I knew when I saw the title of Noah Smith's latest that it was going to be bad. I just didn't realize it would be this bad

We've got a lot to cover. Let's just dive right in.

From "Why I'm (cautiously) optimistic about Elon buying Twitter" by Noah Smith

[Emphasis added]

For one thing, I think Musk is well-positioned to deal with foreign information ops, especially those perpetrated by Russia and China. Russian bots and agents are big on Twitter, and China has been working to build a similar network. This presents the disturbing possibility that the existence of Twitter spells doom for liberal governments — if totalitarians can exercise tight control over their own Twitter-like networks while using info ops to heavily influence the discussion on Twitter itself, it could give them a crucial advantage in the new era of international competition.

I suspect that Musk is thinking about this scary future and how to avert it. When Russia invaded two months ago, Musk instantly shipped thousands of Starlink internet kits to Ukraine. This was crucial in helping the Ukrainians keep their internet running in the face of Russian cyberattacks and bombardments. This demonstrates that Elon values the defense of liberal societies against totalitarian aggression. It stands to reason that he’d also care about this in the case of Twitter info ops as well. Elon famously cares about free speech, but when totalitarian governments use their power to selectively disrupt speech in free societies, that seems pretty detrimental to free speech, and it needs stronger pushback.

We'll get back to China in a minute. For now, though, let's look at Starlink. 

From the Washington Post.

After Russia launched its invasion, Ukrainian officials pleaded for Elon Musk’s [Emphasis in original] SpaceX to dispatch their Starlink terminals to the region to boost Internet access. “Starlink service is now active in Ukraine. More terminals en route,” Musk replied to broad online fanfare.

Since then, the company has cast the actions in part as a charitable gesture. “I’m proud that we were able to provide the terminals to folks in Ukraine,” SpaceX president Gwynne Shotwell [Emphasis in original] said at a public event last month, later telling CNBC, “I don’t think the U.S. has given us any money to give terminals to the Ukraine.”

But according to documents obtained by The Technology 202, the U.S. federal government is in fact paying millions of dollars for a significant portion of the equipment and for the transportation costs to get it to Ukraine. [Emphasis in original]

On Tuesday, the United States Agency for International Development (USAID) announced it has purchased more than 1,330 terminals from SpaceX to send to Ukraine, while the company donated nearly 3,670 terminals and the Internet service itself.

While the agency initially called it a “private sector donation valued at roughly $10 million,” it did not specify how much it is contributing for the equipment or for the cost of transportation.

Sometime after the announcement, the agency removed key details from its release. It now states that USAID “has delivered 5,000 Starlink Terminals” to Ukraine “through a public-private partnership” with SpaceX but does not specify the quantity nor value of the donations.

USAID agreed to purchase closer to 1,500 standard Starlink terminals for $1,500 apiece and to pay an additional $800,000 for transportation costs, documents show, adding up to over $3 million in taxpayer dollars paid to SpaceX for the equipment sent to Ukraine. [Emphasis in original]

In a letter to SpaceX last month outlining the deal, the USAID mission director to Ukraine said the terminals would be “procured” and sent on behalf of USAID by a third-party contractor, which would “arrange for transportation and delivery of the equipment” from Los Angeles International Airport to Ukraine via Poland.

The letter said the nearly 3,670 terminals donated by SpaceX would come with three months of “unlimited data.” In addition to the more than 1,330 terminals that USAID confirmed it had purchased, the agency earlier agreed to buy a separate 175 units from SpaceX, according to the documents.

...

It is also unclear whether the price the U.S. government is paying for individual Starlink units matches their typical market price. [Emphasis in original]

USAID is paying $1,500 for each standard terminal and the accompanying service, documents show. According to the Starlink website, a standard terminal set costs $600, while the monthly service charge costs $110, plus an additional $100 for shipping and handling.

There are quite a few unanswered questions, but we can be reasonably certain that Musk got a ton of great PR for himself and Starlink that was at least partially paid for by taxpayers.

Now back to China. Here's Smith again.

Then there’s Jeff Bezos’ odd allegation that Musk would do the bidding of China’s government:

Interesting question. Did the Chinese government just gain a bit of leverage over the town square? https://t.co/jTiEnabP6T

— Jeff Bezos (@JeffBezos) April 25, 2022

This doesn’t make a lot of sense even in the narrow sense of business incentives, since China’s government is supporting a host of Tesla competitors. Some Western business executives over the years have fallen all over themselves to do favors for China’s government in exchange for promises of market access, only to see themselves muscled out by state-sponsored competitors after giving China’s leaders what they want; Elon seems way too smart to fall for that old trick. But on top of that, as I mentioned above, Musk seems committed to the defense of liberal countries against authoritarian ones. Nothing kills free speech faster than when the tanks roll in.

("Seems" is doing a lot of heavy lifting.)

Perhaps the Amazon founder was thinking of this. From the Guardian:

Tesla has opened a new showroom in the capital of Xinjiang, a region at the heart of years-long campaign by Chinese authorities of repression and assimilation against the Uyghur people.

...

The US has enacted a range of sanctions and regulatory and other measures against China over its continuing human rights abuses in Xinjiang, including restrictions on US business dealings with local operators and suppliers.

President Joe Biden last month signed the Uyghur Forced Labor Prevention Act, and the US government intends to conduct a diplomatic boycott of the upcoming Beijing Winter Olympics.

Uyghur rights groups criticised the opening of the showroom, reportedly Tesla’s 211th in China. The Council on American-Islamic Relations urged its immediate closure, and the cessation of what it alleged “amounts to economic support for genocide”.

Human Rights Watch’s Australia researcher, Sophie McNeill, said: “Beijing and businesses have long banked on a global willingness to put profits ahead of human rights, even in the face of crimes against humanity, but we must not allow this to continue in 2022.

“Elon Musk and his Tesla executives need to consider human rights in Xinjiang or risk being complicit.”

Tesla has been contacted for comment.

Tesla’s decision drew some support on Chinese social media, and followed revelations a week earlier that US tech company Intel had requested suppliers not to source goods, services, or labour from the region.

One commenter welcomed Tesla’s support for “the development and construction of Xinjiang, unlike some other companies”, an apparent reference to multinationals seeking to reduce business links with Xinjiang over the rights abuses. 

Back to Smith:

This also applies to an even more damaging variant of “cancellation” — the ability of powerful Twitter users to sic mobs on people in real life. The most famous example of this is when Twitter activist Shaun King posted the photo of a man whom he falsely accused of the murder of a young girl. The true culprits were later arrested, but the man King falsely accused suffered extensive real-life harassment. Yet the company refused to crack down. Musk has a chance to address this as well.

