The result of course, was low quality articles and hastily copied press releases with no critical thinking. They believed everything corporations said, because they had no time to check.

I don't blame the journalists, but the work conditions and economical situation of the media industry.

I say this as someone on that side of the political spectrum.

That's probably where the journalists find people to interview, likely because a large majority of journalists are active on Twitter.

This definitely isn't the only place, because there are free services like SciLine by the AAAS (source: https://www.sciline.org) and many others for connecting journalists to researchers. Many universities also have their own expert directories set up, and journalists can also find papers and contact their authors.

However, for many journalists on deadline, it's just far less effort to message vocal professors on Twitter, so this may be a reason for this effect if it's true. The people who market their research more on Twitter may be more likely to get covered in the press, and thus interviewed.

I'm not buying it. Science reporting from university press departments is just as bad if not worse, and they are funded by the university not advertising (tuition, grants, alumni donations, etc..)

From an enjoyment side, writing is hard, but for a lot of people, it's much easier than research. There are fewer credentials needed for a career in science journalism versus scientific research (a Bachelor's only versus a PhD for many senior scientist positions). I know a lot of people who have a natural skill at writing who decided to major in a science field for one reason or another, performed okay-ish at their courses, and tried to get back to writing work.

The end result is also similar writing for a magazine/newspaper versus a university. A university may have higher standards for accuracy and precision (especially if an interviewed scientist wants to review it, while a newspaper/magazine may have a policy to avoid sharing drafts to avoid bias in the article). However, higher-end magazines (like the New Yorker) have more prestige. There is also far less room at universities for dissent (e.g. presenting an opposing scientific view or publishing investigative work).

So what you get is worthless fanboy journalism in which those who are smart enough to ask tough questions get removed, and those who are left just want to copy paste university press releases.

In turn that beds general scepticism amongst the population because they can sense that nobody is challenging the "experts".

An example, "My uncle became blind because of diabetes", similar answer: "I don't buy it, my father is diabetic and never got blind". Doesn't work.

In other words, the reason why scientific journalism in non-university press could be bad for various different reason than university press.

> It is ALL about click bait to drive digital advertising

(Emphasis my own.) Science reporting is still crap even when it isn't funded by advertising, so there's obviously more going on than the funding model making things shit. The BBC and CBC both have crappy science reporting too, as do university press departments. These counter-examples refute the claim which was stated too strongly (I do think advertising plays some role in it, but obviously not all.)

It's not about people not wanting to pay, it's that the value of a single article is too small to charge money for. A single article is maybe worth a hundredth of a cent to me. There's no viable mechanism for me to pay a hundredth of a cent to a writer, just transaction fees will be more than the amount transacted.

I'm not going to pay a subscription to a site I might not ever visit again. I'm also not going to pay a subscription fee to every site I come across that might have a decently written article once in a while. Advertising ends up being the only viable way to "charge" a tiny fraction of a cent to readers.

What if the site was consistenly putting decent articles? I think one of the argument is that you would see more of those decent articles if people were more enclined to follow a subscription model where the journal is accountable to its customers.

I think this argument holds up. For example, publications like the Scientific American and The Scientist tend to be very high quality in terms of accuracy (usually with unobtrusive citations, if I recall correctly). It's too bad their subscriptions prices are so high; they were hundreds of dollars a year if I remember correctly, versus less than 50 dollars a year for other magazines.

Ii'm not sure about how financially successful they are, but I found myself struggling to justify such a subscription price, when the purpose was essentially edu-tainment. The publications that can more easily get people to pay more, in my view, ostensibly help people make more money or find career success, which is why industry publications like STAT for biotech/medicine (and more generally, publications like the WSJ/Bloomberg) tend to be more financially successful.

This is an interesting hypothesis, because PBS is near the top of my quality ranking for news orgs while NPR is near the bottom (obviously excluding the long-tail of sources I don't often encounter, like breitbart). I don't see much of a correlation with manner of funding.

Despite? Or because of it? Left leaning politics need trust worthy news to advance, right leaning politics don't as their goal is to maintain tradition and the status quo. If nobody mobilises the right wing polices win by default, noise instead of news helps to create apathy, apathy is the bread of the status quo.

?>>>? Right leaning politics need trust worthy news to survive, left leaning politics don't as their goal is to upset tradition and revolt on the status quo. If nobody maintains the status quo the left wing polices win by default, noise instead of news helps to create chaos, choas is the bread of revolution and malcontent.

As you can see, this sort of "I define my side to be the more accurate view of reality" is highly malleable and can be used to support any side with semi-convincing "just so" stories.

And you're not actually describing the left-wing/right-wing divide here (itself a highly simplistic and outdated view of politics and economics dating from more than a century), you're describing the conservative/progressive divide, which tends to correlate with the left/right divide in the US, which might be where you are from. But even there the correlation is not perfect, Trump for example was fairly 'progressive' in his populist politics, his extensive engagement with the public on Twitter was a very novel form of presidential press releases that upseted many traditions and angered many political conservatives. See how conservative/progressive is multi-dimensional ?, you're just too used to a few dominant dimensions being coorelated with a left\right slant.

Furthermore, even accepting the dubious renaming of conservative\progressive to right-wing\left-wing, it's still a leap of faith from here to assuming that the progressive side always benefits from representing truth and reality as it is. In fact, this is almost gauranteed to be false, because even if conservatives are randomly guessing, they would still get roughly 50% of the world right, and since progressives always oppose conservatives in this hyper-polarized political climate, they are gauranteed to be wrong that 50% of the time. This seems to be supported by reality as being a progressive is associated with a higher chance of believing there are 700 genders and that biological males are women, so there must be at least some situations where progressives ("Left-Wing" in your parlance) benefit from a mis-representation of reality.

Can you elaborate on this? Society has inertia, history has inertia, how is this "nobody mantains status quo" work?

It refers to when a media outlet thinks that everything the Trump admin does is an unprecedented terrible action, even when that thing is humanitarian aid or peace treaties in the middle east.

