Computer simulations and knowledge

Computer simulations help science view and test models and theories. They help with the visualization of otherwise abstract knowledge. They can accurately give us a general idea of how a model behaves given controlled parameters. Is it dangerous to rely on such computer-generated programs?

Article: http://phys.org/news/2016-04-science-wrong.html

With credit due to Jeremy Gibbons, The Conversation.

The article describes the shift from pre-modern science, where theories had to somehow be tested using real life instrumentation, to modern science, where computer programming is almost absolutely necessary to test theoretical knowledge. One problem that the author poses is that theoretical solutions come up to theoretical problems, which can be unrepresentative of what is going on. Another problem is that the scope of models is too limited to account for natural variables. In other words, data raised from models only work under very specific conditions.

But, the problem of importance to us is a third problem, which is that we tend to trust computer programs too easily, because of the accuracy and speed of computers. This and the fact scientists (generally seen as very smart people, geniuses) can be considered authorities when speaking about topics of their expertise. Both factors make the majority of people believe in whatever highly educated people and their models describe, regardless if the theories are wrong or the computer simulations are inaccurate, filled with bugs and errors.

I would like to point out a very important concept: conscience. Doing things with conscience and considering aspects outside the box can lead to create or modify in more complete and beneficial ways.

We should always consider what implications our actions (not just scientific models) have over our environment.

Is it acceptable to rely on experiments that cannot possibly factor all variables? Lab testing tries to minimize the contact with external agents that can skew experiments, striving for ideal conditions so that tests can be repeatable under the same circumstances. The same principle should apply to computer modeling.

In the end, we have to do whatever we can do, even as little as that might be, to transform our world into a better place (Peter Singer). Knowledge is key for human progress, as it empowers us to change our way of thinking. All knowledge can be used to transform our environment into a better place.