I interact with journalists quite a lot and I have specific preferences. Not just for articles, but for behaviour. And journalists do behave pretty strangely at times. 

This account comes from talking to journalists on ~10 occasions. Including being quoted in ~5 articles. 

Privacy

I do not trust journalists to abide by norms of privacy. If I talk to a friend and without asking, share what they said, with their name attached, I expect they'd be upset. But journalists regularly act as if their profession sets up the opposite norm - that everything is publishable, unless explicitly agreed otherwise. This is bizarre to me. It's like they have taken a public oath to be untrustworthy.

Perhaps they would argue that it’s a few bad journalists who behave like this, but how am I supposed to know which the good and bad journalists are? The general advice in many tech, EA, rationalist spaces seems to be “be very careful when talking to journalists”. So it doesn't seem like it’s a rare bad apple.

And even the ‘good’ journalists think that there are somewhat arcane rules about what can and can’t be shared from public conversations, to the extent that I have to weave the following magic spell even when talking to the ‘good’ ones.

“Hi can we agree this is all off the record and then I can agree specific quotes to be on the record? I almost always give quotes, but I want the ability to veto anything that goes next to my name”

And sometimes journalists I respect say no. Their standard negotiating position is that they get to take what they want from a private conversation and publish it how they want. Imagine what the bad ones do. 

This is not the typical norm around privacy and I don’t agree to the inversion of it. I ask family and close friends whether I can tweet things they say even if I am pretty sure. But with journalists, even the ‘good’ ones, I have to specifically ask that they don’t take my words and tell them to 100,000s of people, far more people than most people have ever communicated with at once^[1]. 

And worse, if challenged on this, I'm not confident they'll acknowledge it. It's not just the norm breaking, it's the gaslighting. Oh us? No, we're bastions of democracy. We’re the good guys. I have two fairly vague recollections of journalists mangling my quotes badly and several more where articles had misleading titles. If this was a friend I would be deeply upset, but in those cases there were low effort apologies, as if this happened all the time. It is crazy to watch. 

Preserving meaning

I do not trust journalists to accurately convey what I said or what I meant. If I tell a journalist X, I have to go to great lengths to ensure that if my name is published it is attached to X and not some related statement. And it is even harder to ensure that the title of the article matches the content.

There are entire organisations staffed with people who regularly mislead, many of whom think they are noble freedom fighters. I don't understand how this façade is maintained. I understand why we respect The Economist, and surely the people who work at Fox News must know they are charlatans, but does The Guardian really deserve the high standing that its journalists get, given the misleading pieces it sometimes publishes. The Telegraph and Times are at least as bad. 

I think my message to journalists is, “if you can't commit to uphold some reasonable standard of ethics, then do a different job”. I know that jobs in journalism are hard to come by, but many of you are clever, hard working, insightful individuals. You have other options. So either be the journalists you think you are or stop publishing misleading quotes under clickbait titles. No one forced you to do this. If you choose to, it's on you. 

If I am to give reporters quotes, I require them to take responsibility for their headlines.

I have done this twice. One journalist was happy to accept responsibility and I gave them a quote, another wasn't and I didn't.

And again, if there are journalists who behave badly, I can’t tell the difference. I don’t know which papers are good or bad, who is trustworthy, who everyone hates. If you don’t teach me this information, the median journalist is untrustworthy to me and I will not trust any of you.

Reporters vs pundits

At this point, I'd like to make a distinction between what I think of as reporters and pundits.

Reporters discuss the world as it is, or will be. They write news articles and, in my preferred world, they would write forecasting articles as well. They are commentators at a football match, the engineers of the journalism world, reporting how things are, or realistically might be.

I also see another group I call pundits, who are spotting patterns and writing about them. Are men getting weaker? Does Trump speak to a deeper truth? What's going on with Iran? They give concepts that one might use to interpret the world, but accuracy isn't really the point. It's about ideas. I take these people seriously, but not literally. They are there to add patterns and colour, sometimes to speak truth that reporters might miss. They are like pundits at a sports match, with colourful stories, but dubious analysis. Rather than engineers, pundits are more like mathematicians. They write about a world only tangentially^[2] related to the real world but they provide tools that might be useful later.

