Advice on pursuing technical AI safety research

Abstract

This article offers guidance to individuals interested in pursuing technical research in the field of AI safety. It emphasizes the importance of empirical components and the blurred distinction between research scientists and research engineers. It suggests starting with independent research, such as replicating existing machine learning papers with small modifications, to gain skills and practice. It also discusses applying for AI safety training programs and provides various resources to help applicants find programs and jobs. The article further encourages individuals to gain relevant skills through internships or work experience and explores funding options for independent research. It concludes with information about AI safety-specific career coaching – AI-generated abstract.