Superintelligence for Humanity

Governed cognitive infrastructure

Common context, memory, and coordination protocols that allow humans, agents, and institutions to reason together under accountable rules.

Tacit expertise and provenance

Technical and institutional primitives for capturing situated judgment while preserving authorship, context, consent, revocability, and auditability.

Plural reasoning and disagreement

ACI must preserve calibrated dissent, uncertainty, local epistemic authority, and diversity of heuristics instead of collapsing every view into one opaque answer.

Benchmarks and pilot collaborations

Focused experiments that test coordination, agency, trust, institutional learning, and cross-domain intelligence transfer rather than model scale alone.

ACI as a distributed learning system

The series frames ACI as a socio-technical system in which humans, models, tools, memories, incentives, governance, and communication channels jointly shape collective behavior.

Governance as architecture

Provenance, consent, revocation, audit trails, access rights, and local authority are treated as design requirements, not policy annotations added after deployment.

Resource-bounded coordination

Useful collectives must allocate finite attention, compute, trust, and experimentation under uncertainty, while remaining robust to faults, correlated errors, and adversarial behavior.

Agentic AI and collective reasoning

Recent work on agentic AI, reasoning models, and societies of thought motivates a serious examination of how structured diversity, debate, and coordination can improve machine and human reasoning.

01

Deep discussions

Invite focused exchange among researchers, builders, institutional partners, and civic leaders on the core assumptions, constraints, and failure modes of ACI.

02

Working workshops

Convert the discussion into artifacts: definitions, architecture diagrams, governance protocols, benchmark designs, and concrete pilot proposals.

03

Preparation for ACI

Prepare for ACI as shared cognitive infrastructure that protects human agency while enabling plural, accountable human-AI collectives to learn and act.