What am I up to now?

April 30, 2024

October, 2021

October is my favorite month. My birthday, Halloween, Sukkot, MLB postseason, a spike in fiction releases, all add up to a great few weeks. Now, in Sierra Leone, I’m realizing how much I underrated the weather changing at this time of year.

Temperatures don’t change here. On clear days, which are increasingly common as the rainy season ends, there’s no difference between October and February. This has absolutely thrown off my internal clock. While beach weather everyday is pleasant, I’m looking forward to being somewhere with a bit more variability in seasons.


  1. Reading
  2. Listening
  3. Writing
  4. Eating


I have an increasingly hard time picking books to read. Last year, I had my favorite book bloggers, and I went through all of their recommendations, and skimmed off the ones I was confident that I would enjoy. In some sense, I’m jealous of my friends who keep their Goodreads account up-to-date. Whatever my book-choosing algorithm is, Amazon’s is likely better.

However, I do have a philosophical problem with Goodreads’ book-rating system. A five-star score can be useful, as a single dimensional measure of a books quality: one friend mentions that he only reads books with a 3.9 or higher average rating. But this dimension only captures some of what makes me want to read a book. One issue is that some of my favorite books are just terrible. My book choices don’t always optimize for book quality. One of my favorite books to re-read is “1632”. If you’re interested in how West Virginians could reinvent a modern industrial base using technology from Early Modern Germany, these stories are for you. By some sci-fi, an Appalachian mining town is time-traveled to the middle of the Thirty Years’ War, and the rednecks “bring the American Revolution to Europe: 150 years ahead of schedule!”Actual tagline. The writing is overburdened with fictional and historical characters and technical minutiae. Take one plotline: the local hippie (“Stoner”) becomes invaluable to the war effort against the Hapsburgs due to his background in “pharmaceuticals.” There are pages and pages of details on the making of sulfa drugs and analgesics. Unsurprisingly, the median Goodreads reviewer leaves this book three stars.For an opposite view, see Gleech’s sort by controversy, the books for which his rating most disagrees with the Goodreads average. All of them are, in his opinion, overrated, i.e., he rates them low and the average reader rates them highly.

There’s an extensive literature on ratings systems.And if there weren’t, Amazon’s secret squad of economists would create one. Across Amazon product reviews, Uber ratings, and movie scores, anything below 4 is considered sub-par. But intuitively, a 3 should be average. I sometimes dream about systems where raters have to provide uniformly distributed results. Instead, ratings are right-skewed, making it difficult to distinguish the great from the excellent. This makes no sense to me, and has apparently irritated me for years: here’s a screenshot from a friend’s instagram in 2017, when I was already complaining!

This month, I also spent hours reading memos from this list compiled by Sriram Krishnan at Andreessen Horowitz.

In their memos, Zuckerberg, Jobs, and Gates shine as convincers. The acquisition of Instagram, one for the history books, wasn’t decided by accountants or in board rooms, it closed because Mark Zuckerberg kept writing emails to Kevin Systrom. The emails are well-written, of course, but more interesting is that Zuckerberg took so much time to write them. Another tech memo which provided a good perspective was Jobs on iBooks. Amazon has firmly won the e-books war,Unlike some, I don’t think this is a bad thing. but it wasn’t a sure thing. Jobs aggressively lobbies James Murdoch (!) to have HarperCollins sign a deal with Apple to distribute their ebooks at a substantial markup from Kindle prices.

The reason we are doing this is that, with our experience selling a lot of content online, we simply don’t think the ebook market can be successful with pricing higher than $12.99 or $14.99. Heck, Amazon is selling these books at $9.99, and who knows, maybe they are right and we will fail even at $12.99. But we’re willing to try the price as we proposed.

Krishnan calls this email thread “infamous,” so there’s history which I haven’t read. If anyone has reading recommendations on the history of the ebook market, please send it along! “Book Wars” by Thompson is on my reading list, but ironically not worth $28 for the Kindle version.

Krishnan’s list has just five government memos, of which two are from political campaigns and one is from Du Pont Chemical company. The political ones are fascinating: read Hillary Clinton’s 2008 chief strategist on how to take down a junior senator from Illinois. Obama was “all sizzle and no steak,” he “lacks American roots,” and he’s “unelectable except perhaps against Attila the Hun.”

