You Did No Fact Checking, and I Must Scream

I am not one to jump on the "This is AI" train. But the "author" responding is absolutely AI, I would bet vast sums of money on it. It reads exactly like chatgpt.

The author of this post reponds like AI

Wow that is pretty bad. The comments in particular don't really make sense and are also dripping with AI mannerisms. Gross.

What's even worse: Some of the responses to the authors AI generated responses read like prompts. Case in point:

https://news.ycombinator.com/item?id=45610341 https://news.ycombinator.com/item?id=45611835 https://news.ycombinator.com/item?id=45610450

The last one hits hard:

  Can you provide the numbers for Manhattan, or SF centre?

I am distraught at how AI is conditioning us with its sycophancy.

A decent newspaper can afford this because it also has a fact checker, a copyeditor, a line editor, and an expectation that a journalist will be fired[1] if they systematically fuck up the substance of their writing. It’s difficult to find a decent newspaper.

[1] Or otherwise not employed—newspapers perfected not treating their core workforce as employees decades before everyone else.

Gell-Mann Amnesia?

https://en.wikipedia.org/wiki/Gell-Mann_amnesia_effect

That's why it's called Churnalism: https://en.wikipedia.org/wiki/Churnalism

Also, if any news orgs are listening: when you are regurgitating a press release about, say, a report or scientific paper, please make it your house style to hyperlink to the report or paper. That way I can see your sources and judge the claims for myself.

Also, people who write reports or papers and then make press releases: please upload them to your own damn websites, and make them easily findable by the public. Don't just email the press release to your pals in the media, and not put your words anywhere else.

The average newspaper has grossly declined in quality IMO.

But there are some good investigative journalists out there.

Arguably, all the smart and careful journalists have moved to the weekly or monthly format. Economist, The Atlantic, and the like.

Average newspapers have average content.

But there are good newspapers just like they are good <any category of thing>.

Although good newspapers still have bias, but as a reader, you can correct for bias. You can’t correct for sloppy fact checking.

Like in archery, if you always land in the same spot, you can “reverse bias” the result back to bullseye. If you land all over the place, there’s nothing you can do.

The only problem is that good newspapers cost some money.

I was a news junkie for several years (now cured).

I was mildly obsessive about fact checking. And oh wow, it is bad.

My takeaway was that people who casually read the news (e.g. newspaper, scanning headlines on their favorite news site, etc) are the most misinformed.[1] The one who doesn't follow the news knows he is ignorant and doesn't know the inaccurate information. The one who follows it heavily, and with an eye towards gaining knowledge (and not following a tribe) will develop the skill to sift through the crap.

[1] Well, OK - those who obsessively follow only the news in their bubble are probably worse.

Perhaps there's also less stringent editing on the "Showbiz & TV" or "Culture" sections of the paper than the "News" section. I mean, papers in general are working leaner than they should. Hopefully, they put the editing focus on what's most important, but still, being lazy even in a lighter section does reflect poorly on the entire publication.

> In many case I think there should be moments where people just stop for moment and do most basic sanity checks.

Yes!

And two of the most common indicators of that are if you read something that either makes you thrilled or furious.

If I might be more optimistic, I think people may actually care about a statement being rooted in reality, but people may not be likely to slow down and engage in suspicion of something they would not expect to be false. (Though the size of that window may be its own problem!) If I see someone claiming they have the cure for cancer, then I consider it a bit fantastical and want to investigate further. If a supposed quote from an older actress talks about her time doing Shakespeare, then it doesn’t really proc any doubt in me; I’m offering a baseline of trust to the publisher that forwarded that information along to me that this information is factual and not someone’s strange fanfiction about her life. I can appreciate that the author doubted it because a quick scroll of the blog shows that he’s got an interest in stagecraft and so it bumped up against his expertise, but I don’t think that I would have seen the quote myself and done the same… maybe I am one of those sub-median ignoramuses you mention. I agree that people uncritically eating up sensational news is a problem, but this is like, pretty straightforward in-memoriam news that I’d hope to not have to doubt.

