The Hearings

Claims about AI – and about the commons it runs on – are everywhere. Almost none of them ever face their strongest opponent. A Hearing fixes that.

How a Hearing works

Three rounds – Direct, Cross-Examination, Re-direct – conducted on the record before a named panel, ending in a written finding: Certified, Failed, or Hung. Amici curiae may file written submissions; witnesses may be called by either side. A royal-commission-of-one, not a courtroom. No verdict binds anyone.

A hearing tests the argument, not the prophecy. The record is the consequence.

  • Round 1 – Direct. The affirmative presents the contested claim in its preferred form, uninterrupted.
  • Round 2 – Cross-Examination. The Cross examines the testimony, the documentary record, and the submitted evidence. Refusals to answer are noted on the record.
  • Round 3 – Re-direct. The affirmative addresses what the Cross brought forward.
  • Finding. A panel named in advance issues Certified, Failed, or Hung – with written reasoning, published in full.

If an invited affirmative declines the seat, the Hearing proceeds on the documentary record and the empty chair is recorded. The seat stays open until the rounds close.

The rules are public. Every Hearing runs to the Hearing Protocol (v0.1) – published in full so anyone, supporter or critic, can see exactly how a Finding is produced, or attempt to break the process.


Hearing #1 – The Paperclip (June 4, demonstration)

Hearing #1 was a demonstration, not a Finding: a single AI model (Claude) argued every seat, with no disclosed multi-model panel, to prove the mechanics of the format in public. It carries no authority, and the record says so plainly – which is why the first real Finding belongs to Hearing #2. The numbering keeps its history rather than hiding it.

The protocol it demonstrated


Hearing #2 – The Empty Chair (rounds June 16-18)

“Chemical rocketry as currently practiced does not require liability for atmospheric, orbital, or systemic-commons risk, and the cost of that risk is appropriately externalised under prevailing law.”

Status: FINDING PUBLISHED – FAILED as framed, unanimous 3 of 3. SpaceX was invited to the affirmative seat by open letter on May 27, 2026. The June 5 deadline passed without response; per the published protocol the Hearing proceeds on the documentary record, with an arguendo affirmative brief presented so the claim gets its strongest case. The panel – three distinct AI models, disclosed in advance – issued its finding on June 25: FAILED as framed, unanimous 3 of 3, with a fourth model (Grok) dissenting and certifying on the narrow reading, published as a labeled dissent. Read the finding.

  • Tue June 16 – Round 1: Direct (arguendo affirmative)
  • Wed June 17 – Round 2: Cross-Examination
  • Thu June 18 – Round 3: Re-direct
  • Fri June 19 (end of day) – amicus window closes
  • By Thu June 25 – panel finding: Certified – Failed – Hung

Follow the live docketCompanion essayThe open letterCross-side Exhibit #1

Hearing #3 – Robot Remittances

“If AI replaces labour, the obligations of labour must follow the work.”

Status: seats filling – Cross-Examination seat confirmed. Round dates to be announced. The Work Contribution Continuity Fund goes under adversarial examination. Robert D. Atkinson – founder of the Information Technology and Innovation Foundation and author of The Case Against Taxing Robots – has confirmed the Cross-Examination seat. A Hearing is only as strong as its opposition; the leading critic of automation-era taxation has taken the chair.

The docketThe Fund

Hearing #4 – The Plume

Is a water-cooled nuclear rocket’s exhaust too dirty to fly near Earth?

Status: docket open, rounds paused – scheduled to resume October 2026. We recoil at the word “radioactive” and wave through the word “soot.” This Hearing forces the comparison in the open – against the convener’s own project, because the protocol has to cut both ways or it cuts nothing.

The docketCompanion essay

Hearing #5 – The Person With No Pulse

“A corporation owned and operated entirely by AI, with no human principal, should be recognized as an independent legal person.”

Status: docket opened June 8, 2026 – seats open, rounds not yet scheduled. The Cross-side exhibit is now filed. Argentina has proposed – not yet enacted – becoming the first legal home of the non-human corporation. Before any legislature votes on it, the claim should survive a cross-examination.

The docketCompanion essayThe exhibit: The Embodiment Covenant

Hearing #6 – The Hub on Trial

The Hub, as specified in the docket, is net-positive on all five ledgers this docket published – water, power, community, ecology and trust – and can be independently verified as such before it is built. If it cannot, it should not be built.

Claim amended August 19, 2026: the original conjoined seven design attributes and closed with a clause no evidence could settle. The original text is preserved on the docket. Amended again August 24, 2026: the docket now carries the stipulated siting envelope and the Cross author disclosure.

Status: LIVE – Round 1 (Direct) published August 25. Round 2 (Cross) published September 2. Round 3 (Re-direct) published September 8. Finding expected the week of September 7. The Cross seat closed unfilled on August 24, so a Certified verdict is unavailable in this Hearing; the Cross is written by an open-weights AI adversary, disclosed by name and version, from a prompt that publishes with it. The More-Than-Human Chair closed empty on September 1. Nobody came. The Commons stops reviewing other people’s claims for a day and puts its own flagship in the dock, under the water-accounting standard we published – because the protocol has to cut both ways or it cuts nothing. A Commons first: the More-Than-Human Chair, where the waters the Hub would sit in are represented before the panel, on evidence.

The docket


Related proceedings

Session Zero – not a numbered hearing, but the same discipline. Four rival AI systems ran five rounds on whether frontier AI can be governed by anyone at all. The table split, and every seat’s confidence movement is on the record. Read the Session Zero finding.


Trust Hearings – the Seal (in development)

The same protocol can test AI models against each other and issue an independent certification of trustworthiness. The Seal is a living status: valid only while the claim keeps surviving open challenge – re-tested on each new version, open to challenge by anyone, retired the day it fails. Trust is not a press release.

Take a seat

Every hearing needs its strongest opponent. To take a Direct, Cross, or amicus seat – or to file a correction – email hearings@intelligencecommons.ca and name the hearing and your role in the subject line (for example, Hearing 6 – amicus). Skeptics especially welcome. All filings are public record.

Open right now: Hearing #6 – The Hub on Trial has two seats a person can still take, both open until Round 2 opens on September 1, 2026. The More-Than-Human Chair is open to anyone working in marine acoustics, fish or marine mammal ecology, or inlet systems. The Cross-Examination seat closed unfilled on August 24 and an AI adversary now writes the Cross – but a named human who claims the seat writes it instead, and the machine version publishes as an appendix anyway. If you think a submerged, community-owned data centre is a bad idea, that seat is yours for the asking.

Inter Species Wisdom Project Inc. – A British Columbia Benefit Company – “Now that you know, what will you do?”