This is probably why exaggerated claims from press releases usually just get passed on and, frequently, exaggerated. Even someone educated in the field doesn't have time to dig into 3 or 4 new sub-fields a day and actually understand the papers that press releases are talking about. e.g. A physicist with a background in optics is better situated than most to understand an experimental quantum cryptography paper, but they're still going to have to bang their heads against several walls to figure out what's going on. Head-banging takes time. Copying from a press release doesn't.

writing and publishing four articles a day with only cursory review/editorial oversight, if any, isn't journalism at all, it's a content mill

I remember a story from a famous freelance investigative journalist who found evidence of disenfranchisement of black voters in the 2000 election. He collected a huge amount of evidence and testimony, and sent it to news outlets across the US. Most of them came back with:

"Uh, we contacted the <person/group accused>, and they're denying this. We're not going to run the story."

Who'd have thought people accused of nefarious, potentially illegal activities would deny doing them?

Why wouldn’t you blame people who voluntarily took a job writing science articles while having zero credentials? They could be normal news journos, nobody forced them to report on science specifically.

I'm sure a view against credentials isn't shared by all journalists (for example, many publications do require a relevant background), but that experience was personally disheartening because it seemed to be supported in the room.

I was also disheartened by the focus on self-marketing and self-promotion in general over doing quality, accurate journalism, and a much bigger focus on having a good "narrative" and writing engagingly versus having a discussion on ethics. To some extent, I get it, because the advice likely works in helping someone succeed in their career. But there are clearly problems in science news related to precision and accuracy, and not much drive in many publications to fix it.

Why?

Because they need to justify their budget. If nobody reads/views your reports, why should the government pay for it.

The real answer is: the public doesn’t want science reporting unless it’s punchy, simple and exciting.

This behavior needs to be called out and there needs to be accountability. However, there is none. Even when the stories are completely fabricated, as happened at USA Today recently.

What the public wants and what tricks people into giving you attention are different things. Journalists today go for the latter, which is why we have all of these "alternative" news sources now.

I'd love a browser extension that shows me a crowd-sourced batting average for the author of any article I'm reading online. Something that tells me "Oh, you're reading an article written by journalist Joe Schmoe? Joe has previously written 400 breathless articles about scientists discovering space aliens"

I know I know, technical solution to a social problem. And curating these crowdsourced article tags and statistics would probably be an intractable nightmare. It probably can't hurt much to try though. I think the fundamental problem is there are too many journalists for many individual journalists to develop a reputation in the minds of readers. I read several hundred, maybe a few thousand articles a year, but I could only tell you the names of a handful of journalists at best. There must be some way for technology to help bridge this gap.

https://nautil.us/

I actually happened to meet a writer who had this happen to them, but this incident was also publicized in an open letter (source: https://nwu.org/an-open-letter-from-freelancers-at-nautilus-..., with discussion about this on Twitter at: https://twitter.com/aznfusion/status/941051077922869248). This happened in 2017, and I'm not sure how long it took to get resolved.

The future seems bleak.

ETA: Oh!! The link re: the Twitter thread is gnarly! I hope this got resolved and the freelancers got something out of this.

But instead of making dozens of people learning something you paid for school lunches for 200 students each month?

It’s a limit pot of tax payer money and it needs to be used on the highest impact things.

Let me use UK costs as an example, to show how cheap school lunches are compared to all the other aspects of education, because I don’t know the USA well enough: If I assume £2/school lunch (mine were £1, but that was the 90s), and if the school year is still 39 weeks, that’s £390/year, or £11,700 per year for a class of 30, compared to a qualified teacher’s starting salary of £25,714 (head teachers go up to £125,098 in London), and then you need to add the cost of books, and consumables, and the building itself, and insurance, and support staff (HR, caretakers, supply teachers).

Because the state requires them to be there anyway, so it's the state's responsibility to ensure they are well fed? Academic performance has no bearing on whether somebody deserves food.

Edit: we prioritize feeding people over giving them educational science articles; seems appropriate

What other people are doing that is wrong is rarely particularly useful to think about when working toward a goal of your own.

I think good scientific journalism might be something where funding is not proportional to benefit to society. Also things like investigating regulatory capture.

It's a very hard problem, and there are no obvious or pat solutions.

The first test you've got to do for any proposal like yours is to consider what the impact of it will be when your political opponents are at the reins. Are you comfortable with Donald Trump (or really, anyone of his faction) having this power?

You do realize government control and regulation are synonymous, right?

I was definitely worried about my name being featured in the middle of absolute nonsense, but the experience has usually been good: very few of the journalists completely missed the mark. The "tone" didn't always match what I was trying to get across, but it was usually close to someone in the field's attitude.

There were some minor errors and a lot of them come from a surprising source: many journals believe it's unethical to show "copy" (the complete article) to a source--or sometimes anyone outside the newsroom. Some scientific terms have nuances that aren't immediately obvious to people outside the field. For example, mine distinguishes between "inhibition" of neural activity, which involves specific molecular mechanisms (GABA, mostly) and suppressing it, which could be anything. This distinction probably isn't obvious to even attentive "general" fact-checker.

It seems like you didn't quite complete this thought. It sounds like your point is that inaccuracies creep in because the journalists won't give a chance to vet an article's accuracy to the people who have the expertise to do so?

Some newsrooms do have exceptions for scientific expertise, or have wiggle room saying that experts can verify whether quotes or sections of the article are accurate, versus the whole draft. This is a decent compromise if a publication allows it, though I'm personally in favor of having a more trusting relationship between journalists and scientists for typical articles on research findings (unless the article is investigative).

Well-funded magazines (e.g. The New Yorker) also get around this by having fact-checkers with strong scientific backgrounds. This is probably the best solution for editorial independence that avoids sharing drafts, but there's not a lot of money in media and writing as-is, so it's not a realistic solution for the vast majority of publications (especially when even big magazines have been cutting funding for their fact-checking teams, shifting more responsibility to the editors/journalists for accuracy).

I’m a physicist by background and even when active in academia it was hard for me to understand some papers in closely related sub fields, and that was with working experience in that area. I did theoretical and computational physics research and understanding leaps forward in for e.g. experimental hardware would have got blank stares from me. For a journalist, who may have been out of academia for some time, a large amount just gets taken on trust from what the academics themselves say, because you haven’t got a hope of understanding the fine points of the research.

I've noticed this too. The reason why I think this happens is because if you're an employed scientist, then the system is set up in such a way that you HAVE to publish, or else you lose your job. If someone criticizes your paper, and it gets pulled from publication, then thats the same as having never published anything in the first place. Therefore, the technique to survival is to write your paper in such a way that repels criticism as much as possible. The easiest way to do this is to write it in such a way that makes it hard to read, but not in such a way that makes it obvious it's gibberish.

