Royal Society cautions against censorship of scientific misinformation online • Hacker News Board

Re: Royal Society cautions against censorship of scientific misinformation online

#851

Post by mike_hearn » Mon, Jan 31, 2022, 4:08 PM UTC

mrtedbear wrote: post #841

mike_hearn wrote: post #671

Earlier quoted context omitted.

"There are ways to mitigate this, but since the model was collecting outcomes to create a statistical distribution, rather than a single deterministic number, it didn't need to." This is the justification the academics used - because we're modelling probability distributions, bugs don't matter. Sorry, but no, this is 100% wrong. Doing statistics is not a get-out-of-jail-free card for arbitrary levels of bugginess. Fi…

At the time, critics of the model were claiming the model was buggy because multiple runs would produce different results. My comment above explains why that is not evidence for the model being buggy. Report 9 talks about parameters being modeled as probability distributions, i.e. its a stochastic model. I doubt they would draw conclusions from a single run, as the code is drawing a single sample from a probability d…

Your comment above is wrong. Sorry, let me try to explain again. Let's put the whole fact that random bugs != stochastic modelling to one side. I don't quite understand why this is so hard to understand but, let's shelve it for a moment.

ICL likes to claim their model is stochastic. Unfortunately that's just one of many things they said that turned out to be untrue.

The Ferguson model isn't stochastic. They claim it is because they don't understand modelling or programming. It's actually an ordinary agent-like simulation of the type you'd find in any city builder video game, and thus each time you run it you get exactly one set of outputs, not a probability distribution. They think it's "stochastic" because you can specify different PRNG seeds on the command line.

If they ran it many times with different PRNG seeds, then this would at least quantify the effect of randomness on their simulation. But, they never did. How do we know this? Several pieces of evidence:

1. There are no confidence intervals in Report 9, only different scenarios. See for yourself: https://www.imperial.ac.uk/media/imperial-college/medicine/m...

2. The program is so slow that it takes a day to do even a single run of the scenarios in Report 9. To determine CIs for something like this you'd want hundreds of runs at least. You could try and do them all in parallel on a large compute cluster, however, ICL never did that. As far as I understand their original program only ran on a single Windows box they had in their lab - it wasn't really portable and indeed its results change even in single-threaded mode between machines, due to compiler optimizations changing the output depending on whether AVX is available.

3. The "code check" document that falsely claims the model is replicable, states explicitly that "These results are the average of NR=10 runs, rather than just one simulation as used in Report 9."

So, their own collaborators confirmed that they never ran it more than once, and each run produces exactly one line on a graph. Therefore even if you accept the entirely ridiculous argument that it's OK to produce corrupted output if you take the average of multiple runs (it isn't!), they didn't do it anyway.

Finally, as one of the people who forensically examined Ferguson's work, I never accepted Guptra's either (not that this is in any way relevant). She did at least present CIs but they were so wide they boiled down to "we don't know", which seems to be a common failure mode in epidemiology - CIs are presented without being interpreted, such that you can get values like "42% (95% CI 6%-87%)" appearing in papers.

Re: Royal Society cautions against censorship of scientific misinformation online

#852

Post by mike_hearn » Mon, Jan 31, 2022, 4:15 PM UTC

immibis wrote: post #786

mike_hearn wrote: post #730

Earlier quoted context omitted.

This is the sort of mis-use of logic that has led to so many problems. "Lockdowns" is shorthand for "people not coming into contact with other people in a way that spreads the virus" It's shorthand for a set of government policies that were intended to reduce contact, not eliminate it, because "not coming into contact with other people" is impossible. People still have to go to shops, hospitals, care homes, live with…

you're exposed to the virus normally 10 times per week every week. With lockdowns that drops to 5 times a week. In this scenario, lockdowns work fine, but are insufficient! Perhaps with stringent use of N95 masks during contactless food ration delivery it can be reduced to 0.05 times per week. Then the pandemic ends and we go back to normal. Also, if you only go out half as often, that's not exactly a lockdown either…

You don't appear to realize that food gets to your front door via a large and complex supply chain that involves many people doing things physically together at almost every point. Your definition of a lockdown is physically impossible even for cave men to sustain, let alone an advanced civilization. It isn't merely a matter of "works but insufficient".

This kind of completely irrational reasoning is exactly why lockdowns are now discredited. The fact that the response to "here's lots of data showing that lockdowns didn't work" is to demand an impossible level of lockdown that would kill far more people than COVID ever could simply through supply chain collapse alone, really does say it all.

