• Josh

Scott Alexander vs NYT: Meta-Analysis, Part 2

See here for Part 1, including Scott Alexander, Fredrik DeBoer, Jacob Falkovich, and Elizabeth Spiers.


Minor correction: rereading Jacob Falkovich's take and some of the things he's tweeted since, I moved him a bit further left on the "Call To Action" axis (see end of this post).


Article #5: Scott Aaronson

(A grand anticlimax: The New York Times on Scott Alexander,

but I recommend you see also On standing up sans backbone)


The NYT article about Scott may not have happened without Scott. No wait, Scott A. and Scott A. Wait! I mean, the piece about Scott Al. may not have happened without Scott Aa.

I spent many hours with Cade [Metz], taking his calls and emails morning or night, at the playground with my kids or wherever else I was, answering his questions, giving context for his other interviews, suggesting people in the rationalist community for him to talk to, in exactly the same way I might suggest colleagues for a quantum computing story. And then I spent just as much time urging those people to talk to Cade.

Scott Aa. had previously worked with Metz, and found him to be a trustworthy journalist; he had no reason to mistrust when Metz said he was interested in Rationalists and, in particular, why the community had gotten COVID so right when the rest of media/government was getting it so wrong.


When the story came out (with no mention of the COVID angle, by the way), Scott Aa. was not pleased, and in his post he details 14 ways the story was misleading or wrong (up from Scott Al.'s 4). This, in spite of Scott Aa. having talked with Metz about all of these details, only to have the final piece end up missing the point about all of them.

The trouble with the NYT piece is not that it makes any false statements, but just that it constantly insinuates nefarious beliefs and motives, via strategic word choices and omission of relevant facts that change the emotional coloration of the facts that it does present. I repeatedly muttered to myself, as I read: “dude, you could make anything sound shady with this exact same rhetorical toolkit!” ... [W]ere I ever tempted to bang my head and say, “dammit, I wish I’d told Cade X, so his story could’ve reflected that perspective”—well, the truth of the matter is that I did tell him X! It’s just that I don’t get to decide which X’s make the final cut, or which ideological filter they’re passed through first.

He remains agnostic about how much of the bad piece was due to Metz, and how much was due to NYT editors.


In spite of being upset about what happened, Scott Aa. originally planned not to change his behavior with journalists in the future; in the end, he may have acted rationally based on the knowledge he had and could not have known that this would be the outcome. However, in a follow-up post, he announces that he won't work with Metz again without some kind of explanation that sufficiently exonerates him. From one of the more rational non-Rationalists around, who updates very carefully and Bayesian-ly, this is a pretty strong statement.

I now feel like to work with Metz again, even just on some quantum computing piece, would be to reward—and to be seen as rewarding—journalistic practices that are making the world worse...

Overall, Scott Aa. seems more upset than Scott Al. or Jacob or Elizabeth, but less upset than Freddie (from Part 1).


One Sentence Summary: The NYT article was bad, and people who contributed to it (including Scott Aa. himself) were misled into helping it be written; this seems to imply untrustworthy actors at the NYT, if not directly nefarious and bad-faith ones..


(From here on I'm going back to calling Scott Alexander "Scott", not "Scott Al.")


Article #6: Mike Solana

(Okay fine, we're fighting)


According to Mike, the NYT article is more of the same in an ongoing war between tech and media, which nobody wants to admit is an actual war.

The endless cycle is thus: a hit is published, tech fights back, media fights back, tech fights back, the blue check media gang goes nuclear and accuses tech of targeted harassment for publicly commenting on the actual, literal words they are printing, mea culpa (“we’re all wrong here!”) and a prayer for peace. Then, it’s straight back to the garbage dump. I hate it here.

This piece isn't only on the SSC/NYT situation, but covers also several other recent skirmishes in this war (ignored here). It is all just routine now, a back and forth between two camps--something like Tech and Media--that hate and can't seem to abide one another.


Mike is not mad in particular about this piece because this is just what NYT does, and nobody should pretend to be surprised by it.