This isn’t the pinnacle of government memo writing. Some government memos probably contain too much CYOA to be interesting. On the other hand, as Krishnan notes, memos “typically seem to come into the public domain in one of three ways: through being really old, being part of some lawsuit/legal process or, sadly, being part of a hack.” The most interesting ones come from hacks, and these are the ones Krishnan is least likely to put on his website.

One of the most exciting war books I read this year was mostly a summary of memos and conversations about the Iraq war decision. “Leap of Faith” by Michael Mazarr collected memos from the White House, Pentagon, and NSC, which become characters in the story of Iraq. Donald Rumsfeld comes off the worst in his memo writing: one “urgent” (and unexplained) memo penned from his desk in 2002 reportedly read, in its entirety, “Sunni vs. Shia??” Rumsfeld was also deeply possessive of his briefings:

If people began taking notes, he would sometimes tell them to stop. In one such session, according to someone present, [Condoleezza ] Rice said she wanted to keep her copy of the slides. Rumsfeld refused and reached for them. The two were actually pulling the slides back and forth until Rice appealed to Bush, who was sitting at the head of the table: “Mr. President, I need to keep these.” Bush merely shrugged and left the room. Rice eventually surrendered her copy, but sent a staff member to the Pentagon to find another.

We probably undervalue the memo. The most important one on Krishnan’s list is from IBM’s Thomas Watson, aimed at adjutants working in the Pentagon during WWII. The focus is on “completed staff work.”

“Completed Staff Work” is the study of a problem, and presentation of a solution, by a staff officer, in such form that all that remains to be done on the part of the head of the staff division, or the commander, is to indicate his approval or disapproval of the completed action.

Everyone I’ve ever worked for has been a busy person. They don’t read what I write to learn, they read to act. When I was consulting, a memo could change policy or strategy. Now, working for researchers, a memo (occasionally taking the form of a literature review) informs experimental design or field planning. A good memo isn’t information, it’s action.


Here is a list of records I listened to for the first time this month and liked (no star) or loved (star).

Also, Krishnan’s list of memos includes this (undated, but presumably mid-2000s) Dave Goldberg memo on the future of the music industry. The memo was sent to Sony CEO Michael Lynton, and was released in the aftermath of the massive Sony Pictures hack. The document is fantastic for the facts it uses to frame the industry,Catalog provides 50% of the revenue and 200% of the profits of recorded music. its forecasts (all eventually proven true),Physical distribution is going away—it doesn’t need to be eliminated prematurely but it needs to follow digital and not drive it. and the final recommendations for how Sony can place itself in the market.If Netflix wanted to pay 200MM per year to give all of its 40MM subs a music subscription—that should be encouraged, not scoffed at. The goal for digital is for subscription and ad-based services to become the predominant means for people to access music. I learned more in these two and a half pages about the music industry than reading industry news.


I received many useful comments on the draft of my paper on the long reflection, thank you to everyone for reaching out! I also found a pretty egregious coding error in my paper on the role of start-up cities in emigration reduction. Working with cross-national opinion data is hard. Both of these projects are back on the drawing board for now.

In the first days of October, I’m presenting a new paper on the natural resource curse at APSA 2021 in Seattle. The slides are here, and we’re finishing up a full working draft in the early weeks of the month. This draft is my writing sample for PhD applications in the Winter, so I would more than appreciate any comments! Please reach out if you’re interested, or have even a second to take a look.

I’m also spending an inefficient amount of time writing and re-writing the personal statements for my PhD applications. One mentor who’s recently been through the process remarked that, at this point, everything else about the application has been locked in (GRE, GPA, LoRs, CV). Even though the personal statement has a minimal role in admission decisions, it intuitively feels like the last bit of the application under my control.


I moved apartments in early September, and, as part of the welcome package to my new home, I inherited about 25 cans of chickpeas. I have boiled, toasted, roasted, baked, crushed, smushed, blended, and devoured these cans, with the goal of working through Wikipedia’s list of chickpea recipes.

The most common response I’ve gotten, when I bring up my chickpea binge, is that I must re-try all of these recipes with dried chickpeas. This won’t be happening: no increase in quality is worth the 12+ hours of planning involved in rehydrating and cooking dried chickpeas. But I appreciate your concern.


September, 2021

July, 2021

June, 2021

May, 2021

What am I up to now? - April 30, 2024 - Joseph Levine