English Speaking Generalist

$20–30/hrrate as listed

micro1 recruits fluent English speakers to complete short written and conversational tasks that help train and evaluate AI models, with no specific degree or industry background required. Observed pay for generalist work sits around $20–30/hr, billed against completed and approved tasks.

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

Generalist evaluation work is the broadest tier on micro1's contributor bench. On a given day you might compare two model responses and explain which is better and why, rewrite a weak answer into a strong one, write original prompts designed to stress-test a model, or record short spoken responses for speech-related datasets. Tasks arrive in batches through the platform, each with its own rubric and a stated time expectation. You are graded on rubric adherence and reasoning quality, not on volume alone — reviewers read your written justifications, and sloppy or padded explanations get sent back.

Because there is no single domain attached to this role, the variety is real: consumer advice, everyday reasoning, summarization, tone and formatting judgments, factual spot-checks, and instruction-following edge cases. Contributors who do well tend to be strong, fast writers who can hold a rubric in their head and articulate a preference in two or three clean sentences. If you have a specialty (law, medicine, a language, a trade), flag it — micro1 routes contributors into higher-paying specialist queues when the work appears.

What the screen looks for

Written English at a native or near-native level — grammar, register, and the ability to write plainly without filler.
Reading comprehension under pressure — instructions are long and the edge cases are deliberate.
Judgment you can explain — the platform cares less about picking the "right" answer than about whether your stated reason matches the rubric.
Honesty about uncertainty — flagging that you cannot verify a claim scores better than confidently guessing.

Expect an AI-led interview: a short spoken conversation, some scenario questions, and often a sample task. Micro1's screening includes identity and proctoring checks, and using an AI assistant to answer for you is an automatic fail.

Logistics

Fully remote and asynchronous. Work is contract, hourly or per-task depending on the project, with no guaranteed weekly minimum — batches come and go, and quiet stretches are normal. Most contributors treat this as flexible part-time work of 10–25 hours a week, choosing their own hours. You will need a reliable computer, stable internet, and a usable microphone for any voice tasks.

What they look for

AI training
AI Response Evaluation
Critical Thinking & Analytical Reasoning
Attention to Detail
Documentation Interpretation

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.

Is English your native language or have you lived and worked in an English-speaking country for at least 5 years?
English fluency (native or near-native) · required
Do you have stable broadband internet (minimum 5 Mbps download speed) available where you work?
Reliable internet · required
Do you have access to a usable microphone (built-in, headset, or external USB) for recording short spoken responses?
Microphone for voice tasks · required
Can you work flexible hours and accept that task batches come and go with no guaranteed weekly minimum hours?
Availability for asynchronous work · preferred
Do you have deep knowledge in any specific field—law, medicine, a language, engineering, trades, finance, or another domain—that you can demonstrate?
Specialty or domain expertise · preferred

0 of 5 answered

Included with this listing

Interview prep — 10 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 the candidate reads instructions fully before starting, checks their work against the rubric, and can self-correct when they've misunderstood a criterion.
A strong answer: A strong answer names re-reading the rubric once before starting, identifying the specific criteria that apply to each comparison or judgment call, and doing a final check of the rubric before submitting—catching your own drift toward habit instead of the stated rule.
Where people lose it: Candidates who say they 'skim' instructions, assume they understand after one read, or rely on pattern-matching from prior tasks will miss edge cases that the rubric specifically flags. Micro1 batches are designed to have deliberate gotchas; skimming loses points fast.