Hacker News new | ask | show | jobs
by Bancakes 1648 days ago
To quote baseball umpires, "It ain't bothin' till I call it." If something works, I want to see why. If something doesn't work, I want to see why.

You want people to believe in global warming, coronavirus, and healthy foods, make global warming, coronavirus, and healthy food test kits. I hate to burst your bubble, but when the rubber hits the road, science journals are just text on paper, disconnected from reality. Gravity, radiation, magnetism, etc were discovered through experimentation and demonstrations. We live in a dark age of irreproducibility.

Take health industry and cybersecurity industry for example. The former publishes controversial results every weekend, and the latter has us trust closed source systems due to appeal to authority.

Anecdotal data isn't much but it's your data, and that sure is better than wrong data.
1 comments
JoeAltmaier 1648 days ago
It's pretty much identical to wrong data. One data point is nothing. It's a comprehensive mistake to give it any weight.

At least for complex subjects like health, weather etc. "I saw a snowflake! Global warming has to be wrong!" is not any kind of conclusion.

link
Bancakes 1648 days ago
What's the alternative? "5% of crops died in a country halfway around the world, that's used to dead crops. You should buy a Tesla now." "Humans have been eating saturated fats for 2 million years but half rancid PUFA are healthier."

No, I can look at and trace the proofs in any mathematical publication. I can look at and trace any open source code. If something works, I should be able to test it. And it's your flaw to assume people will play along with the onus.

link