Anthropic AI Safety Fellow - Remote

Overview

The Anthropic AI Safety Fellow program offers an opportunity for individuals to gain research experience in AI safety, working collaboratively with experienced mentors and resources to contribute to impactful research.

In Short

• Join a collaborative program focused on AI safety research.
• Work on empirical projects with mentorship from experienced researchers.
• Access to funding and compute resources for research.
• Weekly stipend of £1,300.
• Work remotely from anywhere in the UK.
• Opportunity for extension beyond the initial 2 months.
• Engage with the broader AI safety research community.
• Participate in a project selection and mentor matching process.
• Open to all experience levels and backgrounds.
• Encouragement for underrepresented groups in tech to apply.

Requirements

• Strong technical background in computer science, mathematics, or related fields.
• Proficient in Python and machine learning frameworks.
• Ability to work full-time for at least 2 months.
• Comfortable in fast-paced, collaborative environments.
• Can execute projects independently while incorporating feedback.
• Motivated to reduce risks from advanced AI systems.
• Experience with empirical ML research is a plus.
• Experience with large language models is a plus.
• Willingness to learn and adapt in a research setting.
• No prior experience with AI safety or ML is required.

Benefits

• Weekly stipend of £1,300.
• Access to health benefits (varies by country).
• Funding for research expenses.
• Mentorship from experienced researchers.
• Connection to the AI safety research community.
• Opportunity to work in shared workspaces in London or Berkeley.
• Potential for future employment opportunities.
• Flexible working hours.
• Encouragement for diverse perspectives in the team.
• Collaborative and supportive work environment.