The New York Times isn’t publishing one-off hit pieces. At least, in the narrow context of tech coverage, it is obvious many reporters at the Times, and across the press broadly, confuse the attention they receive for provoking controversy with righteous affirmation. They think, in general, they are doing good work — not just well-reported work, but morally good work...

While he stops short of fully supporting #BlockTheNYT (at least, he didn't explicitly update his stance from a previous post), he closes with skepticism of their credibility and good faith.

[F]ine, whatever, we’re in fight. We’re fighting. But let’s call it that... let’s dispel with the bullshit “objectivity” frame and robustly, openly disagree. We want different things, so what? You think you maybe kind of hate me, okay. Just do me a favor and tweet it.

One Sentence Summary: The NYT piece was bad in a way that most tech journalism is bad, because Media sees Tech as a threat; this is just the latest strike in a long-standing war between the two.


Article #7: Noah Smith

(Silicon Valley isn't full of fascists)


In a smog of people decrying political partisanship with angry words, Noah is a breath of statistically-literate fresh air. His central thesis about the NYT article is clear:

To put it bluntly, I think the article both draws on and feeds into the mistaken stereotype that Silicon Valley is full of right-wingers.

Noah argues that the NYT imagines a world that looks like this:

When in reality it looks more like this:

And the rest of his post looks at the data to show how Silicon Valley, SSC, and Rationalism are nothing like safe havens for right-wing politics.


1) On Silicon Valley politics:

Though Silicon Valley founders tend to be more skeptical of regulation and unions than the average Democrat (as you might expect given their jobs), they are overwhelmingly Democrats. On social issues (gay marriage, abortion, gun control, etc.) they are much more liberal even than the average college-educated Democrat. They also strongly favor government redistribution, which you might think would go against their class incentives. And most importantly, they score lower on racial resentment and lower on the authoritarianism scale than the average Democratic base voter. In short, tech entrepreneurs are standard liberal nerds.

2) On SSC politics: SSC likely has no more than 10k readers, and only ~40% of his readership works in a tech-adjacent field; on the other hand, there are nearly 400k tech workers in the San Francisco Bay Area.

In other words, Slate Star Codex was almost certainly a niche interest within the tech industry.

There's no way SSC is holding sway with anything like a majority of Silicon Valley, let alone Tech more generally.


3) On Rationalist politics:

Like SSC readers, Rationalists are not primarily in tech ("the only major Rationalist figure I could find who is actually in tech is Eliezer Yudkowsky, who is sort of an A.I researcher" (lol at "sort of")), but like Silicon-Valley-ites most public Rationalists have left-leaning politics. Based on Scott's complicated stances around e.g. BLM and feminism, Noah reluctantly gives him the label of "conservative". (Of course given that Scott is "a pro-gay Jew who has dated trans people and votes pretty much straight Democrat" and in 2016 endorsed literally "anyone but Trump", I doubt he means "conservative" as in "right-wing" or "Trumpian".)


Noah's upset to the extent that the NYT gets this general picture wrong, and he retorts with his own analysis of the relevant statistical facts. He doesn't appear to see this as part of any big failure mode in media, so no larger call to action is necessary.


One Sentence Summary: The NYT piece is wrong in its broad picture of the relationship of Tech, Rationalism, and SSC readership.


Article #8: Will Wilkinson

(Grey Lady Steel Man,

but I recommend you see also his interview with Robert Wright)


Will tells a decidedly different story than most others, because it is primarily told from the perspective of Cade Metz rather than Scott Alexander.

Somebody tells Metz about SSC, he finds it really interesting, wants to write some kind of article... Metz contacts Siskind and at some point he tells Scott that he already knows his real name and at some point Scott tells Metz it’s very important that he doesn’t use his real name... Well, the Times won’t promise, so Siskind actually does [burn SSC to the ground]. This seems super-crazy and the natural journalistic response to it is “What the hell is this man hiding? What’s he so afraid I’ll find on his blog?”

