Job Description
We are pioneering the next generation of Artificial General Intelligence (AGI) and are seeking a visionary AI Ethics & Alignment Researcher to join our San Francisco headquarters. As we look toward the deployment of our systems in 2026, we need a thought leader who can ensure our AI architectures are not only powerful but safe, transparent, and aligned with human values.
In this role, you will act as the moral compass for our engineering teams, bridging the gap between theoretical philosophy and practical machine learning implementation. You will define the safety protocols for our flagship models and contribute to the global discourse on the future of technology.
Why Join Nexus Horizon Labs?
- Work on the cutting edge of AI safety and alignment.
- Shape policy for the AI ecosystem in 2026 and beyond.
- Competitive compensation and equity packages.
- Collaborate with world-class researchers and engineers.
Responsibilities
- Develop and implement safety frameworks for large-scale language models and autonomous agents.
- Conduct adversarial testing and red-teaming to identify potential risks in AI behavior.
- Collaborate with product teams to integrate ethical guidelines into the development lifecycle.
- Research and propose novel methodologies for value alignment in complex, dynamic environments.
- Advocate for responsible AI practices within the industry and with stakeholders.
- Write and publish research papers contributing to the field of AI safety.
Qualifications
- PhD in Computer Science, Philosophy, Cognitive Science, or a related field.
- 5+ years of experience in AI/ML research or a strong academic background with significant industry application.
- Deep understanding of machine learning concepts, including neural networks, transformers, and reinforcement learning.
- Proven track record of publishing in top-tier AI or ethics conferences/journals.
- Strong analytical and critical thinking skills with a focus on long-term impact.
- Excellent communication skills to translate complex concepts for diverse audiences.