Michael Darius Eastwood · Independent AI alignment researcher
Published
On 24 March 2026 the author sent a single email to two addressees, Professor Lloyd Demetrius at Harvard and Professor Geoffrey West at the Santa Fe Institute. The subject was a Cauchy convergence argument for metabolic scaling. No reply has been received to date. This article explains what the letter argued and why the silence is treated as silence.
Metabolic scaling laws hold across roughly twenty orders of magnitude of body mass and are among the strongest quantitative regularities in biology. The letter proposed that the convergence properties of the scaling exponents, treated as a sequence indexed by physiological organisation, admit a Cauchy-style analysis in which the residuals decay in a way that constrains the space of admissible mechanistic derivations. The mathematical detail is not the point of this article. The point is that the argument was formal, addressed to two of the two people who could adjudicate it fastest, and sent with the manuscript hashes and OSF DOIs attached.
Professor Demetrius is one of the small number of researchers who has published critical extensions of the metabolic scaling framework, and his 2006 papers with Tuszynski are cited across the programme. Professor West is a co-originator of the framework and remains the most prominent public advocate. If either found the argument obviously wrong, the fastest correction would come from them. If either found it obviously right, the fastest confirmation would come from them. Both outcomes would advance the research; neither has followed.
The rule is explicit and it is enforced by the truth gate. Silence in reply is not treated as endorsement. It is not treated as rejection. It is a null datum. Two busy academics did not answer an unsolicited email; that is a common outcome and its inclusion in the register does not change the mathematical merit of the argument one way or the other. The metabolic-scaling section of the corpus remains flagged with the recompute-required status in the canonical facts file and is not used in external funding materials until that recompute is complete.
It records that the argument was made in advance of any external attempt to falsify it. It records the addressees, the date, the sending address, the SHA-256 of the .eml, and the attachments. It records outreach conducted openly with named senior researchers rather than through anonymous back channels. It records, in short, an attempt to be adjudicated on the merits, and the log-file of that attempt, and the outcome, which is that as of 3 July 2026 no reply has been received.
An honest engagement log is a small thing that makes several other things possible. It lets a reader see who was asked and who was not asked. It lets a hostile reviewer check whether the programme has fished for endorsement quietly. And it lets the researcher himself see, in one register, where the programme is stuck. All three functions matter. None of them requires anyone at Harvard or Santa Fe to have done anything.
From the book Infinite Architects: Intelligence, Recursion, and the Creation of Everything by Michael Darius Eastwood.