🎙️ Charles Hoskinson — “Discord” series (P1–P4) · governance logic

ConclusionGovernance fails not because of the participants but because of the venue's design. Build a purpose-built space (first a moderated Discord), iteratively agree on ① a definition of growth → ② executive functions → ③ strategy & budget, and legitimize it via constitutional amendment. To force participation, seat a political party as the “teeth” — it registers as one DRep (likely the ecosystem's largest) and Hoskinson himself becomes a DRep, committing to vote NO on every funding proposal and every governance action from anyone who does not take part in that space.

Diagnosis ↔ Solution

The 3 conditions for effective conversation never hold on X → governance impossible. The space restores them (+ a foundation).

1. Empathy

✗ On X, ego, attack and attribution kill any understanding of the other side.

✓ Chatham House Rule — no attribution

ideas are used without exposing who said them

P3 4:24

✓ ZK anonymity — ideas over identity

prove membership only; no retaliation against ideas

P3 5:55

✓ Facilitation (human/AI)

a pre-meeting defines personas & the ideal outcome

P3 9:04

2. Shared goals

✗ There is no agreed definition of growth, so there is no common ground.

✓ Pillar ① define growth → constitutional preamble

the one shared goal, made supreme law

P3 20:48P4 6:55

✓ Governance circles + statistical inclusion

representative sampling by domain expertise

P3 7:44

✓ Knowledge graph → Cardano LLM

a single shared source of truth (public good)

P3 18:36

3. Aligned incentives

✗ On X the reward is influence (win-lose); nothing rewards cooperation.

✓ Teeth (auto-NO) + carrot (rewards)

non-participation is punished, service is rewarded

P3 17:09

✓ Token gate (ADA holders)

skin in the game

P3 15:33

✓ Pillars ③④ strategy + budget (FOs, annual)

funding flows to one yearly, structured decision instead of constant proposals

P3 23:11P4 13:10

Other — foundation & rollout

✗ What makes the above durable and legitimate.

✓ Pillar ② executive branches → constitution v2

2 → ~5 branches (+ exec, AI, dApps rep); Lessig problem

P2 17:21

✓ Nomic — rules evolve by vote

a referendum after each cycle (v1→v2→v3)

P2 15:18

✓ MVP rollout: Discord → Cardano-native

Stage 1 concept → 2 closed beta → 3 open; later fork a decentralized SNS onto the chain (12–24 mo)

P3 0:30P3 34:10P3 40:42

✓ Legitimacy: DRep vote + amendment

①+② proposed as one amendment; the party institutionalizes it

P3 49:58

▸Detail — the diagnosis (why X can't do governance)

X = a broadcast medium — reward is spectacle / dopamine / influence. Good for reach, bad for strategy, collaboration, building.

P1 0:56P1 5:43

↓

The 3 conditions for effective conversation never hold: empathy ✗, shared goals ✗, aligned incentives ✗ → only win-lose.

P1 11:35P1 13:30

↓

→ effective conversation impossible → GSC (goals/strategy/collaboration) impossible → governance impossible.

P1 15:11P1 16:04

↓

⟹ a purpose-built space (heterotopia) is required. You don't do surgery on X; you design the environment for the purpose.

P1 43:02P3 37:32

Heterotopia (a purpose-built space)

Foucault's concept: a space designed for a purpose. The library metaphor — a library where someone bangs a pot doesn't work; a venue works only with a fitting social contract + enforcement.

P1 0:33P1 3:56

X is a broadcast medium

Its purpose is to spread messages; rage-bait drives engagement (a colosseum). Reach has value, but it's wrong for strategy, goal-setting, collaboration and building.

P1 1:43P1 3:14

Why Intersect couldn't be that space

When SIP 1694 was pushed, X already dominated the discourse — a step that was missed. Membership/initiation models suit social media poorly; legitimacy was hurt when one founding entity (CF) didn't join at launch. P4: it crammed ~25 roles into one.

