Hi everyone! I'm Tom Chivers, and I'll be doing an AMA here. I plan to start answering questions on Wednesday 17 March at 9am UK: I reckon I can comfortably spend three hours doing it, and if I can't get through all the questions, I'll try to find extra time.
Who I am: a science writer, and the science editor at UnHerd.com. I wrote a book, The Rationalist's Guide to the Galaxy – originally titled The AI Does Not Hate You – in 2019, which is about the rationalist movement (and, therefore, the EA movement), and about AI risk and X-risk.
My next book, How to Read Numbers, written with my cousin David, who's an economist, is about how stats get misrepresented in the news and what you can do to spot it when they are. It's out on March 18.
Before going freelance in January 2018, I worked at the UK Daily Telegraph and BuzzFeed UK. I've won two "statistical excellence in journalism" awards from the Royal Statistical Society, and in 2013 Terry Pratchett told me I was "far too nice to be a journalist".
Ask me anything you like, but I'm probably going to be best at answering questions about journalism.
If you were given several million dollars to structure a media vertical on "news that actually matters"*, what would you do differently from Vox's Future Perfect?
*By EAish lights
How nervous should we be about talking about/recommending action on AI risk?
I think a lot of people in the EA community worry that AI risk is "weird", sufficiently weird that you should probably be careful talking about it to a broad audience or recommending what they donate to. Many would fear alienating people or damaging credibility. (Especially when "AI risk" refers to the existential risks from AI, as opposed to, e.g., how algorithms could cause inadvertent bias/prejudice)
A thought experiment to make this more concrete: imagine you were organising a big sponsored event where lots of people would see 3 recommended charities. Would you recommend that (say) MIRI would be one of the recommended charities?
What books or articles (or movies, podcast episodes, etc.) would you most recommend to a random thinker (or, say, a random science journalist) if they wanted to level up their rationality and/or their altruistic efficacy?
What do you think people in journalism think of EA, if anything? Are there any key errors that EA is currently making or would be likely to make in the future when dealing with journalists?
I think there are a lot of journalists who think EA is very wise and sensible, and there are a lot of journalists who think it's all neckbeard rationalist techbros, or whatever dismissive term they might use. I think the first journalists are right and the second journalists are wrong.
(TBF I think the largest group of journalists is probably the ones who've never heard of EA, and don't write about things that are anything to do with EA.)
Key errors: ah man. I don't really want to advise after I got it so badly wrong with the Scott Alexander/NYT stuff, and also I don't feel I know the EA community well enough to say what they do now. But I do think they could do with finding a few more media-savvy, personable spokespeople who can help you get your stuff into the media when you want it there. I'm always surprised that, say, 80,000 Hours doesn't act more like a think tank, trying to get journalists reading their latest work. But maybe if EAs start playing the game like think tanks do, they'll end up drifting away from their purpose and start chasing headlines and so on, and that would be a shame.
How one should go about learning how to write high-quality material? And what is the way to get it published?
Are there journalists or outlets you think EAs or rationalists should especially be following? Particularly ones who might not already be on our radar?
If you were given several million dollars to run a news organisation with positive impact, how do you think that would compare to GiveWell's top charities? Do you think there is a way for funding a news org to be an impactful donation?
The majority of the public (I guess) still gets all their news from a couple of sources. None of the major news orgs have good fact checking and most have blindspots. Do you have any thoughts on how a news organisation could provide an accurate view of the world whilst gaining a large readership?
If you could re-write The Rationalist's Guide to the Galaxy today (and you weren't too worried about making the book too long), what are the ~three things you'd add that aren't covered there?
1. this is a totally fair question! But … hmm. I am (these days) rarely "assigned" work, but obviously I do have to get my editors to agree to let me write things, and I can't just write "buy antimalarial bednets" every week forever. That said, if I did, no one would read me, so I need to write interesting things that get an audience. So there's a theoretical tradeoff between "say what's important" and "say what will get read", and the greatest impact means finding the right balance.