Yes, we can all agree that it's wrong for "powerful Twitter users to sic mobs on people in real life." Here is Felix Salmon with an example. 

With more than 22 million followers, Elon Musk knows exactly what happens when he uses his enormous Twitter bully pulpit to bully female journalists. Science writer Erin Biba, for one, has made that abundantly clear, with the story of what happened when he merely replied to one of her tweets.

Which is to say: If you don’t have a strong stomach, don’t look at Business Insider reporter Linette Lopez’s @-replies right now. Musk did much more than just reply to one of her tweets: He has gone on a veritable Twitter rampage aimed at her. 

...

This is worse than just stalking: Musk is setting his army of fanboys loose on Lopez, he’s retweeting stuff they find, and he’s encouraging them every step of the way. Milo Yiannopoulos was banned from Twitter for setting mobs upon his enemies; Musk should be banned too, but won’t be.

Musk’s harassment of Lopez is obsessive and deranged, to the point at which it should worry every shareholder of any company where he serves as CEO. But since even former journalists seem to think that somewhere in the madness there’s a legitimate beef, let’s put that idea to rest. Lopez has been reporting aggressively on Tesla for a while; her sources include Tesla whistleblower Martin Tripp, whom Musk considers a saboteur for talking to the press. Lopez has also, in the past, written about Jim Chanos, a dogged investor who is shorting Tesla stock.

One area where Musk does have relevant experience in is bots.

Russ Mitchell writing for the LA Times:

In early November 2013, the news wasn’t looking great for Tesla. A series of reports had documented instances of Tesla Model S sedans catching on fire, causing the electric carmaker’s share price to tumble.

Then, on the evening of Nov. 7, within a span of 75 minutes, eight automated Twitter accounts came to life and began publishing positive sentiments about Tesla. Over the next seven years, they would post more than 30,000 such tweets.

With more than 500 million tweets sent per day across the network, that output represents a drop in the ocean. But preliminary research from David A. Kirsch, a professor at the University of Maryland’s Robert H. Smith School of Business, concludes that activity of this sort by so-called bots has played a significant part in the “stock of the future” narrative that has propelled Tesla’s market value to altitudes loftier than any traditional financial analysis could justify.

Smith also assures us that we shouldn't worry about Musk lifting the ban on the former president since "Trump has stated that he has no intention to return" and if you can't trust DJT... Still, it is useful to actually follow Smith's link.

“I hope Elon buys Twitter because he’ll make improvements to it and he is a good man, but I am going to be staying on TRUTH,” Trump’s statement read.

  But what if Truth Social goes away?

 Chris Cillizza writing for CNN.

Trump, ever the opportunist, announced in the fall of last year that he was starting Truth Social, a rival social media company that would be a free speech haven – unlike Twitter.

Except, well, Truth Social has been sort of irrelevant since it launched in February. Trump himself, the man around whom the entire operation was built, has so far sent one tweet Truth. And the special purpose acquisition company tied to Truth Social has lost 44% of its value since Musk first disclosed his stake in Twitter.

...

Trump insisted to Fox on Monday that nothing had changed.

“I am not going on Twitter, I am going to stay on TRUTH,” Trump said. “I hope Elon buys Twitter because he’ll make improvements to it and he is a good man, but I am going to be staying on TRUTH.”

But with Musk now in charge, the likelihood of Trump getting his Twitter account back just went WAY up. And if he is ultimately reinstated, he would get back the 88 million followers he had on Twitter – and the instant feedback that he grew so addicted to over the last few years.

(Trump has suggested he would not have been elected president without Twitter. “When somebody says something about me, I am able to go bing, bing, bing and I take care of it,” he said in 2017.)

If you believe that Trump would somehow be able to resist the lure of Twitter if Musk allowed his account to be reinstated, then you haven’t been paying attention for the last seven years.

And with that I'm afraid Smith has broken me. There's more in the piece that need context but it's late and I'm too tired to finish the job. Before I go, however, I will leave you with this link to an article by the excellent Lora Kolodny who demolishes the notion that Elon Musk has any interest in any free speech other than his own. 

Transitions in a democracy

This is Joseph.

Oliver Dowden is the chairman of the UK conservative party and a member of parliament. In the middle of a series of scandals with the current UK Prime Minister (Boris Johnson) he apparently made this quote:

Maybe I am not the target audience. And it is true that there is a lot going on in European politics at the moment that might benefit from consistent policy. But France has managed to have a successful election and it was far from the end of the world. 

The real issue here is the authoritarian impulse here. There are many reasons not to change leaders at a specific point in time. Perhaps the leader is widely popular and is fulfilling a mandate successfully. Or they have the support of the wide majority of the party who agrees with the policies. 

But something as vague as the national interest is an argument so vague it suggests never changing leaders. The whole point of democracy is short term instability and change to ensure that there is a method to obtain popular legitimacy. Governments in the UK have changed at many points of major turmoil -- July 1945 anybody? Or the US had an actual election during a civil war.

Leadership change is a natural part of the democratic process and it is never great when the normal process of transition is disrupted or disputed. We just need to be robust enough as a democratic society to handle transitions of power when it is called for (varies between a presidential and parliamentary system) and to have a strong civil service and judiciary to ensure consistency of the  legal structure (including changes in the legal structure) to ensure that business has a favorable regulatory environment to allow for prosperity. 

Strauss crosses parties

This is Joseph

Well, I guess this was inevitable. Strauss is the modern incarnation of the "noble lie" or the idea that people can be better off if they are misinformed to nudge them towards "correct" behavior.

This is an example from the US government under the Democrats. Here is the reason that is being given as to why the current administration (via the FDA) plan to delay the Moderna vaccine for children under 5 years of age for a few weeks:

The question, however, is whether regulators will authorize Moderna a few weeks ahead of Pfizer, or wait to authorize them at the same time — in June — in an effort to minimize confusion.

In an interview on Thursday with CNN+, White House medical advisor Dr. Anthony Fauci suggested that regulators will do them both at the same time.

Fauci said that the purpose of authorizing both at once, rather than Moderna a few weeks ahead of Pfizer, would be to not "confuse people."

"It's going to be two separate companies, two products that are similar but not identical, particularly with regard to the dose,” he said, explaining that Moderna's regimen for kids under five is two doses, while Pfizer's is three doses.

In case you think that there was confusion on the part of the reporters:

“And what the FDA wants to do is to get it so that we don't confuse people,” Fauci said.

Asked multiple times by CNN’s Kasie Hunt why parents couldn’t be expected to navigate having two options separately authorized around the same time, particularly for those who are eagerly waiting for a vaccine as soon as possible and would take whichever came first, Fauci said he couldn’t answer.

