Artificial intelligence doesn't build itself. Behind every large language model, every image recognition system, and every voice assistant is a vast amount of human work — evaluating outputs, providing feedback, creating training examples, and teaching machines how to think more like us. These roles are collectively known as human data jobs, and they represent one of the fastest-growing employment categories in the world.
The Basics: What You Actually Do
At its core, a human data job involves providing the data, feedback, or evaluation that AI systems need to improve. The specific tasks vary enormously depending on the role, but they generally fall into a few categories:
- AI Training: Creating prompts, writing responses, and generating the examples that language models learn from. This is the bread and butter of platforms like Outlier, Remotasks, and Scale AI.
- Data Annotation: Labelling images, categorising text, or tagging audio so that AI systems can learn to recognise patterns. Think: drawing boxes around objects in photos or classifying sentiment in customer reviews.
- AI Evaluation: Comparing two AI-generated responses and judging which is better, more accurate, or more helpful. This is the "RLHF" (Reinforcement Learning from Human Feedback) work that made ChatGPT possible.
- Expert Consultation: Domain specialists — lawyers, doctors, engineers — providing their expertise to train specialised AI models. This is the premium end of the market, with rates reaching $225/hour or more.
Who's Hiring?
The ecosystem is large and growing. Major AI labs like OpenAI, Anthropic, Google DeepMind, and Meta all rely on human data work, but they typically outsource through intermediary platforms. The platforms you'll encounter most often include:
- Outlier (by Scale AI) — The largest employer of AI trainers, with thousands of active projects across writing, coding, and domain expertise.
- Alignerr — Focused on STEM and technical evaluation work, often at higher pay rates than general platforms.
- Turing — Matches engineers and developers with AI training projects, particularly in code generation and review.
- Ethos — A premium expert network connecting domain specialists with $100-225/hour consultation opportunities.
- Micro1 — Vets and places AI trainers and engineers through a rigorous screening process.
How Much Can You Earn?
Pay varies dramatically based on your expertise and the type of work:
- Entry-level annotation: $15-25/hour — Basic labelling and categorisation tasks that require minimal specialisation.
- General AI training: $25-50/hour — Writing, evaluating, and creating training data for language models.
- Technical/STEM training: $40-80/hour — Coding tasks, mathematical reasoning, and scientific evaluation.
- Expert consultation: $100-225/hour — Domain specialists providing feedback in law, medicine, finance, and engineering.
"AI training is fundamentally affecting the economy; it's an entirely new job sector." — Ali Ansari, CEO of Micro1
Is This a Real Career?
Yes — and it's getting more real by the day. The Wall Street Journal recently reported on the surge in AI training jobs, noting that the industry is "giving rise to new engineering and training jobs" even as it transforms others. Goldman Sachs estimates the broader AI market will reach $1 trillion in annual spending within the next few years, and a significant portion of that flows directly into human data work.
The roles range from flexible side income (10-20 hours per week on your own schedule) to full-time careers with benefits at the platform companies themselves. Many people start with part-time annotation work and progress to higher-paying expert roles as they build experience and reputation on the platforms.
How to Get Started
The barrier to entry is lower than you might think. Here's the typical path:
- Sign up on 2-3 platforms — Don't put all your eggs in one basket. Each platform has different projects and availability.
- Complete the screening — Most platforms have an assessment or qualification test. Take it seriously — your score determines which projects you're offered.
- Start with available work — Accept what's offered initially, even if it's not your ideal project. Building a track record matters.
- Specialise over time — As you gain experience, you'll qualify for higher-paying, more specialised projects.
- Leverage your expertise — If you have domain knowledge (law, medicine, finance, engineering), highlight it. Expert roles pay 3-10x more than general tasks.
