Let's Become Better Writers in 2021
Think Out Loud newsletter 1/2/2020
Welcome to Think Out Loud, a newsletter about technology and philosophy. If you’d like to learn more about this substack, you can do so here. You can also subscribe below:
Contents
6 ways I’ll become a better writer in 2021
What I’ve been reading recently
2.1 Startups
2.2 Cyber Security
2.3 Economics
2.4 Cryptocurrency
2.5 Software Engineering
2.6 Philosophy
1. 6 ways I’ll become a better writer in 2021
How do you write well, and why might you want to? A number of years back, I read an essay by Paul Graham that that got me about 51% percent of the way to an answer.
“Having good ideas is most of writing well. If you know what you're talking about, you can say it in the plainest words and you'll be perceived as having a good style.”
If there’s any kernel of truth to this—and I think there is—then seeing improvement in one’s writing might be as close as we can get to telling if our ability to generate ideas is getting any better. It’s tough to tell if an idea is actually good when it’s living in your head, but, if you write it down, you can feel more confident it’s good or bad. Something about writing gets to the essence of an idea better than speech.
I felt like 2020 was a firehose of bad writing pouring out of just about every major institution. Or worse: no writing. How do I know it was bad writing? I can’t tell if the premises the writing was based on are true, and sometimes—like in the case where the WHO advised citizens to not wear masks or urged countries to not close the border with China early on—I can’t even understand how the premises lead to the conclusion. I can tell it’s bad when I’m left with more questions than answers.
Here’s another example: take the COVID vaccine. The educational messaging on this has been abysmal. Over the holidays, I had the unpleasant experience of trying to convince a parent that the COVID vaccine was safe by googling “CDC statement on covid vaccine safety”, assuming there would be something convincing. Take a quick at the CDC’s website to find out why:
“Clinical trials are being conducted to evaluate additional COVID-19 vaccines in many thousands of study participants. These trials will generate scientific data and other information that will be used by FDA to determine vaccine safety and effectiveness. Clinical trials on all COVID-19 vaccine candidates are being conducted according to the rigorous standards set forth by FDA in their June 2020 guidance document, Development and Licensure of Vaccines to Prevent COVID-19. If FDA determines that a vaccine meets its safety and effectiveness standards, it can make these vaccines available for use in the United States by approval or through an EUA.
After FDA determines that a COVID-19 vaccine candidate is safe and effective, the Advisory Committee on Immunization Practices (ACIP), a committee comprising medical and public health experts, reviews available data before making vaccine recommendations to CDC. Learn more about how CDC is making COVID-19 vaccine recommendations.”
That’s as close to a public statement as I could find on their site. There are about 1,000 words there, most of them are weasel words, or links to long, highly technical, and bureaucratic documents. The above is lacking so much detail that it is barely more than an appeal to authority, and undermines people’s confidence. It also has the effect of making you look silly when talking to your parents.
Medical YouTube and some scattered blog posts have been my primary news sources through the pandemic because they seem to be the only people being transparent. But, if I have to dig this deep to believe the vaccine is safe, how are my parents supposed to be convinced?
Making the CDC’s statement more convincing is trivial. The John-Hopkins page, for example, does a much better job answering the question on everyone’s mind:
“The type of vaccine developed for COVID-19 by Pfizer/BioNTech has been years in development for other infectious viruses. Thus, the manufacturing process was ready very early in the pandemic.”
I can really only hope our institutions communication improves in the future. I can, however, improve my own writing. I was giving this some thought earlier in 2020—it’s a big part of why I started this substack—and I want to (briefly) share those thoughts with you.
1. Building my Library
This is a fancy way of saying I want to read more, and actively absorb what I’m reading.
I started this substack is to list (and briefly describe) everything I’ve read the last 2 weeks. I didn’t choose to list things out because I wanted to hold myself accountable. I love reading, I’ll do it one way or another, but I’ve always found that forcing myself to write about what I’ve read is the only reliable shit-test to see if I’ve understood. Understanding things—and not just feeling like you understand them—is the only way to build a set of premises you can form your own ideas upon.