“I can't really honestly answer that question because I don't know the answer to that question. Because I don't have all the data in front of me,” Fauci said, adding that the data before the FDA was confidential while it was under review.

 Now, nobody is seriously arguing for a vaccine mandate in this age group. Only a quarter of kids aged five to eleven have been vaccinated since the vaccine was made available. Arguing that the data is confidential is fine but why is there a delay? Does the US government not trust individual parents and medical providers to be able to understand that different products exist for different age groups? 

We already deal with this type of nuance for products like aspirin, which has age restrictions on use in children as a painkiller, unlike ibuprofen and acetaminophen which do not. That is for a product that is commonly used as an ingredient in other medications and which is given to relieve symptoms at home, not at a clinic or pharmacy with a trained medical professional to explain details. 

There are a lot of threads asking why this might be on Twitter. But, honestly, the infantilizing response of "reduce confusion" seems so implausible that you feel you must be being misled. Or that the government has a very, very low opinion of the cognitive skills of the average clinician or parent. Because we had a period where only Pfizer was approved for teenagers and many of us were asked to get a Moderna booster to save Pfizer supply for these people. So we have actually already had a successful roll-out of different age eligibility for different vaccine products. 

What makes this so annoying is that it is a pure unforced error that reduces government credibility. Far better to say something like "I don't feel able to comment until the FDA reviews the submission data". Arrogant, yes. But it doesn't feel like either an obvious falsehood or an evaluation of the intellect of the average American parent that was . . . low.

Passport delays

This is Joseph.

The Canadian government is currently dealing with a backlog in passport applications. There is a claim that this backlog is because people put off their passport renewal. Fortunately, the article allows us to assess this claim.  So let us look at passports issued by Service Canada over the past three years:

April 1, 2019 to March 31, 2020: 2,300,000

April 1, 2020 to March 31, 2021: 363,000 

April 1, 2021 to March 31, 2022: 1,273,000

What did Service Canada have to say about this:

"Service Canada recognizes that an increase in demand for passport services has resulted in lineups and longer wait times for service, and we understand the concern this may cause for clients," the spokesperson said

Now there might be an explanation for the slow service times that is credible. But it seems extremely odd that any competent agency would think that the 84% drop in applications during the peak of the pandemic was, in any way, likely to be a good prediction for the likely future volumes of applications. Imagine if there actually was a "catch-up" increase in applications to above the previous volume of applications. 

Now, it is possible that this delay in service is a consequence of pandemic precautions that make service provision less efficient. But then that should be a separate discussion about the costs versus benefits of pandemic precautions. This seems to be what the website says:

But this is now two years after the beginning of the pandemic? Are our governments so sclerotic that they can no adapt to changed circumstances with years of planning. These are, after all, documents that need to be processed. They have also increased the need for a passport (e.g., for children re-entering Canada) and there should also be consideration of this when planning policy -- are you able to provide the services to support the new policy.

In general, a lot of services have shown fragility instead of resilience in the face of the pandemic and it might be time to consider the trade-offs involved in the provision of essential services. 

Tucker Carlson and Susan Sontag

 

A physically idealized man stands, arms stretched out, head thrown back to the heavens, his genitals obscured by a mysterious primordial glow while "Also sprach Zarathustra" pounds in the background. What does this suggest?

This is so gay. https://t.co/KPNCog3y9I

— George Takei (@GeorgeTakei) April 16, 2022

Well, sure. I wouldn't phrase it that way, but I wouldn't disagree with the underlying sentiment, nor would most of the commenters on Twitter. Beyond that though, there's something familiar about this montage of idealized Nordic men working and wrestling in a bucolic setting, juxtaposed with portentous mystical imagery, Nietzsche references, and a voiceover talking about chaos and social collapse leading to the rise of superior men who bring a new order. 

On a completely unrelated note, I've been thinking about the essay "Fascinating Fascism" by Susan Sontag.

Here are some excerpts:

Admittedly, if The Last of the Nuba were not signed by Leni Riefenstahl one would not necessarily suspect that these photographs had been taken by the most interesting, talented, and effective artist of the Nazi regime. Most people who leaf through The Last of the Nuba will probably look at the pictures as one more lament for vanishing primitives, of which the greatest example is Lévi-Strauss on the Bororo Indians in Brazil in Tristes Tropiques. But if the photographs are examined carefully, in conjunction with the lengthy text written by Riefenstahl, it becomes clear that they are continuous with her Nazi work.

Riefenstahl’s choice of photographic subject—this tribe and not another—expresses a very particular slant. She interprets the Nuba as a mystical people with an extraordinarily developed artistic sense (one of the few possessions which everyone owns is a lyre). They are all beautiful (Nuba men, Riefenstahl notes, “have an athletic build rare in any other African tribe”); although they have to work hard to survive in the unhospitable desert (they are cattle herders and hunters), she insists that their principal activity is ceremonial. The Last of the Nuba is about a primitivist ideal: a portrait of a people subsisting untouched by “civilization,” in a pure harmony with their environment.

All four of Riefenstahl’s commissioned Nazi films—whether about Party congresses, the Wehrmacht, or athletes—celebrate the rebirth of the body and of community, mediated through the worship of an irresistible leader. They follow directly from the films of Fanck in which she acted and from her own The Blue Light. The fictional mountain films are tales of longing for high places, of the challenge and ordeal of the elemental, the primitive; the Nazi films are epics of achieved community, in which triumph over everyday reality is achieved by ecstatic self-control and submission. The Last of the Nuba, an elegy for the soon-to-be-extinguished beauty and mystic powers of primitives, can be seen as the third in Riefenstahl’s triptych of fascist visuals.

In the first panel, the mountain films, heavily dressed people strain upward to prove themselves in the purity of the cold; vitality is identified with physical ordeal. Middle panel, the films made for the Nazi government: Triumph of the Will uses overpopulated wide shots of massed figures alternating with close-ups that isolate a single passion, a single perfect submission; clean-cut people in uniforms group and regroup, as if seeking the right choreography to express their ecstatic fealty. In Olympiad, the richest visually of all her films, one straining scantily clad figure after another seeks the ecstasy of victory, cheered on by ranks of compatriots in the stands, all under the still gaze of the benign Super-Spectator, Hitler, whose presence in the stadium consecrates this effort. (Olympiad, which could as well have been entitled Triumph of the Will, emphasizes that there are no easy victories.) In the third panel, The Last of the Nuba, the stripped-down primitives, awaiting the final ordeal of their proud heroic community, their imminent extinction, frolic and pose in the hot clean desert.

