Fifty Sources, in Reading Order

Start here (the arguments)

1. Ken Goldberg, Closing the 100,000 Year "Data Gap" in Robotics, Science Robotics, Aug 2025
2. Rodney Brooks, Why Today's Humanoids Won't Learn Dexterity, Sep 2025
3. Sergey Levine, Sporks of AGI
4. Sergey Levine, Language Models in Plato's Cave
5. TRI et al., A Careful Examination of Large Behavior Models
6. Lin et al., Data Scaling Laws in Imitation Learning
7. Figure, Master Plan
8. IEEE Spectrum, Walden Robotics / Tedrake interview
9. IEEE Spectrum, Humanoid robot scaling
10. The Robot Report, Unitree IPO analysis

Architectures and models 11. Chi, Song, Tedrake et al., Diffusion Policy 12. Zhao, Finn et al., ALOHA / ACT 13. Open X-Embodiment / RT-X 14. Octo 15. OpenVLA 16. CrossFormer 17. Wang, Chen, Zhao, He, Heterogeneous Pre-trained Transformers 18. Physical Intelligence, π0 19. Physical Intelligence, π0.5 20. Physical Intelligence, π*0.6 / RECAP 21. Physical Intelligence, π0.7 22. Google DeepMind, Gemini Robotics 1.5 23. Google DeepMind, Gemini Robotics 2 24. NVIDIA, Isaac GR00T N1 25. Figure, Helix and Helix 02 26. Generalist AI, GEN-0 and GEN-1 27. Skild AI, Omni-bodied 28. Meta, V-JEPA 2

Data collection 29. Chi et al., UMI 30. DexUMI 31. Fang & Agrawal, DEXOP 32. Mobile ALOHA 33. TWIST2 34. DROID 35. AgiBot World Colosseo 36. RoboMIND 37. Ego-Exo4D 38. EgoDex 39. Sunday Robotics, No Robot Data

Evaluation, simulation, safety 40. SIMPLER 41. RoboArena 42. BEHAVIOR Challenge 2026 43. lbm_eval 44. Isaac Lab / Newton 45. MuJoCo Playground 46. Drake 47. ISO 10218-1:2025 48. ISO/TS 15066:2016

Structure and abstraction 49. Garrett et al., Integrated Task and Motion Planning survey 50. Mao, Wu, Tenenbaum, Building Intelligent Agents with Neuro-Symbolic Concepts

---