[2023] Good death

My highlights:
If a protocol is alive, it lives through the activity of its participants, and dies by metaphor.

A World is something like a gated garden. Has borders, laws, values, dysfunction, can grow up, has members who live in it, manifests evidence of itself in its members, emissaries, symbols, tangible artifacts and media. Yet it is always something more. A world is an ecosystem of avatars, grown through their time and attention. Not every protocol can die because not every protocol lived, as a world, in the first place.

When we use death as a metaphor, we reveal something is gone. It cannot be recovered. On these moments of grief we can see–perhaps more than at any other time–what characterizes the life of these worlds. Protocols live through attention: they die from neglect.

“When a child loves you for a long, long time, not just to play with, but REALLY lives you, then you become Real.” “Does it hurt?” asked the Rabbit. “Sometimes”.

Emotional investment is not obviously quantifiable, but it has reality.

Sometimes a world will appear to be dying, then come back to life. While a world is dying, one can argue about its relative deadness or debate different death criteria. In general, the end of the dying process takes the shape of a decision, consensus death.

Death-as-a-process: death is not a simple binary transition, it is a process, dying, that may result in a protracted liminal state between alive and dead. Possible disagreement or ontological uncertainty about relative liveness or deadness.

Death-as-a-decision: with domain-specific criteria determined by technology, authority and culture.

In the late 1800s, diagnosing death was considered medically difficult, with some doctors insisting the only inarguable marker of death was the beginning of putrefaction. Public anxiety spread about being buried alive. Over the course of the 19th century, as death became increasingly a medical and not a religious event, doctors sought ever-more precise ways to judge the moment of death. In the late 20th century, it became possible to diagnose brain death separately from the collapse of other physical systems. A body can continue in brain-dead stasis for years. Clear consensus around brain death as legitimate death are not global universals.

In most jurisdictions, there are protocols for eventually coming to consensus abut whether an absent or missing person is dead. In this intermediary period, the missing or absent person is, while likely physically dead, still legally alive. The converse can also happen, where someone is declared legally dead but is in fact alive, sometimes as part of a scam or identity theft and sometimes by administrative error. Useful edge cases for thinking about what it is exactly that dies.

Rite of passage: a transition from one state to another and three phases: separation from the previous state, liminality between states, and incorporation into the new state.

Torajan families care for the mummified bodies of their relatives, sometimes keeping them inside the home and referring to them as not dead but merely sick until the final funeral rite is concluded. Bears some connection to the Christian purgatory, or embalming as a way to extend the period before burial.

A second death is when an individual’s name or memory is no longer know among the living. Second loss acknowledges that second dead has been complicated by the digital.

A social decision-making process where the ontology of what actually dies is flexible. Death in a protocological context dose not (and cannot) account for the personal, subjective experience of death. We cannot ask a protocol whether it feels sick, we can only observe it symptomatically.

Who has been given authority to decide that something is dead; what rituals or protocols are undertaken to cement its deadness? In more decentralized worlds there may be no clear decisionmaker, introducing a fundamental anxiety.

There are plenty of dead blockchains. The end state that gives most clarity–all severs shut down–is very rarely reached.

Some blockchains clearly die. 99Bitcoins’ death indicators are mostly economic, and others look at official communication output–but they are missing something central to worlding. Any sufficiently healthy world will spawn activity in peripheral forums where the protocol is discussed. It will spread and animate spaces outside itself. The life of the world is not only in the world.

DAO deaths should be the subject of further study, especially as cases where there are real stakes to the decision in the form of financial assets that should be redistributed to participants, spent, or otherwise handled.

0xPARC: we inhabit Worlds–both physical and conceptual. We learn how to work and behave within them. We engage in tribalism, spatial reasoning, and territorialism, even within Worlds that live entirely in our minds. We have a sense for the boundaries of Worlds and their rulesets. The physical artifacts of such Worlds are not the reason that these Worlds are alive.

Perhaps the worst case for a world is to linger indefinitely between life and death with no emerging death-decider. One of the important functions funeral rituals can serve is reconstituting society without the lost member-- the various social roles of the dead person must be taken over by others or changed. An incomplete funeral moment can impede the grief process. Linger in liminal state between life and death in social media has adjacent harms from the lack of an end state.

The difficulty bomb had to be postponed five times, as it became clear that shipping PoS would take longer than expected. People began to see the difficulty bomb as a way to renegotiate miner compensation. A recurring reminder of death in a living world.

The community debates protocol ossification, whether current governance processes for protocol updates should wind down with an end state of no further changes.

To think about death in a digital context is to consider the archive. What kind of archive, cemetery, or memorial is appropriate for dead worlds? A first impulse might be to store all of the data that encompasses the world. Would we have captured what matters about the world? A protocol must be enacted, and a world must be lived in.

The Github Arctic Code Vault has only a shallow copy. Who wrote this code and why? What did they use it for? What kind of people worked on it and what did they disagree about?

Indeed, the thing that dies is the thing that cannot be archived. No such pretense persists with a memorial. A memorial exists as a testament to what cannot be archived. There is something uncapturable, too big, or that we want to stand in for a larger symbolic meaning.

Death itself is not the enemy. It is when a protocol cannot be discarded or changed that it becomes more dangerous. Forgetting–mortality, ephemerality, deletion–is what enables health, forgiveness, and grace.

How can we characterize the ultimate endgame of technology? Is the maximally extended body immortal?

A blockchain makes a claim about time and permanence: that it is ongoing, that it is unidirectional, that the past is immutable, that it is singular. In most cases this will turn out to be false.

The Internet Archive’s About page makes explicit the proposition to grow without end: Our mission is to provide Universal Access to All Knowledge. This mission is, of course, the source of its sacredness.

This premise–the archive that only grows and its concurrent relgiosity–is a specific sacredness related to a specific culture, one with a deep progress narrative and a growth-based economic system. The act of deleting the archive could be as sacred as keeping it forever.

Agrippa. How a culture dependent on complex mnemonic systems handled an excess of remembered material? Crowding: too many images overlapping one another in a given location, images that are too much alike, will confuse and even cancel one another out. Post-event information can modulate memory, impair recall, or create a memory of an event that never took place or of a sdetail that was never part of the original information.

Your protocol will die. What constitutes the life of your protocol? What symptoms of death indicators could be used to asses its health? These indicators may not reside in the protocol itself. Identify the decisionmaking processes that exist in relationship to your protocol and whether they would withstand an existential question.

A decision-making process can be assessed by whether it is sufficiently robust that the losers of any governance outcome retain faith in the decision-making process itself. The death decision may be the ultimate challenge.

Killswitches can reveal structures of authority and perhaps even become adversarial vectors warning “who controls this killswitch also controls the system’s survival.”

What ritual acts or archive practices will render legible the meaning of the protocol to the future?

Death is the sanction of everything that the story-teller can tell. He has borrwoed his authority from death.

May it die like a dandelion.