Re: Royal Society cautions against censorship of scientific misinformation online

#853

Post by rapind » Mon, Jan 31, 2022, 4:23 PM UTC

lettergram wrote: post #113

AussieWog93 wrote: post #87

Just throwing my 2c into the well, as someone who used to be highly "pro-science" but lost confidence in much of academia and the validity of scientific research in general after starting a PhD and seeing how the sausage is made. The biggest problem science is facing is not an external threat from a rabble of ignorant science deniers, but the complete degradation of quality within the scientific institutions themselv…

> Most research is flawed or useless, but published anyway because it's expedient for the authors to do so. For the record I now know a professor at one of the premier institutions in the world, who is a total fraud. Their research was fraudulent in grad school. Their lab mates tried to raise concerns and nothing happened. That person graduated, with everyone on the committee knowing the issues. Then they got a premi…

I agree that there is an integrity problem in some areas of research (due to incentives), however the Eric Weinstein reference is laughable. I'd consider him a prime example of pseudo science for dollars / influence with zero credibility. Not saying he can't be on the right side of an issue some times, just that he's a particularly untrustworthy source, and the way he gets to his conclusions is just... wow.

https://youtu.be/0TI0jtr6APw

Re: Royal Society cautions against censorship of scientific misinformation online

#854

Post by raxxorrax » Mon, Jan 31, 2022, 4:24 PM UTC

scottLobster wrote: post #828

s1artibartfast wrote: post #161

I see the root cause of all this debate as a failure in scientific communication. When it comes to complex scientific decisions with significant impacts on the public, nuanced and detailed justifications are required. Instead, we often get simple declarations written by PR departments at juvenile reading level. This spawns chaotic and poorly conducted debate in all walks of life. What is required is a robust and tran…

Really? I'm speaking from a US perspective, but my opinion is the communication out of the CDC has too much nuance and is too concerned about being scientifically accurate. They forget they're speaking to a largely scientifically illiterate crowd, and that crowd sees the constant back-and-forth not as scientific progress but as a reason not to trust the institution. Never mind that it's often delivered with a tone of…

I believe this would have been an even worse strategy for policy acceptance. We were and still are dealing with unknowns and these unknowns are able to justify wearing mask as a precautionary measure without a (not so) noble lie being necessary. Society isn't a military unit and has profoundly different dynamics. Some will ignore this advice but the repercussion of that are very likely managable.

This type of PR will not work with a modern information infrastructure anymore or at least significantly less effective.

> handed ammo to the anti-vaxxers by the truckload

It is a stupid image of an enemy and policy was crafted with a steady look on these groups. That is a mistake in leadership if there is one. You have provided them with legitimacy without anything for you to gain. A complete waste for nothing and you elevated your opposition without benefit to you. That is neither competent leadership nor politics. Figuratively not looking them in the eye would be better. They are beneath you and don't even warrant attention. Few politicians would have even used the term, at least the smart ones.

That said, since I oppose mandatory vaccination I am an anti-vaxxer myself in the eyes of many.

Re: Royal Society cautions against censorship of scientific misinformation online

#855

Post by b112 » Mon, Jan 31, 2022, 4:26 PM UTC

nautilius wrote: post #805

b112 wrote: post #323

Earlier quoted context omitted.

Or, such narratives are created, just as narratives that all politicians are corrupt, that meritocracy is wrong, as an attempt to weaken western democracies.

Not arguing one way or another, but what is the evidence?

Most nations spend a notable portion of their defense budget, on disrupting opponents via whisper campaigns, and dividing an opponent from within.

This has been going on through all of recorded history.

Disrupting belief in any important part of your opponent's society, is key. The goal is to cause societal disfunction, an inability to work cooperatively.

Why fight a war, if your opponent destructs from within? Or, almost as good, loses some capacity to make war?

Also good, is if your actions cause your opponent to change regimes, and become aligned with your goals.

Re: Royal Society cautions against censorship of scientific misinformation online

#856

Post by jimkleiber » Mon, Jan 31, 2022, 4:29 PM UTC

toolz wrote: post #507

jimkleiber wrote: post #315

Earlier quoted context omitted.

Does it render masks "ineffective" or "less effective"? I think I see so much absolutist language whereas I wish we would use more relative language.

What is this "it" you're referencing specifically? I'm not sure which point you're asking about, but any of the points aren't like to have great data. Not many masking studies go outside of modeling and the lab and the ones that do aren't that detailed (that I can recall).

Maybe I just read too deeply into your statement:

> People who don't want masks aren't getting fitted properly for n95's, they aren't shaving, they aren't behaving consciously in the many ways you can behave that render masking ineffective.