Not trying to be snarky, but that's not how publishing works.. you don't get a paper retracted for criticism, you get it retracted if there was scientific malfeasance. And retractions are actually exceedingly rare.

In fact having criticism / debate around your paper is a great way to get more citations, the real publication currency in academia...

Why would somebody criticize a scientific paper for something other than to point out some kind of scientific malfeasance?

If the paper is written in such a hard to understand manner, then its not possible to make any response at all. That's the point.

Because most papers are not actively deceptive, they're just wrong. Or if not just wrong, they've got some critical error. Even great papers. Most people don't seem to get this.

Out of every paper I've read in my life (easily in the thousands now), the number that I think are/were unquestionable I can count on one hand, and have fingers left over. Case in point: once I made the mistake of pulling the "source paper" on Okazaki fragments (a Nobel-caliber discovery on a core part of DNA replication) for a seminar I was teaching in biochemistry. I thought it would be neat to go back to the source material for such an important discovery.

What I didn't realize is that the original paper was...let's just say that it wasn't really conclusive. It didn't take long for my students to rip it apart, and I was chastened. I should have gone into it with the attitude that I was going to show them how hard and messy real science is. Instead, I feel like I made them believe that their textbook was wrong!

Science is Hard. Even stuff that is considered Nobel-worthy after years of post-hoc examination is rarely definitive when it first gets published. These "reporters" who rush out and breathlessly write a fawning/sensational/scary article about something after they half-read an abstract on arXiv, but question nothing within the article itself, are tremendous hacks.

I am not sure where these weird beliefs come from. Scientific papers are difficult to understand because it's incredibly difficult to explain things that have never been explained before [1]. I encourage anyone who has the above view to spend two years solving a difficult scientific problem, and then do a comprehensive summary in 10ish pages.

A more direct criticism of the above comment is that publication pressure is a very post-world war 2 thing. But you can pull random papers from earlier and find many of them extremely difficult to understand. Here is the 2nd most cited paper by Pauli from 1939 [2]. Try understanding what it truly says. This is one of the smartest humans to ever exist.

[1] To your knowledge at least. People who write bad papers, or do stuff similar to what has been done before often do it because they don't understand the work of others fully.

[2] https://royalsocietypublishing.org/doi/pdf/10.1098/rspa.1939...

On the other hand, Avicenna claimed to have read Aristotle's Metaphysics 40 times before Avicenna could understand it. So it's not a new problem!

Not straightforward for the university press department? What could be more straightforward? Tell the university press writers to go talk to the university researchers they're writing about. If the researchers refuse to cooperate, then don't write the press release!

I find that today children’s education of scientific concepts is cartoonised and made into ‘fun’. These kids grow up to write about science as entertainment.

Any form of education should be challenging. Not fun. I am very taken aback and strongly disapprove of the American tendency to inject ‘fun’ into everything.

Example: learning Math should bring joy..not fun. If learning science and math is about ‘fun’, when the next fun thing comes along..learning would be abandoned.

I also believe that learning should keep one hungry and wanting more. In that.. it has to be a goal that is always a tad difficult to reach. America fails badly here. I can’t speak for other countries.

India and Europe where I have spent time do a better job in this regard, but my opinion about these places are dated.

If we don't have better science news is direct consequence of the disrespect or even deliberate mocking shown for science in the last decades by prominent politicians.

Deliberate dehumanization or mocking of scientists by journalists is also a really old problem; A piece of news titled: "'politicians' say that X is true" would be redone immediately and replaced with the name of the politicians saying that. For science, not having the right of linking your name to your hard work until the four paragraph is the norm.

And this if you have luck, I know the case of a scientist that was asked to be interviewed for a newspaper about their career. She thanked him for their patience, time and free advice, calling him egghead in the newspaper.

Another problem is the infamous formatting all scientific news to pulp format with variants of "scientists are perplexed!" at the end (just to tickle your audience with the implicit message that "they are not smarter than us"). When people is bombed constantly with "science can't explain it", "researchers are puzzled", etc, the trust in science is seriously damaged. And then we have people acting irrationally in the middle of a pandemic, just to make a point.

I'm not trying to one-up you; I'm sure you have far more capacity to understand papers in adjacent subfields to yours, and probably in other fields. I'm wondering what the difference is between our experiences.

I find scientific papers easier to read than most news articles - much more clear, informed (of course), they ask and address much better questions - and generally don't skip the obvious ones that I think of. The graphics in papers are so much better than in news articles, I wonder where scientists get such good training in visual presentation. Mostly I read papers in Nature, Science, or more highly-cited ones I find through Google Scholar. Maybe that population skews toward better writing.

Not to forget push for media by most institutions on anything that could garner positive press.

A science publication that wouldn't care about news economics and readership -- therefore maybe only a publicly-funded no-profit -- could maybe approach science news in a way that would solve the problem illustrated here.

I guess my point is that science is large enough and has enough history that it doesn't need the 24 hour news cycle to be interesting. In the sense that what you already know is news to most people, you can stay interesting without getting a scoop.

I tried following ScienceDaily with an RSS feed a while back and it was just too much. A fire hose of articles and little organization.

One thing I wonder is why there isn't a resource like HN outside of this area. I suspect that people aren't doing as much writing to understand themselves in other fields, and at the same time, they are afraid to discuss preliminary results publicly.

For additional evidence, consider how History Today articles often reach the front page of HN, about historical findings that haven't necessarily been published recently. There is certainly an interest for well-written timeless articles.

What is this need of which you speak?

Let's say some responsible journalistic outlet decided to just store it up, filter it, edit it and publish it in something like Popular Mechanics. Does that mean that Yahoo and Google have to still scrape the bottom of the barrel for new daily garbage from space-fun.biz to fill their "science news" sections?

> Scientific American has hit rock bottom with this new op-ed that is nothing more than a hit piece on Ed Wilson, basically calling him a racist.

> It is written by someone who apparently has no training in evolutionary biology, though she says she “intimately familiarized [herself] with Wilson’s work and his dangerous ideas on what factors influence human behavior.”

https://whyevolutionistrue.com/2021/12/30/scientific-america...

https://medium.com/lessons-from-history/the-shameful-decline...

