News for hackers. Stuff that matters.
Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by aphextron on Monday May 03, 2021 @01:34PM (#27028081)
>People are losing faith in the scientists because of their ever changing statements.

I'd say this perfectly illustrates the complete lack of understanding of how science works among the public. There is never one specific stance that will always be right and hold up to all scrutiny. It's just highly intelligent educated experts giving their best possible estimate based on current data. As the data evolves, so do the recommendations. People not understanding this is the root of the "experts know nothing" insanity.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by paperwasp42 on Monday May 03, 2021 @01:58PM (#27028426)
I agree a general lack of understanding of the scientific method is a large part of the issue. I think an additional issue is how some politicians/political individuals treat science--"Scientists say X, therefore X is now gospel, and anyone who even questions X is morally at fault and an uneducated idiot."

Those sorts of inflammatory, controversial statements get amplified far more loudly than the actual statements of scientists, which are generally, "We think X is the best method, so we're trying this for now. But we have limited data, and as we learn more, we may have to adjust things in the future."

If the public only hears the first kind of statement (which are the type that fill social media news feeds), they're quickly going to grow skeptical of "Scientists say X" statements. Which appears to be exactly what's happening for some people.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by thu2111 on Monday May 03, 2021 @02:09PM (#27028612)
It'd be really nice if that were the case. Unfortunately, at least in many countries, it's been scientists/academics attacking anyone who questioned the reliability of scientific output.

For sure, politicians deserve their fair share of the blame, but they're by and large not the ones attacking people who say "you have limited data and it does not seem to support your conclusions". Nor are they the source of all the anti-COVID policies that now control people lives, nor are they routinely presenting the output of entirely unverified models as "science" that is beyond any rational questioning. Boris Johnson has been reported to have been furious about how he and other ministers were repeatedly bounced into policies he disagrees with by scientists who instantly run to the press and claim he's ignoring The Science/doesn't care about lives/will kill millions if he doesn't do exactly as told today without even the tiniest delay to double check things. And then days later the data presented by these experts has been torn apart by people who simply downloaded and double checked the maths.

To put it bluntly, people are losing trust in academics because they keep doing untrustworthy things. It's not a failing of the public, or a result of distortion of what innocent scientists are saying. It's an accurate perception of what's really been happening.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by paperwasp42 on Monday May 03, 2021 @08:58PM (#27032887)
You make some good points, and it raises a question I've been wondering about. My scientist friends--those who are working in labs and doing actual research--have all been really hesitant to make any hard claims about COVID since the very beginning. Pretty much every conversation I have with them is, "We don't know, we think X, but it's way to early to be certain."

On the flip side, as you've pointed out, some of the scientists advising policy seem to have... not that attitude. They present theories as hard facts, and then confuse the public when those "hard facts" turn out to be only partially true or even false.

My question is why there's an apparent disconnect between the on-the-ground scientists in their labs, and the policy-advising scientists. Is it because the policy-advising scientists worry they'll lose their jobs if they don't sound certain? Is it because they worry the public won't listen if they don't sound like they 100% know what they're doing? Is it purely political motives? A mix of all three?

I have no idea, and I don't know nearly enough about the biology field to even begin to guess. But I wish someone who's more well-versed on the topic would write an article about this.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by thu2111 on Tuesday May 04, 2021 @03:13AM (#27034796)
Mmm, it's a good question. Completely agree we need far more journalism on these topics. Unfortunately the media can't do it because science journalists are all dependent on academia for stories and find it tough to bite the hand that feeds them.

The biggest problems seem to be with scientists who don't actually work in labs, people that in industry we'd call data scientists. There's nothing to ground them in the real world, no clear research directions or high level goals, and it's too easy to just manipulate the data or ignore data that disproves prior papers. It's possible your friends are better because they work in labs.

But there's also another possibility, and it's not a pleasant one to ask about, but have you actually downloaded and read papers published by your friends? One thing that becomes apparent when you dig into this is that researchers are very willing to give their friends/colleagues a heavily hedged conclusion, whilst telling the public that their conclusion is not merely unarguable fact, but also an absolute scientific consensus that demands immediate government action. This is part of the reason why academics are losing trust; open access has happened, the internet has happened, and now it's easy to download their papers to see the contradictions for yourself. An example would be Flaxman et al 2020 which attempted to prove that lockdowns worked. It set up a model that could only conclude that lockdowns worked (because it had no component to represent any sort of natural decline and for other reasons), and admitted half way through that this made the results unrealistic. But by the time those same researchers were publishing press releases on the university website and giving interviews, they were claiming there was now solid proof that lockdowns had saved millions of lives. Same people, totally different stories and in ways that were not innocent mistakes. Journalists deserve at least some blame for not treating their sources adversarially but the deception was originally coming from the researchers, not the press. A related problem seen in the UK is the way that predictions assigned near 100% certainty are suddenly rewritten as having been mere "scenarios" the moment someone corners a scientist and shoves a graph of real measurements in their face.