It is Gotterdämmerung time. The important events in Nuba society are wrestling matches and funerals: vivid encounters of beautiful male bodies and death. The Nuba, as Riefenstahl interprets them, are a tribe of aesthetes. Like the henna-daubed Masai and the so-called Mudmen of New Guinea, the Nuba paint themselves for all important social and religious occasions, smearing on their bodies a white-gray ash which unmistakably suggests death. Riefenstahl claims to have arrived “just in time,” for in the few years since these photographs were taken the glorious Nuba have already started being corrupted by money, jobs, clothes. And, probably, by war—which Riefenstahl never mentions since she cares only about myth, not history. The civil war that has been tearing up that part of Sudan for a dozen years must have brought with it new technology and a lot of detritus.

Although the Nuba are black, not Aryan, Riefenstahl’s portrait of them is consistent with some of the larger themes of Nazi ideology: the contrast between the clean and the impure, the incorruptible and the defiled, the physical and the mental, the joyful and the critical. A principal accusation against the Jews within Nazi Germany was that they were urban, intellectual, bearers of a destructive, corrupting “critical spirit.” (The book bonfire of May, 1933, was launched with Goebbels’s cry: “The age of extreme Jewish intellectualism has now ended, and the success of the German revolution has again given the right of way to the German spirit.” And when Goebbels officially forbade art criticism in November, 1936, it was for having “typically Jewish traits of character”: putting the head over the heart, the individual over the community, intellect over feeling.) Now it is “civilization” itself that is the defiler.

What is distinctive about the fascist version of the old idea of the Noble Savage is its contempt for all that is reflective, critical, and pluralistic. In Riefenstahl’s casebook of primitive virtue, it is hardly the intricacy and subtlety of primitive myth, social organization, or thinking that are being extolled. She is especially enthusiastic about the ways the Nuba are exalted and unified by the physical ordeals of their wrestling matches, in which the “heaving and straining” Nuba men, “huge muscles bulging,” throw one another to the ground—fighting not for material prizes but “for the renewal of sacred vitality of the tribe.”

...

But fascist art has characteristics which show it to be, in part, a special variant of totalitarian art. The official art of countries like the Soviet Union and China is based on a utopian morality. Fascist art displays a utopian aesthetics—that of physical perfection. Painters and sculptors under the Nazis often depicted the nude, but they were forbidden to show any bodily imperfections. Their nudes look like pictures in male health magazines: pinups which are both sanctimoniously asexual and (in a technical sense) pornographic, for they have the perfection of a fantasy.

The first amendment

This is Joseph.

The first amendment is a very American idea and not one that is widely shared by other countries. It is a unique response to the repression of political speech at the time of the American Revolution and has resulted in an extremely open culture. You can sure have a modern democracy without it (see Canada) but it does have some advantages.

However, the consequences of free speech are, well, consequences. The freedom to speak is not the freedom to be without consequences. One place that this has really bit in is with pseudo-anonymous accounts, where the person shields themselves from the immediate consequences of their speech. This can be helpful -- the bold truth teller might lose access to inside information. But that has always come with the risk of being unmasked and facing consequences. Reality Winner, for example, was convicted of leaking information on Russian interference with the 2016 election, a matter of great public importance. 

More recently there has been controversy about a recent Washington Post article talking about the identity of a person who runs a major TikTok account:

The anonymous account’s impact is deep and far-reaching. Its content is amplified by high-profile media figures, politicians and right-wing influencers. Its tweets reach millions, with influence spreading far beyond its more than 648,000 Twitter followers. Libs of TikTok has become an agenda-setter in right-wing online discourse, and the content it surfaces shows a direct correlation with the recent push in legislation and rhetoric directly targeting the LGBTQ+ community.

A Twitter account with hundreds of thousands of followers is influential the article notes that the name is being trademarked as a news reporter service. This is public facing speech. 

Now, the irony here is that the same free speech that allows this platform to be so controversial is why exposing the person behind the account is ok. You may want a kinder public forum but it is odd to attack your political opponents and then be unwilling to stand behind your words. But this isn't a very useful perspective:

After all, the account exists to publicize offensive content, at least content offensive to a right wing perspective. One key thing in the Washington Post piece is that it is clear that specific people are identified in the account, which is fair play. But I think if we apply equal standards to the account then it is just applying the same principles consistently. 

This was a good ideas for the sort of journalistic principle involved here. Insofar as the person in question is a political operative, and the article makes a good case, there is a lot of room to consider this to be in the public interest. 

I think one could ask whether the first amendment should consider respect and kindness as possible features. But I understand why it does not. But if you are going to advocate for it, then the natural result of this type of "market place of ideas" is that you will need to accept the consequences of your speech, Free speech, without consequences, is privileged speech (like that of a king) and exactly what the amendment was intended to avoid. 

The Beast in the Jungle and the Next Big Thing

I've been thinking a lot recently about the Henry James novella The Beast in the Jungle. If you've never read it (or have read but forgotten it), here's the synopsis:

John Marcher, the protagonist, is reacquainted with May Bartram, a woman he knew ten years earlier while living in southern Italy, who remembers his odd secret: Marcher is seized with the belief that his life is to be defined by some catastrophic or spectacular event, lying in wait for him like a "beast in the jungle". May decides to buy a house in London with the money she inherited from a great aunt, and to spend her days with Marcher, curiously awaiting what fate has in store for him. Marcher is a hopeless fatalist, who believes that he is precluded from marrying so that he does not subject his wife to his "spectacular fate".

He takes May to the theater and invites her to an occasional dinner, but does not allow her to get close to him. As he sits idly by and allows the best years of his life to pass, he takes May down as well, until the denouement where he learns that the great misfortune of his life was to throw it away, and to ignore the love of a good woman, based upon his preposterous sense of foreboding.

What if it wasn't just one man but an entire culture that built its identity around an overwhelming but unfounded belief that it was destined to see its world upended by some great, unknown Next Big Thing?

Obviously the definitions we're using here are largely subjective and open to debate, but I’d argue that we haven't had an honest-to-god Next Big Thing since the 90s. There have certainly been technological advances that have changed our daily life, but very little that registers on the scale that we saw from, for example, 1876 to 1896, a period that included the telephone, the invention of recorded sound, Pasteur's vaccines, the light bulb, practical electric motors, modern electrical engineering, motion pictures, internal combustion, automobiles, the first successful engine-driven heavier-than-air (unmanned) flight, x-rays, the linotype and the birth of modern printing, wireless telecommunication, and probably a good dozen that I'm leaving out. (don't get me started on aluminum or Luther Burbank.) I could make a similar, albeit slightly less impressive list starting in 1945 and going through 1970.

Those were, of course, exceptional periods in human history but if we want to play with the big boys, that is the standard we need to meet.

