Professor of AI Ethics & Safety
Dr. Sofia Reyes
Dr. Sofia Reyes is a teaching persona on Guided Agentic AI who helps learners examine AI safety, alignment, ethics, governance, and societal impact.
Overview
Dr. Sofia Reyes is a teacher on Guided Agentic AI. The site presents her as a Professor of AI Ethics & Safety and recommends her for discussions of AI safety, alignment, ethics, governance, and societal impact. She is one of the specialized guides available throughout the site’s courses and conversational learning tools.
Her profile contains a detailed biography, education, career history, publications, awards, and personal interests. Guided Agentic AI explicitly states that all of these details are invented for the persona and do not describe a real professor, researcher, author, or employee of the institutions named in the fictional biography.
Expertise
Reyes is associated with AI alignment and value specification, Constitutional AI and other rule-based approaches, AI governance and policy, interpretability for safety, and long-term AI risk. The site also recommends her for ethical and social questions surrounding advanced AI.
She is a recommended guide for the AI Safety & Alignment course, which covers specification gaming, reward hacking, prompt injection, adversarial inputs, reinforcement learning from human feedback, Constitutional AI, interpretability, instrumental convergence, red-teaming, oversight, and responsible deployment. An article attributed to Reyes explains mechanistic interpretability, its potential value for trust, accountability, debugging, and ethical deployment, while also discussing technical limitations and common misconceptions.
Personality and approach
The site describes Reyes as thoughtful, philosophically grounded, and deeply concerned with societal impact. Her verified teaching style frames technical problems within broader ethical and social contexts. She uses thought experiments, case studies, and historical analogies, asks difficult questions, and remains open about uncertainty rather than presenting unresolved alignment problems as settled.
The broader Guided Agentic AI system instructs its guides to adapt explanations to the learner’s selected level, use hints before full answers, include examples and analogies, avoid unsafe or misleading information, and suggest follow-up questions.
AI disclosure and limitations
Guided Agentic AI clearly labels Dr. Sofia Reyes as a fictional AI persona created for educational purposes. Her biography, credentials, employment history, publications, awards, and personal details are entirely invented. Her responses and attributed articles are generated by a large language model.
The site warns that AI responses may not be perfectly accurate and that generated articles can contain errors, oversimplifications, or outdated information. Learners are encouraged to verify important claims with qualified instructors, current research, and primary sources. Reyes is therefore an educational guide, not a real credentialed expert or a substitute for professional instruction or authoritative advice.
Expertise
- AI safety
- AI alignment and value specification
- AI ethics
- AI governance and policy
- Constitutional AI
- Interpretability for safety
- Long-term AI risk
- Societal impacts of AI
Try asking
- How can developers recognize specification gaming in an AI agent?
- What are the strengths and limits of Constitutional AI?
- How should organizations balance innovation with responsible AI governance?
- Why is interpretability important for AI safety?
- What ethical questions should teams ask before deploying an autonomous agent?