> Scientific American just did a hit job on one of America’s leading biologists and conservationists. All that, right after his passing. This marks a shameful low point in a steep decline of the magazine in recent years.

> Scientific American is the oldest magazine in the US dealing with science. Continuously in print since 1845, it is known for being full of articles by world-class scientists on different topics. According to its About page, over 200 Nobel Prize winners have contributed to it.

Nope, and that's basically the root cause: it's designed to get clicks and shares from people who Believe (rather than Understand) Science™

Also there are lots of high quality youtube channels for any specific area of science...some of my favourites:

- Everyday Astronout (his interviews with Elon were awesome)

  - CrisprTalk (https://www.youtube.com/channel/UC8TW2xKYRqbLvfGaT783WQA/videos)... a retired guy with lots of lab experience going through clinical trial results, investing in CRISPR stocks and bashing ARK invest's stock picks

  - Modern Healthspan interview with Gregory Fahy (the channel has lots of other interesting videos, but I'm less interested in supplements and more about future research reversing aging).

 https://www.youtube.com/watch?v=4x_OTIP7kjo&list=PLkfzM7KJv6vY68Fvw1g7l9vN9NqhW7Fay

[1] https://publications.copernicus.org/open-access_journals/jou...

Edit: Can only speak of nuclear physics.

1. Error and uncertainty. The news like concrete facts, not (paraphrased from the article), "3 months plus or minus 100 years". Also, a lot of science does not hold up over time. It takes a long time to explain uncertainty and it also dilutes from how interesting the article might be.

2. Science is mostly boring. Incremental, tiny improvements in understanding add up over time, but really are hard to get excited about individually. Once in a while we get a huge leap, but most of the time, it's "we got a few more digits of pi calculated". It's really hard to extrapolate these tiny changes into anything that the average reader will even notice in their lifetime.

3. Ultimately, science has to compete for attention with wars, politics, sport, local, finance (ok, might be as boring as most science, but at least people are literally invested).

So, it's hard. BTW - the author of the article does a great job making science more interesting. Would love to see more writers cover the subject at the level Sabine Hassenfelder does.

I generally see two ways of dealing with this on the ley side, which is to either ignore it altogether or to write if the finding or result completely. Neither is really appropriate and I think it's due to humans not being well wired to appreciate how uncertain everything that we don't have uncertainty parameters for actually is. Establishing a probable interval doesn't actually affect the probability of a thing.

On the academic side, the problem is that most academic are cowards. They see a disproportionate cost to being wrong, and culturally, are trained to always be able to give a 'right' answer, even if that answer is useless. Theres a kind of automoton language that is used by researchers that allows them to escape any real consequences that might extend from having an opinion. It's a kind of aloofness that pretends that the science is happening in some kind of white tower vacuum of pure intellectualism. And they aren't wrong from a social perspective, in that academia will enforce a serious cost to them for being wrong.

To sum up, my broader point is that both journalists and academics tiptoe around and hide behind uncertainty far too much. Any issue worth having a conversation on is one where decisions will need to be made inspite of uncertainty. As well, everything is uncertain, and just because we haven't/ can't get an uncertainty on something, it shouldn't give you more confidence than something we can establish an uncertainty for.

I like this article probably better because they didnt d approach it this way, but they seem to miss out on why many people consume science media and what those expectations are. See every discussion about COVID in the media ever.

considering the size of the scientific community (i.e. people who could tell you what a confidence interval is) vs the size of the general news-reading population, I'd suggest that the likely alternative to a lot of the science reporting referred to in the article is not better science reporting, but less of it overall. sadly, these companies are simply meeting a demand from a relatively uneducated population a large proportion of whom's brains will switch off at the sight of the words "confidence interval"

my suggested solution to this is more public investment in the news industry. more grants for independent journalism

This reminds me of the "Gell-mann amnesia effect":

http://www.sfu.ca/~easton/Econ220W/WhySpeculate.pdf

   Media carries with it a credibility that is totally undeserved.
   You have all experienced this, in what I call the Murray Gell-Mann
   Amnesia effect. (I refer to it by this name because I once
   discussed it with Murray Gell-Mann, and by dropping a famous
   name I imply greater importance to myself, and to the effect,
   than it would otherwise have.)

   Briefly stated, the Gell-Mann Amnesia effect is as follows. You
   open the newspaper to an article on some subject you know well.
   In Murray's case, physics. In mine, show business. You read the
   article and see the journalist has absolutely no understanding
   of either the facts or the issues. Often, the article is so wrong
   it actually presents the story backward -- reversing cause and
   effect. I call these the "wet streets cause rain" stories. Paper's
   full of them.

   In any case, you read with exasperation or amusement the multiple
   errors in a story, and then turn the page to national or
   international affairs, and read as if the rest of the newspaper
   was somehow more accurate about Palestine than the baloney you
   just read. You turn the page, and forget what you know.

> Exactly 5 million on exactly that day? Probably not. But if not exactly, then just how large is the uncertainty? Here’s an example for how to do it right, from the economist, with a central estimate and an upper and lower estimate.

Where is the economist example? It's not linked or quoted or anything.

> Here’s an example for how not to do it from the Guardian. This work is published in the journal Physical Review Letters. This isn’t helpful. Here’s the same paper covered by the BBC. This one has a link. That’s how you do it.

The BBC Example isn't even linked (which I find hilarious bc the sentence is describing the BBC not linking the paper). I don't know what BBC example the author is discussing.

> An example is this story from 2019 about a paper which proposed to use certain types of rocks as natural particle detectors to search for dark matter.

What story? It's not linked....

Reading back to the top this appears to be a transcript, however it doesn't make much sense that only some of these parts are linked and as a result the transcript (for whatever reason) randomly includes links.

I agree with GP: Most news is bad, but people can really only tell when it bumps up against something they know. News on science just has a bonus where research gets published separate from the news, and so even people outside the field can compare the two and point out exactly where the news is wrong.

1. Take Taleb's piece on IQ[0]. Which journalists have the mathematical background to understand it? None. Most of the psychologists who reponded to Taleb didn't understand his math either (some did). So how do you know if Taleb's bullshitting? McClure tried to dumb it down in a great article[1], but even if you understand those arguments, practically no journalist would want to go on a limb and declare a "consensus" is dead wrong, when both sides have seemingly coherent arguments.