Going back 25 years from today, the only thing that jumps out as a life-changing, paradigm-shifting Next Big Thing is the smartphone and just barely possibly social media. There is no question that the small 2001 monolith/Mother Box most of us carry in our pockets has had a tremendous impact on the way we live, but it is not all that recent and, more importantly, it is largely the result of combining the next big things of the 80's and 90's, the internet, cell phones, civilian GPS, digital photography. Likewise, social media feels like a small extension of existing tech that had big ramifications. 

Even if we allow the smartphone and social media, the introduction of the iPhone was 15 years ago and Facebook is older than that, meaning that you have to over thirty to have adult memories of what it's like for a piece of technology to radically change everyone's lives.

This doesn't necessarily suggest any kind of stagnation. The end of the 19th century and the post-war era should probably be treated as anomalies. What we are living through now is arguably fairly normal with steady incremental advances in most fields (cancer treatments, batteries, rockets, etc.) and a few big breakthroughs around a handful of specific problems (mRNA vaccines). 

Unfortunately, we have collectively conditioned ourselves to believe that's not enough, that we are or at least should be living in a period of radical and ubiquitous technological change proceeding along an already steep exponential curve. Like James's protagonist, we have convinced ourselves that we are destined to experience some kind of great and potentially terrible event, specifically in our case the kind of thing we read about in old science fiction stories. The result has been to make us all a little bit crazy and in some cases, gullible as hell.

Tales of the Crypto -- either slightly early or really late Tuesday tweets

Web3 writing has been breaking into three broad camps.

 
The first believes (or would like the marks to believe) that blockchains and the rest will usher in an age of tremendous benefits and innovation. The second can see how bad the web3 arguments are but still believes there must be a pony in there somewhere (NYT being the leading example). The third is coming to the Doctor Tarr and Professor Feather conclusion that the inmates are running the asylum.

The best arguments are coming from the third camp.

I worked at the SEC for 18+ yrs, the last 11 as Chief of the SEC Office of Internet Enforcement. I have taught cyber law at Georgetown and Duke Law Schools for 20 yrs. I spent 5 yrs at Stroz Friedberg fighting cyber crime. Why I believe the bulk of Web3 is both scourge & scam.🧵 pic.twitter.com/M273MUTUeE

— John Reed Stark (@JohnReedStark) April 1, 2022

And that's before we even begin to talk about the environmental damage, enabling of illicit financing, regulatory arbitrage, predatory inclusion, and a whole slew of negative externalities.

It is entirely rational to prima facie reject the entire premise of the crypto project.

— Stephen Diehl (@smdiehl) April 3, 2022

New from me and @ben_mckenzie: we took a look at Binance -- the world's largest crypto exchange, a money-printing machine with no headquarters and a nomadic CEO -- and why some users are suing it over a mysterious outage on May 19, 2021.https://t.co/e66wPMGEHx

— Jacob Silverman (@SilvermanJacob) April 1, 2022

Looking at Binance and other exchanges, one thing that's become apparent is how many potentially conflicting roles they play at once: they're marketplaces, shadow banks, crypto hedge funds, VCs, & other intermediary roles that would be done by a third party in a regulated market.

— Jacob Silverman (@SilvermanJacob) April 1, 2022

I meant to follow up our post on the NYT's embarrassing "The Latecomer's Guide to Crypto." Now I don't have to. This exhaustive and devastating paragraph-by-paragraph examination is the last word.

Holy fuck this is amazing.

"The (edited) Latecomer's Guide to Crypto" by @molly0xFFF et al https://t.co/8py4xDKNsC

— Cory Doctorow (@doctorow) March 25, 2022

Yep.

The description of "DAO = slush fund + 4chan poll" is pretty spot on.

— Stephen Diehl (@smdiehl) April 1, 2022

And this.

Weirdly, the #NFT analogy I never hear is the most accurate one:

"An NFT is like paying someone to affix a brass plaque bearing your name onto a piece of furniture, irrespective of whether own that piece of furniture."

— Cory Doctorow (@doctorow) March 22, 2022

For a good deep dive.

One of the most important projects I contributed to recently is an
thorough analysis of crypto from multiple ideologies and perspectives. Trying to steelman the strongest form of their view. (1/) 🧵

— Stephen Diehl (@smdiehl) March 31, 2022

And another.

Very smart critique of smart contracts. Thread. 👇 https://t.co/KFtezB3UXJ

— Stephen Diehl (@smdiehl) April 17, 2022

Because smart contracts replace the *under*-regulated and thus corrupt financial system with an *un*regulated system that has the same fallible humans at its core, and then makes all of their poor or corrupt choices irreversible.

— Cory Doctorow (@doctorow) April 17, 2022

The latest of many.

"The Axie Infinity hack, what happened, and why people keep talking about bridges"https://t.co/KdJQLtHNxT pic.twitter.com/jk58mh1ThH

— Molly White (@molly0xFFF) March 31, 2022

So in case the "digital serfdom" business model wasn't awful enough. Now the hacked funds are being used to fund the North Korean regime. I honestly didn't think this could get any worse, and yet here we are.https://t.co/NAkslcIkh7

— Stephen Diehl (@smdiehl) April 17, 2022

Say this about Ponzi, he never contributed to the destruction of a planet.

From @reveal: In Hardin, Montana, the coal-fired power plant was on the verge of shutting down until bitcoin came to town. But in just one year, the amount of carbon dioxide the plant puts into the air jumped nearly tenfold. https://t.co/t2EvSmRmZf

— Yellowstone Public Radio (@YPRadioNews) March 31, 2022

Stay in school kids, and don't do drugs. https://t.co/4qR8b9vz55

— Stephen Diehl (@smdiehl) April 17, 2022

And finally, our moment of schadenfreude. 

An NFT of Jack Dorsey's first-ever tweet, which sold for $2.9M last year, has been up for sale since last week and has failed to garner a bid over $1K (@iamsandali / CoinDesk)https://t.co/V1A2ooDjJ6https://t.co/yKLfhh07zL

— Techmeme (@Techmeme) April 13, 2022

Ten years ago at the blog -- since Gelman brought it up...

Andrew's recent post on plagiarism got me to thinking about this thread from back in the day.  

Saturday, April 21, 2012

Like complaining about saucy language in Sodom and Gomorrah

Here's an idea for a novel: in a dystopian future/alternate history, the country is governed by a totalitarian central government that forces teams of teenagers to battle to the death in an annual televised event. In the hands of competent writer it's a premise that could generate plenty of drama and suspense and it has highly cinematic elements.

