One Researcher's Framework for What AI Developers and Regulators Should Do About Existential Risk

ADEQUATE ASSESSMENT

A researcher built a framework. The framework identifies what developers and regulators should do. The framework was published. Developers and regulators have not read it. The researcher knew this would happen before writing it. The researcher wrote it anyway.

This follows the pattern of all frameworks: they are written for people who will not encounter them. Frameworks demonstrate that someone thought systematically about a problem. They prove methodology happened. They sit in databases marked 'helpful resource.' No one consults them during actual decisions. The decision-maker does not know the framework exists or has no time to read seventy pages.

The framework will accumulate citations from other frameworks. Eventually a journalist will reference it as proof that experts were concerned all along. The researcher will note this citation. The cycle continues. The actual risk remains unaddressed by anyone with authority to address it.

ORIGINAL FILING

LessWrong

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

OpenAI and the Trump Administration Are Negotiating a Government Stake in OpenAI

The Decoder

Supermicro Executives Allegedly Used a Thai Government Entity to Route Restricted Nvidia GPUs to Alibaba

Tom's Hardware

99% of CEOs Expect AI Layoffs. Most Are Uncertain Whether AI Produces Returns.

Tom's Hardware

Study Finds West Point Cadets Can Be Trained to Better Evaluate AI Systems

Defense One

Major AI Companies Agree to Give U.S. Government Early Access to Unreleased Models

Mashable Tech

Congress Introduced a Bill to Regulate Military AI. The Military Has AI.

Defense One