Dario Amodei

HE QUIT OPENAI TO BUILD THE CAREFUL VERSION

Amodei ran research at OpenAI, then left with a crew of colleagues to found Anthropic on a single bet: frontier AI should be powerful, interpretable, steerable, and safer. The safety-first lab became ChatGPT's chief rival.

1. 2011

   PHYSICS BECOMES NEURAL CIRCUITS

   Amodei completed doctoral work at Princeton on the collective behavior of neural circuits, then continued as a postdoctoral scholar at Stanford School of Medicine.

   SOURCES: Hertz Foundationfirst-party

2. 2016

   OPENAI PUTS HIM ON THE RESEARCH THROTTLE

   After earlier neural-network research including Google Brain, Amodei joined OpenAI and eventually became vice president of research, co-leading work behind GPT-2, GPT-3, interpretability, safety, and human-preference training.

   SOURCES: OpenAIfirst-party

3. Dec 29, 2020

   HE QUITS OPENAI AND BUILDS THE RIVAL LAB

   OpenAI announced Amodei’s departure after nearly five years and said he and several colleagues were planning a new research-focused project.

   SOURCES: OpenAIfirst-party

4. 2021

   THE SAFETY BET GETS ITS OWN LAB

   Amodei and colleagues founded Anthropic at the beginning of 2021 to build large AI systems that were more steerable, interpretable, robust, and reliable.

   SOURCES: Anthropicfirst-party

5. Mar 14, 2023

   CLAUDE ENTERS THE CHAT

   Anthropic introduced Claude as a next-generation assistant built around its research into helpful, honest, and harmless AI systems.

   SOURCES: Anthropicfirst-party

6. Nov 22, 2024

   THE SAFETY LAB GETS AN $8 BILLION CLOUD ALLIANCE

   Anthropic expanded its AWS partnership, making AWS its primary cloud and training partner and bringing Amazon’s total investment in Anthropic to $8 billion.

   SOURCES: Anthropicfirst-party