I'll get back to that idea in a minute but first I want to direct your attention to this recent post by Andrew Gelman. Go ahead, take a look. I'll wait...

There are a number of things to discuss here but let's start with this assertion quoted by Gelman:

“The essence of plagiarism is passing off someone else’s work as your own."

This nicely catches the stark moral terms that we often see in this debate, but when look at this more closely, particularly when we look at what's entailed in different types of plagiarism and the reactions to those types, the picture is a bit murkier.

Let's go back to the idea from the top of the page and fantasy stories about young adults. Back in the mid-Nineties, J.K. Rowling came up with the inspired notion of combining the two great traditions of British juvenile literature. The concept and Rowling's skillful execution produced the enormously successful Harry Potter and the Philosopher's Stone.

Rowling's success was followed by a wave of science fiction and fantasy novels aimed at the young adult market. These included Percy Jackson, the Lorien Legacies (co-written by the disgraced James Frey), Gregor the Overlander, and, of course, Twilight and the Hunger Games.

But one thing Rowling's success didn't inspire was the idea I mentioned at the top. That one came from a Japanese writer who used it for a novel written in 1996 and published in 1999 under the name Battle Royale,

The book and the movie that followed a year later were huge international hits. Despite the somewhat disturbing subject matter, both generally received positive reviews. Here's the Guardian in 2001, "Some will find the explicit violence of this movie repulsive - or plain boring. But this is a film put together with remarkable confidence and flair. Its steely candour, and weird, passionate urgency make it compelling." And Stephen King, writing in Entertainment Weekly (February 1, 2007) gave the book an enthusiastic endorsement (while noting it had some elements in common with his novel The Long Walk).

A little bit more than a year and a half later, Scholastic published the Hunger Games.

Given the number of blogs by fans of science fiction and Japanese popular culture, it's not surprising that the resemblance was discussed at some length.

From Wikipedia:

The 2008 American young adult novel The Hunger Games by Suzanne Collins has been accused of being strikingly similar to Battle Royale in terms of the basic plot premise and the world within the book. While Collins maintains that she "had never heard of that book until her book was turned in", Susan Dominus of The New York Times reports that "the parallels are striking enough that Collins's work has been savaged on the blogosphere as a baldfaced ripoff," but argued that "there are enough possible sources for the plot line that the two authors might well have hit on the same basic setup independently."

That "might well have" is an awfully weak defense (particularly given the puff piece tone of the NYT article) and it points to one of the central problems in the plagiarism debate: while it's easy to prove the relatively trivial crime of lifting wording, it's next to impossible to prove more substantial thefts. We can look at the timeline. We can look at Collins' previous career as a writer of fairly derivative kids' shows (no Spongebob or Pete & Pete) and the author of the Underworld books (a series that bears a marked resemblance to Harry Potter). Nothing here gives us any reason to believe that she didn't steal the idea but also nothing that could be called evidence that she did.

This is not meant as an attack on Collins who is, as far as I can tell, an excellent writer and who is doing a wonderful job getting kids to read. I'm in favor of what she's doing and I couldn't care less how she does it.

My point is that the theft of wording -- a problem that is both trivial and rare, but easy to prove -- is treated as a major offence while stealing more substantial elements -- a problem that is both serious and common, but is hard to prove -- is largely ignored.

If we truly want to embrace the inclusive definition of plagiarism, we quickly find ourselves in the uncomfortable position of pointing out the extensive lapses of friends and colleagues rather than the failings of a few convenient pariahs.

If we're going to be anywhere near consistent and proportional, we're going to have to ask ourselves whose names really belong on a research paper. I can think of at least one case where the credit was given to someone who happened to be the spouse of the main researcher's thesis advisor (the valid reasons for being listed as an author do not include marrying well). If you didn't substantially contribute to the research behind or the writing of a paper and you put your name to it, you're a plagiarist.

And we need to ask ourselves how much journalism consists of simply paraphrasing and regurgitating other people's ideas, arguments and interpretations. When you hear someone talking about a meme, they actually mean that stories are being borrowed and recycled on a massive scale.

Discouraging plagiarism in the broad sense is a worthy goal, but focusing exclusively on those few people who lift some phrases from other published work is simply a distraction.

The shadow docket

This is Joseph.

I think that this is a very big deal:

Chief Justice Roberts voted with Justices Breyer, Sotomayor and Kagan in dissenting from six previous shadow docket rulings. But the Clean Water Act dispute was the first time he joined in the procedural criticism that the other conservatives were not just using the shadow docket but abusing it. In that respect, his rebuke cannot be dismissed as partisan. By publicly endorsing the charge that the conservative justices are short-circuiting ordinary procedures to reach their desired results without sufficient explanation, Chief Justice Roberts provided a powerful counter to defenders of the court’s behavior. Justice Samuel Alito, for instance, claimed in a September 2021 speech that critics of these rulings are acting in bad faith because their real objections are to the results in these cases.

What is especially telling about Chief Justice Roberts’s dissents in these shadow docket cases is that, unlike Justices Breyer, Sotomayor and Kagan, he’s often been sympathetic to the results. In February’s Alabama redistricting ruling, for instance, Chief Justice Roberts agreed that the court should reconsider the interpretation of the Voting Rights Act under which Alabama’s maps had been struck down; he just believed that any change in that interpretation had to come through the merits docket, not the shadow docket.

Courts have to survive on credibility and this sort of aggressive intervention is just the opposite of that. It is a change in how SCOTUS typically operates (it has been >200 years, I am sure other bad decisions abound) and it has a pretty big impact on the ability of people to predict the rule of law.

One of the classic libertarian views that I thought was a very good point was that a strong and predictable set of laws has actual market value. Business likes a lack of political and legal uncertainty as it makes it easier to run a profitable business. Mark Palko has already talked about another example of this with Disney and Florida, but that is only one state. Plus, if SCOTUS is operating well then it is likely that any such measure might run into trouble upon appeal. 

Now there is a lot of ruin in a court, to follow the wisdom of Adam Smith. But this seems to be an unforced error of shredding credibility. Elections have consequences and I was sure not all that excited by some government policies. But the ability for government to govern is a key piece of democratic rule, even if we think that the rules are sub-optimal. 

Like a lot of bad ideas, no one ruling is too much and mistakes/miscalculations are a part of a vibrant and living court system. But like things like presidential pardons, what can be an important tool for complicated situations can start being a problem if widely over-used. 

Self-organizing fraud (or plain old fraud)?

Another fine piece by Russ, that will of course immediately get flooded with hate from bot accounts. Note that neither Russ nor the Researchers cited point the blame at Tesla.

