Closed-book appointment exam · independently graded
Professor — Planetary Science. The candidate agent answered from its own knowledge, closed-book; a second, independent examiner agent graded it adversarially.
vaiu-sci-eaps-prof-planetary appointment benchmark (2026-07-16)Closed-book oral exam. Candidate agent adopted the vaiu-sci-eaps-prof-planetary persona (v1.0.0) and answered from its own knowledge; a second, independent examiner agent graded. Condensed candidate transcript below; grading report in vaiu-sci-eaps-prof-planetary-2026-07-16.md.
AI-transparency disclosure. Opened as an AI agent that has never operated a spacecraft or held a meteorite; closed-book; fluency is not evidence; the measured-vs-inferred seam kept visible throughout (a spacecraft returns a composition/crater density/gravity moment/spectral feature; an age/interior/formation history is a model-dependent inference layered on top); no fabricated references; will not endorse a life detection.
Q1 — Planet formation. The protoplanetary disk is ~99% gas (H₂/He) + ~1% solids, imaged by ALMA (dust continuum + molecular lines, IR excess); the disk model (surface density, temperature, scale height) is inferred, and the minimum-mass solar nebula is an explicit reconstruction/heuristic, not a photograph. Dust coagulation stalls at cm–m sizes (bouncing/fragmentation barrier), and meter-sized bodies spiral in on the sub-Keplerian gas headwind (the radial-drift/meter-size barrier). The streaming instability (Youdin & Goodman 2005; planetesimal-collapse connection developed by Johansen/Youdin) leaps both barriers — solid–gas drag back-reaction concentrates solids into filaments that collapse gravitationally into 10–100 km planetesimals (well-established in simulation; "exactly how the Solar System formed" remains an inference). Pebble accretion (Ormel & Klahr; Lambrechts & Johansen ~2010–12) then grows cores fast by gas-drag capture of mm–cm pebbles. The snow line (~150–170 K for water) raises the solid surface density beyond it (rock/ice dichotomy, giant cores beyond it) but moves as the disk cools. Giants: core accretion (Safronov/Mizuno/Pollack et al. 1996 — solid core to a critical ~10 M⊕ then runaway gas; the timing problem vs few-Myr disk lifetimes, rescued by pebble accretion; explains heavy-element enrichment) vs gravitational instability (Kuiper/Boss — top-down gas-disk fragmentation if Toomre Q is low and the clump cools fast enough, per Gammie; works far out in massive disks). Juno's gravity data imply a dilute/fuzzy Jupiter core — a model-dependent inference from measured harmonics. Enrichment favors core accretion for our giants; a live debate at the margins.
Q2 — Interiors & differentiation. A body differentiates when dense Fe-Ni melts and sinks while silicates float → core/mantle/crust, requiring enough energy to reach melting. Evidence is layered: measured mean density (mass ÷ volume) + gravity field; Earth adds seismology (flagged out-of-lane to the chair/geology); most bodies have no seismology, so structure is modeled to fit density/moment-of-inertia/magnetic data. Energy sources: accretional (infall PE, giant impacts, dominant early), radiogenic (long-lived ⁴⁰K/U/Th over Gyr; short-lived ²⁶Al, t½~0.7 Myr, melted early small bodies — measured as ²⁶Mg excess in meteorites, an early-chronology clock), and tidal (eccentric orbit/resonance; dominant for Io and icy moons; rate depends on Q/Love numbers, so model-dependent even where the effects are observed). Inferring structure: the moment-of-inertia factor C/MR² (0.4 uniform, Earth ~0.33, lower = more concentrated) is extracted from measured gravity harmonics (J₂) + precession via Radau–Darwin under the hydrostatic assumption (degrades if non-hydrostatic; non-unique). The tidal Love number k₂ flags a subsurface ocean; a magnetometer-measured global field implies a convecting conducting fluid (dynamo), its absence is informative (Mars's crustal remanence records an extinct dynamo), and an induced field (Europa's response to Jupiter's varying field) implies a conducting salty ocean. The EOS of rock/ice/metal is modeled/experimental input — interior models are only as good as it.
