Comment by tripletao | Hacker News Reader

Comment by tripletao | Hacker News Reader

Parent Full thread

The 1977 flu is uncontroversially accepted to have arisen from a research accident, probably either an inadequately attenuated vaccine or a challenge trial in human subjects. The death toll is typically reported as 700,000, though I don't think that's a very good number and I can't find the methodology. Many fewer people died in the 1977-78 season; but the virus has continued to spread, so the cumulative toll is much higher (and continues to increase).

https://muse.jhu.edu/article/936217

So how many people have to die before this is no longer an extraordinary claim? Bhopal managed to change chemical manufacturing standards with "only" thousands or tens of thousands of deaths; but the 1977 flu is somehow completely forgotten.

Perhaps the death toll is so high that people simply can't believe it? YouTube's fact-checkers recently removed an unambiguously factual description of the 1977 pandemic, ignoring appeals well-referenced into the peer-reviewed literature with no stated justification.

https://www.spectator.co.uk/article/how-i-fell-foul-of-youtu...

The argument is not "research accidents categorically do not happen" it was "are they the more common event and therefore the more probable explanation, absent anything else"

Depending how we count, there have been a dozen or so pandemics comparably novel or deadly to the 1977 flu in the last fifty years. Of those, at least one was research-origin.

So are you really saying p ~ 1/12 is an "extraordinary claim"? Is a 6'1" American man "extraordinarily tall"? There's unfortunately no standard map between English phrases and numerical probabilities, but I think most people would understand a much lower probability. If you really want to define it that way, then "extraordinary evidence" is likewise diluted to the point that the circumstantial case (emergence in Wuhan, DEFUSE) brings research origin easily into contention.

> Depending how we count, there have been a dozen or so pandemics comparably novel or deadly to the 1977 flu in the last fifty years. Of those, at least one was research-origin. So are you really saying p ~ 1/12 is an "extraordinary claim"?

Yes, I think it is extraordinary to extrapolate one event in 1977 to a "once every fifty years" rate

Can you quantify what you believe is a correct prior then, and explain how you got that number?

I hope you're not going to count every natural spillover since prehistory in the denominator. The technology to culture and freeze an influenza virus didn't exist before ~1930, and the technology to genetically enhance a sarbecovirus didn't exist before ~2010. The absence of pandemics with such origin before that means nothing. No one had ever suffered a cancer induced by an X-ray tube before 1904; but that doesn't mean the risk wasn't there, and Edison's assistant still died horribly.

Apply the same reasoning with the start cutoff at 2010 then

I said fifty years because that roughly covers the period during which an accident similar to the 1977 flu was possible. Perhaps I should have said longer, since influenza was first cultured in 1931; but we also need some time in the freezer for the circulating virus to diverge. I don't think much changes if we say a hundred years instead.

Can you explain why 2010 would be a reasonable start cutoff? That doesn't make any sense to me, since it excludes most of the time that a research-origin flu pandemic was possible. We obviously haven't had a research-origin novel sarbecovirus pandemic before maybe SARS-CoV-2; but when new technological developments occur, the most similar old technologies are our best model. Nobody had ever died in a plane crash before the Wright brothers, but anyone familiar with unpowered gliders could predict the risk.

Reply on news.ycombinator.com