An eighth-day field guide to the embodied AI scramble

What it does

Octoday Hub is a weekly-updated knowledge index for embodied AI, collecting 319 selected papers, 151 open-source tools, 274 companies, and 538 job postings alongside textbooks and courses. It is maintained by a community of enthusiasts and organized as a set of navigable markdown files rather than a web application. The goal is to bridge academic research and industry practice by giving developers a single coordinate system for a field that sprawls across robotics, vision, and reinforcement learning.

The interesting bit

The project brands itself around “Octoday” (星期八), the mythical eighth day of the week that breaks the seven-day cycle and supposedly offers a fresh perspective beyond routine. That framing turns a conventional awesome-list into a self-described “coordinate origin” for a discipline that currently lacks a canonical map. The value proposition is curatorial authority: the maintainers want this to be the first bookmark you open, not the last.

Key highlights

• 319 curated papers across embodied foundation models, manipulation, locomotion, navigation, simulators, and surveys, plus a 2026–2027 conference and competition calendar.
• 274 company profiles (171 domestic, 103 overseas) and 538 active job listings with direct links to openings at firms like Boston Dynamics, Figure AI, and Unitree.
• 151 tools and open-source projects covering the full chain from simulation to real-robot deployment, plus 14 books and 16 courses from Stanford, CMU, and Tsinghua.
• Content is CC BY-NC-SA 4.0 licensed and accepts community PRs, though the README does not specify a formal editorial board beyond open contribution.

Caveats

• This is a curated reading list, not a framework or dataset—expect markdown tables and hyperlinks, not installable code or reproducible benchmarks.
• The README advertises weekly updates but does not describe a vetting pipeline for verifying job postings or paper quality beyond community PRs.
• The non-commercial Creative Commons license means commercial reuse of the compiled content is restricted.

Verdict

Developers and researchers trying to map the fragmented embodied AI landscape will find this a useful first port of call for papers, tooling, and industry signals. Pure builders looking for a single installable framework should keep scrolling—this is a reading list, not a library.