A friend who used to work in a hedge fund sometimes talks about Eric Weinstein. Weinstein was a talking head in the late 2010s and often seems like a crackpot to me. But for a time, he ran Peter Theil's hedge fund. How could a crackpot do that? Well, my friend says that some hedge funds have two types of guys - careful methodical guys and wild idea guys. 

At a hedge fund people are looking for a way to predictably make money. Often they want a way that the world is connected that no one else has noticed. And in those days, you supposedly had both wild idea guys and careful guys. The wild idea guys thought all sorts of things, trying to spot patterns in correlated data. Take the (real) correlation between the age of Miss America and murders by hot air. This is a silly example but you might suggest that an older beauty pageant queen means you should sell your shares in ironing products. A wild idea guy might come up with something like this, then you got your careful folks to check if there is such a connection. 99 times out of 100 there wasn’t.

But 1 in 100 times, there was. And this is where the hedge fund makes its money, enough to cover the cost of looking up all those bogus ones. The above example is one I could find a graph for, I’m not suggesting that a hedge fund would actually have looked this particular correlation, but I imagine they did look up some pretty bizarre ones. Eg Pizza sales from branches near the Pentagon really might suggest that the US is about to go to war.

Anyway, Eric Weinstein was, supposedly, a great ideas guy. He would come up with all these wacky correlations that his careful folks would check over. Enough of the ideas were good that the hedge fund made a lot of money. And I don’t know if this story is true. Like the thing it describes, the story is a pattern, not an accurate representation of the world. I wouldn’t be surprised to learn that Weinstein is a crackpot and he was just a lucky crackpot. 

But the rule of thumb stands. There might be ideas people and accuracy people. And this is what I want from pundits and reporters (Weinstein would be a pundit).

I would prefer a pundit who is blindingly insightful 1 time in 10 and a crackpot the other 9, to someone who has a pretty similar worldview to me - I already know what I think.

But when looking for reporters, I want someone who cares about the overall accuracy of what they write - that I take away the correct overall interpretation. Ideally I’d like them to check the pundits too. Robin Hanson and Bryan Caplan are great pundits, but I want some more careful economists to redo their major works and tell me whether the consensus agrees about prediction markets or the state of education.

And it might surprise you, but I wish we had more, solid, trustworthy reporters. Everyone wants to be a pundit with the bombastic ideas that explain it all. But I would enjoy many careful, well-calibrated takes on AI, US politics, Taiwan, etc. Being accurate in underrated. 

Accuracy

In my experience reporters (by the above definition) are not as accurate as I would like. They prioritise precision and justification of their writing, potentially compromising the accuracy of their reporting. 

Remember the difference between accuracy and precision from your school science classes imagining someone doing archery. Accuracy is about getting near the centre of the target. Precision is about being closely clustered. It’s therefore possible to spend a lot more time focusing that your shots are close together than that they are actually on target.

Here is the classic graph courtesy of Momemtrix.

In the example here, I’d say accuracy is whether the reader comes away with a view of the situation that matches reality. And precision is whether the reporter sticks close to facts they can justify or to a single coherent narrative. 

Many reporters seem to believe their job is to ensure every sentence they write is precisely true, rather than aiming for an accurate piece overall. They want their arrows to get close to what they intended for them, somehow missing that this can leave the overall picture very mistaken. How can this happen? Well if you only let yourself write true sentences, then there are a lot of sentences you can't write - those containing "maybe" and "perhaps". And a story that contains no “maybe” sentences might miss the point entirely.

To give an example, look at the reporting around SpaceX killing bird habitats. I included a link to a google search. Find an article at random. Such an article will likely discuss the process of some environmental agency. It will talk about rockets and perhaps have quotes from Elon Musk. It will discuss bird habitat destruction, or similar. But it won't discuss that more birds have probably been killed by the cats of the journalists than by SpaceX rockets. Reporters don't like this kind of ‘probably’ and don't see the comparison as relevant.

https://x.com/krishnanrohit/status/1810032112273350978 

To me they have missed the wood for the trees. Their job, in my eyes, is to inform me about the world. And the most basic kind of information is “should I be reading this at all?”. 