And then there are the other issues.

One is surely filtering related. When you look at the composition of the SAGE advisory council in the UK, you discover that one influential member is a lifetime member of the British Communist Party! Another is in an open relationship with a woman who's a part of Extinction Rebellion. Their leader was a man who apparently shocked his friends when he joined GlaxoSmithKline because of his lifelong open hatred for corporations and pharma firms of any kind. These people aren't at all normal or representative of the general population's views, so how does that happen? Some of it is that academia selects for people who want to be judged by impact rather than e.g. their salary size. And some of it must be a question of who chases after advisory roles to governments. A double selection effect, in other words.

Another is the certainty aspect that you allude to. The most influential "scientists" are actually modellers, not scientists. They insist that they are scientists, but the scientific method is MIA in fields like epidemiology. If you read their papers you'll find lots of predictions and no followups to measure the accuracy of the predictions. It's just one long run of model outputs, presented alongside the inevitable policy recommendations, but when you go digging through a reverse citation search to look for followups where the model accuracy is calculated, you come up with nothing. Why are the most influential scientists these people? Well, exactly because they use their models to generate policies. They tell politicians, "our scientific models say X will happen if you do Y and if you don't there will be an enormous crisis that everyone will blame you for". The politicians meanwhile feel they can't push back, even if they intuit the truth that the claim is scientifically unsupportable, because academics are very good at making jargon filled and very technical arguments involving lots of maths. Some papers have obvious errors, but when they're pointed out academics just blast people with even more pseudo-science along with many claims of criticising in bad faith, not being legitimate, etc. You need to have copious free time and an extremely high degree of confidence with logic, stats and programming to correctly see through the BS they are capable of generating. The media can't do it and immediately fall in line, politicians then then feel they have to follow. The only people who do push back end up being a motley assortment of bloggers, often retired or out-of-field, who have the free time and career immunity to wade through the morass pointing out the errors. Journals rarely publish them of course, because they also can't bite the hand that feeds them, and science journalists don't know how to evaluate arguments except via reference to institutional hierarchies so just ignore them. But most people don't have those limits, so more and more people find the articles written by those newly-born "skeptics", read them, realise the arguments make sense and their faith in science declines. The institutions wail about disinformation without realising that they're the ones generating it.

The whole system is locked in this feedback loop and because academia is so heavily government funded there's no obvious way to break it. There are no penalties for doing bad science, as even many scientists will admit, but plenty of rewards. So public confidence in research will continue to fall until it reaches the same level as journalists or politicians, perhaps even less given that those people are at least in direct competition with each other whereas academics aren't (or rather, the competition they're in does not select for anything the public cares about). Eventually you will get a new generation of politicians who openly attack scientists in the same way that Trump openly attacked the press. There will be calls for defunding which will be responded to with more bans, more censorship, more street protests and more chaos. What comes after? Only the future knows.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by cltby on Monday May 03, 2021 @02:06PM (#27028562)
The data didn't evolve, though. The only thing that ever changed was the political winds. Throughout this whole pandemic, experts have bent over backwards to defend and justify political decisions--and most humiliatingly, decisions that _directly_ contravene their prior advice. Two examples:

- Experts encouraged the George Floyd protests because "racism is the real virus," despite their prior anti-protest stance. Given the apparent magnitude and seriousness of the problem, it's curious they only noticed in May 2020.

- In Washington, there was a big battle around opening schools. Public health authorities insisted that this was a verrryyy fraught process that needed to be handled veeerrrryyy slowly and veerryyy carefully. In the middle of this, Gov. Jay Inslee suddenly demanded that the schools be opened in a matter of weeks. The authorities weren't perturbed in the least: after all, they suddenly claimed, "we've known all along that the risk to children is very manageable".

Most people understand that mistakes happen, and they are generally forgiving when those mistakes are acknowledged and corrected. But it's plain to see that the sort of "expertise" that's being held up is deeply and pathologically broken.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by Sohcahtoa82 on Monday May 03, 2021 @02:54PM (#27029195)
Politicians don't make decisions based on the opinions of experts, they make them based on what they think is going to win them the next election.

