Ten hypotheses, put to the test
Ten claims from the new Desktop sets were selected for one property: whether a decisive test could be run now. One held inside its model, four held in part, five failed.
| # | Hypothesis | Source set | Verdict |
|---|---|---|---|
| 01 | Nodules as galvanic oxygen cells | B01 · 99 Field Atlas | not supported |
| 02 | Ryugu amino acids from cyanide polymers | D05 · 99 Field Atlas | partly |
| 03 | Radiative cooling with water co-generation | G01 · 99 Field Atlas | partly |
| 04 | Terahertz links through ice | H05 · 99 Field Atlas | not supported |
| 05 | Marine snow as an acoustic lens | B04 · 99 Field Atlas | not supported |
| 06 | The lunar glass sunshade | H16 · 29 Original | partly |
| 07 | DNA storage repair has a lesion-density optimum | H27 · 30 Unconventional | partly |
| 08 | Dendritic shot noise is a Bayesian resource | H28 · 30 Unconventional | not supported |
| 09 | Phonon-locked time crystals | E04 · 99 Field Atlas | supported |
| 10 | Galactic-tide capture of interstellar objects | D08 · 99 Field Atlas | not supported |
Nodules as galvanic oxygen cells
\n not supported\nB01 · 99 Field Atlas
\nThe nodule couple measured in the dark-oxygen paper reaches 0.95 V. Seawater splitting at 2 °C needs 1.62 V once the paper\'s own 0.37 V overpotential is included. The mechanism fails its own kill condition by 0.67 V.
\n
\n Ryugu amino acids from cyanide polymers
\n partly\nD05 · 99 Field Atlas
\nβ-alanine is 0.57–1.04× glycine and glycine sits 12.9× above the C4–C7 trend: not Fischer–Tropsch. That supports cyanide chemistry, but the data cannot separate polymer hydrolysis from Strecker synthesis.
\n
\n Radiative cooling with water co-generation
\n partly\nG01 · 99 Field Atlas
\nAcross 1,098 nights at three sites, dew is 0.04–0.12 L/m²/night and the electricity from the same latent heat is 0.01–0.03 W/m². Water is marginal; power is one to two orders below 1 W/m².
\n
\n Terahertz links through ice
\n not supported\nH05 · 99 Field Atlas
\nCompact cold ice absorbs 0.76 cm⁻¹ at 0.3 THz; a 60 dB, 100 m link allows 0.0014 cm⁻¹. The claim is 551× over budget at the bottom of its band and 7,382× at the top. Maximum reach: 18 cm.
\n
\n Marine snow as an acoustic lens
\n not supported\nB04 · 99 Field Atlas
\nAggregate sound-speed contrast runs from −0.81% to +1.12%; eight of twelve cases are faster than seawater, which anti-guides. The 2% contrast the claim needs never appears.
\n
\n The lunar glass sunshade
\n partly\nH16 · 29 Original
\nA 10 g/m² cloud over Earth\'s disc is 1.28 × 10¹² kg, 64× the claimed 2 × 10¹⁰ kg budget. Read as flake material at 1.16% cover it is 1.48 × 10¹⁰ kg and consistent — a units ambiguity, not a physics error.
\n
\n DNA storage repair has a lesion-density optimum
\n partly\nH27 · 30 Unconventional
\nThe collision mechanism creates a threshold at 20–116 lesions per 10 kb, but storage damage takes 11–1,096 years to reach it. Over-repair is harmful only for very old libraries.
\n
\n Dendritic shot noise is a Bayesian resource
\n not supported\nH28 · 30 Unconventional
\nA Bayes-optimal decoder loses accuracy monotonically with noise (0.591 → 0.587 at σ = 1). A threshold detector peaks at σ = 0.75 — stochastic resonance, not Bayesian sampling.
\n
\n Phonon-locked time crystals
\n supported\nE04 · 99 Field Atlas
\n256 oscillators lock to a common subharmonic above K ≈ 2σ: thresholds 0.2, 0.5 and 1.7 against predicted 0.2, 0.6 and 2.0. Supported inside the phase model, not in diamond.
\n
\n Galactic-tide capture of interstellar objects
\n not supported\nD08 · 99 Field Atlas
\nAt 1,000 AU the tide changes an ISO\'s speed by 8 × 10⁻⁶ m/s against an escape speed of 1.33 km/s. The ratio is 6 × 10⁻⁹ — too weak by four to eight orders of magnitude.
\n
\n