VotesWith
September 2026

The Great Resign-Alignment

One resignation. Remarkable alignment.

Sep 8 · 8:04 PM ET · resignation

“I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives.”

Jacob Coxon · @hilbertspaess · ex-OpenAI / Anthropic pretraining

Open the thread → See the clock → Follow the money → Who spoke →

Views, coverage & the campaign

  • Associated Press — Independently reported the resignation; thread crossed roughly 100 million views overnight.
  • WSJ exclusive — Press exclusive landed shortly before Coxon’s X thread (exact clock pending).
  • Peter Wildeford — Shared the WSJ story ~2 minutes before Coxon’s thread. Source →
  • Parker Thayer / CRC — Mapped early quoters and funding overlaps; argued the rollout looked like a well-funded PR / campaign push. Source →
  • Ethan Perez — Described Coxon as a senior researcher Anthropic had tried to recruit for ~two years (not an unknown account).
  • Evan Hubinger — Endorsed the substance of Coxon’s concerns. Source →
  • Elon Musk / Bill Ackman — Counter-frame: Musk “Seems like a setup.”; Ackman reshared Thayer. Musk → · Ackman →

The Coxon Connection

Tap a node · edges light · outsiders dim

Ties & links

What he argued

Both labs racing

Said OpenAI and Anthropic are racing to self-improving superintelligence and gambling with people’s lives.

Stakes vs incentives

Argued OpenAI hadn’t internalized the stakes; Anthropic understood them but kept racing because they thought no one else would act responsibly.

Endgame warning

Called entering an “endgame” from a private Slack hubristic — and floated a temporary ban on capability improvements.

Ask to researchers

Urged lab researchers to speak up and call for different conditions.

Democrat Republican Independent Candidate Named a bill

Who responded

Officials and candidates · party + state/district on every tile · tap to jump to their card

Democrat Republican Independent

Jump down for the full quote, bill ask, and source link.

Context. The Ban Artificial Superintelligence Act was announced Sep 3 — six days before Coxon’s thread. Later posts mostly pointed at bills that already existed.

Where the money goes

Tap a box to open citations. Shared donors show overlap — not orders.

Tap a box to light up its money ties.

Grant amounts

Primary public figures from SFF recommendations and Series A reporting.

Funders

Organizations

    17
    Official posts
    11
    Named a bill
    3
    Main funders
    6
    Key groups
    Layer

    Money

    Tallinn, Moskovitz, and SFF sit behind several of the groups that spoke early.

    • Series A ties to Anthropic
    • SFF grants to Encode, Futures, AIPI, IAPS
    Layer

    Groups

    Policy and research shops that quoted Coxon early or employ early quoters.

    • Encode · AI Futures · IAPS
    • Coefficient / Good Ventures
    Layer

    Public posts

    Officials later named bills — many of which were already on the table.

    • Morning through evening, Sep 9
    • Party colors on each Politics card
    Counter-frame

    Elon Musk

    Sep 9 · 9:54 PM ET — pushed a setup frame on the resignation thread.

    “Seems like a setup.”
    Open Musk’s post →

    Who mapped this first

    Other posters and reporters pulled the early quote chain, funding ties, timing, and the counter-frame before this VotesWith brief. Open each source for their full write-up.

    Primary sources and filings cited across this brief. Open each link for the original.

    Funding & filings

    Key posts & reporting

    Official responses (Politics)