Past the second watershed. Of the six balances: 1 lost, 1 tipping, 4 not assessed. Position at least 78 of 100. The balances not assessed can only move it further.
- Biological degradation Not assessed Does this tool damage the living world, or the body, faster than either can recover? Not assessed: nobody has looked at this yet, so it is not counted as holding.
- Radical monopoly Tipping Can people still meet the need another way, without this tool? Tipping: the tool has begun to cause this problem, without going past the limit.
- Overprogramming Not assessed Does using this tool require training or certification that people cannot give each other? Not assessed: nobody has looked at this yet, so it is not counted as holding.
- Polarization Lost Does this tool move power or advantage to a few, at others’ cost? Lost: the tool is clearly past the limit.
- Obsolescence Not assessed Does this tool force people to replace what still works? Not assessed: nobody has looked at this yet, so it is not counted as holding.
- Frustration Not assessed Does this tool give back less than it takes, counting everything it takes? Not assessed: nobody has looked at this yet, so it is not counted as holding.
What it does
Sorts what other people and organizations post into one endless stream, ranked by a platform’s software for what will hold your attention.
Where it stops
It doesn’t. The ranking is set by the platform, and its goal is engagement, not what you asked to see. You cannot see the rules or change them.
How easily you can stop using it
Hard in a specific way: your audience, contacts, and history live on the platform and usually cannot move with you. Repair is impossible, since the tool is closed.
What is scored
- The unit: The ranked feed of a large commercial social platform: the software that chooses what each person sees. Left out: feeds in the order things were posted, feed readers (see Current), and the fediverse.
- The job it is tested against: Holding attention, which is what it is built for. It does that reliably. The platforms promise something else, to show people what matters to them, and it is not tested against that promise here; see the open questions.
- Where and when: The large platforms of the 2020s. Two of the six balances are assessed, from how the tool is built and not from measurement.
This record is ours, not Illich’s. He did not write about social media. It applies two tests adapted from Tools for Conviviality (the definition of convivial tools, chapter II, p. 32, and the second watershed, chapter I) to a tool that did not exist when he wrote: does the tool leave people free to use it as they choose, and does it demand more of them than it gives back?
Why it is here
The telephone record asks what happens when machines fill a convivial network faster than people can. The feed is a case where the network is not convivial to begin with: the company runs the lines and decides what people see.
The six balances
Scored as set out in How to evaluate tools. Tested against what it is built for, holding attention, the feed does its job (see “What is scored”), so the scoring starts from the six balances. Two can be scored from how the tool is built. Four need evidence this record does not have yet.
- Biological degradation: Not assessed.
- Radical monopoly: 1. Your audience, contacts, and history live on the platform and usually cannot move with you.
- Overprogramming: Not assessed.
- Polarization: 2. One company sets the ranking for everyone who uses the feed. They cannot see the rules or change them.
- Obsolescence: Not assessed.
- Frustration: Not assessed. The ranking aims at engagement, not at what you asked to see, but the cost of that to the people using it needs evidence.
Where the threshold falls
Past the second watershed, at 78. One balance scores 2, and any balance at 2 puts a tool past the limit: 70, plus 2.5 for each of the three points in its total.
It is a provisional judgement. The two scores rest on how the tool is built, and nothing here is borrowed from a source yet.
For a feed reader built the other way, with no ranking, no unread count, and nothing sent off the device, see the Current record.
Open questions
- What evidence would show that a feed has crossed Illich’s second watershed, as opposed to being merely annoying?
- Is a chronological feed, or one you can switch to, a different tool? Where does it sit?
- Which sources should this record rest on? Candidates include platform transparency reports and independent audits.
Sources
- Illich, Ivan. Tools for Conviviality. New York: Harper & Row, 1973. https://archive.org/details/illich-conviviality.