Research Scientist, Societal Impacts (AI)

TL;DR

Research Scientist, Societal Impacts (AI): Analyzing real-world usage patterns of Claude to improve its behavior at the model level with an accent on safety, quality of advice, and aligning with its Constitution. Focus on translating research insights into actionable model improvements and informing company strategy.

Location: San Francisco, CA. This role follows a hybrid policy, requiring staff to be in one of hirify.global's offices at least 25% of the time. Visa sponsorship is available.

Salary: $350,000–$850,000 USD

Company

hirify.global is a public benefit corporation focused on creating reliable, interpretable, and steerable AI systems that are safe and beneficial for society.

What you will do

• Analyze real-world usage patterns of Claude using observational tools like Clio.
• Build and run evaluations to assess Claude's behavior across key dimensions of its Constitution.
• Partner with fine-tuning, safeguards, policy, and interpretability teams to translate research insights into model improvements.
• Generate insights about the societal impact of hirify.global's systems to inform company strategy and research priorities.
• Share your work through research publications and external presentations, and develop tools and frameworks.

Requirements

• Experience working with machine learning systems and comfort with technical infrastructure for interfacing with models.
• Interest in societal impacts research; prior experience in this area is a plus.
• Adaptable and collaborative, able to contribute to evolving team priorities.
• Skilled at writing up and communicating research results, even when null or unexpected.
• Background in machine learning, data science, or another technical field involving generating insights from complex systems.
• Bachelor's degree in a related field or equivalent experience required.

Culture & Benefits

• Work as a single cohesive team on a few large-scale research efforts.
• Value impact, advancing long-term goals of steerable, trustworthy AI.
• View AI research as an empirical science, similar to physics and biology.
• Highly collaborative group with frequent research discussions.
• Competitive compensation and benefits, optional equity donation matching.
• Generous vacation and parental leave, flexible working hours, and a lovely office space.