Write and review Lean 4 proofs for a leading AI lab, formalize informal mathematics and judge whether a model's proof actually proves the right statement. A W-2 part-time employment position through Cincinnatus LLC, at least 20 hours a week and up to 40, paying $90–110/hour.
Write research questions, answers and evaluation material that benchmark AI models for a leading lab, 10 to 20 hours a week. Open to postgrad students or people with 2 to 3 years of experience, but you must own an Apple Silicon MacBook to do the tasks.
A short, one-off evaluation task for mental health professionals: compare pairs of simulated clinical conversations and judge which is more realistic. The ad says it takes up to 30 minutes. Remote contract at $100/hour, for clinicians with a relevant degree and practical clinical experience.
US-based pharmacists and pharmacy technicians listen to short audio clips of spoken medication names and score whether each is pronounced accurately and clearly, using a rubric. Remote contract at $70/hour for an AI healthcare project, with a calibration exercise first.
Economists with a Master's or PhD and a year or more at a top research institution (World Bank, IMF, the Fed, a graduate school) work on a research project for a leading foundation-model AI lab. Remote contract at $120–150/hour, at least 10 hours a week for a minimum of four weeks.
Condensed matter PhDs create, solve, review or audit research-level problems for CritPt, a public benchmark testing whether frontier AI models can do real physics research. Nineteen narrow research areas, from bosonization to SYK. Remote hourly contract at $80–110 per hour.
AMO physicists create, solve, review or audit research-level problems for CritPt, a public benchmark testing whether frontier AI models can do real physics research. Seven narrow areas, including levitated optomechanics, cavity QED and ultracold atoms in optical lattices. Remote hourly contract at $80–110 per hour.
Researchers who have published on stochastic autocatalytic growth create, solve, review or audit research-level problems for CritPt, a public AI physics benchmark. A single narrow area: chemical master equations, branching processes and reaction-network moments. Remote hourly contract at $80–110 per hour.
High energy and nuclear theorists create, solve, review or audit research-level problems for CritPt, a public benchmark testing whether frontier AI models can do real physics research. Seven narrow areas, from AdS/BCFT to quasi-PDFs and dark photon searches. Remote hourly contract at $80–110 per hour.
Mathematical physicists create, solve, review or audit research-level problems for CritPt, a public AI physics benchmark, where the standard of proof sits closer to mathematics than physics. Four narrow areas, from hypergeometric identities to Fefferman-Graham geometry. Remote hourly contract at $80–110 per hour.
Hands-on structural, thermal, mechanical design and dynamics engineers review and write hard engineering problems about real hardware (loads, margins, heat transfer, vibration, tolerances) for an AI research initiative. 5+ years with 3 recent hands-on. Remote contract, $100–120/hour.
Hands-on systems, integration, reliability and manufacturing test engineers review and write hard engineering problems about real hardware (V&V, qualification, FMEA, root-cause analysis) for an AI research initiative. 5+ years with 3 recent hands-on. Remote contract, $100–120/hour, weekly via Stripe or Wise.
Embedded firmware, FPGA/RTL, flight software and hardware test automation engineers review and write hard problems about software that runs on real hardware (timing, race conditions, interfaces, bring-up) for an AI research initiative. 5+ years, 3 recent hands-on. Remote contract, $100–120/hour.
Hands-on RF, power electronics, analog/mixed-signal, PCB and signal integrity engineers review and write hard problems about real hardware designs, measurements and bring-up for an AI research initiative. 5+ years with 3 recent hands-on. Remote contract, $100–120/hour, weekly via Stripe or Wise.
AI safety testing in Telugu and English: provoke and document jailbreaks, bias and harmful output from chat models, and judge the quality of their Telugu answers. Evaluation skill matters more than security experience. Remote hourly contract at $16–22/hour, weekly pay; H-1B and STEM OPT excluded.
Remote hourly contract for financial analysts, accountants and business analysts who already use Excel on their own Mac. $60–70/hour, paid weekly via Stripe or Wise. Mac Excel specifically, not Windows; you need your own Microsoft licence and a display above 2.5 megapixels.
Broker-dealer and RIA compliance staff review mock periodic reviews, Reg BI rollovers, marketing pieces and alternatives books, producing rubric-graded compliance deliverables for AI training. US-based, about 15 hours a week, listed at $100–130/hour. Needs 3+ years in US securities compliance.
Odia and English AI safety work: probe chat models for jailbreaks, bias and harmful output and judge whether their Odia answers hold up. One of the lowest-resource languages in the family and the busiest Indian variant, with 88 hired this month. Remote hourly contract, $16–22/hour.
Author AI evaluation tasks from real permit drawings, documents and site photos, with the correct correction letter, review decision or markup as the answer. For ICC-certified plans examiners and third-party plan reviewers with 3+ years. US only, $45–60/hour.
Turn everyday accounting work into AI training data: design scenarios from your own practice, review AI outputs for GAAP/IFRS accuracy and judgment, and write structured feedback. Any specialty, from audit and tax to bookkeeping and forensics. US only, 3+ years and a CPA, CA, ACCA, CMA or EA required. $80/hour.
Red-team AI models in European and other non-Brazilian Portuguese plus English: jailbreaks, prompt injection, bias and manipulation, logged as reproducible safety data. Brazilian Portuguese is explicitly excluded. Remote hourly contract at $29–45/hour, paid weekly.
Remote hourly contract for Android developers who already run Android Studio on their own Mac. $55–65/hour, paid weekly via Stripe or Wise. The software is free; you need a Mac with a display above 2.5 megapixels (MacBook Pro 14" and 16" screens qualify). Tasks are not described.
Remote hourly contract for biostatisticians, epidemiologists and applied economists who already run Stata SE on their own Windows PC. $45–55/hour, paid weekly via Stripe or Wise. Your own Stata SE licence and a display above 2.5 megapixels are required. Tasks are not described.
Evaluate frontier AI responses on grey-area and policy-sensitive topics (misinformation, political persuasion, self-harm, violence, cyber, biosecurity), apply safety rubrics and write structured feedback. 5+ years in trust and safety, journalism, policy, research or security. US, UK and most of Europe. $60–70/hour.
New roles land most days. Get them on Telegram or Discord as they are added, or read how the platforms pay before you apply.