P1 8:01P4 4:06

No definition of growth (most important)

The single most important of '3 things that guarantee failure'. Price/MAU/devs/block-fill/SPO count — definitions proliferate. 'What the definition is doesn't matter; the community settling on one does.'

P4 6:04P4 6:55

Price-max vs growth-with-principles

'I don't care about price' is a misread (price ties to security/utility). But the goal shouldn't be price-max (unattainable, and it discards what makes Cardano special). The KPI is growth with principles — reject anything that harms decentralization even if price rises.

P1 24:38P1 27:09

The economics of attacks

Opposition = professional rage-baiters, under 100 on social-network analysis, yet they dominate the negative discourse. The point of suspicion is the rage, not the content. The aim is to demean / destroy / divide.

P1 9:01P4 16:06

▸Detail — full space design (all layers, tools, rollout stages)

Confidentiality

Chatham House Rule

Information is usable, but the speaker's identity/affiliation is not revealed = no attribution. Stops the 'leak & expose' attack on immature ideas.

P3 4:24P4 40:06

Anonymity

Zero-knowledge / Semaphore (anon voting)

Only 'I'm a member of the governance Discord' is provable; identity stays anonymous = no retaliation against ideas. Port Semaphore (V4) to Midnight; v1 uses existing tools.

P3 5:55P3 11:11

Structure

Governance circles (Holacracy)

Order/process matters. Sample from a statistically significant (representative) group into circles by domain expertise.

P3 7:20P3 7:44

Facilitation

Facilitators (human/AI) + pre→convo→post

Facilitators create goal-orientation; a pre-meeting defines info/personas/ideal outcome. Structure pre→convo→post→broadcast = auditable.

P3 9:04P3 17:46

Record

Governance knowledge graph → Cardano LLM

Structured records become a knowledge graph = a single source of truth (public good) = a Cardano LLM, integrable into every wallet.

P3 18:36P3 19:08

Entry

Token gating (ADA holders)

ADA holders only (a Cardano Collab.Land). Output is broadcast periodically (newsletter/journal/log).

P3 15:33P3 15:11

Incentives

Teeth (stick) / carrot

Stick = the party's auto-NO. Carrot = rewards for service (NFT badges). DReps anchor a proof of participation.

P3 17:09P3 20:31

idea flow / decision core

Pentland's social physics: a core of most-aligned actors + inflow of outside ideas. More idea-flow → higher success, fewer group-think failures.

P1 19:05

Pre-mortem

assume failure and predict the causes (avoids group-think).

P1 20:02

Teardown

ARC forks other chains to take them apart privately and learn their consensus/trade/dApp/ledger logic.

P1 22:56

AI (TDD for org design)

transcript → business-model canvas, pre-mortem 12 causes of death, make the AI say 'as a VC, invest? No' = enforced epistemic hygiene (red→green).

P2 41:27

Stage 1 — concept (now)

Input Output working group + a few others. Establish purpose, RACI, staff/moderators/tools.

P3 0:30

Stage 2 — closed beta

Core entities + selected people beta-test the tools + a hackathon has the community build tools + recruiting/marketing.

P3 34:10

Stage 3 — open

Rule-compliance required; many violators banned at first; survivors debate the growth metric. v1 rules drafted by Emurgo/IO.

P3 35:33P2 48:27

Cardano-native migration

Drop perfectionism, start on Discord → later fork a decentralized social network (Mastodon etc.) and bind it to the chain. Not permanent.

P3 40:42P3 41:32

🏛️ Legitimacy = DRep vote + constitutional amendment. Propose ① the growth metric + ② new governance functions as one amendment → form the party → the party institutionalizes the amendment and protects the channel. A 12–24 month blueprint. (P3 49:58 / 56:22 / 56:58)

Q&A

Q. The “teeth” (auto-NO): what exactly, and how is it justified?

Hoskinson becomes a DRep only if a party forms, then commits to vote NO on every funding proposal and governance action from anyone not in the space. Justification: public money should carry a duty of accountability + growth; he frames it as “forcing commitment, not control.”

