Hacker News Reader

Comment by ubercow13 | Hacker News Reader

Modelling disease spread seems like the kind of scientific endeavour that one can dedicate one's whole life to becoming an expert in, and many people presumably do. Considering that, why should anyone give much credence to a random medium article on the topic by a 'silicon valley tech executive'? (lol)

There are 10's of thousands of scientific research papers [1], by scientists, on COVID-19 already, and this guy is just some tech exec who knows how to play with a jupyter notebook. I see no reason to pay much attention to him.

[1] https://twitter.com/freereadorg/status/1236104420217286658

View on HN

Because the models the government's epidemiologists are using explicitly don't account for all possible actions.

Back when this was taking off in China, a bunch of epidemiologists came out and said it was completely uncontrollable there. They were wrong, because they hadn't accounted for the extreme actions the Chinese government was willing to take; it was not even a possibility in the models. The same story repeated itself for Hong Kong, Taiwan, Singapore, and South Korea. They defeated the models by taking effective actions that were "out of the training data" for epidemiology. Given these 5 examples, there's no reason other countries can't defeat the models too.

The US we are trying as hard as possible to track the exponential pandemic curve as closely as we can.

Though the US government certainly isn’t communicating those details to the general public. Have you seen the CDC website? Why do I need to go to The NY Times website to get a summary of updates in a timely manner?

You don’t think professional epidemiologists have not thought of these? There are many papers in the literature that discuss this.

Clearly they have. I'm just saying that the word of each individual epidemiologist is not absolute law. They can disagree, because they base models on different assumptions. As a result, many, many predictions shouted far and wide by epidemiologists in the past months have turned out wrong.

We don't know yet how sustainable Chinese actions are. People are still in lockdown. Covid19 could come back when the lockdown is relaxed.

You are missing the fact that it doesnt matter if it "comes back" or not, what matters is that the rate at which it "returns" doesnt overflood the healthcare systems.

This idea, UK policy makes hae will absolutely collapse the NHS.... But who knows, maybe that's their idea all along, to collapse the NHS, and after this is done, they will start passing laws to "reform it" and sell it by the kilo to US private healthcare firms....

These may be legitimate criticisms of the UK government's epidemiologists - I'm not in a position to judge - but I'm not sure it explains why we should be listening to Tomas Pueyo on the matter.

Well, you don't have to. There are plenty of actual epidemiologists unhappy with the government's actions. Just because Pueyo was arbitrarily chosen to represent one side doesn't mean professional support for that side doesn't exist.

Honestly, choosing a bad representative for a position is a classic media tactic for unfairly discrediting that position.

Agreed, it would have been nice for them to invite an actual epidemiologist who disagrees with the UK strategy - there are many of them, some have already publicly stated this - unfortunately this is par for the course in the UK media.

Hey, at least it isn't bringing Nigel Farage on to talk about Coronavirus. (As the BBC did last week.)

Reply on news.ycombinator.com