What the work actually involves
You will spend most of your time producing and critiquing artifacts, not writing essays about them. A typical task set might ask you to draft a registrability opinion on a borderline suggestive/descriptive mark, then judge two model attempts at the same opinion and explain in writing which analysis of the likelihood-of-confusion factors holds up. Other days it's a Section 2(d) office-action response, a fair-use memo under the four factors with post-Warhol reasoning, a redline of an assignment clause that fails to address intent-to-use restrictions, or a portfolio tracker in Excel with renewal deadlines, class coverage, and status columns that a paralegal could actually use.
Deliverable craft matters as much as legal accuracy. The lab is training on document, spreadsheet, and slide-deck tasks, so formatting, citation form, table structure, and clean slide hierarchy are part of the graded output — a correct analysis in a sloppy file is a weak sample.
What the screening looks for
- Verifiable practice history. Bar licence, years in trademark/copyright, and specifically whether you personally owned filings and responses rather than supervising from a distance.
- Depth under follow-up. The AI voice screen asks a doctrinal question, then pushes on the edge case. Confident but shallow answers surface quickly.
- Evaluation judgment. Can you articulate why one draft is better, in terms another reviewer could apply consistently — not just "this one reads better."
- Tooling fluency. Comfort building real spreadsheets and decks, not just prose documents.
Logistics
Fully remote, fully async, no fixed hours. Commitment is flexible between 5 and 20 hours per week, with room for more. Work arrives in batches, so availability in practice means being able to pick up and finish tasks within their review windows rather than being online at set times. Pay is $100/hour as observed on this listing; rates on AI evaluation platforms vary by task type and are not guaranteed.