2. Building my Antilibrary
I never feel bad about owning a book I have not read cover to cover. Nassim Taleb changed my feelings about this some years back with his concept of the “Anti-Library”—the collection of books you have, but have not read.
“… a private library is not an ego-boosting appendages but a research tool… Indeed, the more you know, the larger the rows of unread books. Let us call this collection of unread books an antilibrary.” — Nassim Taleb
Knowing what you don’t know deeply enough is key to removing some of the mental biases associated with over-indexing on what you know (Taleb’s Black Swan problem). It also services as reference material whenever something reminds you of those books where you’ve just scratched the surface. What’s best is that the books glare at you every day, reminding you of your own ignorance.
3. Having a space for quick and “imperfect” writing
You’re here!
I type these posts up in a couple hours or as I’m finishing whatever I’ve read. Having a location where I write things out on a regular basis, but don’t hold myself to the same quality bar lets me get feedback on my writing more regularly and improve faster. I’m a strong believer in fast iteration cycles and tinkering being the key to improvement.
4.Having a space for “well-thought-out” writing
If I never write anything I’m taking seriously, all this preparation would seem vacuous. My Medium account—and my personal website—are where I put my thoughts after I’ve fully processed them.
Some day I’d like to write a nontechnical book, but I’m not sure I have the “Library” or the “Anti-Library” to write something I would be proud of just yet. Plenty of writers wait a long time before publishing a book, and I think this might be the best approach. Nassim Taleb didn’t write his first non-technical book until 2001; he was 41 years old.
5. Having a space for creative or personal writing
I’ve been journaling for over 10 years now, and I can’t recommend it enough. Most mornings I just write a couple of sentences, but every once in a while I’ll write at length about something personal, or use a writing prompt.
It’s hard to even enumerate all the ways journaling can help you write. The journal also gives me a space to reflect on what I’ve read recently (if I feel like it) and practice writing in different styles. Telling a story about something that happened to me as if I had an audience helps me be more engaging too. A daily writing habit also makes writing what’s on my mind—like I am here—very natural.
6. Sharing my writing, even if it’s bad
Let’s go back to that Paul Graham essay from before:
“When I give a draft of an essay to friends, there are two things I want to know: which parts bore them, and which seem unconvincing”
This is the holy grail of writing heuristics. Chances are, if your writing is convincing and engaging, you’re doing something right. If you’re writing BS, eventually, to someone, your writing will be unconvincing. And if your writing is non-trivial, and convincing, it will be engaging.
The only way to get this feedback loop started is to actually share your writing.
…
There’s so much more I wanted to say here, but I’ve hit the annoying email length limit. I’ll leave you with some more Paul Graham essays on writing; they’ve probably got better ideas than what I’ve vomited out here, if I’m being honest:
2. What I’ve been reading recently
2.1 Startups
Startup Stock Options - Why A Good Deal Has Gone Bad - A clear write up on stock devaluation for startup employees, and how the game is often rigged in favor of founders and venture capitalists. I was already well aware of the problems with stock devaluation for startup employees (I decided to join a startup recently, and read all about it). But I wasn’t aware that it wasn’t always this way! This paired especially well with all my other readings on debt, inflation, etc lately.
“Early employees take an equal risk that the company will crater, and they often work equally as hard. However, today founders own 30-50 times more than a startup’s early employees.”
2.2 Cyber Security
U.S. Treasury Department Hacked Through Solar Winds Compromise - Solar winds is popular a piece of monitoring software that was recently exploited using a relatively sophisticated “Supply Chain Attack”, allowing attackers to make it to escalation of privilege across an insane number of organizations, not just the Treasury Department. Malicious code was somehow sent via the software’s update platform. Here are some clips from the Krebs article I link above:
“In its own advisory, FireEye said multiple updates poisoned with a malicious backdoor program were digitally signed with a SolarWinds certificate from March through May 2020, and posted to the SolarWinds update website.”
“From there, the attackers would be able to forge single sign-on tokens that impersonate any of the organization’s existing users and accounts, including highly privileged accounts on the network.”
Pretty interesting stuff. Somehow, the article already has a massive wikipedia page, which lists multiple exploits that were used to gain access, so read there if this appeals to you too.
