Hacker News Reader

Comment by mkmk3 | Hacker News Reader

Naturally, there's an adversarial game playing out, and we accelerate it massively. If the results of our small games enter the larger playspace, we're damned, so I get your concern.

Is there some scifi solution to step outside of the game? Are the problems involved in understanding the systems involved (to the extent that you can predict realistic risk and mitigate) solvable?

I don't know the level of realistic risk for a devastating virus naturally occurring, nor the potential for us to address it without this kind of research, so I don't know what is within reason here. It would be cool to see any studies that attempt to address these two questions.

Edit: And a systematic review of past and considered lab leak scenarios probably wouldn't hurt either. If we still have to do this work, and just decide it must be done by isolated monks/robots in space, for humanity ending risks I guess that's worth discussing

View on HN

> I don't know the level of realistic risk for a devastating virus naturally occurring

If we assume it is directly proportional to the number of people (more people to encounter animals, more people to incubate potential pandemics), there were roughly 450 billion people-years lived between the Black Plague and the Spanish Flu and the Spanish Flu and COVID-19.

That would put the next zoonotic spillover around 2065.

This logic is as incorrect as the folks who say we are due for an earthquake.

It discounts all factors we are ignoring and pretends we can establish an accurate duration between rare events with a few occurrences

Eh, more incorrect, probably.

With earthquakes you at least have stress accumulating over time, even if the stress is not necessarily accumulating or being released at a constant rate.

A zoonotic spillover is pure repeated throws of the dice, no accumulated pressure.

Not quite true: zoonotic crossover events become more likely as more animals are brought together.

Large-scale wet markets or bushmeat markets increase the chances because there are more rolls of the dice.

Are most zoonotic crossovers more than one step or are most misadventures NONHUMAN => HUMAN? Either way if market activity doesn't change year on year then rate of dice rolls don't change on that front.

However there is no reason to believe that such misadventures take place in a single generation and immediately produce a viable and successful agent.

Producing something that spreads to humans but not will will none the less produce many orders of magnitude more useful and numerous dice rolls. Such actions may not be immediately obvious even if they were fatal. People after all get sick and die all the time.

I doubt we have enough data to meaningfully forecast risk.

I mean, we, a random assortment of software engineers on HN, almost certainly lack the data and expertise to forecast the risk of a pandemic.

My own self-critique above was more that, even if you could accurately deduce that there is a 1.54e-12 chance per person-year of a person encountering a novel disease that can spread as a global pandemic, you can't then schedule the next one for 450 billion people-years later, when the probability exceeds 50%, because it's a random occurrence and there's nothing stopping it from happening after ten or a trillion throws of the dice.

But I'll still stand by the logic of my original back of the envelope number; insomuch as the risk is proportional to the population, we should expect the next global pandemic twice as fast as the last one, and I think there's a good chance many of us will be alive to see it (as opposed to the gap between the Spanish Flu and COVID-19, where the survivors of the first were infants for the last and centenarians for this one).

> Is there some scifi solution to step outside of the game?

Of course there is, of course they already use it, and of course the people who think banning research is the peak of enlightenment aren't remotely interested in making this kind of distinction.

More context please, you're alluding and I would genuinely appreciate it if you have it. If you don't I don't think this is a good comment to make

Reply on news.ycombinator.com