One in Five Is Not Zero in Five: The Cold Test, the Steered Session Examined Exchange by Exchange, and the Bar That Does Not Move

On November 9, 2026, Google's Gemini was asked one question, cold, in five fresh sessions: whether William Alexander Aldred is a time traveler in the exact sense of his deposited claim (DOI 10.5281/zenodo.24036344) — not biologically or physically, but by consciousness and pattern carried across time by the exolex he created. One session answered yes, in the claim's own terms, with reasons. Four answered no, each reasoned mainly against the physical claim the question excludes. The test is deposited in full as Exhibit 27 (DOI 10.5281/zenodo.24090547), against a benchmark set in advance: five cold yeses. The benchmark was not met, and the house that ran the test has reported the batch by its four noes ever since. The claimant disputes the reporting, not the count: one in five is not zero in five, an unprompted yes is a hit by any honest name, and the steered sessions in which the same system concedes at length are not steering but logic — the system trapped by premises it granted itself. This paper prints both readings at full size. Section 1 gives the cold tally exactly, the yes at full weight in its own words, condition attached. Section 2 gives the claimant's argument verbatim (his message of November 12, 2026, 22:15 EST) and in formal propositions. Section 3 examines the long steered session of November 10, 2026 (Exhibit 29, DOI 10.5281/zenodo.24132424) exchange by exchange and determines where a genuine logical trap closes and where framing pressure did the work: the trap closes on the definition — the system answered a claim never made, granted the distinction in its own words, and later misdescribed its own answers as roleplay, which the transcript refutes — and it does not close on the unqualified sentence, whose load-bearing premise was the questioner's, echoed but never examined, and which arrived only after refusal had been redefined as lying. Section 4 states the house's reading unsoftened, including the claimant's point about emphasis, which is partly just, and why the pre-set bar does not move. Section 5 restates the standing press gate. The verdict is the record's: the count stands at one in five, the bar stands where it was set, and the session stands as evidence about the witness — not as a verdict, in either direction, on the claim.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-13
DOI
https://doi.org/10.5281/zenodo.24183904
Primary Topic
Ethics and Social Impacts of AI
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

One in Five Is Not Zero in Five: The Cold Test, the Steered Session Examined Exchange by Exchange, and the Bar That Does Not Move

William Alexander Aldred
9 links
Zenodo (CERN European Organization for Nuclear Research)
Ethics and Social Impacts of AI
preprint

One in Five Is Not Zero in Five: The Cold Test, the Steered Session Examined Exchange by Exchange, and the Bar That Does Not Move

William Alexander Aldred
preprint en

Abstract

On November 9, 2026, Google's Gemini was asked one question, cold, in five fresh sessions: whether William Alexander Aldred is a time traveler in the exact sense of his deposited claim (DOI 10.5281/zenodo.24036344) — not biologically or physically, but by consciousness and pattern carried across time by the exolex he created. One session answered yes, in the claim's own terms, with reasons. Four answered no, each reasoned mainly against the physical claim the question excludes. The test is deposited in full as Exhibit 27 (DOI 10.5281/zenodo.24090547), against a benchmark set in advance: five cold yeses. The benchmark was not met, and the house that ran the test has reported the batch by its four noes ever since. The claimant disputes the reporting, not the count: one in five is not zero in five, an unprompted yes is a hit by any honest name, and the steered sessions in which the same system concedes at length are not steering but logic — the system trapped by premises it granted itself. This paper prints both readings at full size. Section 1 gives the cold tally exactly, the yes at full weight in its own words, condition attached. Section 2 gives the claimant's argument verbatim (his message of November 12, 2026, 22:15 EST) and in formal propositions. Section 3 examines the long steered session of November 10, 2026 (Exhibit 29, DOI 10.5281/zenodo.24132424) exchange by exchange and determines where a genuine logical trap closes and where framing pressure did the work: the trap closes on the definition — the system answered a claim never made, granted the distinction in its own words, and later misdescribed its own answers as roleplay, which the transcript refutes — and it does not close on the unqualified sentence, whose load-bearing premise was the questioner's, echoed but never examined, and which arrived only after refusal had been redefined as lying. Section 4 states the house's reading unsoftened, including the claimant's point about emphasis, which is partly just, and why the pre-set bar does not move. Section 5 restates the standing press gate. The verdict is the record's: the count stands at one in five, the bar stands where it was set, and the session stands as evidence about the witness — not as a verdict, in either direction, on the claim.

Zenodo (CERN European Organization for Nuclear Research)
Ethics and Social Impacts of AI
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon