Open source and open standards in the age of cloud AI

Tim O'Reilly considers how to extend the values and practices of open source in the age of AI, big data, and cloud computing.

When the term “open source” first took hold 20 years ago, we lived in an age dominated by personal computers. The internet was shaking things up, but it is easy to forget how peripheral it still was. Amazon was only three years old, and Google was born that year. The architectural assumptions of the PC era suffused the definition of open source; the ability to obtain the source code, modify it, run it on your own machine, and redistribute your changes to others was actually how we got new functionality on our machines. Control over proprietary software was the key to monopoly lock-in. Today, network effects and data accumulation are the sources of lock-in. The ability to study, modify, and build on source code surely still matters, but much of the software infrastructure we depend on today is so tightly woven with accumulated user data and business processes, and at such scale, that source code access alone doesn’t level the playing field in the way it once did. How are we to extend the values and practices of open source in the age of AI, big data, and cloud computing? Join Tim O’Reilly to find out.