That said, in reality, it's not often as though I have two brilliant ideas for a column in the same week and I have to choose between them; it always feels like a miracle that I usually manage to have at least one. And since I don't always have a really clear idea about what will have the most impact, I rarely feel like I'm writing something I don't think matters.
(I guess when I do something fun, like this about Warhammer, it doesn't "matter" in some sense that it probably won't save lives. But I enjoyed writing it, and people enjoyed reading it, and maybe it gave a few people an idea for how to have a hobby. I doubt it was the most impactful thing I could have done, but I still feel it was net po... (read more)
Why don't more journalists make concrete, verifiable, quantitative forecasts and then retrospectively assess their own accuracy, like you did here (also see more examples)? Is there anything that could be done to encourage you and other journalists to do more of that?
If you haven't spent time on calibration training, I recommend it! Open Phil has a tool here: https://www.openphilanthropy.org/blog/new-web-app-calibration-training. Making good forecasts is a mix of 'understand the topic you're making a prediction about' and 'understand yourself well enough to interpret your own feelings of confidence'. Even if they mostly don't have expertise in the topic they're writing about, I think most people can become pretty well-calibrated with an hour or two of practice.
And that's a valuable service in its own right, I think. It would be a major gift to the public even if the only take-away readers got from predictions at the end of articles were 'wow, even though these articles sound confident, the claims almost always tend to be 50% or 60% probable according to the reporter; guess I should keep in mind these topics are complex and these articles are being banged out in a few hours rather than being the product of months of study, so of course things are going to end up being pretty uncertain'.
If you also know enough about a topic to make a calibrated 80% or 90% (or 99%!) prediction about it, that's great. But one of the nice things about probabilities ... (read more)
Out of curiosity:
No big story! We changed editors between the hardback and the paperback, and the new editor decided that the book wasn't really centrally about AI. She felt that it was more about the rationalist community, so she wanted to change the title to reflect that. I know sod-all about the publishing industry and I trust her judgment better than mine, so I said "fine".
If I'm starkly honest with myself I think it's probably because the sales weren't great and she thought this would help, but I don't know.
The positive impacts I guess will be if it sells more copies, but one negative impact is that a few people thought I'd written a new book. I hope no one accidentally buys a second copy but I imagine it might happen.
(It's not that unusual, I don't think. At least two of my journalist friends have books with different paperback titles to the original hardback titles.)
What do you see as the consequentialist value of doing journalism? What are the ways in which journalists can improve the world? And do you believe these potential improvements are measurable?
What would your advice be for talking to the media about EA? (And when to figure out whether to do it at all!)
How would you frame the message of EA to go down well with a large audience? (Eg, in an article in a major newspaper). How would this change with the demographics/political bias of that audience? Do you think it's possible to convey longtermist ideas in such a setting?
Being ahead of the curve on COVID-19/pandemics seems like a major win for EA, but it has also been a major global tragedy. How do you think we can best talk about COVID when selling EA, that is both tactful and reflects well on EA?
Honestly, I think this is all about finding journalists who you trust. The Vox lot, for instance, you know they're not going to write a "ha ha look at the weirdos who want you to donate money to stop humans going extinct" piece, even if they don't agree with the specific position you're supporting.
"How to frame it" comes down to the same thing. In the end, assuming that you give interviews to journalists, the journalists will be the ones framing it, so choose journalists you trust and feel you can speak freely to. I know that's kind of unhelpful advice - "be good at choosing people you trust" - but it's really important.
That said. There are some general tips if you're, say, writing research papers (and press releases for those research papers). One of them is including a nice clear list at the top of things that your paper doesn't say: if your paper finds a correlation between doing crosswords and brain health, for instance, it's worth saying "this does not mean that crosswords prevent alzheimer's" at the top. That's been shown (Chris Chambers at Cardiff did some great research) to avoid misunderstandings without reducing press coverage. So I guess that could be relevant. With long... (read more)
What are your thoughts on solutions journalism? Does it have much traction among science writers you know? Do you personally use it or promote it as a framework for writing?
Do you think this is a good/bad idea?:
I have the hunch that EA and solutions journalism could be a good match. E.g. EAs in journalism could join the solutions journalism network and seek solutions journalism angles to their editors. EA projects that think they would be well-served by public media coverage could seek to build relationships with strong solutions journalists and make... (read more)
What are the best changes (in terms of tractability and importance) that you think could take place in the journalism industry in the next 20 years, and how can people help make them happen?
What's the biggest bottleneck on the positive impact of your work?
How did you make the choice to go freelance?
From your Twitter, it appears that you think a lot about covid-19. So why is the UK response to covid-19 so bad*?
Sometimes my American friends will blame US covid failures on US-specific factors (eg, our FDA, presidential system, Trump). But of course the UK is a (by international standards) culturally similar entity that does not share many of those factors, and still appears to have outcomes at least as bad if not worse. So why?
*I admit this is a bit of a leading question. My stance is something like With the major asterisk of vaccinations, it appears that UK outcomes of covid are quite bad by international standards. Moreover, we can trace certain gov't actions (eg "eat out to help out") as clearly bad in a way that was ex ante predictable. But feel free to instead respond "actually you're wrong and the UK response isn't so bad due to XYZ contextual factors!" :)
The UK's outcomes have certainly been bad! Absolutely no argument there.
Before I go into why I think it was, I will say that I suspect there's quite a lot of randomness in these outcomes, and it's not straightforward to say it's because of political responses or whatever.
But that said I think I am pretty comfortable saying that the political response has, in fact, been bad. I think failure to lock down early not once but twice (maybe three times?), the eat out to help out nonsense, the CONTINUED insistence on "washing hands" and so little emphasis on meeting outside, ventilating areas, etc, is really bad and stupid. I think a lot of it stems from bad political leadership (the cabinet mainly chosen for loyalty over brexit rather than talent, a PM who never wants to deliver bad news and always wants to say "I've saved Christmas"). And it also probably stems from years of chipping away at state capacity, having a really centralised state with a paradoxically weak centre, etc.
I am, however, quite low-confidence in all this stuff. I think that this set of politicians has been bad, and a randomly selected cabinet from the last 50 years would on average do better, but I don't know.
Thanks for doing this AMA!
(I'm perhaps most interested in your independent impression - i.e., what you'd believe befo... (read more)
Right guys I'm going to have to stop for now because I need to go and help with my kids' dinner. I'll try to answer some more tomorrow. I'm really sorry I didn't do this yesterday: it totally slipped my mind among all the various things I'm doing at the moment (I have a new book out today!). Sorry if I haven't got to you yet
If you could snap your fingers and change some things about journalistic norms, what kinds of articles tend to get written, how articles and news sites are structured, etc., what would you change?
Since after writing your previous book, do you still follow EA or rationalist content? (eg this Forum or LessWrong)? If so, what things do you find the most helpful?
What are two problems/bottlenecks you wish EAs spent more time thinking/working on?
How do stats get misrepresented in the news? What can you do to spot it when they are? :)
What's UnHerd? Is there anything unusual about it, or should I approximately treat it as "typical news outlet that happens to host your content"?
Thanks for doing this -- I'm a big fan of your book!... (read more)
I'm interested to hear what you think this post about how media works gets right and gets wrong. In particular: (1)
How could those reading this better support you? Is there anyone you'd like to be in touch with?
I think it's better to put every question as a separate comment since that allows us to up or downvote them individually. Sorry to be that guy.
OK I think that's all the questions! Thanks for that and sorry again for the delay. I hope it's been interesting! Best, Tom
Do you have an explicit theory of change for your work as a whole, or for specific projects/books/articles?
Thanks for the work you do I consider you a top tier journalist.
Thanks particularly for making testable forecasts in your articles.
Thanks for doing this also. I know it has benefits to you, but coming to a forum for an AMA is a cool practice.
Right! Sorry. I got a bit swamped but am going to have a go at answering questions now! Sorry about the delay.