▸detail

The condition escalates P1 → P3 → P4:

P1 / P2 56:15

A party is a product of the process: the position that wins the 'what is growth' convergence becomes the platform, the winners become members. The party registers as one DRep = likely the ecosystem's largest.

P3 12:34

Hoskinson becomes a DRep ONLY if a party exists. Condition: auto-NO on any FUNDING proposal from those who don't participate in the Discord (or whatever space is finally chosen).

P4 26:01

Going further: 'if you delegate to me, I commit now to vote NO on EVERY funding proposal AND EVERY governance action from anyone not in the Discord.' Will move to a better channel if one appears.

Justification

Taking public (treasury) money should carry not just audit but a duty of accountability and growth. Accountability = being able to remove those who break commitments. 'Participation is no longer optional (do or die).'

P3 13:31P3 46:17

His rebuttal

'Where did I say IOG owns/runs/controls it? It's an opinion, not a dictatorship — delegators approve it. Every step needs an on-chain vote, so it's not a power grab. Forcing people to commit, not forcing them.'

P4 36:11P4 26:25

Q. Who can join? What is banned?

All ADA holders are welcome if they follow the rules. Transparency-litigation and personal attacks = instant ban. Objections must come with a working alternative.

▸detail

It's not 'No Homers'

Not exclusion of specific people. ALL ADA holders are welcome — on condition of following the rules.

P3 31:41

No transparency-litigation / personal attacks

Not the venue for demands or harsh interrogation — instant ban (do that on Twitter/Reddit). Ethical concerns only as a proposal to 'add a transparency tool'. (operating-room metaphor)

P3 35:33P3 37:32

Duty to offer an alternative

You may object, but (1) state your claim accurately, (2) give a working alternative and explain step-by-step why it's better. 'I've waited 2 years; no one offers one.'

P1 49:00P3 28:25

Q. How does it connect to existing governance (constitution, Intersect, Pentad, CIP-1694)?

The 2025 constitution already allows amendment; Pentad / VC funds handle scoped tasks but not the constitution. He frames this as the last, hardest part of Voltaire.

▸detail

minimum viable governance — The constitution was ratified in 2025; we can vote (incl. amend). Next step = add growth + new functions to the preamble. P4 10:22

Pentad — Founding entities + Midnight Foundation + Intersect handle finite, scoped tasks. Constitution/ADA-holder rights/roadmap are NOT Pentad's job (giving it that power re-centralizes). P4 4:28

VC funds (Draper-style) — Already exist but the community has no say. Needs input on the thesis / coordination with elected roles — but weak. P4 22:09

CIP-1694 / Voltaire — 'The last and hardest part of Voltaire = ecosystem maturity.' SIP 1694 laid the foundation for this decision. P4 29:32

Commercial portfolio — RealFi / Pogan / Project Kaye·Blockfrost / Midnight / Midgard. The very difficulty of funding/integrating these is the architecture's root problem. P4 20:12

Nearby tech event — Laos / Ouroboros Leios testnet on June 23. 'Solved the blockchain trilemma in 10 years.' P4 7:46

Q. Why is Hoskinson driving this? His motivation?

He frames himself as a “revolutionary” after 11 thankless years, regained passion in the moderated Midnight Discord, and fears Web2.5 (KYC/custodial) is winning.

▸detail

•Self-description: 'revolutionary', 'glutton for punishment', '11 years of mostly thankless work'. P2 1:01:01
•'The twin goal of being liked/loved is the most dangerous; you need the courage to be disliked.' P2 1:01:46
•Got passion back in the well-moderated Midnight Discord (~49,000). 'The only place I can be young again.' P1 52:00P2 1:10:43
•Industry pessimism: memecoinification, Web2.5 (federated/KYC/custodial) is winning. 'If we fail, you won't even be able to run your own node.' P2 1:09:19P4 30:38

P1 = diagnosis · P2 = process design · P3 = functional requirements · P4 = rebuttals/teeth. Citation P3 12:34 = Part 3, 12:34.
Source: transcripts of 4 public Charles Hoskinson livestreams. His stated vision, not consensus or fact. Timestamps approximate.