The Covid-19 pandemic emphasizes the speed with which scientific publications are being published, as well as the spread of misleading information in blogs and even scientific papers, giving rise to the problem known as fake news 4. Prashant Pradhan and his Indian partners released a paper on January 31, 2019 where they offered evidence of unusual parallels in the coronavirus sequence with HIV gp120 and Gag proteins, implying that it was a fabricated virus 5.

Serge Horbach of Radboud University Nijmegen submitted a study titled "Article about the pandemic: medical journals are significantly accelerating their publication process on Covid-19" at the pre-publication stage (keep in mind that this is a work that has not yet been peer-reviewed). After reviewing 669 publications published in 14 medical journals, he stated that the time it takes for most articles associated with Covid-19 to be published in scientific journals has been reduced by half 6.

We must keep in mind that not all ideas emerge through the scientific process. Let me recall a few classic non-scientific examples, such as the Archimedes bath incident (287–212 AD). It recalls the account of King Hiero II of Syracuse (306-215 BC), who wished to see if his crown was truly composed of gold. Archimedes attested to this and discovered a solution by bathing in a tin can after noticing a correlation between the amount of water displaced and the body mass.

Another example is Alexander Fleming's (1881–1955) discovery of penicillin in 1928, which paved the way for the development of antibiotics while working with bacterial cultures. When he returned from his vacation, he discovered a petri plate that had been accidently contaminated by a green mold (Penicillium notatum), and with his scientific eye, he recognized the significance of this discovery. This type of chance finding is now known as serendipia.

This was not an isolated occurrence. Other examples include the discovery of X-rays in 1895 by physicist Wilhelm Conrad Roentgen while experimenting with electrons in vacuum tubes, radioactivity in 1896 by Antoine Henri Becquerel, LSD by Albert Hofmann in 1943, aspirin by Felix Hoffman in 1894, and Isaac Newton deducing gravity after falling an apple on his head.

In addition, early notions or ideas are frequently incorrect. Consider the instance of Albert Abraham Michelson and Edward Morley 7, who prepared and carried out an experiment in 1887 to measure the relative speed of the earth with respect to the ether and discovered that their hypothesis was incorrect. Albert Einstein benefited from this observation. Therefore, the data should be the guide to scientific publications. Therefore, the data should be the guide for scientific publications. This approach is not new, as discussed below.

The father of philosophical and scientific empiricism, Francis Bacon (1561–1626), pointed out in his work Novum Organum ("New Instruments") in 1620 that scientific knowledge should not be based on preconceived notions that must be based on empirical data 8, therefore inferences must be drawn from these data, i.e., Bacon argued that science should be technical rather than based on theory or speculation. Furthermore, he argued that knowledge should be constructed by observation and prioritization according to logical principles 8.

Bacon advocated inductive reasoning, the process of drawing conclusions from observations. A notable example of this is Johannes Kepler's (1571–1630) work on planetary motion, until Isaac Newton (1642–1727) was able to publish the laws of universal motion in Principia in 1686 2.

Recently, Chris Anderson (former editor of Wired magazine from 2001 to 2012) published an article titled "The End of Theory: Data Deluge Makes the Scientific Method Obsolete" 9. Basically claims that data and supercomputers will replace the conventional scientific approach, eliminating the need for new hypotheses and theories. In this context, the author believes that we have seen a tendency that may lead to a theoretical science, but this is a topic that has to be researched further in future study.

Anderson emphasizes Craig Venter's case of genome sequencing, in which he effectively sequenced entire ecosystems using ant knowledge gained by sequencing species by using supercomputers and sequencers to generate vast volumes of data 9.

Recently, the possibility of deriving mathematical equations capable of describing the dynamics of a system without considering any operational hypotheses has emerged thanks to the large amount of information and advances in computational programming 10, i.e., a procedure that reverses the scientific method by not using an initial hypothesis.

This initiative was led by Schmidth and Lipson, who used symbolic regression and genetic programming 11, and to date it has been applied to nuclear fusion 12, seismology 13, climate change 14, and recently drugs that can be used to fight Covid-19 15.

As a result, data management is driving the potential of data-driven hypothesis 16 which is an approach that allows us to generate a dynamic solution without any theoretical basis; that is, the objective function capable of explaining the system's behavior is unknown. In other words, this concept is similar to a reverse scientific procedure, with the emphasis on data.