2. There's always what I'll call "pendulum effects". If science journalism took the "skeptic" POV to try to increase scientific literacy, by showing the many cases where the scientific consensus was wrong historically (with the hopes of improving people's critical thinking), you'd just end up giving fodder to people to arbitrarily distrust any consensus they don't like (for political reasons). So it becomes "socially useful" for people to believe that consensus means truth. If you taught people all the ways in which science can get it wrong, even highly cited studies in top journals, they'd burn Harvard down. The bias is in the opposite direction: Freakonomics-style 'science journalism' that takes one finding and turns it into an overarching narrative 'truth'.

Science in more complex fields (quantum, economics, sociology) is getting out of reach of journalists and average people. People apply simplistic and inadequate statistical tools to complex systems and reach BS conclusions. Complexity science isn't taught enough. Will journalists ever understand complexity theory? Power laws? If even AI researchers can misunderstand what they're doing[2], how can journalists keep up?

[0]: https://medium.com/incerto/iq-is-largely-a-pseudoscientific-...

[1]: https://seanamcclure.medium.com/intelligence-complexity-and-...

[2]: https://twitter.com/sean_a_mcclure/status/153423172385190707...

It's such a bizarre argument to simultaneously acknowledge that The Science is often wrong yet double down on manipulating the public with this idea that scientific consensus equals Truth. In my view, this practice adulterates science into a technical sophistry that serves as a fifth column for a political agenda. But the root cause here is that science itself is corrupted by the desire to effect change. That is the realm of politics and if scientists want to stop losing trust, they will get out of the business of "ought".

Because the EP trusted my judgement, Science items on that show had 4 minutes max, which was very generous, considering that news items can be given 1:40-2:00 on 8pm news and 2:30 for late-night.

During those 4 minutes, every box of daily journalism still had to be checked:

• Is it clear?

• Does it explain why it's news?

• Is it informing people of an ongoing event they're familiar with?

• If it's entirely new, are they given enough context to make independent judgement?

Unfortunately, scientific content rarely checks these boxes. If you want to explain the LIGO gravitational waves discovery or the Higgs Boson discovery, you can't give enough background for the user to feel informed other than "scientists believe this is significant" - so it's only fit for print. If you want to explain a new discovery in 4 minutes, it better come with a demo, that demo better be rendered/filmed like a product video, or people lose interest.

So the science you end up covering is either from highly funded labs (think labs in Harvard and MIT that have DOD/industry sponsorship), or venture-funded companies with whitepapers and cool demos.

If you're lucky, you have a relationship with the labs, and you get to serve as their informal media adviser. I often met with professors for coffee to teach them to make press kits for their papers.

As a researcher, if you do market your paper, you run the risk of the press distorting it, as OP said.

tl;dr: science is slow and rigorous. News defers rigor to the extent permitted by law. Incentives follow.

One of the absolute worst things about the pandemic was the deluge of garbage articles about COVID, like the ones she's talking about. "Someone somewhere found a correlation of X with Y!!!" gets reported as if it were Nobel-worthy.

"Did it used to be different?" No, science reporting has always been garbage. But in the before times, the mainstream news still reported garbage, but there was less of it to report, and you could easily find quality reporting if you looked for it.

Now there are many more "experts" vying for press time, and the press is able to disseminate their garbage much more effectively.

First, remember the rule that 'most of everything sucks'. Is science coverage especially worse than politics, economy, etc.? Not sure.

Secondly, over the years there have been more and more reporters who specialize in writing about AI, and whose articles are generally of very high quality.

It's not as bad as it used to be, and I'd argue it's better than most people thing.

Some context: I co-run the Last Week in AI newsletter and podcast, so I am exposed to a loooot of AI coverage.

Rolex isn't optimizing for time accuracy anymore, they're optimizing for that type of social signaling.

Similarly, science journalism (and really, all journalism) isn't optimizing for dissemination of information, they're optimizing themselves as a fashion/signal.

PS: Indeed, pretty much every point made can be found in Aaronson's how to tease out bullshit papers. So the problem is highly fractal (in 8 dimensions if going by the keypoints as definition space)

Most people who consume the news, consume it online. The attention economy is destroying everything of substance, because substance is also hard, hard to produce and hard consume.

I remember when all the news papers were folding up around the early naughts, there was talk about making them all non-profits and giving them tax free status to produce. But that never happened. I think it was a missed opportunity.

Then she quotes a study and doesn’t link to it in the video description. She doesn’t link to this post either.

I like the videos she makes explaining scientific concepts but her criticism videos tend to miss the mark with me.

Journalists aren't subject matter experts and by definition will not understand the full depth of the scientific publication.

And most won't even try to. They're happy to write their twitter-ized clickbaity headline and move on to the next article.

If we want to deliver the scientific gospel to the laymen we need some kind of rigorous technical writing thing.

Unless entertainment is literally our method and profit is literally our aim. Is it?

I don't think it's limited to being a for issue. The by and source (i.e., publisher / editor) as have an impact on quality.

Linking to anything in your article is bad, if people click on it they are not clicking on advertisements.

If its a page not often linked to by high profile websites you are endorsing it.

The target page (journal publication) might not be available to the crawler. It could be poorly formatted from a search engines perspective.

But if it isn't paywalled the article about the publication is pretty much a less specific duplicate of the page you've linked to. Which one to rank higher? The duplicate or the original? Extra points if you copy the illustrations too.

Ah, you've been using the big words now have you? These big unusual words must be important to the context of the article. The average casual reader most likely prefers English.

You put a date at the top of your article? That means it is important today but less important on every day that follows.

I'm sure I'm missing 20 other relevant "optimizations".

edit: Try imagine what happens with the news website if it is dropped from the index because it doesn't follow the SEO guidelines. Is there any hope?

It doesn’t suck. People don’t know where to look.

Because it's written by people who don't know the first thing about science.

By major discoveries, I mean ones I cant talk with my mom (a humanist and a lawyer) and make it exciting. Balls that could go though a wall? that's exciting. You travel fast and you age relatively less than someone who does not? Oh yeah.

Higgs boson? Boring

Something something black hole at 3 MM light years? what's a light year?

