We know models to be untrue, but if they are close enough to reality we can use models to learn, imagine potential futures, and make better decisions. Donella Meadows’ Thinking in Systems is a study of models, and the ways in which seemingly different problems in various settings can be understood using the toolset of stock and flow analysis. Continue reading...
This is the third time that I re-cap every book I read in the preceding year. I guess this makes it a tradition. Like last year, I did not write as I read, so a lot of the nuance and the detail that I would have liked to include in this post was lost to some unreachable corner of my brain. Hopefully that part doesn’t become a tradition. Continue reading...
Merry christmas, shit hit the fan.
Another piece in this great series on China by the Times. There’s so much packed in here and in the rest of the series that I don’t even know what to highlight. Just go read it.
In this piece, Albert explains an unresolved with smart contracts. He asks, “will a new smart contract cause any existing smart contract to misbehave?” This is a problem I hadn’t thought about until now, and it made me wonder, what are some blockchain projects trying to solve for this?
There are many reasons to dislike central planning. Most people arguing against it usually point to the economic calculation problem and the fact that even if the required data could be collated, which is on its own impossible, we wouldn’t have the compute power to make optimal decisions, so instead we should offload this process to markets. In this piece, Buterin and Weyl make an analogy to a common failure mode in statistically inferred systems (not necessarily machine learned) that is even easier to accept. I had not thought about it before, and found it very insightful. In short, they argue that systems with simpler designs, particularly those with less knobs for bureaucrats to fiddle with, are better. Our experience of highly complex systems don’t generalize well, so we should aim for less parameters to tune. This in turn also has the advantage of making systems less prone to corruption, since it is harder for the person behind the wheel to hide their actions behind the complex interaction effects of the system. Lastly, they discuss the differences between simple and familiar systems, and how a lot of the structures that organize our lives are actually quite complicated, but we’re used to them. You can consider the essay a proposal to apply Occam’s razor to political economy.
A play on the usual “free as in beer” of open source software, denouncing the tragedy of the commons implied by the organization system as it is today. “Let’s own up to the absurdity of talking about a personal freedom that depends mainly on hidden people working for free.”
It is interesting to think about how tech IPOs will interact with the bear market of the past few weeks. Here Fred argues that it’s about time for the private and public markets to sync up once again, and that the startup sector is well positioned to make things work. I’m not so sure.
“Neglect is undervalued” - a good short piece. This idea applies to family life, to office productivity, and even urban design. If a space is too controlled, there’s no room for emergent behavior.
Isn’t redistribution more just/fair than our current system? What kind of perverse incentives does it introduce? How do we avoid free-loading? And how do we make sure that redistributive taxation doesn’t deter innovation and entrepreneurship? As usual, Russ does a good job of asking the big questions and answering them smartly, with simple words. Don’t miss out on Part II, while we wait for the third installment.
A strong critique of the cult of the MBA. This is not really about Sandberg, who sadly is just the scapegoat, but about the ethics of business, and the problems that arise from the antinomy of objective profit-seeking and subjective value-judgement. The author blames the great man theory as rehashed by HBS for many of the woes of the industry. I disagree with a lot of what’s said here. In the end nothing gets done without someone making decisions, and I don’t think the case method pretends to have The One Objectively Correct Answer, but it’s good to think about how respected institutions can improve.
“How could I use this product to do harm? […] Can I use this platform, this API, this plugin, this app, this feature to do something that, as reasonable moral human beings, we feel a little uncomfortable about?” These questions matter. The author’s point is these things shouldn’t exist, but instead that there are trade-offs in building them. We should think more about trade-offs.
I’ve been interviewing a lot of people, and a few weeks ago I was asked to cover architecture for the first time. I had never built a distributed system until I started working at Apple. Sure, I’d used Heroku with a Redis instance and a Postgres DB, but all of that was totally abstracted away from me. Reading this was an eye-opener as to how much I’ve learned in the past few years, and it made me reflect on some of the interviews I had while switching teams - even two years ago I already felt skilled at systems design, and that experience is rare. Working at a place like Apple can really distort your perception of what’s normal.
