There’s a well-known WC Fields quote, “If you can’t dazzle them with brilliance, baffle them with bullshit.” When it comes to Machine Learning (ML), there can be a tendency to dazzle with bullshit, to make grandiose claims about the capabilities of magical algorithms. Yet attempts to convey more correctly the real brilliance of Machine Learning often leave people baffled. This can be looked at as a UX problem: there is an enormous gap in understanding between those designing ML solutions and those using them, and it is a more fundamental gap than exists naturally between designers and users of other software systems.
This talk will be both an introduction to Machine Learning and an examination of why it presents some interesting challenges to UX professionals.
The talk will begin with an introduction to Machine Learning, providing examples of where it is used today. It will explain the relationship between Machine Learning and Artificial Intelligence and talk about some of the problematic ways the latter is being reported on in the media and how this contributes to a general misunderstanding of the field by non-practitioners. The talk will present a very basic introduction to how ML actually works, without getting too technical. It will discuss issues around ethics and dealing with biased data and present some common pitfalls. Seeing where and how ML can go wrong is an excellent way for non-practitioners to get a better handle on what it is and how it works. It also helps people to be able to assess claims being made about ML. The talk will also discuss the importance of the idea of uncertainty in Machine Learning.
Katherine Bailey – Acquia