Physical Sciences Research Assistant

$30–50/hrrate as listed

micro1 is contracting physics, chemistry, and materials science researchers to author advanced evaluation tasks and 35+ item grading rubrics used to benchmark AI models. Work is remote, output-based, and paid per accepted task, with observed rates in the $30–50/hr range depending on your speed and task complexity.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work involves

You build the problems that AI models get tested against. Each task starts from a real technical situation — an experimental dataset that needs error analysis, a computational result that needs interpretation, a materials characterization problem with ambiguous signals — and ends as a fully specified evaluation item with a documented reference solution. You are expected to source authentic artifacts: CSVs, instrument output, spreadsheets, published figures, technical PDFs. Textbook-style problems with clean numbers are explicitly out of scope.

The rubric is the deliverable that takes the most care. Project specs call for 35+ discrete scored items per task, covering numerical accuracy, unit and significant-figure handling, methodological choices, interpretation quality, and whether the model respected scientific protocol rather than pattern-matching to a plausible answer. Rubric items need to be independently checkable by a second reviewer who did not write the task, which means your reasoning and assumptions have to be written down, not implied.

What the screen looks for

micro1 runs an AI-led interview before any task access. It probes for verifiable research specifics — what instruments you actually ran, what your analysis pipeline was, where your data was noisy and how you handled it — and follows up on your answers rather than accepting them at face value. Expect to defend a quantitative claim, explain a method you named, and walk through how you would distinguish a correct answer arrived at by wrong reasoning from a correct answer arrived at properly. Clear written English matters, because rubric text is read by other reviewers and by automated pipelines.

Logistics

Fully remote, contractor engagement, asynchronous with no fixed hours
Compensation is per accepted task; the $30–50/hr band is an observed effective rate, not a guarantee, and depends on how fast you produce work that passes review
Minimum weekly task submission requirements apply
Hiring moves fast — roles often fill within 48 hours, with first tasks expected 24–48 hours after onboarding
No prior AI or ML experience required

What they look for

Methodological rigor
Rubric fidelity
Experimental reasoning
Protocol fidelity
Quantitative rigor
Technical documentation

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you hold a Bachelor's or Master's degree in physics, chemistry, materials science, or a closely related discipline in the physical sciences?
Degree in physical sciences · required
Do you have a minimum of 2 years of hands-on research experience or applied technical project work in a physical sciences domain?
Hands-on research or applied technical experience · required
Are you able to complete onboarding and begin your first assigned tasks within 24–48 hours of acceptance into the role?
Availability to start within 24–48 hours · required
Are you able to work independently on a remote, output-based contract with a minimum task submission requirement per week?
Remote work and independent output · preferred
Can you write clear, professional technical documentation and protocols in English?
Proficiency in written English documentation · preferred

0 of 5 answered

Included with this listing

Interview prep — 12 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether you've genuinely authored rigorous experimental design (not just followed someone else's), and whether you can articulate the logic connecting method to outcome—critical for authoring tasks that test AI reasoning about experimental validity.
A strong answer: Name a specific domain (e.g., materials characterization, fluid dynamics, synthetic chemistry), describe the core question and why your chosen method was necessary rather than optional, and explain at least one decision you made to strengthen validity (controls, measurement precision, sample size rationale).
Where people lose it: Describing a project you assisted with or executed under close supervision, without showing you understood or chose the methodological decisions—the role requires you to author tasks, not reproduce them.