- The known unknowns of Covid-19
- The blue checkmark correlates with not distinguishing correlation from causation. (or does it actually cause the ignorance?)
- Someone at DHS hasn’t thought through the second order effects of their stupendously inept J-1 policy change.
- The best essay I read this week was written in 1944.
- Zombie starfish and mutant crayfish. Halloween must be near.
- The true value of Twitter: one misspelled, 2-like post can lead you down some magnificent rabbit holes. Bonus: An introduction to mathematical oncology, which is part of this JCO compilation.
- Median time from application submission to drug approval was 3.3 months (range 0.4 to 5.9 months) with real-time oncology review. Hard to get much faster than that, so why do some economists still want to speed up review?
- Another top physician-scientist leaves academia for industry, and announces it in style.
- Nassim Taleb doesn’t know that DOs and MDs are nowadays functionally equivalent but Twitter quickly sets him straight.
- Thread of the week: Sarajevo assassination 1914 — what really happened.
It’s good to keep a list, else you forget there’s more to Twitter than outraged mobs misunderstanding sarcasm.
- More tweets = more citations (says more about the academic enterprise than about Twitter if you ask me).
- Vincent Rajkumar on the trials (haha) and tribulations of starting a randomized clinical trial in the US.
- Should we add courses on dealing with medical insurance to the medical school curriculum? No.
- The no-meeting crowd is good company to be in (as long as we don’t meet!)
- World’s prettiest libraries (with only one clunker on the list).
These 250 pages on the many ways that cancer care in America is broken should be read by everyone with even a passing interest in oncology, and must be read by every heme/onc fellow or fellow-to-be. Malignant reminded me of the best days of my own fellowship, when the then-program director Tito Fojo would eviscerate an article — these tended to be poorly thought out phase 3 trials of one TKI or another that somehow made it to the New England Journal — with a few slides made at the last minute.1 But this is not just a rehash of those lectures, nor is it the best of Prasad’s prolific Twitter feed, nor an overview of his billion meta-science articles and editorials. It is instead a series of lectures — enough to fill a semester — that takes bits and pieces of the above and adds quite a bit of new to make something better.
It is not the easiest of narratives to follow. This is understandable: cancer research, policy, and outcomes are as intertwined as the molecular pathways Prasad valiantly tried to avoid, and mapping their connections will inevitably result in a crazy wall. There are nominally four parts to the book with four chapters each, because you had to put it together somehow, and the parts make sense. Even so, more than once I was wondering what exactly a particular vignette had to do with where it was in the book, and wanted to put it somewhere else. But the feeling goes away quickly — Prasad’s style is entertaining, the puns are clever2, and there isn’t a superfluous paragraph in site.
To that last point — if anything, the book is too short. My pet cancer peeve, the disconnect between bioplausability and reality, and the many misuses of animal models to inform clinical trials, was barely mentioned when it could’ve easily made a whole chapter. Same for grant mechanisms, which did get a page and a half — that one half is a figure — but left too many things unexplained and uncovered, particularly for the lay audience. And as to Prasad’s big advice that the federal government should take over running clinical trials from private companies, well, it’s nice to put some pie-in-the-sky proposals out there, but something that is so against the grain should be more fleshed out.3 Or maybe mention some more feasible proposals, in the line of Vincent Rajkumar’s plea to cut down the number of people with veto power over a randomized controlled trial. I could go on, but I’d rather not spend too much time on what can be refuted by a single sentence: “Write your own damn book”.
Written by Vinay Prasad, 2020
-
And I mean this quite literally: you could see him cropping screenshots two minutes before journal club. ↩
-
My favorite involves a marinating chicken and curry. ↩
-
There is, of course, the argument that someone who’s never run a clinical trial has no right to comment on the quality of those that are out there, nor to propose how they should be done. Rather than resorting to modern arguments against gatekeeping, I will echo my grandmother: I don’t have to lay eggs to know when there’s a rotten. ↩
The decade in which one’s three children are born will have to rank among the best ones ever, no matter what else happened. But then add marriage, a move to America1, completing 25 years of training2, getting a dream job, becoming an uncle — twice — and not to forget, starting this blog, and, well, it is hard to imagine things getting any better.
