Everybody talks about machine learning and a coming sea-change in IT. While it can be hard to filter out the hype, a lot of computing is moving to an abstraction level where a classifier is used as pervasively as a switch statement. What about offensive research? Is it an automation-resistant cottage industry? Or will “AI” (for some definition thereof) play a role? Perhaps some definition of “AI” already does?
This talk will discuss what we can learn from AlphaGo, and list a number of areas where some form of machine learning seems likely to make progress in the next years – although I have no idea how many.