[chihuahua]

I find this article quite fascinating. It provides a bit of a glimpse into how science can work, and how difficult it is to gain valid knowledge on some topics. It's also a reminder that "science" isn't just a simple process that you follow to find the truth. Sometimes people ask the wrong question, or draw the wrong conclusion from experiments. It seems that research related to nutrition and health has a fair amount of this.

It reminds me of my own experience earlier this week where I needed to call a certain service. The API has 2 parameters. I thought that one was required and one was optional. But whenever I called the service with 1 parameter, it failed. Eventually I thought "maybe both parameters are required" and called the service with 2 parameters, and it worked. So I concluded "both parameters are required". But later I discovered that the service was just flaky and it was a coincidence that it started working just at the moment when I added the second parameter. After that experience it's difficult to know what to trust, when everything could be due to randomness and unreliability.