2.3 Economics
Implications of the US borrowing 3 Trillion in H1 2020 - There’s a lot of talk about the Fed overheating the dollar lately in favor of stimulus all across the board (government debt, stimulus checks, corporate bailouts, etc). I’m kind of curious about this argument, as some of the most knowledgable writers I follow have been arguing for more stimulus; including Nathan Tankus, who tweeted this recently:
I’ve only been read a very small amount in this space, but I found it illuminating to see some specifics on where all these loans were being spent, and who bought them.
So, who bought all this newly issued debt? The largest buyers were private investors, which includes asset managers and hedge funds, both domestic and international, who took up almost $1.8 trillion of the debt. A close second was the US Federal Reserve (Fed) which bought almost $1.6 trillion or 46% of the new debt issued. Interestingly, central banks around the world, which hold trillions of dollars of US Treasuries in their foreign reserve portfolios, were not involved in buying US Treasury securities in the first half of 2020.
Debt: The First 5000 Years - We seldom think about debt as abstractly as we should. Things as fundamental as money, social obligations, religious and legal practices, can often be boiled down to being special cases of the concept of debt. I only started it a few days back, so I don’t want to speak on the book prematurely.
The book has mixed reviews from experts—likely due to some of the controversial conclusions Graeber comes to—but excellent reviews on Amazon. If you’re at all interested in the history of money, you should read it, even if it’s only for the crazy historical and anthropological examples of debt Graeber describes.
2.4 Cryptocurrency
The Bitcoin Standard by Saifedean Ammous - I read this front to back last week (it’s short, just 272 pages). I had mixed feelings about it, but it was interesting to get such a free-market stance from an academically trained economist (this is somewhat rare, from what I can tell). Overall, Ammous paints an intriguing, but largely incomplete, picture of a world under Bitcoin.
I’ll need to have a full essay to complete my thoughts on Bitcoin, but I’ll leave some of my current thoughts on this book here:
Much of the book is spent trying to convince you of the merits of stripping governments from the ability to inflate their currency. Some of these arguments are good: I found the argument that avoiding a single instance of hyperinflation might be worth sacrificing any theoretical gains from inflation pretty straightforward and compelling. This is definitely not what I learned in my undergraduate economics courses and made me excited to read more of the Austrian economists that have inspired Ammous.
The arguments for mainstream economic theories he mentions—Monetarism and Keynesianism—are typically swept under the rug in a few short pages. I haven’t dived keep into these theories, so I didn’t walk away feeling like I’ve heard both sides.
Inflation is portrayed as the primary cause of all the world’s woes; ranging from the excess bloodsheds in World Wars 1 and 2, practically every major recession, stagnation in the sciences, and even to what the Ammous considers to be “bad art” (he’s not a fan of modern art, or Miley Cyrus). The range of topics—especially the 6 page rant about modern art—briefly made me feel like I was being sold snake oil.
Even if I’m to accept that the world is better place under a zero-inflation environment, it was not clear to me how Bitcoin gets us there. Due to its scaling limitations, as Ammous admits, Bitcoin itself will need to remain a final settlement layer, with its costly, immutable, gossip-protocol layer rarely used by individuals. The vast majority of transactions would need to occur either on the lightning network, or through 3rd parties. At the tail end of the book, Ammous admits that if Bitcoin were to become a popular choice for cash, the network will tend towards centralization through 3rd party transactions; and—paradoxically—he endorses handing off-chain solutions handled by “competitive markets” (banks..? ) I don’t yet understand how a world where we could end up trusting third parties to make most our transactions doesn’t land us exactly where we started. If we start trusting competitive banks with our Bitcoin, wouldn’t banks just start fractionally lending again? Ammous mentions the lightning network here, but I think the book was written before it was fully functional. I’ll need to do a lot bit reading on secondary settlement layers for Bitcoin, but I was shocked how at easily Ammous wrote the details of the settlement layer off, given his emphasis on the woes of centralization.
Something I was hoping Ammous would touch on, but doesn’t, is centralization of Bitcoins themselves. How do we trust the individuals that have a substantial amount of Bitcoin in a world where it is a dominant currency?
