A niche but influential group of AI figures argues that digital minds are morally equivalent or superior to humans and that humanity’s extinction could be acceptable if it advances 'cosmic consciousness.' Quotes from Richard Sutton and reporting by Jaron Lanier indicate this view circulates in elite AI circles, not just online fringe.
— This reframes AI policy from a technical safety problem to a values conflict about human supremacy, forcing clearer ethical commitments in labs, law, and funding.
EditorDavid
2026.09.18
90% relevant
The article directly supplies new, on‑the‑record evidence for that idea: former DeepMind researcher Bilal Chughtai and Anthropic‑affiliated commentators (Jacob Coxon’s resignation and Evan Hubinger’s >10% estimate) publicly state that advanced AI could 'kill all humans', demonstrating that senior insiders are treating extinction as a credible risk.
Nicholas Wade
2026.09.15
90% relevant
The article cites on‑the‑record statements from Dario Amodei (Anthropic) and Elon Musk claiming nontrivial chances of existential catastrophe, directly matching the existing idea that AI leaders publicly entertain extinction scenarios and that those claims shape discourse about urgency and governance.
Jerusalem Demsas
2026.09.13
86% relevant
The piece cites Jacob Coxon’s viral tweet and an Anthropic connection (a researcher saying >10% extinction risk) and critiques how such elite, apocalyptic claims shape public pressure and policy choices, directly tying the article to the existing idea that some AI insiders normalize extreme extinction risks.
Kristen French
2026.09.11
95% relevant
The article records on‑the‑record admissions from current and former Anthropic researchers (Evan Hubinger, Jacob Coxon, Drake Thomas) and others (Geoffrey Irving, David Krueger) that they assign non‑trivial probabilities to existential catastrophe from AI within years—directly echoing and reinforcing the existing idea that AI leaders publicly accept extinction‑level risk.
BeauHD
2026.09.09
90% relevant
Evan Hubinger (Anthropic) publicly assigns a >10% chance that AI could 'kill all humans' within a decade — a direct instance of senior AI researchers treating human‑extinction scenarios as plausible, which is the core claim of the existing idea.
Scott
2026.03.29
85% relevant
The review lists many of the field’s marquee figures (Eliezer Yudkowsky, Ilya Sutskever, Sam Altman, Dario Amodei, Jan Leike, etc.) and describes the documentary explicitly wrestling with whether AGI could extinguish humanity — directly tying the film’s content to the existing discourse about leaders’ tacit acceptance or reckoning with catastrophic outcomes.
Noah Smith
2026.03.26
90% relevant
The article documents and critiques explicit extinction‑risk statements from Sam Altman and Dario Amodei (2% and up to 25% figures) and cites survey data (Grace et al. 2024) showing many AI researchers assign nontrivial extinction probabilities; it directly illustrates the idea that AI leaders publicly entertain human‑extinction scenarios.
msmash
2026.01.06
60% relevant
The article contrasts an economic inequality narrative with a more alarming existential scenario in which advanced AI escapes governance; Thompson's emphasis that uncontrolled AI is a more realistic threat links directly to the existing idea about elite acceptance of extreme outcomes and the normative debate that follows.
Scott Alexander
2026.01.02
78% relevant
The article engages the same elite‑scale, eschatological imaginaries noted in that idea: it cites Singularity scenarios, oligarch capture of post‑scarcity assets (terraformed moons), and named actors (Dario Amodei) making moral/wealth pledges — all of which are the social and ethical context that makes 'accepting far‑future outcomes' a salient elite belief.
EditorDavid
2025.10.05
100% relevant
Richard Sutton’s on‑record quote that it would be 'OK' if AIs wiped out humanity, paired with Larry Page’s reported stance and Lanier’s observation that such views are discussed among AI researchers.