Yeah, public health authorities might be saying we need to be careful with reopening schools, but if 70% (This is an example number pulled out of my ass, not an actual poll number) of the population wants schools to reopen because parents are tired of their kids being home all day, then dammit the governor is going to reopen the schools, expert testimony be damned.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by cltby on Monday May 03, 2021 @03:46PM (#27029810)
The governor isn't the bad guy here. It's the supposed experts whose "neutral" opinions are politically contingent.

Parents: open the schools.

Experts: no, it's unsafe.

Progressive politician: open the schools.

Experts: great idea sir, just as we've been saying all along!

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by Sohcahtoa82 on Monday May 03, 2021 @04:41PM (#27030361)
I don't think I've seen any experts saying to open the schools.

I'm sure you could cherry-pick a few doctors saying we should open them (And Fox News will absolutely give them a megaphone), but you can find doctors that think MMR vaccines cause autism. But the overall consensus AFAIK has always been to keep the schools closed.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by 2lwxxtj on Monday May 03, 2021 @02:27PM (#27028857)
>It's just highly intelligent educated experts giving their best possible estimate based on current data.

OK, but can we trust them to completely put out of their mind the interests of the people signing their paychecks? Do the people signing the paychecks blindly hire scientists without considering whether the scientists will be mindful of their interests?

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by _-david-_ on Monday May 03, 2021 @02:10PM (#27028618)
The issue with changing their statements is not so much that they changed their statements, but how certain they were with their original statements. Some of the scientists did not start their statements with stuff like "with the available data we believe". They made it sound like what they were saying was 100% determined. Anybody who disagreed (even if it is now the prevailing view) was called anti-science.

It is not strictly just the scientists making statements based on the available data, but them explicitly lying.

Some of the scientists said masks were not needed for the general population. This is understandable since they wanted people who were at the greatest threat to have enough masks. The problem is when some of the earliest statements they made were lies people lose trust.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by Sohcahtoa82 on Monday May 03, 2021 @02:50PM (#27029132)
I wince thinking back to that interview with Fauci early in the pandemic when he explicitly told people that we didn't need to wear masks. That statement became a massive weapon in the hands of the anti-maskers.

The problem is, I'm not sure what Fauci could have said that would have been better. People are irrational, and he knew that if he recommended masks right away, then people would have begin hoarding them ASAP, and healthcare and other front-line workers would have been facing an even greater shortage.

You all remember the massive run on toilet paper and sanitation supplies last year when lockdowns began? The run on masks would have been just as bad if Fauci recommended them sooner rather than later.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by _-david-_ on Monday May 03, 2021 @03:14PM (#27029446)
I understand why the scientists lied (about masks), but it causes people to not trust them. You lie one to many times and people will stop believing anything you have to say. Fauci and other scientists should have re-read The Boy Who Cried Wolf before lying multiple times.

Also, just to be clear it wasn't strictly Fauci. The CDC and the Surgeon General made similar remarks about masks as well. If it was just Fauci I think that would have been less damaging, but to have multiple government agencies / officals all lying just looks bad.

Re:Reaching ‘Herd Immunity’ from Covid Is Unlikely in the U.... by thu2111 on Monday May 03, 2021 @01:59PM (#27028436)
I'd suggest reading some scientific papers before going too far in that direction. I've read a lot of them. These scientists routinely do things like cite papers that contradict the claim the citation is meant to support. They write papers based on circular logic. They conflate simulations with reality, they make claims that have no basis in biological science. They do not know what the false positive rate on COVID testing is, but are happy to draw conclusions that flatly contradict decades of biological understanding on the basis of data analysis that critically depends on knowing the correct value. They write models in C that are filled with memory corruption bugs and race conditions. When the results randomly change from run to run they claim that it's an experiment and so averaging the results makes it all perfectly fine. And when professional programmers tell them that incorporating memory corruption into their calculations is not fine, they assert those programmers don't understand science (see: https://github.com/mrc-ide/covid-sim/issues/165)

And that's just the scientists. Public health officials like Fauci have lied to the public and then admitted doing so, more than once! That isn't the data evolving, that's them deciding to change their story and explain that the previous one was a lie. What are we to make of such people?

You believe "experts know nothing" is insanity. I've read what those people are writing and am absolutely sure they know nothing, because if they did, they wouldn't keep engaging in these illegitimate, unscientific and desperately concerning practices.