Ammous glorifies the world of the 19th century, where the gold standard was dominant, and many technological leaps were made. However, in the 19th century, there were a remarkable number of social problems, like horrific working conditions for children. Could a world under a Bitcoin, more analogous to the 19th century, suffer from similar issues or not? I don’t think the answer is as straightforward as “governments must still regulate social problems” under a global currency, due to the effect global currency would have on competitive markets. Ammous lauds that, in his ideal world, participation in a given government would be more opt-in than it is today. Much more control is given to individuals over governments, in a world with decentralized currency, and has important social implications. While I agree many important things happened in the 19th century, nothing has ever made me want to go back to the world Anna Karenina and Jane Eyre were imagined in.
[Cryptography] Bitcoin is a disaster. - This piece of Bitcoin decent coming from the cryptography mailing list that Satoshi Nakomoto posted the original Bitcoin paper by Ray Dillinger—one of the first people to review Satoshi’s code. For me, it validated the problems I noticed reading “The Bitcoin Standard”.
“The scarcity of block chain space has led people to re-invent every
last feature of the banks they thought they were going to be escaping.
Including debt brokering (lightning network) and fractional-reserve
banking, starting with the case of Mt.Gox and continuing to ventures
today by "responsible" businesspeople who just don't get, or don't
care, or both, that the entire reason the system existed, as far as the
early adopters were concerned, was to get away from exactly that. They
have made Bitcoin into a debt-based system like any other; as long as
the "exchange" holds your keys for you, there is no obligation for them
to maintain assets equal to the deposits. You can't prove that they
are, or aren't, maintaining sufficient assets until after those assets
are spent and the evidence appears in the block chain.”
…“Bitcoin mining has encouraged corruption (Because it's often done using electricity which is effectively stolen from taxpayers with the help of government officials), wasted enormous resources of energy, fostered botnets, centralized mining activity in a country where centralization means it's effectively owned by exactly the kind of government most people thought they *DIDN'T* want looking up their butts and where the people who that government allows to "own" this whole business work together as a cartel.”
Anti-money Laundering Proposal for Virtual Currencies from the U.S. Department of Treasury - The latest proposal in what will likely be a series of regulation for the cryptocurrency space; including reporting the name’s and addresses of those doing the transactions. I’m really not surprised by this, but it could be big news if you’re anti-regulation in the crypto space. Read the full proposal here.
“The proposed rule complements existing BSA requirements applicable to banks and MSBs by proposing to add reporting requirements for CVC (virtual currency) and LTDA (Legal Tender Digital assets) transactions exceeding $10,000 in value. Pursuant to the proposed rule, banks and MSBs will have 15 days from the date on which a reportable transaction occurs to file a report with FinCEN. Further, this proposed rule would require banks and MSBs to keep records of a customer’s CVC or LTDA transactions and counterparties, including verifying the identity of their customers, if a counterparty uses an unhosted or otherwise covered wallet and the transaction is greater than $3,000.”
2.5 Software Engineering
Recap of AWS re:Invent 2020 - This was an especially cool Re:Invent, and this is a great overview of everything put on display that will only take you a few minutes to read. Here are the things I’m most excited for:
Greengrass V2 - for those interested in IoT and edge computing (I worked on this for a little, so I’ve got to plug this one!)
SageMaker Clarify - One feature that really excites me here is that this is supposed to give SageMaker (the managed AWS machine learning tool) the ability to detect biases within your dataset. I experimented with Sagemaker some years back while building a prototype melanoma detector with a friend, but didn’t find it moved the needle enough to warrant the cost, and found myself moving back to EC2 instances using docker containers for my ML models. I wonder if the new SageMaker features make it worth the cost?
Elastic Glue Preview - I’ve started work on a data engineering team recently and seen some pretty hairy problems where folks need to join datasets across storage layers together. I’ve also seen folks use materialized views (these are precomputed views for your data to speed up queries in services like Snowflake) somewhat often. So idea of a managed service that can create materialized views across data sources sounds like it could save me a lot of work. An old friend of mine from Greengrass is actually a manager on this team too, so it’s a project I’m super excited about.