On other hand if you are able to compile "facts" to an article. You should as well be able to verify them from second source. And trivially fast in modern world. I mean if you synthesis information from one or more sources. Being able to verify them from one more source should not be huge leap.

Then again, maybe it is just AI generated. Which really makes future look lot worse.

I got the impression the "you" in the title refers to journalists, who should have all of the skills you list. Confusing, as he then refers to the reader as "you" at the end, but I'm pretty sure he's berating the professionals and encouraging everyone else to try and do the job they are not.

I agree though, that the general population can't reasonably be expected to do a better job of it than the professionals, so I can't imagine that exhortation having much effect.

The checking doesn't require anything more than what the original writing required.

There is no reason to try to excuse it

The analogy is invalid and casts doubt on those self proclaimed incredibly rare critical reasoning skills.

You need to be curious to fact check. Anyone can be curious. This is different from "being tall".

OK, if you're reading an (alleged) interview with an actress where she, a nonagenarian talks about her 40s, but it turns out she was in her 30s, gasp.

However, if someone in the news section, keeps calling several US cities a warzone, over and over again, with no evidence, ehh, the hardest part about fact checking this is overcoming any personal biases or prejudices you might have.

The article is addressed to journalists, who have not only the necessary skills but also a professional obligation to provide truthful information

Most people don't, sure. But anyone who calls themselves a journalist or has gone to j-school sure as hell better. That's literally the point. That's what journalism school teaches, or at least should; Not how to repost crap from other crap. It's simply not an excuse for an organization like the bbc.

Agreed. It seems like a supposed quote? If my elderly mom wrote that she went to a NY Mets baseball game in 1960, and Wikipedia says the Mets weren't formed until 1962, is a newspaper supposed to determine that the quote is not legit bacause my mom mis-wrote or was confused?

The fact-check discovered the information to be so implausible that it would arguably have no value to publish even if she did say it (being completely mistaken about all the facts of her own life).

> The BBC pulled that article/episode

Pulling after publishing isn't much better. The damage has already been done. The story has already been absorbed by viewers. The viewers are already spreading the thing they saw to each other. The information narrative has already been influenced.

It is blatant dereliction of responsibility to publish something in the first place without checking and relying on later outcry to go "oopsie daisies".

> It sucks being a journalist. Donate to your local investigative newsroom.

That sounds like absolutely atrocious living conditions for the journalists employed by such a news room.

Honestly, at that point - look for a different job!

I'm not saying this to be mean, I am just picturing myself in such a working relationship of being employed by an organization working in an industry which is hemorrhaging valuation because of various developments like AI (but even before that it was going downhill).

You will be happier doing something else! At that point, learn a trade or similar. You'll be better paid, have a more stable lifestyle and will feel happier long term, even if you love being a journalist now.

Sorry to be such a downer, but hearing messages like that gives me flashbacks to people getting exploited - and I bet ya that the owner of that newsroom will not be suffering like the journalists. You're sacrificing your own happiness in life for another person's wealth gain - just because you thought it was "worth doing" in your 20s. Because yes, money maybe doesn't buy happiness, but it sure as hell gets you an incredible amount of stress if it's absent.

So you're saying the pay is poor, and your coworkers deliberately fuck you over for their own gains. That pretty much sounds like a job you should stay away from by any cost. Do something else, something productive that doesnt lead to a bunch of lies being sold as truths.

How much does it cost to check one fact? If you check just one, and it's wrong, it's probably not worth treating the whole as an accurate piece of reporting. If BBC had checked just one fact in the slop, they could easily decide that it's not worth republishing.

> Once told an editor in-writing three times not to add something and he did it right before publishing.

This should be illegal.

If people were able to propose laws and vote on them directly, it would be, be a landslide.

The number of people who benefit from this is tiny compares to the number who are harmed. But it's nearly impossible to pass things like this because people vote for parties and are therefore several levers removed from influencing actual laws.

Uhm is it possible to mitigate second point by publishing encrypted article in social network along with is sending it to the agency to assign it timestamp? Or maybe in any blockchain if one does not like social networks.

Then if journalist does not like how mangled was his piece on publishing he can disclose encryption password to show everyone what he actually wrote in the article?

Factcheckers are QA. If there is anything for them to do, then the publisher has failed by not providing the writer with tooling for establishing unit and integration tests for all information within an article.

The other interesting thing about the article is just that the entire nature of fact checking, references and cross references, especially over time, requires physical documents.

References to electronically stored documents are about as ephemeral as it gets. And certainly the wayback machine which I would guess is running on donations and largesse, isn't going to cut it for document archival purposes.

It's ironic that we're entering an era where we can store massive amounts of scanned media on a scale we can hardly fathom with modern hard drive density technologies, and yet linking and referencing of that data over a substantive time frame is effectively impossible currently.

Combined with the loss in a sea of what used to just be noise but now is increasingly mendacious/ manipulative/ weaponized noise... And it feels like history itself is being consumed by a cacophony of data noise from the internet

I had this idea of a document DNS at one point when I was working on some records management system at a company, so that various important control documents could reference each other without being system, dependent, vendor dependent etc.

I suppose that is just DNS at some level, although it's probably much too granular for actual DNS servers.

It would provide a mechanism to audit and validate that a document is still properly electronically linked.

I have no idea how you do that one. Even the core domain names shift constantly due to corporate acquisitions and the like.

> to grab and keep eyeballs

yes, but also to manufacture consent for the priorities of the rich and powerful

"But the press of today is a for profit endeavour."

It is worth paying attention to the significant rise in prominence of non-profit newsrooms, particularly in the USA.

Some notable examples:

The Baltimore Banner https://en.wikipedia.org/wiki/The_Baltimore_Banner Founded: 2022

ProPublica https://en.wikipedia.org/wiki/ProPublica Founded: 2007

The Texas Tribune https://en.wikipedia.org/wiki/The_Texas_Tribune Founded: 2009

The Marshall Project https://en.wikipedia.org/wiki/The_Marshall_Project Founded: 2014

I'm particularly excited about the Baltimore Banner, who are only a few years old but are earning sizable subscription revenue now (it's healthy for them not to be too dependent on donors).

For-profit media is definitely a problem, but Jeff Bezos didn't buy the Washington Post and Elon Musk didn't buy Twitter because they thought they were more profitable than any other investments they could have made.

I believe they did it because they wanted the power that owning a media outlet can provide in order to help protect their actually profitable businesses.

It certainly helps that they have their own revenue streams so that they're not just money down the drain. If the Post loses $100M per year, but Amazon keeps making Bezos $50B per year, that's fine, probably costs him less than the depreciation on his yachts or jets.

You’re using multiple definitions of “free” here. One is freedom in the Lockean sense, the other is freedom from the properties and consequences of an emergent system. It’s a bit like saying you are free to choose your own mate and have kids without government involvement but you’re still a slave to natural selection.

The concept of the free press does not guarantee that the truth will proliferate, it merely attempts to avoid the problem of the state defining what truth is. It’s an attempt to select the least worst option because no one knows of a perfect solution or even if one exists.

The free press has always been for profit!

What's changed is that the profit used to come from advertising. Since everyone read the news, they could charge a lot for ads.

Those days are over, and news now bubble up from social media. That kinda works, but it's far from ideal.

To me the 2019 "Covington kids" incident showed how broken the media had become. All the prestigious media, from NY Times down, reprinted a viral Twitter thread as front page news without any fact check.

The reported "facts" were completely wrong, and even if they had been right, some random kids being rude in a park should never be national news.

Bu that's the news world we live in now.

> But the press of today is a for profit endeavour.

For me, the press today is a for influence endeavour. Most journalists have a POV the majority of topics they write about they express that POV with how they discuss the topic. For example, which people they quote, generally only ones that agree with their POV. If they present an a opposing view they always couch it and phrase things to push the reader to discount that view. If they preset a supporting view they phrase it in a way to make it sound trusting and authoritative.

To put it more simply, most journalists are trying to change the world to see things their way.

Worth noting this is far from a modern problem. Google "yellow journalism".

Blaming society for the poor state of journalism is tempting but ignores that the root of the problem lies from within. Financial institutions and other journalists demand information dense journalism to do their jobs and have no problem paying for it, so this is what they receive. Most regular people view news as a form of entertainment and have no problem with sacrificing their attention, and this is what they receive.

The press has always been for profit, it was never a charity. What I see today is a mix of trying to maximize profits (which is different from merely making a living from it), and it being more difficult nowadays to make money from diligent journalism, mostly due to how the internet works.

Especially the actor one, understanding why someone would take the trouble to check that is almost as hard as understanding why someone would take the trouble to lie about it. Should we expect it of media? Probably.

The "payload" in this article, the thing he wants to spread debunking of, is the indeed false claim that Euan Blair's son Multiverse's company got a government ID card contract.

But looking into it, that company seems very odd. Can you really get a billion pound valuation and investments from tons of powerful people from placing school leavers into apprenticeships?

Sometimes I wonder if PR companies spread false stories about companies to pre-emptively discredit the true stories that have yet to be told.

Aiui the New Yorker has always been the most rigorous in its fact checking.

People used to pay $2 a day for a newspaper, what's $2 a month gonna do?

> My preference would be for consumers of news to pay for that news.

I think that this is a fundamentally wrong headed idea. I might go farther and say that you're not going to create a new model for news, you're going to discover a new model, or you're going to accept the model that you know works, and figure out how to make it conform to your values without breaking it.

The facts: people who produce a lot of what we call news want it to be read. They write it because they have made a value judgement that it has importance, and that it should be reacted to by people. They should be backed by people who share those values - and those people should be paying for the news to be made and circulated to anyone who they can convince, beg, or trick into spending their time to understand what the producers of the news (both financiers and journalists) think is important.

This is the current actual model of the news, even though it still masquerades as a strange public service model adapted from network television news where OTA channels were required to do something in return for their use of the public airwaves. This was never the print model, which is that you have a boss, and you do what he says.

The other model of the news was never in danger. People will pay for sports and celebrity news and photographs because it is entertainment. People will even pay for crime news (of the titillating type, like violent street crime, rape and murder) because it also usually is entertainment masquerading as public service. None of that is in danger. It's important to be specific about what you're talking about.

I think that the way to save the endangered part of the news is the same way we need to save everything - more collaboration tools for individuals to form into groups. I think we see this forming with things like Substack, Patreon and Locals, etc., and what we need is to make this more decentralized, and not routed through a few gatekeepers all additionally gated by banks and credit card companies. 500 people should be able to start a newsroom of 10 people with their spare pennies, and get what they believe and what they think is important to the widest audience possible. The Substacks, Patreons, etc. have shown that it is not hard to get that many people together to pay a few bucks for something they believe in, or simply like.

If there's a tl;dr: you shouldn't be paying to read journalism; if anything, people should pay you to read their journalism. I wish I could take money from 80% of the garbage I wasted my time reading. We need microdebits, not micropayments.

For me, it seems like the solution is some sort of seamless micropayments solution for individual articles. I don't want to subscribe to 30 different outlets, but I'd pay $1-2 per article for a good piece of journalism from those outlets.

The problem is that micropayments are expensive. 2.9% and 30 cents is 32.9% on a dollar transaction (and basically all of it if you charge 50 cents to read an article). I've seen some cryptocurrency attempts at a solution, but I think a more viable solution would be a single account you periodically top up, and some aggregator that distributes payments to outlets in bulk to minimize fees.

I've looked at others' attempts in this space [1][2][3]*, but none of them seem to have taken off and I'm not sure why. It seems like a win for publishers, unless those micropayment news readers end up cannibalizing their subscriber base.

[1] https://readwithacta.com/

[2] https://www.supertab.co/

[3] https://brave.com/brave-rewards/

* I think Brave's approach of replacing ads with their own and paying in their own crypto is atrocious FWIW