Alan Turing on Embodied Intelligence (rodneybrooks.com)

Rodney Brooks resurfaced a 2011 essay written for a 2012 Alan Turing centenary conference that revisits Turing’s largely overlooked 1948 essay Intelligent Machinery and argues for the primacy of embodied intelligence. Brooks highlights that Turing explicitly distinguished embodied versus disembodied approaches, calling embodiment the “sure” route to a thinking machine—yet he favored disembodied work for practical reasons of his era. Turing’s embodied vision described robots with cameras, microphones, actuators and remote digital “brains,” and he proposed three modes of search for building intelligence: intellectual, genetic, and cultural. Many early AI priorities (games, language, cryptography, mathematics) followed Turing’s disembodied suggestions and shaped the field for decades. This matters now because modern robotics, cloud compute, and social/developmental learning have made Turing’s embodied and cultural ideas practical again. Brooks notes that contemporary platforms (PR2s, “Mekabots,” offboard/cloud brains) and research into social interaction as a learning signal realize Turing’s blueprint: embodiment plus cultural immersion yields richer, grounded learning than disembodied benchmarks alone. For the AI/ML community, the piece is a reminder that architecture choices (sensorimotor coupling, social scaffolding, developmental learning) are not just philosophical but technical levers that influence generalization, sample efficiency, and the kinds of intelligence systems can attain.