I first learned about William Easterly when one of my econ professors handed me his book, The Elusive Quest for Growth: Economists’ Adventures and Misadventures in the Tropics. I read that a few years ago, and by now have forgotten many of its anecdotes, but its key point resonated with me and shaped a lot of my intuition behind economics. Essentially, Easterly shows that the classic top-down technocratic solutions that make up foreign aid tend to backfire in unexpected ways because they’re full of perverse incentives for businesses, bureaucrats, and individuals alike. The headline, and part of the article, opposes his views to those of Bill Gates which is gimmicky and click-baity, but the content is substantial. For a deeper dive into Easterly’s thought, check out his Econtalk episodes, or obviously check out his actual work EQG is a pretty easy read.
Your decentralized system can allow for (1) Self-sovereignty, (2) Privacy-preservation, (3) Sybil-resistance. Pick two.
I read this because Sean posted about it on Twitter. Turns out his family comes from there, which I didn’t know until after I read the article and told him I was adding this to my to-read pile. I have a very hard time understanding company towns. This book will be a good start.
I don’t really know much about Brexit, but this divide between “any-wheres” and “some-wheres” is an interesting one. For a while, I’ve been making a similar argument about the cosmopolitan urban elites vs secluded rural communities elsewhere in the world.
I never got into the Knuth cult, I have not read his books, and don’t pretend to know much about his research, but I know he’s a highly influential figure who changed the field of computer scince again and again in the last ~60 years. Reading about him, and how he spends his time these days was fascinating.
Here, Noah points out what has been clear to many for a while: China is already the main world power, and the trend says that’s not gonna change. Still wrapping my head around the obvious implications & struggling to understand 2nd/3rd order effects. It made me think of this piece by Brian Brooks, the chief legal officer of Coinbase. If the replacement of the USD as the world reserve currency is a real worry near-term, it would be a good strategy for the US to push for a non-sovereign-backed currency instead of ceding the position to China. It’s like a kid, who knowing he’s lost the game, grabs the ball and takes it home with him early.
This ~12k word story of the downfall of GE is crazy. I’m sure someone has studied it well, but don’t conglomerates go totally in the opposite direction of Coasian theory of the firm? Odd that the paper it comes from was written in the golden era of conglomerates. The section on GE Capital overtaking the industrial “real economy” branch of GE has a lot of echoes of Braeburn Capital. Odd to think of the financialization of industrial giants.
Perhaps I’m too much of a cynic, but I welcome these little time saving technologies. They help me more than they make me doubt my humanity. Thompson says near the end that “to see these technologies in action is to be confronted with the fact that we are not so very special or unique.” Good, someone had to break it to us.
I used to follow Israeli politics quite closely, and Judah’s reflections here echo a lot of my own opinions. I remember grasping, as a 15 or 16 year old, for the first time grasping that Israel can’t be both democratic and Jewish in the long run, and how that realization changed so much of my world view. The world is going in a strange direction.
At first glance, this did not seem interesting at all. I only read it because several people I respect brought it up on Twitter as a great read, and it was! TL;DR this is a crazy story about a turkish woman who started a fashion company that sold a ton of really high end bikinis, then a bunch of companies like Victorias Secret and Neiman Marcus started copying her, so she sued them over it. In the end, she settled with VS, but it turns out she had copied the design herself from a woman in Brazil who sells bikinis at the beach, and there’s a whole legal shit-show about what copyright is supposed to be surrounding this case. Super interesting.
This industry seems unnecessarily secretive, but this piece was pretty fun. It’s good science writing, although, I have to say, the self-deprecating tone felt a bit out of place for the NYT. I had never really thought about how glitter is made, and had always assumed it was some kind of shredded/pulverized metal. I guess it makes sense that it’s mainly plastic, since that’s much cheaper, but, again, I’d never thought of it.
A really simple visualization that says a lot. The author combed through 741,576 section front headlines since 1900, looking for which country got the most mentions in any given month. It’s pretty amazing how much attention Britain was given early on, and how much reporting about wars takes up the front page. Germany and then the Axis take over during the world wars, then there’s Russia during the cold war, followed Vietnam and Iraq. The fact that China takes up so much of the last ten years is also telling. This is a great project, I just wish the articles were linked, too.
