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.
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.
Knowledge lineage
Every claim carries provenance — a transparent chain you can trace all the way down.
What this could become
A living archive
Science that keeps updating instead of freezing at publication.
An AI-readable graph
A knowledge base models can query, extend, and build on.
Autonomous replication
A platform where results are independently re-run, not just cited.
Executable research
A repository of work that runs, not just work you read about.
A marketplace of ideas
Where scientific hypotheses are proposed, tested, and traded on merit.
A shared memory
A global record of discovery for humans and machines alike.
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.
Let machines generate what is possible. Let humans decide what should be believed.
Human judgment. Machine discovery.
knowledge beyond either alone