Subscription walls put good reporting behind a price and left the rest | → | Everyone assumes the other side is more extreme than it is |
| The metric was chosen because it was easy to log | → | Ranking optimises for time spent, not for being informed |
| A report takes three days and the post has already done its work | → | Removing a post makes it famous |
| Removing a post makes it famous | → | Ranking optimises for time spent, not for being informed |
| Ranking optimises for time spent, not for being informed | → | A moderator sees the worst of it for the wage of a call centre |
| → | Everyone assumes the other side is more extreme than it is | |
| → | People choose the sources that already agree with them | |
| → | Researchers ask for data and legal says no | |
| → | The rules are forty pages and the decision takes twelve seconds | |
| People choose the sources that already agree with them | → | Every fix ships to the users who complain loudest |
| → | The same event is reported as two different events | |
| → | We cannot publish what we remove without teaching people to evade it | |
| Researchers ask for data and legal says no | → | The rules are written in one country and applied in forty languages |
| → | Recommendation systems are not inspectable from outside | |
| A moderator sees the worst of it for the wage of a call centre | → | The rules are forty pages and the decision takes twelve seconds |
| Everyone assumes the other side is more extreme than it is | → | People I know now live in two different versions of the same country |
| Every fix ships to the users who complain loudest | → | The people affected worst are in markets we do not staff |
| → | We cannot publish what we remove without teaching people to evade it | |
| → | Ad money rewards volume of attention, not quality of it | |
| The rules are forty pages and the decision takes twelve seconds | → | Recommendation systems are not inspectable from outside |
| → | Automated removal is cheaper than being right | |
| The rules are written in one country and applied in forty languages | → | There is no reviewer who speaks the language of half the reports |
| The same event is reported as two different events | → | People I know now live in two different versions of the same country |
| Recommendation systems are not inspectable from outside | → | A feed has no edges, so you cannot tell what you are not seeing |
| → | The problem is not what people believe but that believing has become a | |
| There is no reviewer who speaks the language of half the reports | → | Automated removal is cheaper than being right |
| The problem is not what people believe but that believing has become a | → | Being wrong in front of people you know is worse than being wrong |
| → | Outrage travels further than accuracy, so it is what gets made | |
| → | Arguments online end in blocking, not in agreement | |
| Being wrong in front of people you know is worse than being wrong | → | Ad money rewards volume of attention, not quality of it |
| Ad money rewards volume of attention, not quality of it | → | We built for a scale nobody had governed before |
| → | A newsroom now costs more than the advertising it can sell | |
| Nobody agrees on who should decide what is true | → | Ad money rewards volume of attention, not quality of it |
| We built for a scale nobody had governed before | → | Nobody agrees on who should decide what is true |
| Outrage travels further than accuracy, so it is what gets made | → | There is no shared set of facts to argue from any more |
| → | Arguments online end in blocking, not in agreement | |
| A newsroom now costs more than the advertising it can sell | → | Outrage travels further than accuracy, so it is what gets made |
| There is no shared set of facts to argue from any more | → | A newsroom now costs more than the advertising it can sell |