Guides
How AI expert work works, what it pays, how the platforms screen, and how to prepare. Written for professionals entering the field.
Start here
What Is AI Evaluation Work? The Complete Primer for Domain Experts
AI labs pay professionals to review and grade model outputs in their field. Here's what the work actually involves, why it exists, and why your expertise is the product.
Guide · 7 min read
Maximizing Earnings in AI Evaluation: Rates, Tiers, and Moving Up
The spread between entry annotation and expert review is wide. Here's how the pay tiers actually work and the specific moves that shift you into the higher ones.
Guide · 6 min read
How to Pass the AI Interview: What Platform Screens Actually Measure
Mercor and micro1 both screen you with an AI interviewer before any human looks at your file. Knowing what the system scores removes most of the anxiety — and most of the failure modes.
Guide · 6 min read
Mercor vs micro1: A Comparison for Candidates
The two biggest expert platforms differ in screening, role mix, and who they suit. Here's the practical breakdown — including where each one frustrates people.
Guide · 6 min read
Analysis and updates
Why Frontier Labs Pay a Premium for Domain Experts
Generic annotation is being automated away while expert evaluation budgets keep growing. The economic logic behind the split — and what it means for professionals.
5 min read
Hiring Signals: What Expert Platforms Are Recruiting for This Quarter
A roundup of the role types appearing most across Mercor, micro1, and other expert marketplaces — and what their screens now emphasize.
4 min read
Conference Watch: Benchmark Saturation Is a Hiring Signal
Every saturated benchmark at a major ML conference is followed by a harder, expert-written replacement — and those replacements are built by paid domain experts.
4 min read
From RLHF to Expert Feedback: The Research Behind Your Evaluation Job
Your evaluation job exists because of a specific research lineage — from the original human-feedback loop to today's expert-graded reasoning models. A plain-language tour.
6 min read