3 ms·
While I agree with plenty the author wrote, I have seen plenty of people who are great with math, can speak the language, know how to promote their results in p
by dekhn 10y ago
While I agree with plenty the author wrote, I have seen plenty of people who are great with math, can speak the language, know how to promote their results in pubs and conferences, and yet are still completely and totally wrong.
My best example is a smart physics graduate that I went to grad school (in biophysics) with. An open problem at the time was how motor proteins couple the energy in ATP hydrolosis to directed motion. I said one day, "hmm, maybe it works like this..." and she said, "oh no, my advisor and I proved that mechanism was impossible."
A few years go by, we're getting ready to graduate. I ask her, "so, since you spent the last 7 years studying motor proteins, how do they work?" And she told me it was the mechanism I had proposed. I said, "but you disproved that!". And she said, "well, then I collected data, and it turns out our assumptions are wrong."
I constantly end up arguing with quantitative people in my own old field- for example, I used to argue with people who did GWAS, they insisted all their stats were great and perfect, then Ionnides and others showed their stats were abysmal, and they were massively overconfident in their results.
This is not to say all the cranks are right- they are almost certainly wrong. Anybody who attempts to get around the second law of thermodynamics is going to lose, unless there is something truly and fundamentally wrong with statistical mechanics.