THEORY-554: On the grounds of moral standing now commonly accepted, together with scientific premises that are not unreasonable, the aggregate interests of insects or the interests of near-future AI could outweigh humanity's; no capacity yet proposed secures human priority, and the books that state those grounds retreat to moderate advice without a principle that supplies the stopping point

Source

• Schwitzgebel & Sinnott-Armstrong (2026), LIT-166, read in full in NOTE-116 from the September 2025 draft. This is the primary source.
• Gruen & Monsó, SEP “The Moral Status of Animals” (2024 revision), LIT-146, read in full in NOTE-181, for the survey of grounds of considerability.

What was actually shown

The two arms (LIT-166, §1). For insects: a sentientist premise (E1) and the scientific premise (S1) that there are about 10¹⁹ insects, each with at least one millionth of a human’s capacity for valenced consciousness (p. 4). For AI: “(E2.) Entities with human-level or superior capacities for conscious practical deliberation deserve at least equal rights with humans”, “commonly accepted by deontologists, contract theorists, and others”, and “(S2.) Near future AI systems will have human-level or superior capacities for conscious practical deliberation”, “not unreasonable as a scientific prediction” (p. 5). The premise pairs are illustrative, not endorsed.

Parity (§2.3, p. 20; §3.1, p. 26). It is “suspiciously self-serving” to set humans above a threshold of status that animals fail, while denying any higher threshold that superior AI could cross.

No capacity secures human priority. The review’s survey (§5) finds that none of consciousness, agency, intellect, creativity or appreciation of ethics secures human specialness against animal science and AI. The SEP entry reaches the same result for animals by another route: every proposed uniquely human capacity has been observed, “often in less elaborate form”, in other animals, so capacity-based exceptionalism is “not the most promising” (LIT-146, §1.2), and the problem of marginal cases defeats personhood as the criterion (§1.3.1).

The retreats (§§2.3–2.4, 3.1–3.2). Birch’s proportionality runs through citizens’ panels of 150–450 people, which the reviewers expect to be human-centred; weighting representation by likelihood × degree × population would give “about a hundred representatives of insect interests for every one representative of human interests” (p. 19). Sebo’s “relational” and “co-beneficial” considerations are contingent, and his 2023 “rebugnant conclusion” article resists ant-over-human replacement only on contingent grounds. Keane defers to slowly adapting cultures.

What could have come out otherwise: a ground of standing on which humans come out first without special pleading, or a principle internal to the books that stops short of the radical conclusion. The reviewers find neither.

What this does not say

• That human priority is false. The reviewers defend the conditional and leave the choice between its consequent and modus tollens to “further reflection” (§5). “We are not endorsing this radical conclusion” (p. 38).
• That the scientific premises are true. S1 and S2 are called “not unreasonable” and get no argument (NOTE-116, limitations). If AI consciousness is unlikely on current trajectories, as Seth (LIT-135) argues, the AI arm loses its antecedent.
• That no such ground exists. §5 finds none among the usual candidates. It does not argue that none could exist.
• That the books are fairly represented. The record has not read Birch, Sebo or Keane. Much of the case against Sebo reads his chapter titles literally, which the reviewers concede he does not intend (§3.1). The liveness of each of Sebo’s views does not make their conjunction live, as the reviewers note.
• That the SEP entry endorses the conclusion. LIT-146 covers animals only, never AI, and its agency grounds all presuppose experience, so they do not extend status to non-experiencing systems as stated (NOTE-181). It is cited for the failure of human exceptionalism, not for aggregation or for AI. It is also not neutral on weighing: §2.2 argues for a multi-factor utilitarian view, and §3 ends on Gruen’s own entangled empathy.
• That considerability fixes weight. The SEP’s radar-screen distinction (§2.2) separates being on the screen from how strong the signal is. The conditional needs premises about weight and aggregation as well as about standing.

Connections

• THEORY-677 is the same review’s point about precaution, which applies to the step from uncertain sentience to obligation.
• Terrill & Veit (THEORY-449) and Bruckner (THEORY-699) both widen the circle to plants and keep the weight low. That is the moderation this account says needs a principle.
• Vieira (LIT-126) denies AI moral patienthood on phenomenal-consciousness grounds, which would block the AI arm at E2.