Now some reporters might argue that they aren’t doing. We aren’t saying it’s important, we’re just reporting what’s happening! Well I don’t buy it. I reckon most people read the titles of articles and little of the content. And even more read the content for a general view, rather than thinking carefully about what isn’t said. To publish an article is to say that we should care, and that we should update our beliefs towards the headline - an article about SpaceX killing birds implies that SpaceX is killing lots of birds or very valuable birds. To pretend otherwise is wrongheaded. 

Worse, sometimes a reporter will report another reporter, further passing the buck. They will uncritically report or retweet another hack’s phrases, simplifications and accusations to 100,000s of people, with little consideration. When pushed here, they won’t say “it’s a thing someone said” they’ll say “It was said by John Barker, of the Enfield Times. Would such a learned fellow lie?”. Well I don’t know about lying, but John Barker, whoever he is, will probably do exactly the same as this journalist, laundering sayings into happenings without stopping to consider whether he would say those same things in his own voice.

If I put a gun to your head, do you think SpaceX are a significant percentage of rare bird deaths? If not, don’t write about it! Take some responsibility.

I want articles which have sentences like “Trump is 50% likely to win the election”, “China is 15% likely to attempt to invade Taiwan before 2030”. I want journalists to care about their calibration at least a moderate amount, seeking for their 70% claims to be accurate 70% of the time. I want them to have accurate views of the world, write accurate articles and give me an accurate view of specific topics. 

Somehow reporters have focused on "is each specific sentence true?" rather than "is this article accurate?". And  "can I defend why I said this thing?” rather than "will my reader believe more accurate things about the world after they are finished reading?". At scale, this seems pretty disastrous to me. 

If you are a reporter reading this, I suggest you ask two questions:

• Will my reader come away with a more accurate view of the world if they read just the headline, the first paragraph, the whole article,
• Is this article worthy of my reader’s attention? Will they endorse having read it? Is it something that matters or stirs their heart?

In other words:

• Am I misleading my reader?
• Am I wasting their time?

My personal stance

I lose respect for journalists who consistently spread inaccuracies, whether they wrote them or are quoting others. I do not accept mealy-mouthed words about how it's technically what happened or that the quotes person is respected. If you publish something, I consider it a bet by you on it being true. If it is false, I deduct points from your social score. Ideally you would lose money. 

I encourage journalists to use prediction markets and forecasting tools can help improve the overall quality of reporting. These methods allow reporters to better calibrate their understanding of world events and increase the accuracy of their predictions. I can hardly underrate the value of having an accurate model of the world to report relevant goings-on within it.

Journalists exist to help us understand the world. But if you are a journalist, you have to be good enough to deserve the name. And if your colleagues aren’t, help me to know who is good and who is not. If you are a pundit, explain deep truths and teach me ideas that might be relevant in the next 5 years. If you are a reporter, give me an accurate view of the world as it is and will be. 

Currently I deal with journalists like a cross between hostile witnesses and demonic lawyers. I read articles expecting to be misled or for facts to be withheld. And I talk to lawyers only after invoking complex magics (the phrases I’ve mentioned) to stop them taking my information and spreading it without my permission. I would like to pretend I’m being hyperbolic, but I’m really not. I trust little news at first blush and approach conversations with even journalists I like with more care than most activities. 

I will reiterate. I take more care talking to journalists than almost any other profession and have been stressed out or hurt by them more often than almost any group. Despite this many people think I am unreasonably careless or naïve. It is hard to stress how bad the reputation of journalists is amongst tech/ rationalist people.

Is this the reputation you want?

Thanks to Josh Hart for helping me edit this.

1. ^{^}

   It is little wonder that ‘media training’ is commonplace.

2. ^{^}

   Don't think I don't respect mathematicians, it's just that their work can be correct without being relevant to the world we live in. In fact that's kind of the whole bit. Also I think my use of tangentially is really nice here, since, like a tangent often the specific mathematical rule we choose to represent a physically situation is locally accurate and globally inaccurate, like the tangent as an approximation to a curve.