I really enjoyed meeting so many of you at my Politics and Prose event a couple weeks ago. It was meaningful for me to be talking about my books in person again after having to launch my last title during a pandemic.
It was also a great opportunity to thank people for their support of Slow Productivity, which just yesterday landed #4 on the NYT’s monthly business bestseller list. (If you haven’t bought Slow Productivity yet, you should! If you read this newsletter, you’ll love it…)
It’s with all this in mind that I wanted to briefly share two upcoming opportunities to come meet me and hear me talk about Slow Productivity in the DC area next week!
Earlier this month, a group of scientists from universities around the world published the results of an ingeniously simple experiment in the journal PLoS ONE. Every month, for ten months, they randomly selected an article from a journal in their field to promote on their Twitter accounts, which, collectively, added up to around 230,000 followers. They then later compared the success of these tweeted articles with control articles randomly selected from the same issues.
The result? No statistically significant increase in citations in the promoted articles versus the controls. There was a difference, however, in the download numbers: more people took a look at the tweeted citations.
A brief book update: I wanted to share that Slow Productivity debuted at #2 on the New York Times bestseller list last week! Which is all to say: thank you for helping this book make such a splash.
If you still haven’t purchased a copy, here are two nudges to consider: (1) due to the rush of initial sales, Amazon has temporarily dropped the hardcover price significantly, making it the cheapest it will likely ever be (US | UK); and (2) if you prefer audio, maybe it will help to learn that I recorded the audiobook myself. I uploaded a clip so you can check it out (US | UK).
Last March, Sebastien Bubeck, a computer scientist from Microsoft Research, delivered a talk at MIT titled “Sparks of AGI.” He was reporting on a study in which he and his team ran OpenAI’s impressive new large language model, GPT-4, through a series of rigorous intelligence tests.
“If your perspective is, ‘What I care about is to solve problems, to think abstractly, to comprehend complex ideas, to reason on new elements that arrive at me,'” he said, “then I think you have to call GPT-4 intelligent.”
But as he then elaborated, GPT-4 wasn’t always intelligent. During their testing, Bubeck’s team had given the model a simple math equation: 7*4 + 8*8 = 92. They then asked the model to modify a single number on the lefthand side so that the equation now equaled 106. This is easy for a human to figure out: simply replace the 7*4 with a 7*6.
GPT-4 confidently gave the wrong answer. “The arithmetic is shaky,” Bubeck explained.
This wasn’t the only seemingly simple problem that stumped the model. The team later asked it to write a poem that made sense in terms of its content, but also had a last line that was an exact reverse of the first. GPT-4 wrote a poem that started with “I heard his voice across the crowd,” forcing it to end with the nonsensical conclusion: “Crowd the across voice his heard I.”
Other researchers soon found that the model also struggled with simple block stacking tasks, a puzzle game called Towers of Hanoi, and questions about scheduling shipments.
What about these problems stumped GPT-4? They all require you to simulate the future. We recognize that the 7*4 term is the right one to modify in the arithmetic task because we implicitly simulate the impact on the sum of increasing the number of 7’s. Similarly, when we solve the poem challenge, we think ahead to writing the last line while working on the first.
As I argue in my latest article for The New Yorker, titled “Can an A.I. Make Plans?,” this inability for language models to simulate the future is important. Humans run these types of simulations all the time as we go through our day.
I know it’s been a minute since I’ve published one of my normal essays. I’ll be returning to these soon as the chaos of the Slow Productivity launch dissipates.
In the meantime, I wanted to share two quick notes: one about the book, and one about something interesting (but completely unrelated) that several of you have sent in my direction recently…
A note about the book
On Saturday, March 16th at 3:00pm, I’ll be appearing at Politics and Prose on Connecticut Avenue in Washington, DC. I’ll be joined in conversation with David Epstein, the New York Times bestselling author of Range. We’ll talk Slow Productivity and take questions from the audience. (For a preview, see my recent interview in Dave’s excellent newsletter.)
This is my first live event of the book tour, so if you’re in the DC area, I’d love to see you there! (More details.)
(You might also be interested in my most recent essay for The New Yorker, titled “How I Learned to Concentrate,” which discusses how my early years at MIT shaped almost everything I’ve written about ever since. I had fun writing this one: lots of Stata Center nostalgia!)
My podcast producer recently turned me onto a show called Acquired, which features its co-hosts, Ben Gilbert and David Rosenthal, diving deep into the backstories of well-known brands and companies, from Porsche and Nike, to Amazon and Nintendo.
It turns out I was late to this party. In the eight years since Acquired was originally launched, it has grown into a huge hit. The show now serves more than 200,000 downloads per episode. As Rosenthal revealed in a Fast Company profile last summer, they now face the problem of their audience becoming too large for their advertisers to afford paying the full fair market price for their spots.
What interests me about Acquired, however, is less what they’ve accomplished than how they did it. The conventional wisdom surrounding new media ventures is that success requires frenetic busyness. You need to produce content perfectly-tailored to your audiences’ attention spans, master The Algorithm, exist on multiple platforms, and above all else, churn out content quickly to maximize your chances of stumbling into vibe-powered virality.
Acquired did none of this. Gilbert and Rosenthal’s podcasts are very long; the two-part treatment of Nintendo I just finished clocked in at a little under seven hours. They also publish on an irregular schedule, often waiting a month or more between episodes. Combine this with the reality that they largely ignore YouTube and have no discernible social media strategy, this venture should have long ago crashed and burned. But it instead keeps growing.
In 1951, Abraham Joshua Heschel published a monograph titled simply, The Sabbath. It consisted of ten short chapters, comprising of less than a hundred total pages, illustrated with original wood engravings by Ilya Schor.
Early in the book, Heschel establishes the unique importance the Bible places on rest:
“It is, indeed, a unique occasion at which the distinguished word qadosh [a transliteration of the Hebrew term for ‘holy’] is used for the first time: in the Book of Genesis at the end of the story of creation. How extremely significant is the fact that it is applied to time: ‘And God blessed the seventh day and made it holy.‘ There is no reference in the record of creation to any object in space that would be endowed the quality of holiness.”
As Heschel elaborates, this idea was new. In pagan religions, places were holy; a sacred mountain, say, or a deified river. But the Abrahamic faiths found something Godly in a ritual of rest amid the flow of time.
It’s hard to believe that we’re less than two months away from the publication of my new book, Slow Productivity: The Lost Art of Accomplishment … Read more
Someone recently forwarded me an essay from a blogger named Henrik Karlsson. It opens with an admission: “When I started writing online, the advice I got was to publish frequently and not overthink any single piece.”
Karlsson was not alone in receiving this suggestion. As social media erupted into cultural dominance over the past decade, it carried in its wake a force that thoroughly disrupted written media: virality. An article or post that hit the Twitter or Facebook zeitgeist just right could summon hundreds of times more readers than average. Because it was difficult to predict which pieces might ascend to this cyber-blessed state, the optimal strategy became, as Karlsson was told, to publish as much as possible, maximizing the odds that you stumble onto something sticky.
What makes Karlsson’s essay interesting, however, is that he decided to test this hypothesis on his own work. “I’ve now written 37 blog posts and I no long think this is true,” he writes. “Each time I’ve given in to my impulse to ‘optimize’ a piece it has performed massively better.”
Using new subscribers as a metric of success, Karlsson calculated more specifically that spending twice as long on an article yields, on average, more than four times the number of new subscribers.