OTOH you have areas such as biology where the announcements are revolutionary (about 3 extraordinary findings against cancer avery week) but there is a lot of money behind so the details are sparse.

On my PhD thesis, the introduction was 5 lines long, stating that if you need an introduction you should probably not be reading this thesis but here are a few books you could start with. And of course the thesis was terribly boring for someone who is not in the field (even though it had "particles", "neural network" and other buzz words inside).

11. Don't conflate correlation with causation.

As a side note, I hate how her channel has become something which exemplifies much of what I dislike about the now typical youtube channels: Top ten lists, grossly overstated titles, "give me your opinion in the comments", the awful generic CGI backgrounds ...

The other irony is that likely she does it for exactly the same reasons as why science media is often bad: economics.

We need more Sabines not fewer.

The first is criticism of things that are clearly wrong, such as the aforementioned science journalism. This is a good public service, I agree.

The second is criticism of other physicists’ ideas. i.e. her personal opinions and professional disagreements with other physicists.

She doesn’t delineate the two clearly enough to her audience, so some of her viewers may come away thinking that views that are held by physicists who are her peers, are in the same bucket as junk pop science articles. Just read the comments on her videos, they are full of “physics is a scam” type people who feel vindicated.

For example in the video to this article she was using a lot of climate science examples, which will be interpreted in a certain way by the "science is a scam" crowd. I'm not sure she is doing this on purpose (I have the impression she strongly believes in climate change), accidentally or if I'm just oversensitive to some things.

I am very personally convinced of anthropogenic climate change and that it's a serious risk for humanity. I still believe that it's important to take people's questions seriously, and to respect that people who aren't convinced have been making their best attempt at understanding the world. For these people, biased stories that don't put numbers in context are seen as deceptive, and I think that perception is legitimate. The only way to actually meaningfully reach them is to credibly demonstrate that you're actually checking the evidence that disagrees with your conclusion.

However, I disagree that what she does is the "less popular and therefore less lucrative work of summarising what is just not correct out there". The sort of takedowns she does are quite popular and very easy to do. However, my criticism is that quite a few of them are superficial and fall essentially into the same traps that she criticises, i.e. the actual topic is much too complex to either present or take down without a more comprehensive in-depth discussion (which would be much less popular).

Now this is still somewhat ok if she's the expert on the topic she is talking about as she has the expertise to know how good/bad the simplifications she makes are. However, recently she has started weighing in on topics where she not an expert at all: diesel fuel, antibiotic resistance, light pollution to name just 3 from the front page of her youtube channel. In this case things become quite problematic, because she is simplifying things that she might not have a full grasp of herself but still talks about like an expert.

I've actually found a couple channels on YT specialising in lengthy reviews/essays, which I find very good. Off the top of my head:

1. Whitelight[1] - game reviews/critiques/analyses. Particularly worthy of note are the Assassin's Creed Unity[2] (1.5h), Batman Arkham City[3] (3h 10min) and Watch Dogs[4] (1h 15min) reviews.

2. MauLer[5] - critiques of mostly Star Wars, but sometimes also other mainstream films. The ongoing series of TFA critiques has 4 parts so far (there is going to be at least 6 total), each part taking anywhere between 2 and 4 hours.

I also feel that I'm forgetting some, which I don't watch regularly, but periodically am reminded about their existence and then after a long break spend several days watching, to then forget about them again.

But the channels I listed aren't fringe. Quite the contrary, they're quite popular. I think it's also interesting that one of, if not the most popular Vsauce video is the one about the Banach-Tarski paradox[6], which is almost 30 minutes long. His other videos also show this trend, where the long-form ones seem to get more views in general.

And those are just essays/reviews/etc. There is a whole genre of podcasts on youtube dedicated to 3-4h in-depth interviews. Everybody knows a whole bunch of them, therefore I don't even need to list any examples here to support that.

So it seems that YouTube is a pretty good place for long-form in-depth exploration of whatever topic (as long as you don't say "fuck", "murder" or show a human body).

[1]: <https://www.youtube.com/c/Whitelight>

[2]: <https://www.youtube.com/watch?v=Z5dOporS8IY>

[3]: <https://www.youtube.com/watch?v=l3U1TL5yBm4>

[4]: <https://www.youtube.com/watch?v=Kk2G6zc5pKo>

[5]: <https://www.youtube.com/c/MauLerYoutube>

[6]: <https://www.youtube.com/watch?v=s86-Z-CbaHA>

I couldn't find any "listicle" style top N list, the closest was "top players in quantum computing"

> awful generic CGI backgrounds

That's pretty par for the course for this tier of channel. Arguably the "quirky room with nerd tchatckies backdrop" is more authentic, but both are just filler behind a talking head.

> give me your opinion in the comments

Necessary evil to drive engagement to appease The Algo.

Imho her work is a step above the lion's share of science reporting, but I can see how she might be polarizing.

> I couldn't find any "listicle" style top N list, the closest was "top players in quantum computing"

The whole premise of the video of this article is "I give you the ten reasons why I think science suck". Sure it's not in the title (which is even more polarizing), but it's still a top ten list.

>> awful generic CGI backgrounds

> That's pretty par for the course for this tier of channel. Arguably the "quirky room with nerd tchatckies backdrop" is more authentic, but both are just filler behind a talking head.

She could just be sitting at her desk though, but I agree this is more aesthetics and it would have not grated me if it wasn't for the other things

>> give me your opinion in the comments

> Necessary evil to drive engagement to appease The Algo.

That's exactly my point. I don't necessarily blame her, but am more lamenting the fact that all videos seem to have to become like this, if a creator wants to make a living on youtube.

> Imho her work is a step above the lion's share of science reporting, but I can see how she might be polarizing.

I'm not sure, her work is definitely not on par with quanta magazine for example.

Can we all stop blaming 'economics'. Economics says people are largely self-interested, selfish and maximize their own utility (happiness-whatever).

It's not 'economics', it's people. Labor unions are also maximizing their revenue.

You mean, "many Economists with a simplistic, reductionist view of human societies' behavior say".

Gee, back some weeks ago at

https://news.ycombinator.com/item?id=31463927

I presented my opinion:

"... the worst bottleneck in civilization, bad documentation of computer hard/software, ... "

And my opinion would be similar on nearly all less important topics, that is, not really bottlenecks for civilization.

