Work inside a leading AI lab's research team as its insurance and actuarial specialist: QA model outputs, write instruction specs and golden solutions, and build insurance benchmarks. Needs an FSA, ASA, FCAS or ACAS, or an active adjuster, underwriting or broking licence, plus 4+ years. Full-time W-2, hybrid Bay Area, $60–100/hour.
Design marketing tasks and solutions, score AI outputs against rubrics, and refine marketing-specific evaluation guidelines for an AI lab. For marketers with 8+ years at top-tier brands or agencies and prior LLM rubric experience. US, 35+ hours a week, $60–80/hour.
Native Korean speakers in Canada, South Korea, the UK or the US map the layout of real Korean PDF pages and transcribe every region exactly in Hangul (hanja and handwriting included) to train document AI. Remote hourly contract at $33.58/hour; every task is reviewed by a second Korean expert.
AI safety testing in Urdu and English: provoke jailbreaks, bias and harmful output from chat models and judge whether their Urdu answers are accurate and appropriate, in Nastaliq script or Roman Urdu. Evaluation judgment is the core ask. Remote hourly contract at $16–22/hour, weekly pay.
Build retail merchandising, category and operations tasks, write solutions, and grade AI outputs against rubrics for an AI lab. For merchants, category managers and retail operators with 8+ years at major retailers and prior LLM rubric experience. US, 35+ hours a week, $60–80/hour.
Evaluate generative music AI for a leading AI lab: compare AI-made songs head to head on musicality, prompt adherence, vocals and mix, label genre and structure, and check lyrics and vocals. For Thai-speaking producers or engineers with 2+ years' experience. Remote, flexible hours, up to 6 months, $18/hour.
Rate AI-generated music for a leading AI lab: head-to-head song comparisons on musicality, prompt adherence, vocals and mix, genre and structure labelling, and lyric and vocal checks. For Russian-speaking producers and mix engineers with 2+ years' experience. Remote, flexible hours, up to 6 months, $35–49/hour.
Senior US full-stack engineers (6+ years, end-to-end system ownership) build production-grade software on a leading AI lab's pre-release models, integrate tool interfaces and evaluation harnesses, and document model failure modes for researchers. Full-time W-2 through Cincinnatus LLC, 40 hours a week, $90–110/hour.
Remote hourly contract for video editors and content producers who already cut in Adobe Premiere on their own Windows PC. $30–40/hour, paid weekly via Stripe or Wise. Your own Adobe subscription and a display above 2.5 megapixels are required; the tasks are not described.
Evaluate GPU and accelerator kernel development tasks for a frontier AI lab: numerical correctness, benchmarking fairness, task scoping and compile or runtime validity across CUDA, Triton, NKI and Pallas. For US engineers with 3+ years of kernel work in at least two of those frameworks. $70–90/hour.
Turn ambiguous AI program requirements into clear, contradiction-free rater guidelines and rubrics across finance, retail, insurance, legal and sports. For linguists, instructional designers and technical writers with 3+ years and GenAI/RLHF guideline experience. US, 35+ hours a week, $45–65/hour.
Paid pilot for US market access and pricing professionals: set and pressure-test gross-to-net and formulary-tier assumptions by drug class, judge whether coverage and rebating assumptions match real payer behaviour, and write rubrics for AI drug analysis. 5+ years, 10–20 hours. $175–200/hour.
Remote hourly contract for photographers, retouchers and digital designers who already use Adobe Photoshop on their own Windows PC. $30–40/hour, paid weekly via Stripe or Wise. Requires your own Adobe licence and a display above 2.5 megapixels. 9 hires this month.
Audit Kubernetes tasks used to train and evaluate a frontier AI lab's models: cluster-operations scenarios, manifest and Helm correctness, and failure-mode troubleshooting (CrashLoopBackOff, OOMKilled, eviction). For US engineers with 3+ years of production Kubernetes and Go, Python or TypeScript. $70–90/hour.
Design adversarial prompts, find jailbreaks and policy failures, and document vulnerabilities in frontier AI models across cyber, biosecurity, fraud, misinformation and political content. Hourly remote contract at $70–84/hour for residents of Europe, the UK and the US; 5+ years' relevant experience required.
Full-time finance specialist embedded with a leading AI lab: QA model outputs, write instruction specs and golden solutions, and build finance benchmarks across FP&A, IB, asset management, PE, risk or treasury. 5+ years at a recognised institution, VP-level progression. W-2, hybrid Bay Area, $60–100/hour.
Experienced Spanish tutors lead recorded one-on-one lessons with native English-speaking learners so a language-learning platform can capture authentic tutoring conversations. 3+ years' teaching and a degree required. About 7 hours a week for roughly three weeks, $60–65/hour, remote.
Grade AI-generated slides, spreadsheets and documents for real-world data science quality, flagging factual, visual and presentation errors in structured written feedback. Needs 5+ years at a top firm in the US, UK, Canada, Australia or New Zealand. $100–150/hour.
Paid pilot for US epidemiologists: size patient populations for drugs and indications (prevalence, incidence, diagnosed to treated to addressable), judge whether estimates are sound and well sourced, and write rubrics for AI drug analysis. 5+ years and a graduate degree, 10–20 hours. $150–175/hour.
Coordinate the finance raters on a leading AI lab's training-data program: build tracking and escalation workflows, monitor throughput and quality, triage rater questions and keep finance tasks consistent. 5–10 years in finance or finance operations with team coordination experience. US only, 35+ hours a week, $40–60/hour.
An expert-interview listing for engineers who have shipped production search, especially agentic search in the LLM era: a 25-minute conversational interview about relevance, evaluation and real trade-offs, with a possible paid 30-minute follow-up call at $200. No coding, no take-home. Listed at $80–150 per task.
Grade AI-generated slides, spreadsheets and documents for real-world software engineering quality, flagging factual, visual and presentation errors with written feedback. Needs 5+ years at a top firm in the US, UK, Canada, Australia or New Zealand. $100–150/hour.
Native German speakers write prompts on sensitive subjects, classify conversations and flag adversarial phrasing so AI models behave safely in German. Business English and a bachelor's (in progress is fine). Part-time remote at $48–52/hour; Germany or Western Europe preferred, not required; 8 hires this month.
Native French speakers write prompts on sensitive subjects, classify conversations and flag adversarial phrasing so AI models behave safely in French. Business English and a bachelor's (in progress is fine). Part-time remote at $48–52/hour; France or Western Europe preferred, not required; 7 hires this month.
New roles land most days. Get them on Telegram or Discord as they are added, or read how the platforms pay before you apply.