Frequently Asked Questions

Did AI write these papers?

No. An AI scaffold did not independently produce this corpus. Operator-scaffold coupling produced it — a domain expert directing a coordinated system of AI agents through thousands of deliberate, informed decisions.

The operator provides taste, judgment, adversarial integrity, and domain-level fluency. The scaffold provides coverage, enumeration, and reduction pressure. Neither produces the output alone. Removing the operator gives you generic model output. Removing the scaffold gives you normal academic pace. The coupling is the method.

How do you know the papers are correct?

We don't claim they all are. That's the point.

The tiering system exists because generation has outpaced verification. We built the infrastructure to be honest about that gap rather than pretend it doesn't exist.

Can I verify the papers myself?

Yes. The full corpus is available through the corpus browser. We encourage independent verification. If you find errors, we want to know. If you find contributions, we want to know that too.

How is this different from using ChatGPT?

A single prompt to a language model produces a single output. Operator-scaffold coupling runs thousands of directed interactions across dozens of agents, with an expert making judgment calls at every branch point. The operator recognizes plausible nonsense, kills dead-end lines of inquiry, and redirects the scaffold toward productive territory.

The difference is the difference between asking a question and running a research program.

Are you claiming to have solved the Riemann Hypothesis?

No. The corpus contains novel attacks on open problems in analytic number theory and algebraic geometry. An attack is a new angle of approach — a framework, a reduction, a connection to other results. It is not a proof. The distinction matters, and we are precise about it.

Three mathematical results have been confirmed by an independent expert as novel and correct. None of them claim to resolve a millennium problem.

Who is the independent expert?

An independent mathematician with no financial, institutional, or personal relationship with Soulmetric. Details on the verifier profile page (coming May 5).

Why not just submit to journals?

We will. But peer review operates on a timescale of months to years, and the corpus contains hundreds of papers. Traditional channels cannot verify at generation speed. Soulmetric's tiering system is a bridge — a way to be honest about verification status while making the work accessible for scrutiny now, not in 2029.

Is this replicable? Could someone else do this?

Yes — with the right operator. The methodology is transferable to any domain expert willing to direct a scaffold with adversarial integrity. The operator's domain knowledge is the irreplaceable component. The scaffold is infrastructure.

We expect operator-scaffold coupling to become a standard research method within five years. Soulmetric is the first systematic demonstration that it works.

What does "operator-scaffold coupling" mean?

Operator: A domain expert who directs the research. Provides judgment, taste, adversarial integrity, and the knowledge to recognize when the scaffold is wrong.

Scaffold: A coordinated system of AI agents — generators, critics, metacritics, verifiers, formatters — that operates under the operator's direction.

Coupling: The iterative, directed interaction between operator and scaffold that produces output neither could produce alone. Not automation. Not human-in-the-loop. A new kind of research collaboration.

How can I engage Soulmetric for research consulting?

Email kscharp@gmail.com with a brief description of your research problem. Engagements are scoped individually. Typical delivery is days, not months. See the consulting section for engagement types.