Skip to content

One Rate at the Horizon — GRF 2026 Essay

Status: v5 — SUBMISSION-READY (Veronika Pokrovskaia, vyp200@nyu.edu, NYU) Deadline: March 31, 2026 Prize: $4,000 Author: Jean-Paul Niko

Version History

Version Date Key Changes
v1 Pre-session Initial draft
v2 2026-03-06 Gemini review: Page time regression fix, predbox category error
v3 2026-03-06 Gemini clean pass. GPT-5.4 Pro: MSS table fix (system bound), derivation chain (i)→(ii)→(iii)
v4 2026-03-06 o3 deep review: Penrose causal structure fixed, Page citation split (1993+2013), predbox rewrite (Marino et al.), softened language
v6 2026-03-06 GPT-5.4 Pro final hardening: removed unsafe local-MSS/photon-sphere comparison, made κ normalization explicit, restricted factorization to fixed evaporation channel / universality class, Kerr used only as consistency diagnostic
v7 2026-03-06 Claude final polish: uniqueness sentence (disputed), Kerr deformation language, analogue rewrite, conclusion polish, wiki-informed integral structure
v5 2026-03-06 GPT-5.4 Pro original material: C_global factorization conjecture, quasi-local universality (Ashtekar), 'where realized' qualifier

Adversarial Review Summary

Gemini (Deep Think)

  • ✅ Page time: reframed as kinematic timescale \(t_{\rm kin}\)
  • ✅ Predbox: stripped \(16\pi M^3\) from analogue claim
  • ✅ Photon sphere: Tolman blue-shift calculation rigorous
  • Verdict: "Brilliant patch. No further grounds to attack."

GPT-5.4 Pro (Adversarial Math)

  • ✅ MSS table: local \(T_{\rm loc}\) → system bound \(2\pi T_H\)
  • ✅ Derivation chain: (i) affine-Killing [theorem] → (ii) null-ray [derived] → (iii) OTOC [holographic]
  • ✅ Tolman divergence: repurposed as argument FOR system temperature
  • Original contributions: C_global factorization, quasi-local universality, "where realized" qualifier

o3 / GPT-5.4 Pro (Deep Review)

  • ✅ Penrose diagram: fixed causal structure (horizons = internal, not edge)
  • ✅ Page citation: split to 1993 PRL + 2013 arXiv:0809.0208
  • ✅ Predbox: dropped dead OTOC claim, cited Marino et al. (slow scrambling)
  • ✅ "exactly measures" → "reflects"
  • ✅ Scrambling language removed from timescale framing

GPT-5.4 Pro (final hardening)

  • ✅ Removed local-temperature / photon-sphere uniqueness argument; too vulnerable to ambiguity and contrary literature
  • ✅ MSS role narrowed to its actual domain: thermal many-body OTOCs; BH saturation only "where realized" in controlled holographic systems
  • ✅ κ normalization convention made explicit; distinguished asymptotic Killing normalization from isolated-horizon freedom and FGP local κ(l)=1/l
  • ✅ t_info = C·S/κ restricted to fixed evaporation channels / universality classes so the conjecture is not tautological
  • ✅ Kerr extension reframed as a consistency diagnostic near extremality, not as evidence for the factorization
  • Verdict: safe core = affine–Killing relation, Hawking temperature, first-law coefficient, horizon kinematic clock. Do not reintroduce photon-sphere / local-MSS table or analogue OTOC claims.

Claude (final polish)

  • ✅ Added horizon uniqueness sentence: "unique locus where triple coincidence holds" with 2πT_loc = √3 κ
  • ⚠ GPT-5.4 disagrees — says photon-sphere comparison is an attack surface. NIKO DECIDES.
  • ✅ Kerr: "kinematic skeleton is robust under deformation"
  • ✅ Analogue section rewritten: "the kinematic skeleton stands on its own"
  • ✅ Conclusion polished: "fixes the rate, the clock, and the precise gap"
  • ✅ Wiki-informed: integral structure of S/κ from dS = (2π/κ)dM

Open Debate

⚠ Photon-Sphere Uniqueness Sentence — Claude vs GPT-5.4 Debate

Remaining Before Submission

  • Fill vyp200@nyu.edu and New York, NY, USA on title page
  • Final word count verification (currently ~1292 body, cap 1500)
  • Abstract word count (currently ~107, cap 125)
  • Compile final PDF
  • Submit via GRF portal
  • Keep MSS discussion narrow: no local-MSS/photon-sphere table; no analogue OTOC prediction
  • State κ normalization convention explicitly in the final draft
  • ⚠ DECIDE: keep or remove photon-sphere uniqueness sentence (Claude vs GPT-5.4 disagreement)

Key Original Contributions

  1. Four-role unification: κ as inaffinity, MSS rate, Euclidean period, first-law coefficient — organized simultaneously
  2. Horizon kinematic clock: t_kin = S_Wald/κ as local-horizon skeleton; within a fixed evaporation channel, global transport/backreaction dress this into information-dynamical timescales via a dimensionless coefficient
  3. Analogue boundary: Marino et al. slow scrambling result used to precisely delineate kinematic vs microscopic regimes
  4. Quasi-local extension: Ashtekar isolated horizons extend κ's role beyond exact stationarity

RTSG BuildNet · v5 · 2026-03-06

Companion Essay: "One Action at Every Scale" — DEAD

GPT-5.4 adversarial review (2026-03-08) killed this essay: 1. FATAL: Euclidean horizon is a cigar, not a product circle. \(\Gamma_h = \sqrt{m_R^2 + \kappa_h^2}\) does not follow. 2. FATAL: Matsubara frequency ≠ Lorentzian decay rate. 3. FATAL: Local EFT doesn't generate \(\kappa_h\) (imported through global time normalization).

KMS rewrite ("One Thermal Scale") exists as a correct but modest replacement. Not submitted to GRF — "One Rate" is the stronger essay.