So yes, it has been a good ten years. Certainly better than any other ten-year span I’ve had. A good test would be imagining the ten-year-ago me learning the outcome of the single decision he’s made back in 2008 to go for residency training in the US of A instead of a PhD in Germany, and I’m not one to pee their pants from excitement but I do imagine myself coming close, even without knowing the counterfactual.
Is America the greatest country in the world? It was for me, at the time, even as a visitor3. And it may continue being so for decades to come; I don’t see any competitors coming close.4
- The HBO show manages to be more like the comic book than the movie ever was, even though — or rather exactly because — it is nothing like the original while the movie was for the most part a literal shot-for-shot translation of the comic and therefore missed two things that made the comic great: 1. amplifying the anxiety of the day to intolerable levels, and 2. deconstructing its own medium.
- Re: no 1, the original was all over the place time-wise but mostly set in the 80s and the perceived threat was nuclear holocaust. The movie came out in the mid-2000s, during the time of war against terror and existential angst, but was still set in the 80s and the threat again was nuclear holocaust — two beats already missed. The HBO series is all over the place time-wise but is for the most part set in 2019, and the perceived threat is white supremacy. Note the “perceived” and note that it takes some time for the real villains to be revealed.
- Re: no 2, I’ll pick just one example although there are many. The original featured a comic book within the comic book. Of course, in a world in which super heroes are real, escapist media wouldn’t be doing its job just by featuring even more super heroes. So what kids get instead are pirates, and what you as a reader get are panels featuring ships at sea, pirate raids, and the like, interspersed with the “real” story, to great effect. The movie had… breaks in which it showed panels from that same pirate-themed comic book, with the same story line. Only because you’re not mixing two comic books but instead are interrupting a movie to show some drawings, it doesn’t work at all. The HBO series, brilliantly, has a TV show within a TV show, which is, again brilliantly, not pirate-themed. As to what it is, well, that’s one knock I’d have against the show because it’s trying to be cute and funny, and yes a parody docu-drama about super hero origins in the style of American Crime Story is cute and funny, but it’s not in the spirit of the original.
- Another knock against the HBO series is that it coddles the audience, almost as if HBO got complaints about a few of its other shows being too obtuse. A dialogue line was foreshadowed 10 minutes ago in a different dialogue? Cut to the foreshadowing. Characters recognize a clue in something that occurred two episodes ago? Cut to that scene to remind you what happened. A character breaks an egg? Cut to them holding a different egg in a scene from the same episode. Why?
- Yes, it’s petty criticism, but the show is so masterful in so many other ways that the tiny imperfections stick out. Also, it’s also easier for me to list the few things I didn’t like because everything else (Regina King! Jeremy Irons! Jean Smart! The kid actors! The two skinny white guys who I hope will team up for Season 2! The self-conscious wokeness. That soundtrack!) is pitch perfect.
- A half-sequitur: everything I liked about Lost was put in there by Damon Lindelof, and everything bad about it came from JJ Abrams. I didn’t realize that at the time, but their work post-Lost speaks for itself.
- It ties with Westworld Season 1 as the second-best season of the decade, but The Leftovers Season 2 is still my number 1.
Directed by Various, 2019
EconTalk with Russ Roberts is the best interview podcast I’ve listened to, period. Unlike Tyler Cowen Roberts focuses on an issue or two, not the personality being interviewed. He gives fewer if any passes. The effect is that I feel like I’m actually learning about the thing in question, not just getting acquainted with Cowen’s personality du jour. Whether any learning actually takes place at my advanced age is another matter.
