Are your design skills about to become irrelevant in an age of AI? Or can you reinvent yourself in time?You might have heard that smart algorithms are the next silver bullet. Data scientists promise that algorithms will read customers’ data and magically predict their needs, bringing them the information they want, when they want it.I’ve been spending time with the people who write the algorithms, looking at how they work. How is it that data scientists are able to predict future behaviour from just a few variables? When do the people building recommendation engines know that they're sampling the right data? What problems keep them up at night? What does it take to get an algorithm ready to unleash on real people?In this talk I’ll look what algorithms are good for, and where they fall down.I’ll look at the fallacy that underlies much of our thinking about context and how changing the way we think about context helps us understand our role.I’ll discuss what that means for the kinds of design patterns we need to develop in the future.I’ll identify where interaction designers' core skills can help create the products and services of the future - and the challenges we'll face in adapting.I’ll explain how we can influence the work of the people writing the algorithms and sifting the data. And I’ll share some of the tools, methods and documents that I and others have been working on to help interaction designers communicate their knowledge and ideas to the people writing the algorithms.