AI Policy Wiki
Dashboard

Daniela Rus

high confidence · updated 2026-06-06

Andrew (1956) and Erna Viterbi Professor of Electrical Engineering and Computer Science at MIT; Director of MIT CSAIL. Pioneer of liquid neural networks and robotics-ML. Co-founder of Liquid AI. Author of 'Private Physical AI for the Edge' (Digitalist Papers Vol. 2).

Daniela Rus is the Andrew (1956) and Erna Viterbi Professor of Electrical Engineering and Computer Science at MIT and Director of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL), a position she has held since 2012. She is a principal investigator behind liquid neural networks and a co-founder of Liquid AI (2023). Her research spans robotics, embodied AI, and compute-efficient machine-learning architectures.

Research

Liquid neural networks

Rus is a principal investigator behind liquid neural networks (LNNs), compact, causal, energy-efficient AI architectures inspired by the neural dynamics of small organisms; the nematode C. elegans has 302 neurons, and LNNs abstract the underlying neuron-dynamics mathematics at scale. The architecture is described as provably causal, learning the task rather than the task context, and capable of zero-shot generalization across environments. It is designed to run on modest hardware such as phones, drones, and embedded devices, with energy efficiency expressed through a "tokens per watt" metric.

Robotics and embodied AI

Rus's robotics research includes modular robotics (self-reconfiguring robot systems), drone autonomy (frequently used as a demonstration platform for liquid-neural-network generalization), soft robotics (using morphological computation to reduce reliance on control systems), and embodied intelligence (the physical grounding of AI capabilities).

Liquid AI

Rus co-founded Liquid AI in 2023 with Ramin Hasani, Mathias Lechner, and Alexander Amini, all MIT collaborators. The company open-sourced its Liquid Foundation Models (LFMs), small models the company describes as performant and trained on hundreds of GPUs, compared with the tens of thousands used by frontier competitors.

Positions and statements

In "Private Physical AI for the Edge" (Digitalist Papers Vol. 2), Rus articulates the case for private physical edge AI — on-device, privacy-preserving, energy-efficient AI — which she frames as an ARM-equivalent counterpart to hyperscale compute. She argues that edge AI could redistribute the benefits of intelligence more widely than any prior innovation.

Rus's work reframes the scaling question from "more parameters means more capability" toward whether desired capabilities can be obtained with substantially less compute, presenting a robotics- and efficiency-oriented alternative to the frontier-model scaling paradigm. Her arguments bear on environmental, governance, and distribution debates.

Distinctions

Rus is a MacArthur Fellow (2002) and an IEEE Fellow, and received the Engelberger Robotics Award (2017). She has received numerous best-paper awards across robotics and machine learning venues.

Relationships