In a profoundly optimistic but admonishing report, "Towards 2020 Science," the Microsoft Research lab of Cambridge, UK, in association with a diverse array of leading scientists, found that advances in computing will transform the way the sciences — especially the life sciences — are conducted, but also cautioned that a radical educational, governmental, and societal accentuation on science will be required to bring this promise to fruition.

Advances in computation will afford greatly more accurate modeling of complex systems such as population epidemic and weather patterns, helping to improve response to real-time outbreaks and avert potential epidemics of avian influenza, severe acute respiratory syndrome (SARS), and malaria; as well as make possible much more subtle comprehension of climate change.

Biodiversity, energy consumption, personalized medicine, and social systems are other complex concerns pregnant with promise, according to the 2020 Science Group, the name by which the report authors are collectively known. The Group views the "ability to understand and predict how complex systems produce coherent behavior" as the single most important scientific challenge of the next 15 years.

The report observes how computer science is already enabling new kinds of science through the development of "molecular machines" and the widespread encoding of scientific knowledge, the most widely known early example being the Human Genome Project.

Prof. Stephen Emmott, director of Microsoft's scientific research programs in Europe and chairman of the 2020 Science Group, said in a statement, "The weight of human existence on the planet has begun to break down the very systems on which we depend, and it is vital that we increase our knowledge of complex physical and biological systems through scientific advances. This report establishes the necessity of applying the cutting edge of computer science to more quickly find solutions to the challenges we are facing."

One of the most philosophically interesting areas of change is the traditionally human function of formulating and testing scientific hypotheses, which has already become unsustainable in many sciences without the aid of computers because scientists are "no longer able to conceptualize the breadth and depth of the relationships and potential relationships" within and between huge databases now being generated.