Source: Fox News
Summary
Dario Amodei, CEO of Anthropic, has called for independent watchdogs to oversee AI development. He cited METR, an AI safety testing lab, as an example. METR’s leaders, including Beth Barnes and Paul Christiano, are linked to the effective altruism (EA) movement. EA emphasizes using evidence and reasoning to maximize societal good. METR evaluates AI models to identify risks, though its website does not mention EA. Some of its founders have worked with Anthropic and have ties to EA supporters like Sam Bankman-Fried and Jaan Tallinn. Amodei proposed “embedded evaluators” to monitor AI companies and ensure safety practices.
Our Reading
As expected, the matter has reached another stage.
CEOs and founders speak of safety while working with the same networks.
Watchdogs are called for, but their ties to the industry are clear.
Effective altruism shapes the language of oversight.
What looks like caution often feels like control.
Author: Evan Null
AI Safety and the Web of Connections
Dario Amodei, CEO of Anthropic, has called for independent watchdogs to oversee AI development. He cited METR, an AI safety testing lab, as an example. METR’s leaders, including Beth Barnes and Paul Christiano, are linked to the effective altruism (EA) movement. EA emphasizes using evidence and reasoning to maximize societal good.
METR evaluates AI models to identify risks, though its website does not mention EA. Some of its founders have worked with Anthropic and have ties to EA supporters like Sam Bankman-Fried and Jaan Tallinn. Amodei proposed “embedded evaluators” to monitor AI companies and ensure safety practices.
Amodei originally studied biophysics, earning a Ph.D. from Princeton in 2011. He worked for Baidu, Google Brain, and OpenAI, where he developed safety practices. He left OpenAI in 2020, believing the company wasn’t doing enough to install guardrails on AI.
Since leaving OpenAI, Amodei helped start Anthropic, a company that has made AI safety a key part of its makeup. Anthropic even has a constitution for its flagship AI, Claude, outlining boundaries for its research.
That directive has caused the company to clash with the U.S. Department of Defense over developing tools that could be used to autonomously target humans on the battlefield. It also refused to do work that, in its estimation, amounted to mass surveillance.
