5 ms·
To get the 1% figure you have to include an unstated assumption, which is that a novel virus is just as likely to emerge from this lab as from all top 100 citie
by NLips 5y ago
To get the 1% figure you have to include an unstated assumption, which is that a novel virus is just as likely to emerge from this lab as from all top 100 cities put together.
If we try to make a rough calculation from an unknowledgeable starting position, it might be more reasonable to say "just as likely to come from each 1M+ city as to come from each lab [of which there is one]". Then, if we didn't know the geographic origin, the lab probability is ~1% ( = 1/(100 cities + 1 lab)), but knowing the geographic start point you'd get 50%
P(this lab given Wuhan) = P(Wuhan given this lab) * P(lab) / P(Wuhan)
P(this lab given Wuhan) = 1 [this lab is in Wuhan!] * (1/101) / (2/101)
- himinlomax 5y agoFact is, there's nothing that points to Wuhan as being a more likely source of viruses in the first place. It's not a tropical rainforest area, for instance. And indeed, the closest wild relatives we know of have been found much farther south.
- NLips 5y agoI agree? In the above I'm assuming equal likelihood between each lab and each city, rather than equally likely to have come from any lab vs any city. The second is a poor starting point as it suggests having two, 100, or 1M labs would mean no increased chance of coming from a lab vs 1 lab. That's why I suggest the first.