From the course: Introduction to Responsible AI Algorithm Design
Unlock the full course today
Join today to access over 22,600 courses taught by industry experts or purchase this course individually.
Solution: Policy search
From the course: Introduction to Responsible AI Algorithm Design
Solution: Policy search
(upbeat music) - [Instructor] How did you find the question? I think sometimes we read lots of different articles about the potential for AI to cause harm. But it's important to learn from when these groups have tried to implement systems, and I'm sure they mostly have the best intentions and they usually make mistakes. but it's interesting to think about how they started and what they ended up with. So, although I won't name names explicitly here, one of the instances that came to my mind was that there are many different data sets that are being used in machine learning. And sometimes it's hard to know exactly what's in that data. Recently, some companies have started adding responsible guidelines to these data sets and also including where some fields can be trusted and other ones can't. So in a world before where anyone would access data, train a model and then implement it, now they're getting more enhanced metadata…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.