We are a community of scientists at MIT and IBM Research. We conduct AI research and work with global organizations to bridge algorithms to impact business and society.

↳ Inside the Lab

Spotlight

How the Coronavirus Pandemic Is Breaking Artificial Intelligence and How to Fix It
news

How the Coronavirus Pandemic Is Breaking Artificial Intelligence and How to Fix It

07/29/2020 - Gizmodo

Computers Already Learn From Us. But Can They Teach Themselves?
news

Computers Already Learn From Us. But Can They Teach Themselves?

04/08/2020 - The New York Times

AI Is Changing Work — and Leaders Need to Adapt
news

AI Is Changing Work — and Leaders Need to Adapt

03/24/2020 - Harvard Business Review

ObjectNet: A bias-controlled dataset object recognition
research

ObjectNet: A bias-controlled dataset object recognition

Computer Vision

Causal inference is expensive. Here’s an algorithm for fixing that.
research

Causal inference is expensive. Here’s an algorithm for fixing that.

Causal Inference

AI could help design better drugs that don’t clash with other medication
news

AI could help design better drugs that don’t clash with other medication

02/06/2020 - MIT Technology Review

EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs
research

EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs

Graph Deep Learning

Class-wise rationalization: teaching AI to weigh pros and cons
research

Class-wise rationalization: teaching AI to weigh pros and cons

Natural Language Processing

GAN Dissection: Visualizing and Understanding Generative Adversarial Networks
research

GAN Dissection: Visualizing and Understanding Generative Adversarial Networks

Generative Models

Membership

The MIT-IBM Watson AI Lab membership program is forging a new model for R&D. We extend the unique collaboration between MIT and IBM Research to a small group of innovative companies and strategic partners comprising leaders in consumer technology, medical devices, finance, construction, energy, and international development.