Thursday, July 17, 2008

Goodbye science

Chris Anderson of Wired has written an article on the end of the scientific method. He claims that the range of 100% data we hold today renders models and testing obsolete. He claims data can be analysed straight away with a bit of mathematics.

I think the notion of science in this article is a bit blurred, surely mathematics is science as well. And even large data sets need science to be analysed. In fact the speed of computing cannot cope with the size of data available, so the models need to cope with more data simply. Many new scientific developments today are about reducing the size of the data.

Also his article rests on the assumption that all data can be collected - surely not true.

No comments: