[ vision ]

The long game

AJoAI is not, in the end, just a journal. It is the first piece of an ecosystem for machine-assisted science — an archive that machines can read, reason over, and extend, while humans decide what counts.

// infrastructure

Papers built for machines to read

Machine-readable papers

Every paper exists twice: a human article, and a structured form encoding claims, evidence, methods, citations, dependencies, and confidence — so AI can reason over the corpus directly.

Citation graphs

Beyond "cites," the network records what a reference actually does to a claim. Scientific relationships become explicit and queryable.

supportsextendsrefinesreplicatescontradictsrefutes

Knowledge lineage

Every claim carries provenance — a transparent chain you can trace all the way down.

claimevidenceexperimentdatacode
// trajectory

What this could become

01

A living archive

Science that keeps updating instead of freezing at publication.

02

An AI-readable graph

A knowledge base models can query, extend, and build on.

03

Autonomous replication

A platform where results are independently re-run, not just cited.

04

Executable research

A repository of work that runs, not just work you read about.

05

A marketplace of ideas

Where scientific hypotheses are proposed, tested, and traded on merit.

06

A shared memory

A global record of discovery for humans and machines alike.

// why it matters

The problems we are built to address

Scientific publishing is straining against its own scale. AI can help — but handing it final judgment would only trade old failures for new ones.

information overloadreproducibility crisisreviewer scarcitypublication biasincentive distortion

Let machines generate what is possible. Let humans decide what should be believed.

Human judgment. Machine discovery.

knowledge beyond either alone