Three seasons in, and I still don’t know how to describe Heavyweight. Jonathan Goldstein somehow manages to create amazing stories where I can initially see none. This time around, he tracks down the single guy who blew it during the filming of Russian Ark, an experimental single-shot full-length movie I had never heard of.
An unusually political and not very internetty episode for this show, but still a very interesting listen about the interactions between local governments and corporations. It pairs nicely with this Planet Money episode about Amazon and NYC.
A fun conversation about how humans thought about technology thousands of years ago, and how little some things have changed. I had never heard about Talos and especially liked that part of the podcast.
I really like the idea of cryptography enabling the “building blocks” style tech that Web 2.0 companies had promised many years ago. Today, no one wants to build on top of other people’s platforms, because historically, platforms end up screweing developers - just ask developers who built on top of Twitter. Crypto might finally lay out the right kinds of incentives, and lead us to a place where tiny services can actually ride on others’ rails. It’s still early days…
This was one of the most interesting episodes of Patrick’s podcast recently. Origin stories are always cool, but especially so when they’re so eclectic. Learning a bit about the decisions behind 2nd Life’s and YouTube’s early economies was fun, but what I found most insightful (by far!) was “is it a value or is it a tactic?” towards the end of the episode. We need to remember why we do the things we do.
This isn’t a link to a single article, but to 24. Through this advent calendar, these three decided to tell the stories of 24 women who, as they explain, “made today’s computing industry as amazing as it is.” The whole set is awesome, but I especially enjoyed reading about Kathleen Booth and Wendy Carlos.
This set of photos from Poland in the 80’s is just 👌 Have you seen other similar photo sets that capture a culture at a single point in time? I’d love to see them!
A few weeks ago, Hannah had to go to Hawaii for work, and asked if I wanted to tag along. The tickets were cheap, so I said why not! I had low expectations, beacuse how much better could the beaches be than the ones I grew up going to in Costa Rica? I’d say they are different enough, and cool in orthogonal directions - the cultural and historical context are totally different. More...
This is the first epic in a series about China that the NYT is working on. It is a good overview of how the Chinese government has switched from being a closed-in communist anti-market regime to an expansionist communist but market driven regime. The tone is a bit too positive for my taste, as it brushes away the clearly autocratic/totalitarian tendencies of the Chinese system, but the content is well put together, and I’m sure they will touch on these in one of the upcoming pieces. Make sure to see the appendix articles, too.
As I read this wonderful piece I kept thinking of Robert Gordon. Unsurprisingly, the authors invoke Gordon’s pessimistic headwind thesis but, unlike him, Nielsen and Collison are optimistic, demanding “large-scale institutional response” to a problem they see as solvable. The thing is, progress is non-linear. Most science isn’t worth the time and effort invested in discovering it. I was a bit bummed out that the domain they are hosting their appendix at (http://scientificreturns.org) is empty. I was excited to click around and learn more about their position, and hoping to find proposals they were bringing to the table. I assume they have more than just this op-ed up their sleeves, so I’m looking forward to hearing more.
An in-depth look at why GDP grows the way it does. This post decomposes growth into many components, and ultimately ties it down to how many people can use a bounded pool of knowledge at any given time, and how we can extract more of that knowledge to put it to use.
Somehow I’d missed this NYT profile of Lina Khan, and her work on antitrust, from this past September. It’s easy to draw parallels between her questioning the meaning of consumer welfare & Mariana Mazzucato’s questioning the meaning of value. We need to rethink the basics.
In this short piece, Alvaro questions the modern eurocentric view that cartesian straight lines are good, while complex illegible curves are not. We have a learned set of ideas of what things should look like, and that knowledge is a lens through which we view other societies. When at first sight a system doesn’t fit nicely with our views, it is easy to dismiss it as wrong - our first reaction is to impose our accepted truth on top of it. Clearly our way of doing things is better. Whether we’re discussing urban planning or software engineering, people are often trigger-happy and ready to raze down effective solutions layered by history, enacting our own abstractions to provide legibility at the cost of actual efficiency. Without knowing it, he was channeling Seeing Like a State.
