What the work involves

This is application-use data collection, not code review. You work inside Ubuntu 24.04 as you normally would and document tasks at the level of detail a model can learn from: installing and pinning packages with apt, resolving a broken dependency, configuring GNOME Settings and extensions, managing permissions and ownership in Files or from the shell, setting up systemd units, handling display scaling and multi-monitor arrangement, troubleshooting a failed boot or a Snap that won't update. Tasks typically pair a screen recording or screenshot sequence with written step-by-step reasoning — what you clicked, what you typed, why that path and not another, and what the output told you.

Submissions are judged on whether the workflow is genuinely representative of real desktop use. Contrived sequences, steps copied from a wiki page, or screenshots of a fresh VM with no history tend to be flagged. The strongest contributors are people who have lived with the OS long enough to have opinions about Wayland versus X11, about Snap versus apt versus Flatpak, and about where GNOME's defaults get in the way.

What the screen looks for

  • That you are actually on 24.04 now. Expect direct questions about your daily driver, your desktop session type, your kernel and release channel. Answers that describe a VM you spin up occasionally, or a server you SSH into, generally don't pass.
  • Depth under follow-up. A screener will take a routine answer — 'I use apt' — and push into held packages, PPAs, `apt` versus `apt-get` output, or what you do when `dpkg` leaves a half-configured package.
  • Hardware honesty. The resolution requirement is checked. Claiming a qualifying display you don't have surfaces immediately at first submission.

Logistics

Fully remote and asynchronous. Work is drawn from a task queue with per-task deadlines rather than fixed shifts, so contributors set their own hours; most sustained earners report committing 10–20 hours a week. Observed pay for this band is $60–70/hr, stated as reported by contributors and not guaranteed — rates vary by task type, review outcomes and project stage. You use your own hardware throughout.