Decentering the (Online) Self

algorithms and indexing have become so good that anything we put online seems to immediately define you and follow you everywhere, even if it’s exploratory. this challenges openness as a value and makes it very hard to work out loud. it makes it hard to be a subject, to explore yourself. to subjectify and complicate.

additional observations

• non-transitive nature of trust
• centralized models of “you.” not always the goal.
• desire for a central index?
• potential of segmented audiences (family, professional)

the desire is an online place to be in process. to be wrong. to be subjective. and what we want to avoid is heavily curated public content the shiny public blog.

different approaches:

• private index, content private by default, per-content sharing - eg. google docs.
• anonymous public content - very hard. arguably impossible.
• pseudonymous content (“open secret”) - public but dissociated. different email, nothing under your pubic name, otherwise publicly accessible.
• multiple and/or throwaway identities - PGP blogging.
• unindexed/grey content - eg. private mailing list. not publicly discoverable. known audience. hard to link to organically. need to bring people into a cone of trust that extends beyond any one piece of content.
• alternative indexes/networks - Secure Scuttlebut. Interplanetary Filesystem (IPFS). (often with different characteristics).
• balkanized content - locally public; but not universal (eg. intranet). arguably the same as grey content.
• shared identities - collective statements under a shared pseudonym. commander marcos (who is sometimes rumoured to have been played by several people) or Laboria Cuboniks or… Homer??

use cases:

• groups/communities thinking together. this can be a trusted environment, but there’s often a non-homogenous set of interests and intended audiences (eg. data trusts vs. lunar settlement). need for accounts. the non-transitive nature of trust.
• non-scaling - thinking out loud can be loud. we don’t necessarily plan to read everything by all of us, but we do want discoverability. so an email for every post can be a lot.

options for a dissociated presence, depending on how much you care to obfuscate:

• use whois privacy for the domain.
• if your setup requires authentication (eg. github) consider setting up a separate email
• set up your robots.txt to disallow indexing if you want to minimize the changes that your alter ego is itself colonized by identity constraining algorithms .