Machine Learning for Policymakers

Countries around the world are establishing strategies, initiatives, and policies to guide the development and use of machine learning, which is the foundation for modern artificial intelligence (AI). What is essential to understand about machine learning technology and how can it be developed and used responsibly?   

CHM and Google AI present an invitation-only, interactive workshop to decode machine learning for US government leaders. The two-hour virtual event will offer a rare opportunity to interact with a senior research scientist at Google about machine learning technology and key issues of fairness, security, privacy, and explainability. The workshop will address:  

• What is machine learning? How does it work? What can machine learning do?   
• Why is machine learning so prevalent today and what is the importance of data, computing power, and novel algorithms?   
• What does it look like in practice to develop artificial intelligence (AI) in a responsible way? How do you combat unfair bias and address concerns about privacy and security?   
• Is machine learning a “black box”? What level of explainability—the ability to explain why a machine learning model arrived at an outcome—is currently feasible? What are some common misconceptions?   
• How do organizations implement machine learning responsibly?   
• How can machine learning be applied for climate change?  

Following the interactive briefing by Dr. Héctor Pérez Urbina, the second part of the workshop will explore these issues through a case study of machine learning for climate change. The workshop will leave ample time for questions and open discussion of any topics that may interest participants about AI development, responsibility, and policy.  

Agenda 

10 a.m. to 12 p.m. PST 

Access to the Virtual Program

The live program will be livestreamed using Zoom. Registered guests for the virtual program will receive a link to join 24 hours prior to the event date. For more information about attending this program, please email Russell Ihrig at rihrig@computerhistory.org.