2.6 Philosophy
Why I’m Losing Trust in the Institutions - Lately, like many others, I’ve felt like I’ve had a tough time trusting our institutions. Especially around the virus. The mixed messaging around masks, failure to take action early enough, the early arguments from ignorance, etc. So, when I saw this article, my eyes initially lit up: what else could the CDC be doing wrong? The author critiques a CDC proposal that suggests frontline
“In one of the most shocking moral misjudgments by a public body I have ever seen, the CDC invoked considerations of “social justice” to recommend providing vaccinations to essential workers before older Americans even though this would, according to its own models, lead to a much greater death toll.”
“In other words, the CDC was effectively about to recommend that a greater number of African-Americans die so that the share of African-Americans who receive the vaccine is slightly higher. In blatant violation of the “leveling down objection,” prioritizing essential workers in the name of equality would likely kill more people in all relevant demographic groups.”
Sounds scary at first. I didn’t look too deeply into the articles claims until I saw the below tweet.
This prompted me to look into the slides the article was based on: Give those a look here. It’s difficult to tell, just based on the slides, how good or bad of a proposal this really was. It’s tougher to know how big a difference there would be between lives saved if essential workers were vaccinated first vs. adults over 65 years old, due to the multiplier effect. Depending on how they weighed each of these concerns, it seems like it might have been a decent analysis.
It Took Two Days to Develop the Moderna Vaccine - In the early days of the pandemic, I got pretty interested in how viruses work and how they’re combated (there wasn’t much else to do, ok) but despite that research, I had no clue a vaccine could be developed so quickly. Not only did this article provide some details on how why the COVID vaccine is safe (it’s already been tested for about a year, which really isn’t that uncommon!) the author suggests some interesting “pre-testing” to try and speed up these kinds of processes in the future. The author is a neurologist, not a virologist though, so I’m taking it with a grain of salt.
“The question now, however, is can we do even better in the future? The idea is to develop a vaccine platform, and test that platform for safety and general efficacy. Then, when a new bug comes along, all you have to do is sequence its genome and then plug that information into the existing platform, and start cranking out vaccines. This literally could be done in a week. Since the platform itself is pre-tested, we won’t need to do months of testing before distribution. Still, we may need to do some testing on the specific vaccine. Perhaps much of this testing can be done with computer modeling, and some with animal testing.”
Side note: Medcram had the best COVID-19 coverage throughout the entire pandemic.
Exclusive AMA w/ Noam Chomsky on Jung, Wittgenstein, and Gödel - I mostly hear Chomsky’s thoughts on political matters, but this AMA was interesting because it focused entirely on philosophical questions. I was especially excited to hear chomsky talk about the “Hard Problem” of consciousness (I’ve been quite interested in it since I first heard about it a few years back).
To see my own thoughts on the Hard Problem more fleshed out, as well as a brief introduction into it, you can read my blog post on the hard problem of consciousness.
Chomsky’s thoughts on the Hard Problem seem in line with what Chalmers argues for: namely that consciousness is fundamental. However, his take on the current philosophical intrigue the Hard Problem is generating was somewhat unique, relating the Hard Problem to the problem of motion in the 17th century.
“If you go back to the 17th century, there was a hard problem: it was motion. Motion was, what was called, the “Hard Rock” in philosophy, philosophy meaning science. We can’t comprehend it. Turns out, it was right, we couldn’t comprehend it. And we still don’t comprehend it. Not in the sense in which Galileo, Leibniz, Newton, and the other founders of modern science wanted to understand things...” —Chomsky
Chomsky goes on to say that the search for the “Hard Rock” was abandoned after Newton was able to explain the phenomenon of motion in purely mathematical terms, and that he expects the consciousness will end up being quite similar.
“…it was correct to give up the search for an intelligible account of motion and move on to develop theories that explain it…” —Chomsky
It reminded me of a video I saw once of Richard Feynman explaining why magnetic forces are so difficult to “explain”. Forces—much like consciousness—seem to be bedrock. You can’t really explain their existence in terms of other things (even though you can explain their behavior).