As I am currently banned from Facebook, please feel free to copy the article there - with the full acknowledgement, above!
"One of the most frustrating things a person can go through is for their normally stable computer system, tablet, or phone, to suddenly stop working—a bad patch, a bad virus, or just old age. Whatever the case, the level of stress rises considerably when your tool for information stops.
One of the main reasons that software fails is that the software is designed for a specific set of hardware, and usually that software is developed on systems that are higher end so that it can be tested in an optimal situation. The problem is that sometimes software gets put into old systems. Sometimes the software gets used in a completely wrong way.
An interesting example of this is the software used to help predict the fallout of our current coronavirus pandemic. At the Imperial College in London, Neil Ferguson and his team used software that they had used previously to predict other high-risk outbreaks.1 The problem is that the software is flawed—very flawed. When developing software, it is imperative that you run that software through so many iterations that, when you finish, you would be certain that it is going to come up with the same thing over and over. Except that this software, when ran more than once with the same input, would come up with different numbers every time. If you can’t duplicate the results of the original number, then there must be something wrong with the software’s code.
This is basic science 101. In order to have a successful experiment, you must be able to repeat what you did over and over. You would also need other people to be able to duplicate the experiment and get the same results. This did not happen. Not only were they not able to duplicate the results on their own systems, they did not share the code to the rest of the community so that other people could scrutinize their work.
This software is 13 years old and was originally used to predict an influenza outbreak. Neil and his team used this software to predict other outbreaks, and all of his predictions from this software were way over-estimated. In 2002, he predicted that there would be 150,000 deaths from mad cow disease by the year 2080. But the death count was well short of that at 177 in Britain. In 2005, Ferguson predicted that 150 million people would die from the bird flu. Once again, we see that the numbers did not come even close. Over a 6-year period, only 282 people died from contracting it. In 2009, a government prediction based on Neil’s input predicted that 65,000 British people would die from the swine flu, and yet only 457 died.2
One has to wonder why anyone would even listen to a person who consistently gets the number wrong. Yet, Britain and the United States went into full lockdown, costing trillions of dollars—partly based on old code that never worked in the first place.
Now, when you are sitting at your desk, or looking at your tablet or phone, and it starts to slow down and maybe even freeze up a bit, you can feel a little better knowing that at least your gadget isn’t going to cost the world trillions of dollars if it doesn’t finish loading fast enough.
At the Institute for Creation Research, science is important to us. It is equally important to us that people know when bad science is being done. So many people have fallen for the information that Ferguson and his team provided, and so many people fall for science that talks of an old Earth and evolution. A lot of this is because many people fall for the idea of “consensus” science—if certain “elite” scientists say something, then it must be true.3 They will also believe scientists because of data that is cherry picked.4
If nothing else comes of COVID-19, maybe a few more people will stop trusting every scientist they hear and start questioning the “science” that has kept them away from God.
1. Fund, J. ‘Professor Lockdown’ Modeler Resigns in Disgrace. National Review. Posted on nationalreview.com May 6, 2020, accessed May 20, 2020.
2. Williams, T. Telegraph: UK Lockdown a Result of the “Most Devastating Software Mistake of all time.” Breitbart. Posted on breitbart.com May 18, 2020, accessed May 20, 2020.
3. Guliuzza, R. Consensus Science: The Rise of a Scientific Elite. Acts & Facts. 38 (5): 4.
4. Johnson, J. J. S. Cherry Picking Data Is the Pits. Acts & Facts. 42 (2): 9. "
No comments:
Post a Comment