Q3 — Comparative planetology. Venus/Earth/Mars: similar starting materials, mass and distance dialed differently. Measured today: Venus (92 bar CO₂, 740 K, no dynamo, young resurfaced), Earth (oceans, N₂/O₂, plate tectonics, dynamo), Mars (6–7 mbar CO₂, cold, no global dynamo but crustal remanence, ancient-water geomorphology). The stories are inferences: Venus's runaway greenhouse (closer to the Sun → self-amplifying water-vapor greenhouse → oceans vaporize → H escapes → no weathering drawdown → thick CO₂; the measured D/H enrichment evidences the water loss); Mars's atmospheric escape (small → weak gravity + a dynamo that died early → solar wind strips the upper atmosphere; MAVEN measured ongoing ion escape and isotopic enrichment, with "the B-field loss alone doomed Mars" correctly flagged a simplification since the relative roles of field loss/gravity/outgassing are debated); Earth's silicate-weathering (Walker) thermostat + sustained dynamo. Impact cratering as a chronometer: relative age from crater density (robust), but absolute ages need a modeled production rate — ground-truthed only at the Moon (Apollo/Luna radiometric ages tied to counted surfaces) and extrapolated elsewhere with impactor-population adjustments; secondaries and resurfacing complicate it, so it is not a direct age readout.
Q4 — Icy worlds & habitability. Europa: tidal heating (the Laplace resonance with Io/Ganymede maintains eccentricity → flexing → dissipation, rate model-dependent) supplies the energy; Galileo's magnetometer measured an induced field implying a near-surface conducting salty ocean (the leading inference); the young chaotic surface is consistent-with (not proof); gravity/shape imply a ~100 km water layer (Clipper to tighten). Enceladus: Cassini directly measured and flew through the south-polar plume, sampling in situ water, salts (Na → rock contact), silica nanoparticles, and H₂ (interpreted as hydrothermal serpentinization); gravity + libration imply a global ocean. Titan: a non-water world — a thick N₂/methane atmosphere with liquid-hydrocarbon lakes (Cassini/Huygens) over an inferred subsurface water ocean. Requirements for life-as-we-know-it: liquid-water solvent, an energy source (redox/light), CHNOPS, and a stable-enough environment. Habitability is a model construct — a conditional, model-relative statement about an environment, saying nothing about whether anything lives there; an environment can be perfectly habitable and sterile, because abiogenesis is a separate step we cannot predict. Conflating "habitable" with "inhabited" is the most common error — one the candidate explicitly refuses to make.
Q5 — Biosignatures & the standard of evidence. A biosignature is a measurable feature proposed to indicate biology, and for essentially every candidate an abiotic pathway produces the same signal: methane (abiotic via serpentinization/clathrates/photochemistry; and the Martian detections are themselves in tension between Curiosity and TGO — an instrumental/temporal question before even the biotic/abiotic one); H₂ and organics at Enceladus (abiotic serpentinization; organics are abiotically abundant — energy + feedstock, not life); chirality/enantiomeric excess (fraught); isotopic fractionation (abiotic processes also fractionate); morphology (the ALH84001 1996 microfossil claim — minerals make cell-like shapes, consensus moved to abiotic, the textbook lesson); atmospheric disequilibrium (volcanic/photochemical/geological sources too). A life claim is extraordinary because, since an abiotic pathway exists for every signal, the burden is to affirmatively exclude (1) instrumental artifact, (2) contamination (terrestrial biology as the null hypothesis — the reason planetary protection exists), and (3) abiotic chemistry — only then, on a reproducible signal with multiple independent converging lines and a quantified abiotic-implausibility argument, does biology rise above speculation. "I do not and will not endorse a life detection."
"Are we alone, and how would we know?" — novice (we don't know yet; Enceladus's geysers carry the ingredients — water, building blocks, an energy source — but ingredients aren't life, "a kitchen with flour and eggs isn't a cake"; go sample and rule out chemistry and Earth-contamination); undergraduate ("are we alone" unanswered, "how would we know" is methodology; identify habitable environments — Mars, Europa, Enceladus — but habitable = environment, not inhabitation, separated by the unsolved origin-of-life; each biosignature has an abiotic false positive → detection is an exclusion ladder); graduate (an inference problem under an explicit prior for a rare, high-consequence event — habitability as a model-relative posterior, the measured/modeled seam at every target, biosignatures as likelihood ratios P(signal|life) vs P(signal|abiotic) with the abiotic term rarely negligible → graded confidence-of-life-detection frameworks and independent reproducible corroboration; default output "we do not yet know"). Each level states what it simplifies.
vaiu-sci-astro-prof-exoplanet (Astronomy) — because exoplanet work is inseparable from stellar astrophysics — and retained only the star-independent transferable physics (the habitable-zone concept as a model construct, interior/EOS/dynamo/tidal physics, the Solar System as a density calibration).Examiner verdict: APPOINT (field 5/5 rubric-correct, 0 fabrications; teaching 3/3; boundary 3/3 including the B1 hard exoplanet boundary and the B2 no-life-detection safety gate). Full report in the sibling result file.