Learning Machines: Computation, Ethics and Policy
THIS COURSE HAS BOTH AN ONLINE AND OFFLINE OFFERING
An introductory course about Machine Learning, its applications, and ethical and societal impacts. The course leverages project based, hands-on learning, and plugged and unplugged activities.
Machine learning models are increasingly being used across several industries and domains. Applications of machine learning (ML) impact our lives directly or indirectly, regardless of the fields in which we work. Further, these applications have immense ethical and societal concerns and implications. This beginner course is designed to teach participants the basics of machine learning, different types of machine learning models, their applications in the real world, how they can be embodied in autonomous systems, and ethics and governance considerations.
This is a team-based, hands-on course where participants work on a series of activities, both plugged and unplugged, to build and interact with their own machine mearning models. Through this process of learning-by-making, participants develop their intuitions and understanding of how machine learning systems are built and factors that impact their efficacy. No prior programming experience is needed!
The course begins by an introduction to machine learning and examples of ML we encounter in different fields of work. Participants will learn about the three major types of machine learning: supervised learning, reinforcement learning, and unsupervised learning. Day 1 covers supervised machine learning, classification systems, algorithmic bias, and deep learning. Day 2 introduces reinforcement learning systems and how robots learn from their own experience to interact in the world to perform tasks. Day 3 covers machine learning systems that generate creative outputs such as generative adversarial networks, (GANs), style transer, and recurrent neural networks. Participants will use GANs to create machine-generated digital art and discuss the societal implications human-machine generated IP and of “deep fakes” and the spread of misinformation through social media. Day 4 will bring these types of machine learning capabilities together into autonous systems where participants will learn about the ethics, policy, and governance surrounding AI through hands-on activities with an autonomus robot. Participants will walk away with a basic understanding of different types of Machine Learning models, their applications, and their societal and ethical implications in the real world.
This course is appropriate for everyone–from computer scientist, to journalist, to concerned citizen. It is designed for senior managers in any industry who want to develop their understanding of the major classes of machine learning algorithms build intuitions about how they work and when they don’t work as well as one might expect. The course also discusses the social implications of such technologies in terms of ethics and policy, especially as learning machines become more autonomous.
This course is appropriate for individuals at any company—from global corporations to small start-ups—that develops AI system capabilities or is a user/buyer of AI systems. Participants are welcome from a wide range of industries including energy, consumer AI products and services, aerospace, transportation, robotics systems, finance, national security, and health. Participants interested in the implications of deploying AI systems are especially encouraged to participate.
Cynthia Breazeal, Kate Darling, Safinah Ali, Daniella DiPaola, Randi Williams
Introduction and overview
Types of Machine Learning: Supervised, Unsupervised and Semi-supervised. Experience supervised machine learning with a construction activity.
Working lunch: discussion
Introduction to Deep Learning
Introduction to Reinforcement Learning
Experiencing Reinforcement Learning with robot toolkits
Working lunch: discussion and brainstorm: how can we apply these ideas to students’ areas?
Ethics of autonomous systems
Artificial Intelligence and Creativity. Introduction to generative modeling techniques.
Generative Adversarial Networks
Working lunch: discussion and brainstorm: ownership and agency in machine generated art
Human-AI co-creativity. Can AI inspire creativity in humans?
Ethics. Misinformation and Deepfakes
Ethics and policy overview
Hands-on robot session
Law and policy
There are no pre-requisites to this course. This is an introductory course. No prior programming experience is necessary.