You will create specialized training data and review large language model responses in scientific and technical subjects. Your work will help assess whether AI systems explain important concepts accurately, clearly, and completely.
- Write prompts grounded in your area of science or engineering.
- Research topics with help from AI and turn complex knowledge into clear task instructions.
- Rate AI-generated responses for accuracy, relevance, and quality.
- Create or apply criteria for judging specialized concepts.
- Use expert judgment to identify strong, incomplete, or misleading explanations.
What it pays and takes
- Pay: Up to $48 per hour.
- Work arrangement: Independent contractor work with flexible, asynchronous participation. Projects may open periodically throughout the year, and part-time participation may be available during active projects.
- Location: Open to candidates in Austria, Belgium, Bulgaria, Cyprus, Czechia, Germany, Denmark, Estonia, Spain, Finland, France, Greece, Croatia, Hungary, Ireland, Italy, Lithuania, Luxembourg, Latvia, Malta, the Netherlands, Poland, Portugal, Romania, Sweden, Slovenia, or Slovakia.
- Language: Fluent written English.
- Education: Current doctoral study, a completed PhD, or postdoctoral experience in mathematics, chemistry, biology, physics, or engineering.
- Skills: Explain complex research knowledge clearly, assess whether AI understands concepts in your field, and work independently in a remote setting.
- Helpful background: Experience creating AI training data, writing domain-specific prompts, or evaluating language model responses. Academic research and careful, evidence-based reasoning are also relevant.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training work
AI training is the human work behind systems that write, answer questions, and explain ideas. Contributors prepare examples and review model responses, while specialist projects value people with strong subject knowledge.