Drowning in Data : VO to the rescue

Astronomical datasets are growing in size and diversity, posing severe technical problems. At the same time scientific goals increasingly require the analysis of very large amounts of data, and data from multiple archives. The Virtual Observatory (VO) initiative aims to make multiwavelength science and large database science as seamless as possible. It can be seen as the latest stage of a long term trend towards standardisation and collectivisation in astronomy. Within this inevitable trend, we can avoid the high energy style of building large fixed hierarchical teams, and keep the individualist style of astronomical research, if the VO is used to build a facility class data infrastructure. I describe how the VO works and how it may change in the Web 2.0 era.