Maybe you weren't saying not shaving renders an n95 ineffective.

Re: Royal Society cautions against censorship of scientific misinformation online

#857

Post by spion » Mon, Jan 31, 2022, 4:30 PM UTC

jazzyk wrote: post #846

alignItems wrote: post #814

Earlier quoted context omitted.

The problem is that we sometimes have to rely on other people. We can’t all be experts in everything. We can’t all reproduce or witness every experiment. What we can do is delegate the process. There’s a difference between saying “whatever X says is true, because X said it”, and “I trust what Y says because Y has explained the process and methods of how it was derived, and I don’t have reasonable doubt to assume Y li…

What to do when you have two very different opinions coming from experts: - Anthony "I am The Science" Fauci or - Dr Malone (mRNA inventor)? Who to believe? Difficult to say, but these days, sadly, the first (and prudent) thing to do is to follow the money . While credentials matter, the first thing I check when I read a paper/publication is who funded the work.

Doesn't really work. Social media grifting can be really profitable - it brings in audience, clicks, ad money. Even just audience is enough as long as you can come up with something to sell later.

Even if the original paper was an unfunded honest mistake, grifters can find a way to profit from spreading it especially if its novel and sensational.

Re: Royal Society cautions against censorship of scientific misinformation online

#858

Post by gilbetron » Mon, Jan 31, 2022, 4:32 PM UTC

colechristensen wrote: post #33

toolz wrote: post #12

So many "misinformation censorship" proponents seem to have quickly forgotten just how much literally everyone in public health has spread misinformation this pandemic. It's unavoidable being wrong and often the cost of making claims without unimpeachable evidence is higher than the cost of being wrong. Name me a public health official who regularly made recommendations this pandemic and I'll show you misinformation.…

There is a difference between "malice" and "intentional misinformation". Authority figures are not accurately representing risks and rewards of different actions and have deliberately shaped their messaging to achieve outcomes. Direct from the horse's mouth: https://twitter.com/cdcgov/status/1233134710638825473 "CDC does not currently recommend the use of facemasks to help prevent novel #coronavirus. Take everyday pr…

While protecting the supply of masks may have been part of the point, the CDC also believed (along with much of the medical community) that masks did little against viruses because of old research. It took quite a bit of scientific discussion and bickering for them to revise that opinion.

I largely agree with the rest of your points - particularly the outcome we have found ourselves in. Barring the emergence of some horrific new variant (Imagine Omicron with 10x the fatality rate), this is just another bug we can no longer avoid without effectively destroying our existing society - time to move on and live with it.

Re: Royal Society cautions against censorship of scientific misinformation online

#859

Post by jimkleiber » Mon, Jan 31, 2022, 4:33 PM UTC

toolz wrote: post #766

jimkleiber wrote: post #390

Earlier quoted context omitted.

I appreciate you sending the source. I guess I still have the same argument. A LOT less effective does not equal ineffective. Or maybe I just see certain words like "ineffective" as binary words, either it works or it doesn't, and have a reaction to that. It may be a lot less effective to have facial hair while wearing an n95 compared to no facial hair and a properly fit n95, but is that still much much more effectiv…

ineffective is not binary - it depends on your endpoint, which is subjective. If your endpoint is to not catch covid and someone ineffectively wears their mask such that they would not have caught covid if they wore it correctly - then the mask was ineffective even if statistically they _maybe_ had lower chances of catching covid than without wearing a mask at all. and I say maybe very strongly here...we do not under…

Ah, yeah, I think it was just a misalignment on the word ineffective. I think when I hear many people use it, I assume they mean pointless, ineffective for all, whereas in you saying it, it sounds like you meant ineffective in that particular instance based on what the outcome was. I see you seem to have a lot more nuance in your perspective than I may had been assuming. I'm sorry. Thank you for clarifying.

Re: Royal Society cautions against censorship of scientific misinformation online

#860

Post by Chris2048 » Mon, Jan 31, 2022, 4:33 PM UTC

ithkuil wrote: post #804

Chris2048 wrote: post #684

Earlier quoted context omitted.

> otherwise they'll be criticized for being weak politicians, sure, but public officials such as the EPA?

Ok. Replace weak with "shifty" and "unclear". In any case a public communication that doesn't reassure will quickly be replaced by one that does, by escalating the issue to the nearest political figure who takes on the job of providing such assurance.

I would think lying would make someone seem more shifty than making the facts known.

> that doesn't reassure will quickly be replaced by one that does

sure, just doesn't need to be the EPA that does it. pre-empting politicians and doing their manipulation for them is not justified - politicians are not a source of truth.