人择、有效利他主义 (EA) 以及相关的人工智能恐惧和厌恶运动。

The movement’s Fearful Leader.

Here is a source-by-source research digest of what the compiled overview reports primarily. Research was largely generated by Grok, Perpexity and Gemini. Claims are attributed to the named writers. Numbers, quotes, and named links are left as they appear in that file. Extra outtakes are pulled from the same passages, so more of what those people actually said is on the page.

What the packet is

A compilation of 2025–2026 coverage of Anthropic, Effective Altruism (EA) funding, AI “x-risk” organizations, evaluators, and media fellowships. It groups journalists, columnists, cable/web amplifiers, one NYT marketing critique, and three money-mapping researchers.

New York Post

Thomas Barrabi — Nov. 9, 2025

Headline reported: Anthropic’s ties to “cult-like” Effective Altruism and Democrat megadonors are on the Trump administration’s “radar.”

Facts the piece puts on the table:

Early money from EA circles and Sam Bankman-Fried: a $500 million stake in April 2022, later described as FTX customer funds.

Dustin Moskovitz as an early backer who later moved the stake into a nonprofit vehicle.

Amodei’s past anti-Trump language and Harris support.

Reed Hastings on the board; Reid Hoffman among backers.

Open Philanthropy / Moskovitz money into RAND during the Biden AI-order period.

Anthropic reply in the piece: Amodei does not identify as EA; the company praised Trump’s AI Action Plan and won a $200 million War Department contract.

Valuation then: $183 billion.

Outtakes:

Critics quoted treat EA as a “doomer” governing philosophy that wants a small priesthood to control access to AI.

The piece is described in the overview as political more than forensic: Anthropic trying to look non-“woke” in Trumpworld while sitting on an EA-and-Democratic-donor origin story.

Vivi Lin — Sept. 17, 2026

Claim as stated: Moskovitz and SBF did not only capitalize Anthropic; the same network funded the labs that test models and the reporters who cover them.

Traced items:

Moskovitz in the 2021 $124 million round led by Jaan Tallinn.

Moskovitz/Tuna’s Coefficient Giving (ex-Open Philanthropy) as a major Tarbell Center backer.

SBF leading the April 2022 $580 million round; now in prison; stake sold by prosecutors; no current company interest.

Coefficient → Tarbell Center, which she reports pays 80+ reporters embedded at Time, Bloomberg, The Guardian, The Verge (Tarbell insists on editorial independence).

Longview Philanthropy around METR; METR head Paul Christiano as Amodei’s former roommate and OpenAI colleague.

Anthropic demand-generation: Claude Corps (~1,000 paid people in 400+ nonprofits) plus college/community ambassadors.

Brian Chau: those programs used to be openly EA-promotion vehicles and rebranded to “AI safety” after FTX.

Investor Nathan Zhao: early Anthropic had a “strong cult-like culture” and “unquestioning faith” in Amodei.

Chau: Amodei’s regulation talk copies the SBF playbook—scare, then lock in rules that favor the lab already at scale.

Valuation in Lin’s piece: $965 billion.

Outtakes:

The two Post stories together: Barrabi put EA + Democrat megadonors on the White House radar; Lin put the media and evaluator layer on the same org chart.

Lin’s frame is operational: “ecosystem manipulating how you view AI.”

Andrew Orlowski (Telegraph, Spiked, The Critic)

Spiked, Feb. 22, 2026 — “The cult that wants to give human rights to robots.”

January 2026 Claude Constitution, in his argument, is not a safety spec.

It talks about Claude’s “virtue,” “psychological security,” “ethical maturity,” and possible wellbeing—language previously reserved for persons.

He calls that EA ultra-utilitarianism: living humans demoted relative to future people, shrimp, and hypothetical digital minds.

Anthropic as “an EA vehicle that just happens to also make AI models.”