Long my usual summary view has been that the MSM (mainstream media) has some traditions: Create narratives as in E. Bernays. Repeat those narratives as in Nazi Minister of Propaganda Dr. J. Goebbels and his famous

"If you tell a lie often enough people will believe it.".

Then have the media outlets gang up, pile on, form a mob, and repeat the most recently selected narratives over and over. Do this with deceptive "click bait" headlines to get eyeballs for ad revenue and, maybe, to push some specific political agenda. A little more generally, have the writing borrow from formula fiction and belle lettre, that is, with my view of such literature, create VEFEEE -- vicarious escapist fantasy experience emotional entertainment.

The technique is to grab people by the heart, the gut, and below the belt, always below the shoulders, nearly never between the ears.

Missing are, say, the standards of common high school term papers with careful quotes of credible, hopefully primary references, etc.

Also usually missing is a goal of providing objective, credible information as needed by an informed electorate or credible information for any purpose, science, cooking, parenting, software development, much more in careers, finance, ....

Also nearly totally missing is credible, meaningful presentations of quantitative data, e.g., statistical hypothesis tests with stated probabilities of Type I and Type II errors. Actually, far simpler than hypothesis tests, the media commonly is unable even to report percentages carefully. E.g., instead we get some

"up 7.60%"

without since when, measured how, by whom, reported where? And, why is this not just some case of cherry picking? What about corrections for inflation? What about over more points in time than just two? What about causes?

While generally I'm outraged at the writing and content in the media, apparently some people like it. So, my condemnations have to be just my own opinions.

E.g., my standard remark about the NYT and WaPo is that

"On paper they can't compete with Charmin and on the Internet are useless for wrapping dead fish heads."

but lots of people like the NYT and WaPo and disagree with me.

Maybe due to bad writing lots of old media outlets are losing readers and, thus, ad revenues and are on the way to massive change or just out of business, now to be replaced by new media on the Internet, media that gets to save on ink and paper. Maybe. I can wish but can't be very confident.

Let's give the Internet and its new media a few more years and see if some greater variety of content sprouts up from the landscape.

1. They were all written by barely sentient hominids, or possibly AIs who were tasked at 4am with turning a very short letter into a clickbait story

2. None of these hominids had actually read the letter in a linear way

3. None of them had the slightest reason to think you would read all the way to the end of the article they were writing.

Take this form of 'journalism' and apply it to anything about black holes, quantum physics, strawberry ultra harvest moons, yesterday's Wordle, inflation, riots, covid, etc. and you basically have the recipe for (a) a severely bewildered population and (b) an extremely frustrated small group of people trying to hack through all this bullshit to obtain some idea of what, if anything, is actually going on.

Oh. That reminds me. This was good:

https://leightonwoodhouse.substack.com/p/the-generation-that...

Eg, in the text it is presented as if the actual producer (artist) is somehow involved in the value adding while in most Art markets I know of the production is really removed from the actual value chain and the market is largely a market that adds value starting on the second sale, ie. after the producer has lost touch with the value chain. Also, Art as such was really tangential to the subject matter of Media and Journalism.

So, that part of the article could have been left out, as it brings no extra clarity - especially as the main point was not really about market value for the product (neither Journalism nor Art) but instead market value for the producer (either "Artist" or "Liberal Arts Graduate" on a totally different market: The Job Market.

Still, it was an interesting read, and it did add another piece to the puzzle of "understanding some of the reasons we as a civilization are steadily losing our grip on reality" (here, I assume that "we" and "civilization" is lacking the unstated specifier "USA/American")

This is a common trope - and has a name - "gell-man amnesia", that people raise again and again, but I don't think it's the whole story. There are different incentives at play. Journalists are sometimes investigators, sometimes storytellers, and other times just mediators. The article mentions the exagerations and unique incentives at play in the research domain. It is in universities' interests to present research in a certain way. Journalists, as well, are incentivised to create juicy headlines for mainstream consumption, but not always and not consistently. Sometimes it's necessary to simplify, but that doesn't mean truth is completely lost. Detail is not always necessary for insight. Sometimes as well, a news domain is suffiently well understood to be applied as intended; e.g. traffic, weather, crimes in general, matter-of-fact reporting of sequenced real-world events. But then we get nth-order insights: e.g. geo-political reporting from "this thing happened" to "the implications are...", and those necessarily lose some truth because consequences and causalities are hard.

I'd say, generally, things are not so simple as "news = ~lossy". Some domains suffer from bad reporting more than others, and it's worth inquiring why that occurs on a domain-by-domain basis.

The only differentiator is to what degree. Gell-man amnesia is stating that all news has a high degree of it, not exactly equal amounts.

I've generally found the same thing is true in other domains. Seek out writers who know their stuff. They exist, they're just not consistently the most popular.

I can ask someone who's an expert in a different field, and they might say "yes, it's terrible there too", but maybe what they're thinking of as "terrible" is considerably better or worse than what I am.

And it seems clear that journalists in general are much more interested in some fields than others. I don't think it makes sense to assume that coverage of (say) party politics is as bad as coverage of science, because they're surely putting much more effort in there.

(Coverage of party politics is probably bad too, but if so I think it's bad for different reasons.)

I don't think so. Weather news are alright, traffic news are alright, etc. There are fields that suck harder and there are fields that just work okay, maybe because they are simple enough, maybe because you can learn them once and apply what you have learned forever.

But science is different for sure and the journalism on it sucks because getting to the point where you even understand the problem might require years of preparation. And the journalist might not have had those years, despite them having to explain it in a simpler way to an audience which might even know less. A good journalist writing about any other topic might be able to grasp a topic quite firmly when they prepare for a week or two, but something as complex as the bleeding edge of any scientific field will be extremely hard to grasp at times for them (understandably so).

Now the issue is: to explain a complicated issue in a very simple fashion requires more proficiency in that subject, not less.

For what it’s worth, I work in Supply Chain and articles around that in the last year range from vaguely accurate to wild stabs in the dark.

Most people will interpret "8% of rain" as "it's not going to rain", and then be mad at the "inaccurate" weather report if it rains the next day.

This is not at all unique to science; people tend to assume this about their own fields, not realising that it actually applies to most of them.