Whatever Metz's piece was about before went immediately to the back-burner; he certainly must have taken Scott nuking is own blog as evidence of something much much more interesting. So he went digging. Scott himself notes that around this time, Metz "switched to interviewing everyone who hated me and asking a lot of leading questions about potentially bad things I did." SSC readers think this might be a sinister revenge plot, but a good journalist follows their nose, and Metz' smelled something fishy about this man who would insist on his anonymity at the expense of his entire body of work and hard-fought online following.


What did Scott have to hide, that he would give all that up to keep it under wraps? According to Will, this question--not Silicon Valley, not Rationalism--is the real topic of the NYT piece.


And the answer is found in the piece: Scott has some very un-woke writings about race and gender that he would prefer not be released to the world under his real name; Scott appears connected somehow with people like Peter Theil and Curtis Yarvin; Will insinuates that Scott is also probably friendly with Balaji (scary!!); etc. The NYT article doesn't touch these issues as a quick jab hidden away in a long piece "about Rationalism"; these points are the topic of the NYT article!

[T]he most interesting things about Slate Star Codex, from an outside perspective, are (1) that it’s influential in Silicon Valley, and its enthusiastic fans include incredibly rich and powerful people whose technologies and businesses affect all our lives; (2) one day Siskind burnt it all down and summoned a vengeful horde to attack an innocent reporter and assail America’s best newspaper.

This is the piece! Who gives a shit about some weird Rationalism movement? Sure, they're there in the background, but the Main Event is a story about some weird blogger with an army of nerds, with the backing of Big Tech moguls and spooky Alt-Righters, willing to fight for him.


Also at the forefront of Will's piece is a defense of the NYT against those who would slander her. Clearly, NYT >> SSC or ACX!

The professional culture of New York Times is far more concerned to correct for the biases of self-selection than the culture that’s evolved around Scott Siskind’s blogs...
I think it’s important to emphasize that there is simply no sense — none! — in which people who like to talk about epistemology on the Internet are more committed to objectivity and truth than experienced reporters who, in the service of truth, navigate mazes of lies, gaslighting, spin, bullshit and threats for a living.

This we know because people trust the NYT, millions of people around the world do, and its reputation would be tarnished if it didn't hold itself to the highest journalistic standards.


But in his piece, Will also emphasizes that "journalism isn't science", and:

...nearly everyone is motivated to conceal, bend or shade the truth to their advantage, especially to the New York Times — because it is extraordinarily reliable and, therefore, widely trusted. If you can pull one over on a Times reporter and get them to shout your preferred story through their megaphone, a huge number of people will believe it... that’s another reason why it’s literally impossible for the Times, and other outlets with similar reputations for accuracy and journalistic integrity, to not constantly fuck up.

He emphasizes this... quite a lot. Why? It wasn't clear to me until I listened to his interview with Robert Wright. There, he explains that while the Metz piece was totally in good faith and accurate, it maybe wasn't so well-written, it was misleading at times, ehh some parts are "kind of shitty" (e.g. about Charles Murray). Journalists are human, they make mistakes. But Metz did his best and definitely didn't deserve to have hordes of so-called Rationalists sent against him.


In the interview, he summarizes thus:

The main idea really was that, here's a place that's open to all sorts of ideas, but when you open yourself up to all sorts of ideas, some ugly ideas creep in around the edges.

It's totally valid for a journalist to write a piece to illustrate this. (I feel the need to emphasize: Will is a self-described Scott Alexander fan; he finds him very insightful and continues to follow his new site at Astral Codex Ten.)


One Sentence Summary: The NYT piece is an example of good journalism (though maybe not so well-written because journalists are human), on the topic of a blogger who shut down his blog in response to having his name revealed; it basically gets the details right.


Second Halftime


The posts keep coming, and I'll keep summarizing until I get tired; next time I'll include Gideon Lewis-Kraus and Matt Yglesias, which both come highly recommended.


I can't fit a full actual quote in a 2-D graphic anymore, so I tried a 4-word summary of the main point in each case:

See you soon for Part 3.

544 views0 comments

Recent Posts

See All