My top 5 episodes: - Keith Smith on free market health care - Venkatesh Rao on Waldenponding - Adam Cifu on the case for being a medical conservative - Patrick Collison on innovation and scientific progress - Andrew McAfee on more from less
Honorable mentions: Cowen, Holiday, Hossenfelder, Bertaud
Conversations with Tyler are as good as ever. This year’s favorites: - Mark Zuckerberg and Patrick Collison - Margaret Atwood - Masha Gessen - Emily Wilson - Ezekiel Emanuel
(Note that the majority are episodes with women - Cowen has Roberts easily beaten here)
Breaking Smart with Venkatesh Rao I would recommend to anyone who’s enjoyed the above-linked interview Russ Roberts did with Rao on one of the better Breaking Smart essays. It’s 15-20 minutes of Rao performing mental stretching excercises, solo.
Plenary Session with Vinay Prasad is another podcast that shines with the solo performances, but the interviews aren’t half-bad either. That isn’t a surprise, since this year Prasad has talked to David Steensma, Frank Harrell, Adam Cifu, H. Gilbert Welch, and Clifford Hudis, among others. Sadly, the podcast still doesn’t have a proper website, so I can’t link to any of these episodes directly.
- A brief overview of the past, present, and future of capitalism by a Serbian-born and formerly World Bank-employed CUNY professor Branko Milanović, who specializes in income inequality.
- Some parts hit closer to home than others, most of all the idea that you can have a welfare state, and you can have open borders, but that mixing the two is ill-advised. I am also well-acquainted with America’s indirect and informal immigrant tax, a version of which Milanović proposes as one of the solutions to the welfare/immigration dilemma. I am not a fan.
- His big insight before this book was the elephant chart. Capitalism, Alone’s big idea is that communism may not have been the pinnacle of society that Marx and Engels had predicted, but rather a good way of transforming feudal agrarian societies into modern economies. Centralized planning and broad-stroke changes work well up to a point, but the production chains soon get too complex for communism, at which point the invisible hand steps in.
- Second big insight: corruption is hard-wired into how a particular type of capitalism (which he calls “political”, in contrast to the Western “liberal meritocratic”) operates. This is supported by many a “liberal meritocratic capitalism” city and state, their financial services and real estate markets being dependent on the “political capitalism’s” dark money.
- It was an easy read for this non-economist. Recommended.
Written by Branko Milanović, 2019
- The only reason I bought and read this book was the excerpt published in the Atlantic which noted some parallels between 1990s’ Balkans and 2010s’, well, the world, which I was already mulling in my head. Turns out that’s the best part of the book.
- The rest is uneven. Holbrooke was a slime ball of a human and his accomplishments were nil, yet Packer still manages to make the book into a hagiography. Which I guess is an accomplishment.
- Did Holbrooke truly think that his memos would change the world? In the Gervais Principle hierarchy, he was a clueless posing as a sociopath.
- Packer’s account of the Dayton negotiations confirms that the only reason a deal was made was that Milošević wanted it at any cost. The agreement was for Holbrooke to mess up, and he almost did, multiple times.
- There is mention of HBO buying rights to make a show out of Holbrooke’s account of the Bosnian was. I haven’t read “To End a War”, but I like the idea of the Dayton negotiations being the centerpiece of a mini-series, with flash backs to each individual warlord’s (and Dick’s) messy history. Someone please give the idea to Damon Lindelof after he’s done with Watchmen.
Written by George Packer, 2019
I started the decade childless and am ending it with three, so I have missed most of the 2010s’ pop culture. This includes the entire Transformers franchise and most of the Marvel Cinematic Universe (so, not much missed then?)
- Film: “Get Out”
- Blockbuster/action film: “Spider-Man: Into the Spider-Verse”
- Album: “Hamilton”
- Single: “Rolling in the Deep”
- TV Show: “Veep”
- Single Season: “The Leftovers” season 2
- Book Fiction: “The Dark Forest” (or “Death’s End” if you count the publication of the first Chinese edition, but TDF is superior)
- Book Non Fiction: “Antifragile”
- Athlete of the Decade: The Đoković-Federer-Nadal trio, but if I had to pick one then obviously Đoković.
Movies and music were better in the 2000s, but oh what time to watch TV and follow tennis. It’s too early to judge the books (though it’s telling that my favorite was originally written in 2008).