Part-time remote work (20 hours a week on weekdays) with a leading AI lab’s research team: design hard biology tasks and benchmarks on cloning, CRISPR design, functional genomics and cell-based assays, write reference solutions and evaluate AI models’ reasoning.
Remote contract work for generalists on an AI training project: carry out real-world online tasks with AI tools (including purchase and cancellation workflows, with reimbursement procedures), document them with screenshots and give structured feedback on AI outputs. micro1 lists 50 openings.
Part-time remote work (20 hours a week) with a leading AI lab’s research team: design chemistry tasks and benchmarks on route design, reaction optimization and spectroscopic characterization, write reference solutions and evaluate AI models’ chemical reasoning.
Full-time (40 hours a week, no other engagements) remote AI training work for ML systems engineers, open worldwide: design and solve hard tasks on accelerator kernels, profiling, distributed debugging and LLM inference serving, and review ML systems solutions. JAX, XLA or TPU experience is especially valued.
Full-time (about 40 hours a week) remote role helping a leading AI lab evaluate generative video models: set quality standards for color consistency, skin tones, pacing, scene cuts and HDR, curate hard benchmark examples and critique model output.
Full-time (about 40 hours a week) remote role helping a leading AI lab evaluate generative video models: define standards for VFX detail, compositing, simulation, previs and concept art, curate hard examples and critique generated footage frame by frame.
Full-time (about 40 hours a week) remote role helping a leading AI lab evaluate generative video models: define standards for acting realism, blocking, screen direction, eyelines and continuity, curate hard examples and write notes on generated scenes.
Full-time (about 40 hours a week) remote role helping a leading AI lab evaluate generative video models: define standards for lenses, camera motion, lighting and exposure, curate hard examples and critique generated footage shot by shot.
Remote, asynchronous AI research work for physicists: write or verify rigorous multiple-choice questions with step-by-step solutions and references in fields such as semiconductors, photonics, quantum optics, plasma physics and geophysics. 10+ hours a week.
Lead a pod of trainers building terminal-based physics tasks for Terminal Bench Science: review every task end to end (datasets, simulation inputs, environments, reference solutions, automated tests) for physical correctness, reproducibility and grader reliability, and coach trainers.