Abstract
I have chosen first to present the overall subject in all its breadth in a double lecture, i.e. without a seminar. In the first part, I recalled that computer science is founded on four pillars: data, algorithms, languages and machines, accompanied by interfaces with humans or other machines. I went on to explain why the information it manipulates is very different from the objects of study in the classical sciences - matter, energy and waves - and why processing it opens up powerful new possibilities. I illustrated this essential point with examples of photographic and medical image processing, followed by the computerization of scientific instrumentation and the widespread use of digital simulation. I then recalled the importance of Moore's Law (the doubling every two years of the number of transistors per square centimeter of silicon), which has driven the evolution of the hardware infrastructure, enabling the transition from the large systems of the past to much more powerful portable objects like smartphones, which have made computing ubiquitous, as well as the evolution of the associated software infrastructure with the widespread networking of machines, the Web and cloudcomputing. Finally, I've highlighted two more recent trends: the refocusing on data, due to new analysis possibilities enabled by probabilistic, statistical and machine-learning algorithms, and the rise of widespread computerization of everyday physical objects, leading to what is known as the IoT (Internet of Things).
