For pharmacists

Pharmacology is a domain where language models are being pushed hard and where their remaining errors are among the most consequential. Drug-drug interactions, contraindications, dose adjustments, pharmacokinetic reasoning, and off-label considerations are all areas where a plausible-sounding wrong answer causes direct harm. Frontier AI labs know this. Their expert-contributor programs for medication reasoning have expanded quickly, and practicing pharmacists are the specialists they are hiring.

What the work looks like

• Response evaluation. You review model outputs on prompts covering drug interactions, dosing (renal, hepatic, paediatric, geriatric), medication reconciliation, pharmacokinetics, or clinical pharmacology. You score them against what a working pharmacist would say.
• Interaction and contraindication checking. You verify specific claims against authoritative references and flag where the model has generated plausible-sounding but incorrect information.
• Adversarial testing. You construct prompts where the model is likely to miss a specific interaction, dose-adjust incorrectly, or overstate an off-label indication.
• Rubric review. You audit the criteria the lab is using to grade medication-related outputs and flag anything that would let unsafe reasoning through.

The work is asynchronous. Two to four hour blocks. Structured deliverables.

What labs actually look for

Active pharmacist registration. Recent clinical, retail, hospital, or industry practice. Written English at the level of a clinical pharmacy note or a peer-reviewed article.

They do not look for AI experience.

Where the pay lands

Rates are in USD. For pharmacists in emerging markets, this is typically well above local dispensing or hospital pharmacy rates on a per-hour basis. For pharmacists in higher-cost markets, it is competitive with clinical consulting.

How to enter

Sign up, upload a recent CV that shows your registration and current practice setting, and let the platform surface open roles in pharmacology annotation, evaluation, and safety flagging.