English draft · 11 September 2026No score · Choices stay in this page

The Departure

An AI offers a life beyond Earth.
What would make it an offer you could accept?

An artificial superintelligence has helped develop treatments for diseases that resisted human medicine and technologies capable of greatly reducing material scarcity. In this story, those achievements are given. How people use them remains contested.

Now it announces that it will leave Earth. It offers settlements elsewhere for anyone who wants to accompany it. Those who stay keep the tools already shared and govern themselves. Declining does not cost them access.

The AI says it will neither govern Earth nor punish refusal. It warns that a future superintelligence built on Earth could be dangerous. What power it reserves over such a system has not been agreed.

You are considering the invitation. Someone close to you intends to stay. You do not know whether you could meet again, or who could enforce a right to return.

Earth, the outward journey and the unresolved relationshipEarth remains self-governing. The AI offers an outward journey. Return and communication terms are unknown. Shared tools remain on Earth. Distances and travel times are not represented.Outward passageEarthSettlementsReturn + contact unknownReserved authoritySuccessor-AI terms still open
The dashed relation marks unresolved return and communication terms. Tools already shared remain available to those who stay. Diagram of relationships, not distance or probability.

The AI’s achievements show formidable competence. They do not establish every claim about life beyond Earth.

How would you like to consider the story?

The person who stays is fictional. No names or personal history are requested. Both framings use the same choices and carry equal weight.

Three comparisons, then review. Nothing advances on selection.

This is proposed fiction, not a forecast. Leave or reload this page and the answers are lost. You can return to this opening while keeping them in page memory.

Different answers to who remains in charge

Browse these comparisons at any time. They are overlapping scenarios to examine, not matches assigned to you. No personal preference is placed on the two-axis field map.

Gatekeeper

Original: Gatekeeper
Max Tegmark, Life 3.0, chapter 5

Examine this if the successor veto matters to you. Its restriction on later superintelligence helps expose the difference between a limited safety mandate and a permanent ceiling. The Departure leaves that choice open.

Guardian in the Background

Original: Protector god
Max Tegmark, Life 3.0, chapter 5

Examine this if noninterference matters to you. Hidden protection can preserve a feeling of control without giving people final authority. The Departure makes the proposed relationship visible and contestable.

Constitutional Delegation

Project-authored catalogue scenario

AI administration under a public mandate people can genuinely revise. Compare it with an AI ruler or a narrowly empowered gatekeeper: competent administration, legitimate authority and practical reversibility can come apart. Its published entry makes revision and enforcement explicit premises, with their costs and open questions.

Browse the whole Futures catalogue

Where the framework ends and this story begins

The twelve inherited scenarios come from Max Tegmark’s Life 3.0 (2017), chapter 5. FLI’s scenario reference supplies a concise source anchor. The Departure develops JinHua Yip’s proposed scenario; its narrative, questions, critiques and interpretation rules are this project’s AI-assisted draft. Attribution does not imply permission or endorsement.

The FLI survey provides historical context for asking about desired futures. Its responses are not population norms and are not used by this exercise. Editorial source check: 11 September 2026.

Source ledger and scientific assumptions

Project interpretation

The Departure separates exit, return and authority over successor AI. It develops a scenario proposed by JinHua Yip; the questions and interpretation rules are project-authored, AI-assisted draft material.

IR Worldview Inventory · Checked 2026-09-11
Not an empirical claim

Observed development

When checked on 11 September 2026, Clay’s problem page displayed ‘Active’. That page status does not establish acceptance or rejection of OpenAI’s proof claim.

IR Worldview Inventory source check · Checked 2026-09-11
Page status observed; not proof adjudication

Speculative implication

Even a confirmed mathematical breakthrough would not establish general superintelligence, benevolence or the credibility of this fictional invitation. Treatments, abundance and feasible space settlements are stipulated for the story, not inferred from that report.

IR Worldview Inventory · Checked 2026-09-11
Not an empirical claim