The notion of smart cities is inherently connected with the notion of Big Data. It is Big Data that allows more and more intelligence to be added to our existing urban systems. This intelligence then, at least as a goal, is used to serve the needs of the citizens better, making the everyday operations more efficient and adaptive. Many recent successes of supervised machine learning make it an auspicious tool; however, the long-term vision of smart cities clearly requires technology that goes beyond that. The data collected based on the current operation of the system does not in itself contain information about possible improvements. The next generation of smart cities undoubtedly lies with the systems that build towards autonomous and semi-autonomous “knowledge creation.” They can self-improve and adapt to changing conditions and expectations. They must handle situations that were not anticipated during their design. Such construction of knowledge can be illustrated with the Data, Information, Knowledge, and Wisdom hierarchy. It requires collecting and representing the data; creating relevant “events” from this data; generating rules that can combine information from different sources; and finally, the ability to project into the future and reason back into the past. © 2020, Springer Nature Switzerland AG