Hacker News new | ask | show | jobs
by arisAlexis 1338 days ago
It's very clear that you really don't want to believe in the probability. You know nothing is 100%. If you assign 100% for no leak then your bias level is 100%. If you assign some probability then we just disagree on the probability level. Btw, people from the lab live there and shop there.
1 comments
mikeyouse 1338 days ago
Of course there's the possibility - but the probability surely changes if the "market right next to the research lab" actually turns out to be a 20-minute drive away, no?

For people in the Bay Area, this is roughly equivalent distance-wise to a conspiracy theory about an outbreak at Twitter's HQ started by an "adjacent" lab in the Upper Mission that actually turned out to be in Jack London Square in Oakland instead.

link
arisAlexis 1338 days ago
Again, bayesian priors. If there are only 4 labs worldwide messing with exactly these viruses, the combinatorics of this happening 20km from one of those 4 is infitesimal.
link
mikeyouse 1338 days ago
I'm bored of this debate that happens every time one of these charlatans releases a paper that grabs a bunch of headlines and is disproven two days later (as already has happened with the original in this thread).

But "This happening" needs to be defined for Bayes to be of any use. State clearly what you think happened and it can be judged, because as has been proven over and over again, the virus is of natural origin. So the proximity to a lab studying Coronaviruses isn't nearly as interesting given China's history with these viruses.

This is long but in my opinion, worth reading regarding the Bayesian probabilities that actually matter here: https://protagonistfuture.substack.com/p/natures-neglected-g...

link
arisAlexis 1338 days ago
See how your wording is super biased? "Charlatans"
link
mikeyouse 1337 days ago
What would you call non-virologists writing virology papers, releasing them as non-reviewed preprints with a huge PR push and then weeks later quietly admitting that their science was bad and none of their conclusions held up?

What if they’d done that multiple times?

“Bias” isn’t relevant when it comes to calling out bad actors.

link