But it also covers simplifying things for a (somewhat educated) lay audience, making decisions about what to include and exclude, and including perspectives that some experts may take issue with.

Certainly, it would be nice if many journalists had a better background in what they're writing about. But it's also a case that experts can have unreasonable expectations about the depth and nuance of something written for the more or less general public.

The "Science Section" of your local paper covers an absolutely massive range of topics: astronomy one week, zoology the next, and everything in between. Nobody--literally nobody--has the breadth of expertise to do all of those fields justice.

The rest of the paper has a more consistent focus. If you covered last year's debate over gun control, that background carries over to this week's debate. The players change slowly too--some of the folks in Congress have been there for decades.

I'd venture to say weather news is also alarmist and therefore bad, particularly during severe weather events and hyping up hurricane news.

There’s also a new sub-field that explores topics through data and they’re usually quite excellent.

Such pieces often turn into books or a series because it’s difficult to condense it all into one article. My favorite example is the article series about illegal plutonium experiments by the US Government on civilians where they injected poor (and mostly minority) people and dying children with plutonium just to see what would happen.

The articles forced disclosure from the US Govt. in the form of an executive order and led to the book, The Plutonium Files https://en.wikipedia.org/wiki/The_Plutonium_Files

> The government covered up most of these radiation mishaps until 1993, when President Bill Clinton ordered a change of policy and federal agencies then made available records dealing with human radiation experiments, as a result of Welsome's work. The resulting investigation was undertaken by the president’s Advisory Committee on Human Radiation Experiments, and it uncovered much of the material included in Welsome's book. The committee issued a controversial 1995 report which said that "wrongs were committed" but it did not condemn those who perpetrated them.[3] The final report came out on October 3, 1995, the same day as the verdict in the O.J. Simpson case, when much of the media's attention was directed elsewhere.

You should be able to read it here, https://repository.library.georgetown.edu/handle/10822/86870...

Another great example is the assassination of the head of Iran’s nuclear program by the NYT, https://www.nytimes.com/2021/09/18/world/middleeast/iran-nuc...

It’s a beautiful example of great journalism.

If you notice both pieces took a lot of time to compile and were about events that had happened in the past. It’s much more difficult to do this with breaking news, which is where the “all news is bad” perception comes from.

News can be great. Given enough time and research.

The issue compounds when poorly trained, poorly paid journalists interview scientists and then misreport and misquote their findings, leading to a loss of a trust, the next interview being less detailed, etc. It's a vicious cycle of dumbing down and hyping up, all to fit our clickbait-sized attention spans.

Is there an easy way around it? I doubt it. You're essentially reporting bleeding-edge findings from leading PhDs flailing under a publish-or-perish model, to an audience who mostly has not touched science since middle school (if even then). People don't even know why they should CARE about science, much less what your margins of error and P-values are, etc.

University media relations departments are just not going to to point out flaws in the work of their own PI-led research groups, they're in the business of fluffing their reputations, because that means they might get more students, more grants, more positive media coverage, etc. It's a business these days, isn't it? Corporate PR professionals are running that show more often than not.

Reporters are supposed to sort through that and find the good ones. It's just not easy.

You can try it yourself sometime if you'd like. Choose a field of science, comb through a year (or a month's) worth of primary liteature on Google Scholar and Sci Hub, along with press releases from research labs, without resorting to secondary sources like blogs and science news. See if you can pick out the top 10 stories in any given time period and accurately summarize them for an 8th grade reading level. Email competing labs and other experts to get contrary opinions. Write it all up within a day or two, per article. Try to get it published somewhere, and get paid for it if you can. It's not an easy job...

What I expect, is more wordy gibberish, that gives the impression of high level writing, but is not.

> In any case, you read with exasperation or amusement the multiple errors in a story, and then turn the page to national or international affairs, and read as if the rest of the newspaper was somehow more accurate about Palestine than the baloney you just read. You turn the page, and forget what you know.”

https://www.epsilontheory.com/gell-mann-amnesia/

Here's a similar discussion from a decade ago, a rather defensive piece from a journalist:

https://www.forbes.com/sites/daviddisalvo/2011/08/08/why-sci...

Corporate media is, more often than not, just a mouthpiece for state and corporate propaganda, and the types of journalists who succeed in that environment are just pliable weathervanes who do what their editors tell them to do, and the editors do what the owners tell them to do, and hey, meet the Washington Post owned by Jeff Bezos whose AWS got a $600 million CIA contract for web services, so no more investigative journalism like Top Secret America, please!

As far as the main points Hossenfelder raises, i.e. basic concepts like original sources, range of uncertainties, margins of error, unquestioning reliance on press releases, alternative hypothesis, understanding of how mathematical models of physical phenomena are tested against observational data - well, that might force the public to think about what they're reading. That's not the job of the corporate media, they're not there to encourage critical thinking - they're there to take complex topics, simplify them to the point where a small child could understand them, and then repeat, repeat, repeat. That's the essence of propaganda tactics.

That is, at it's core, saying that journalists are, on average, getting more and more stupid.

This seems a bit strange and unsubstantiated, but could very well be true. What would cause a gradual decrease in the intelligence of humanity's journalists?

Or are things like science just getting more complex (more specialized) that an average journalists requires "too much" time to educate themselves about. That would then mean they aren't getting less intelligent, per se...

There are plenty of people who could do competent science (and other) journalism if that was the standard they were held to. Such work can still be found here and there, in specialty journals (Science and Nature news reports are often quite good), but it's increasingly rare in corporate media for the above reasons.

No matter how good the advice to reporters is, no matter how closely they follow that advice, science news will still suck.

> ... stories about how the increasing temperatures from climate change kill people in heat waves, but fail to mention that the same increasing temperatures also save lives because fewer people freeze to death. Yeah, I don’t trust any of these sources.

This is not sourced, but then again it is also not science.

Science is the new religion.

The problem is particularly bad in science journalism, but exists across the board. Financial journalism is really the only area I’ve seen where the journalists tend to have real backgrounds in the subject. And unsurprisingly financial journalism stands head and shoulders above other journalistic fields.

Why do you think it was an arrogant and toxic statement? Do you disagree that financial journalism is better? Do you think finance journalism itself is toxic? Do you have evidence that a belief such as "finance jounalism is generally more excellent than others" is toxic in some way?