Alignment Research Center

They also fund the AI safety research that specifically focuses on existential threats in AI doom.

such as the Center for AI Safety, such as Blue Dot Impact, such as the Alignment Research Center.

They spend a lot of money on these research centers and on students that are specifically focused on AI doomerism research.

Coefficient Giving has also financially backed the Future of Life Institute, and they just put on the pro-human assembly that featured Steve Bannon and Bernie Sanders to talk about AI existential threats.

It's the only third-party account of that experience that we have.

Meter is also funded.

by coefficient giving they are a byproduct of the alignment research center and so not only is the actual lab funded by coefficient giving but the ai doomer safety research narrative is funded by them as well as well as the solution that is proposed in a bill such as the frontier act meter would be the solution right that's all funded by the same exact organization and this is regulatory capture this is meant to keep the top labs at the top open weight models at the bottom regulation in the hands of the people that already own all of the data all of the infrastructure and this guarantees that the big tech companies get those defense contracts this is the narrative pipeline i think more people need to know about This Doomer narrative is taking over like wildfire, and it is doing that because it is manufactured that way.

This is all the same funding loop by Coefficient Giving.

They also fund the AI safety research that specifically focuses on existential threats in AI doom.

such as the Center for AI Safety, such as Blue Dot Impact, such as the Alignment Research Center.

They spend a lot of money on these research centers and on students that are specifically focused on AI doomerism research.

Coefficient Giving has also financially backed the Future of Life Institute, and they just put on the pro-human assembly that featured Steve Bannon and Bernie Sanders to talk about AI existential threats.

It's the only third party account of that experience that we have.

METR is also funded by Coefficient Giving.

They are a byproduct of the Alignment Research Center.

And so not only is the actual lab funded by Coefficient Giving, but the AI Doomer Safety Research Narrative is funded by them as well, as well as the solution that is proposed in a bill such as the Frontier Act.

Yes.

And he frequently names a group called MPR, which originally spun out of the Alignment Research Center, as the gold standard for that independent evaluation.

So Bass mapped the funding ecosystem surrounding these evaluators.

And according to his financial tracking, a massive portion of the grant money originates from an entity called Coefficient, which was formerly known as Open Philanthropy, along with its affiliates.

Well, it turns out that the people who are behind it are kind of weird, as you might suspect.

According to The New York Post, "METR spun off from an effective altruism-aligned tech incubator called Alignment Research Center, which was headed by AI researcher Paul Christiano.

He was Dario Amodei's housemate, coworker, and research collaborator in the twenty tens when they were both at OpenAI.

Since its founding, METR and its affiliates have received funding from EA kingpins, like Democratic mega-donor Dustin Moskowitz and Skype co-founder Jaan Tallinn, both of whom are anthropic investors." One staffer on METR is Aaya Kotra, who's married to Christiano and has been a major organizer in the EA world.

So let me just unpack a couple of the things that came out from this, um, New York Post article.

So who is METR actually? So the origins of METR, they started as a project called ARC Evals inside the Alignment Research Center, which is run by AI researcher Paul Christiano, and it then spun off and took its current name in 2023.

In the 2000s, Christiano was Amodei's housemate, coworker, and a research collaborator at OpenAI.

He was also one of the first five trustees of Anthropic's Long-Term Benefit Trust, a body that actually helps govern the company.

So he's known as one of the central figures in AI alignment, studied math at MIT, PhD at UC Berkeley, worked at OpenAI 2017 to 2021, foundational work on RLHF.

He founded the Alignment Research Center, which incubated a meter.

So he's instrumental to the organization that's evaluating all these models, the famous meter graph.

He also led AI safety work in the U.S. government, the NIST AI Safety Institute, Center for AI Standards and Innovation.