The consensus in physical robotics just took a massive shift\u2014but the real story is far more nuanced than the hype suggests. \u2699\ufe0f\ud83d\udca1<\/p>\n
Six months ago, the prevailing orthodoxy was clear: text\/code LLMs lacked spatial grounding, and end-to-end Vision-Language-Action (VLA) or world models were the *only* path to physical control. Then came Claude Fable 5.2, setting new state-of-the-art scores on RoboDojo\u2014a simulated two-arm manipulation benchmark\u2014and demonstrating impressive zero-shot spatial planning on un-trained tasks.<\/p>\n
Naturally, many rushed to claim LLMs had solved physical manipulation. But if you look closer at real-world physical tests\u2014beyond clean simulation environments\u2014the picture changes fast.<\/p>\n
In a recent informal physical tabletop test, an LLM-guided robotic arm was set up to play checkers. The result? 64 misses and multiple dropped tokens.<\/p>\n
Why the stark disconnect between benchmark performance and physical reality? It comes down to the fundamental divide between high-level spatial reasoning and real-world motor execution:<\/p>\n
\ud83e\udde0 What LLMs Get Right (The High-Level Brain) <\/p>\n
* Strategic Spatial Logic: Fable 5.2 excels at abstract spatial priors (gravity, leverage, multi-step ordering) and long-horizon in-context error recovery. It easily decides *what* moves to make.<\/p>\n