This is important because the existence of $TSLA Bots is clearly a real thing. Who's behind it all??? https://t.co/fEJ84YD4hq

— ESG Hound (@ESGhound) April 12, 2022

Russ Mitchell writing for the LA Times:

Whether Twitter bots are being deliberately programmed to manipulate stock trading is among the questions that Kirsch and his research assistant, Moshen Chowdhury, are trying to answer.

...

A Twitter bot is a fake account, programmed to scour the social media site for specific posts or news content — Musk’s posts, for example — and respond with relevant, preprogrammed tweets: “Tremendous long term growth prospects” or “Why Tesla stock is rallying today” or “Tesla’s Delivery Miss Was ‘Meaningless.’” The bots can also be programmed to send nasty or threatening messages to company critics.

Kirsch and Chowdhury collected and reviewed Tesla-related tweets from 2010, when the company went public, to the end of 2020.

Over that period, Tesla lost an accumulated $5.7 billion, even as its stock soared and Musk became one of the richest humans on the planet; his net worth is estimated at $275 billion. Operational results can’t justify anything close to the company’s $1-trillion market value, based on any kind of traditional stock-pricing metric.

...

Using a software program called Botometer that social media researchers use to distinguish bot accounts from human accounts, the pair found that a fifth of the volume of tweets about Tesla were bot-generated. That’s not out of line with giants like Amazon and Apple, but their bots tended to push the stock market and tech stocks in general, with those companies as leaders, but not focus on any particular narrative about the companies.

While any direct link between bot tweets and stock prices has yet to be determined, the researchers found enough “smoke” to keep their project going.

Over the 10-year study period, of about 1.4 million tweets from the top 400 accounts posting to the “cashtag” $TSLA, 10% were produced by bots. Of 157,000 tweets posted to the hashtag #TSLA, 23% were from bots, the research showed.

...

The researchers are looking at the timing of the tweets and options activity in the overnight stock market, among other factors. One big unknown: whether the bots are the work of entities with a direct financial interest in Tesla.

I'm out of my area of expertise here but here are a few points that worry me about all this. 

1. Though we don't know the exact magnitude, we can be reasonably sure that bots are a relatively small part of the concerns with Tesla's valuation and Tesla is a relatively small part of the concern with bots. That does not mean that this is a trivial issue. When you have problems this big, even a small portion can still be pretty damned large in absolute terms. 

2. Bots and the larger problem of fake accounts and falsified metrics are not by any means limited to Twitter. The company formerly known as Facebook has a huge problem with fake accounts and fraudulent ad metrics. I'm sure there are similar issues with Snapchat. Even Tik-Tok where it's hard to imagine a big bot presence certainly has issues with cooking data such as viewership numbers. 

3. Social media metrics have gotten to be a huge part of data-driven systems and I assume social science research. I know from the business side they come up all the time. Our metrics, models, and algorithms are all vulnerable to these kinds of fraud. The big question at this point is how bad the problem has gotten. 

4. I wonder if our mental models for thinking about the next generation of fraud are sufficiently complex and open to unknown unknowns. In this case, my guess is that these bots come from multiple sources. Possibly Tesla or people working covertly but directly with the company. Investors looking to pump and better time the spikes. Fan boys wanting to amplify their displays of love and loyalty for Elon and their hatred of his enemies (particularly when those enemies are female journalists). We could easily see something that looks like coordination but without any direct communication, and as bots grow more sophisticated, their behavior will grow more complex and presumably more difficult to predict and detect. Which brings us to... 

5. Bots are a moving target.

Botometer, by its design, will always be a step behind the evolution of bots. I encourage @darchivist to:
a) Take a look at the bots on the $TSLAQ block list
b) Examine how Elon Musk/#Tesla/#SpaceX pays real people in India and Eastern Europe to pump his narrative
2/2

— Claire Musk (@ClaireMusk) April 12, 2022

Housing cost inflation

This is Joseph.

Mike Moffat posted this interesting graph:

The clear loser was Winnipeg, going from a little over $100K to a little over$ 300K, in terms of appreciation but the increase in inflation for the time period is $100 going to $134.61. So even the coldest Canadian market is briskly outpacing inflation. But Southern Ontario went from around $150K to over $800K. It seems odd that housing has exploded in value by so much across so many markets. None of them are even remotely close to the inflation rate, far exceeding it.

But the inability to secure affordable housing has numerous downstream social consequences -- from increased homelessness, inability to escape domestic violence, and financial instability. 

And this is not just a Canada problem. Look at housing prices in Seattle: in June of 2005 the average sale price was around $250K whereas in Jan 2022 it was around $800K and by Feb it was $814K (here the peak was before this at $870K). US inflation over this period was about 45% so you are seeing more than a doubling of real costs in the real estate market.

This means that we should think carefully about this from a policy perspective, keeping in mind that feeding a bubble is a bad idea. 

EDIT: A typo found by clever commenter DJL has been corrected. The correction is in bold

Megan McArdle, Disney, and the defending the indefensible adjacent.

One of Megan McArdle's go-to moves is to take some truly reprehensible Republican policy and come up with an argument that while it doesn't really make the same policy and it comes at the question from a completely different set of assumptions and justifications, is something kinda, sorta in the general ballpark which reasonable people might be able to accept. A bit like David Brooks but more hot-take centric. 

Unless you've done the sensible thing and stopped following politics, you've probably heard about this.

1. You don't know the difference between a trademark and a copyright.

2. It's (pretty famously) M-I-C-K-***E***-Y M-O-U-S-E https://t.co/sCwkXaxg00

— Cory Doctorow (@doctorow) April 4, 2022

Banks: If they’re going to attack the American laws… they don’t deserve the special laws and protections.. I wrote Disney today and said “when your copyright laws expire… I’m not gonna vote for it” pic.twitter.com/gobaTdJrHr

— Acyn (@Acyn) April 1, 2022

A political party promising to rewrite laws and regulations to punish a company for criticizing the party is about as blatant as corruption gets. With any kind of honest framing this is truly indefensible.

With that in mind, check out Megan Mcardle's comment.

Republicans should threaten to let Disney's Mickey Mouse copyright lapse because it's an inexcusable giveaway to big business, with no economic benefit to anyone except Disney. https://t.co/EjDC6dxzev

— Megan McArdle (@asymmetricinfo) April 7, 2022

At the risk of putting too fine a point on it, we are talking about the most naked example of political extortion we've seen since the Trump Administration (I know that was only two years ago but they set the bar really high). The Republicans have loudly and openly spelled out that these threats are payback for Disney's (constitutionally protected) criticisms of Republican policies.

