Journalist

Job Summary

The Expert Journalist creates, evaluates, and refines high-quality training data to train frontier AI models. Core daily duties include writing deep investigative prompts and response variations, evaluating model outputs for factual accuracy, executing rigorous fact-checking loops against authoritative sources, editing text to enforce strict style guidelines, and auditing model reasoning for hidden bias or breaches of media law and ethics.

Foundational Capabilities & Prerequisites

• Professional Portfolio: Proven professional experience as a practicing journalist, backed by a strong portfolio of published work, investigative pieces, features, or complex beat coverage.
• Geographic Residency: Must be a legal resident currently operating from one of our approved talent regions: the United States, United Kingdom, Canada, Australia, or New Zealand.
• Fact-Checking & Synthesis: Exceptional proficiency in executing investigative research, conducting interviews, and synthesizing highly complex, technical, or fragmented information into clear, objective prose.
• Media Ethics & Law: A solid, uncompromised understanding of media law, copyright standards, plagiarism thresholds, and journalistic best practices.
• Linguistic Articulation: Exceptional command of written and verbal English communication, showing meticulous attention to detail and sharp editorial formatting skills.
• Operational Habits: Demonstrated ability to work independently, exercise strong personal initiative, employ critical thinking, and hit tight deadlines within a remote framework.
• Digital Literacy: Completely comfortable utilizing modern digital tools for research, secure collaboration, and cloud-based content management.

Preferred Professional Multipliers

• Digital-First Newsroom Footprint: Previous experience writing or reporting for international news agencies, wire services, or digital-first publication hubs.
• Multimedia Adaptability: Familiarity with modern multimedia storytelling architectures, including scriptwriting for podcasts or video production frameworks.
• Polyglot Capabilities: Proficient or fluent in additional languages beyond English, enabling multi-lingual model training tracks.

Technical Workflow Architecture

As an AI Data Engineering Journalist, your contributions will typically guide model behavior across several evaluation tracks:

• Supervised Fine-Tuning (SFT): Authoring pristine, authoritative articles, summaries, and structural layouts to serve as the "gold standard" target texts for model imitation.
• Reinforcement Learning (RLHF): Reviewing two or more model-generated news summaries, ranking them based on clarity, structure, and readability, and providing detailed written justifications.
• Adversarial Fact-Verification: Testing model resilience by feeding it deceptive inputs, tracking its ability to catch factual errors, and correcting its reasoning paths using verified data sources.
• Bias & Ethics Alignment: Stress-testing models against media law violations, libel patterns, and unverified rumors to ensure the system refuses to generate harmful or unethical output.

Key Responsibilities

1. Data Creation, Article Generation & Prompt Engineering

• Research, pitch, and develop original news stories, complex features, and in-depth articles within our training platform to teach models tone, structure, and investigative narrative flow.
• Create clear, compelling, and accurate written content tailored to diverse audiences, serving as primary training data inputs.
• Monitor global current events, societal trends, and specialized beats to feed highly relevant, real-time contexts into fine-tuning datasets.

2. Analytical Fact-Checking & Model Evaluation

• Conduct mock interviews, gather secondhand or firsthand data insights, and verify volatile statements against authoritative sources.
• Evaluate and grade text outputs generated by frontier AI models, auditing their logical reasoning, source attribution accuracy, and structural fluidity.
• Execute meticulous fact-checking procedures, calling out subtle model hallucinations, data exaggerations, and false structural assertions.

3. Structural Editing & Adherence to Style Guides

• Edit and proofread complex textual data blocks to guarantee absolute grammatical clarity, logical consistency, and strict adherence to defined editorial standards.
• Review model summaries for conciseness, structural pacing, and correct heading usage, adjusting text manually to establish training benchmarks.
• Collaborate with data lab editors, technical designers, and cross-functional journalists to continuously improve data generation playbooks.

4. Media Law Modeling & Ethical Supervision

• Apply your deep understanding of media law and professional ethics to flag and neutralize biased, unverified, or legally compromising data models.
• Uphold the highest ethical standards and maintain strict journalistic integrity across all data outputs, ensuring trained models avoid generating misinformation or propaganda.
• Design safety parameters that prevent AI systems from generating defamatory content or violating intellectual property standards.

Skills Required:

• Communication / Pr / Journalism