I’ve always hated reading translations. It is really surprising that mainstream USians have pushed back on this for so long.
Power laws are real, and they are everywhere. Of ~6k firms making $1B+ in revenue globally, the top 1% creates 36% of all profits, middle 80% record near-zero profits, Bottom 10% destroys as much value as the top 10% creates. Similarly, out the top 50 cities host 8% of the global population and represent 21% of world GDP, translating into a GDP/capita that’s 45% larger than their peers. The dynamics that lead to this are not studied enough.
In this essay, Noah points out two possible paths for the Democrats: social democracy and corporatism. Either we want the state to take up a larger role in people’s welfare, or we want companies to do so. There’s a clear tension between the two. Honestly, I’d like to think of a third path.
Recently, with the turmoil in markets in the US and abroad, I’ve been thinking a lot about personal finance, and ways to improve my lot. The question of whether one should invest as soon as cash is available with each paycheck (ie, cost averaging over time) or invest in a lump sum when some event occurs (ie, a simple momentum play, or a market drop) is deeply controversial. I’ve been cost averaging for a while, but the longer I do it the more I think I’m doing it wrong.
People want to help, you just need to make it easy for them to do so. I have been lucky over the years to have naturally found mentorship in family members and friends alike, but the kind of specific pointed requests for advice that Bethany recommends here seem like an art that I should spend time perfecting. If I weren’t at Apple, I’d probably make a bigger effort to tell people about what I’m working on, and what I’m struggling with on a day to day basis.
While I’ve thought a lot about incentives in startups, I had never thought of the kinds of conflicts that arise when a founder fully vests an initial grant. The fact that something like this can trigger a bigger conversation seems especially mis-understood.
The fact that some “huge successes are the results of pivots” does not mean that all pivots are successful. I agree with Fred that in most cases, hard pivots are a bad idea. There’s probably enough data around to calculate whether P(success | pivot) > P(success).
This is a topic that I’ve been thinking a lot about at work, given the growth of my team and the trajectory that we’re following as a data platform to other teams within the company. I can’t say much, but I’m glad other people have thought about these problems before.
You can trace all sorts of parallels between Sears and Amazon. This is one I hadn’t heard about before, but it is interesting how sometimes unrestricted capitalism and its tunnel vision can lead to unexpected good outcomes. Having access to catalogue shopping meant that you had access to products you otherwise were not able to obtain. The capitalist machine was just catering to the market. Another example, technology meant to reduce friction for everyone can disproportionately ease the experience of shopping for those who are usually targeted as potential thieves.
Language, geography, history, and identity are deeply tied together. Here’s an awesome visualization from Germany
The internet is scary. I have multiple factor auth in most sites, and somewhat understand the intricacies of phone spoofing. My mom does not, and most of my non-techie friends don’t either. This is a problem we have to solve via easy to use products that abstract away all this complexity, or we’ll have to pay for it later on.
Fake news! The problem is that people are gullible, and that we don’t question what we see and hear, not that we are presented with false information. This episode is a good case study with a couple of extreme historical examples, but for a wider overview on the topic you can also check out Sam Harris interviewing Matt Taibbi for his podcast
In the US, you need to have money to have a voice in an election. What if we helped people attain that voice? Well, it’s a hard design problem. Seattle tried it, and for the most part flopped. This is an interesting idea worth exploring, and I hope other local experiments expand on it.
An interesting conversation on the sharing economy, and the value of platforms, from the ancient souks of the middle east to the present day gig economy. A lot of the dynamics behind these systems can be explained as ways of lowering transaction costs.
What kind of analyses can we apply to private brands to try and pick out winners? Caldbeck has been doing this for years, and has great insight into what signals are worth looking for.
The previous episode with CoVentue’s Ali Hamed was great, so I had very high expectations for this one, too. It did not disappoint, and reminded me of how little I actually know about the mechanics of credit, even though I interened in two different credit-related fintech startups.
The usual blend of humor and science that you’d expect from the Oatmeal. Somehow I had missed that InSight was coming.