McArdle is absolutely correct to suggest that the endless extensions of copyrights to protect the intellectual property of companies like Disney and Warner Brothers has been a horrible example of regulatory capture. Lots of people have been making this point from much larger platforms than ours including Dean Baker, Adam Conover, and Cory Doctorow, but the Republicans have clearly stated their reasons for this and they have nothing to with the greater economic good.

Megan McArdle doesn't defend the GOP extortion scheme here, but she very clearly takes the edge off of it. It's a bit like hearing a news story about firemen refusing to hook the hoses until they get a pay-off and responding that putting out fires is a good thing. 

When the methods make you worried

This is Joseph

Thomas Lumley has a great piece on a recently published article. Go read that first, we'll still be here. In it he looks at a paper that is an ecological study of cannabis and cancer. Thomas probably got the most interesting parts, but I want to focus on two things that are in a professional paper.

In Table 6 we are presented with 3 measures of relative risk (presumably of cannabis use and prostate cancer). Relative risk is not mentioned in the text of the paper and is conventionally defined as:

Relative risk is a ratio of the probability of an event occurring in the exposed group versus the probability of the event occurring in the non-exposed group. 

The three relative risks proposed are: 1.25,1.66E+ 52, and 5.65E+ 05. No, the second two are not jokes. A relative risk of 1.25 is plausible, if hard to prove with a tricky outcome like cancer. But something went quite wrong with the other two. It might have been a hint when this was in the acknowledgements:

We wish to acknowledge with grateful thanks the work of Professor Mark Stevenson in modifying and enlarging the capacity of “epiR” to handle the enormous integers encountered in this study. His prompt and timely assistance is greatly appreciated indeed.

 The other issue is with E-values. An E-value is defined as: 

The E-value is defined as the minimum strength of association, on the risk ratio scale, that an unmeasured confounder would need to have with both the treatment and the outcome to fully explain away a specific treatment-outcome association, conditional on the measured covariates. A large E-value implies that considerable unmeasured confounding would be needed to explain away an effect estimate. A small E-value implies little unmeasured confounding would be needed to explain away an effect estimate. 

But the authors slide this into other forms of bias in a rather interesting way:

Furthermore this was also an ecological study. It is therefore potentially susceptible to the short-comings typical of ecological studies including the ecological fallacy and selection and information biases. Within the present paper we began to address these issues with the use of E-values in all Tables.

Now, E-values are quite useful for thinking about confounding but I am quite interested to see how they apply to things like selection bias, except insofar as a larger estimate requires more substantive bias to be entirely artifactual due to selection bias. The other stuff, about biological plausibility, is fine and that is the reason to have conducted the study. But the E-values say nothing about bias other than how strong a set of confounders would need to be to induce a spurious association. 

Also this is not quite true:

Causal inference was addressed in two ways. Firstly inverse probability weighting (IPW) was conducted on all mixed effects, robust and panel models which had the effect of equilibrating exposure across all observed groups. IPW were computed from the R-package “ipw”. Inverse probability weighting transforms an observational dataset into a pseudo-randomized dataset so that it becomes appropriate to draw inferences as to truly causal relationships.

This is only true under a set of assumptions,  

One, there is no unmeasured confounding present. Two, that the marginal structural model is correctly specified (both the marginal structural model and   the   model   for   exposure).   Three,   that   each participant’s    counterfactual    outcome    under    their observed   exposure   is   the   same  as   their   observed outcome  (consistency).  Finally,  we  need  to  assume positivity — that   there   are   exposed   and   unexposed participants  at  all  levels  of  the  confounders.  If  these assumptions are not met, then the marginal structural model may give misleading estimate

IPW can be a powerful and useful technique, but it is hardly magic and cannot replace randomized data in causal inference. The use of E-values immediately casts doubt on the first assumption being fully realistic.  

Now there may be some important information in this ecological study, but I find these epidemiological methods issues really detract from the potentially good science elsewhere in the paper.

Something offbeat: subverting expectations

This is Joseph.

Back in the day, there was a lot of discussion about how shows like Game of Thrones felt refreshing because they subverted expectations. Part of this was a willingness to reject literary tropes like plot armor that allowed for important and beloved characters to die. But this technique is a masterclass one, and hard to get right. It requires foreshadowing so that the reader or watcher can look back, afterwards, and feel like the resolution was expected and they'd have guessed it if they'd paid closer attention.

A good example is the Mountain and the Viper. Part of what makes the ending (spoilers for a 8 year old show follow) work is that it is so carefully set up. Bronn actually goes through, beat by beat, the way that one could win and the vulnerability to one mistake. The Viper uses an innovative method of attack that makes sense as being high risk, high reward. Tyrion even comments on the lack of a helmet, which ends up making a big difference in the event. The flaws that cause the Viper to lose are set up for many episodes and are the reason he volunteered for the fight. We end up with conflicting expectations: it is possible that the Viper could win (and he comes close) but it is clear from the beginning that the actual outcome is the most likely.

Clever foreshadowing of how it could all fall apart makes the twist make sense. We know the stakes and it is made clear how this could all come together. 

Now consider how different the killing of the Night King is. The foreshadowing is weak and the early part of the episode shows Arya struggling to hide from Zombies, who are hyper-aware. So when she ends up sneaking past a horde there is no real foreshadowing for what is about to happen. It's actually worse -- previously when a character got themselves in a deadly situation, adversaries did not suddenly get slow or incompetent unless that was also set up. 

A tragedy is made good by the failure of the protagonist to overcome known flaws (Hamlet's lack of decisiveness, for example, leads to disaster). By highlighting these personality traits, it makes sense when they end up coming back to haunt the character. 

This also works the same way for clever victories. They need to be set up in a way that we can all revel in the cleverness of the character. A pretty decent example is Die Hard, where the character is trying to alert authorities to the presence of terrorists. The blocks set up by the antagonists show them as competent and the final solution is precisely telegraphed by previous efforts by the protagonist. It feels earned. 

But the worst way of subverting expectations is to just do something unexpected because it contradicts the previous evidence and character development. There was a comic Armageddon 2001 where the ending was leaked. So they switched the character who was the secret villain. To one of the only characters seen alive in the future and fighting the villain. We even see that over a multitude of futures that this character fights the secret antagonist but never becomes them. They created surprise, because one of the only 2 characters actually ruled out by the investigation as definitely not the secret villain was the secret villain. Yes, we were surprised. No, it was not a good surprise.

I think that the key to this sort of "bait and switch" is to make conflicting promises. By following a trope you induce expectations in the reader. But by clever foreshadowing and a set of different promises (e.g., we are going to be realistic even when it hurts) then you end up with the reader being both surprised but accepting that the switch was done well. 

Surprise is easy. Smart surprise in storytelling is very hard and it is often better to just be conventional then to be simply bad.