After meeting in person with my teammates for work, I realized it’s far easier to joke around and riff in person rather than on Zoom. For this group of people, my primary means of communication was Zoom for the past 6 months and the effects were noticeable.

It has something to do with removing latency and improved feedback with body language. Everyone can hear even if there are overlapping voices. Jokes are more responsive.

See also:

• Self-presentation causes zoom fatigue, less so for in-person meeting with close colleagues.
• Low-bandwidth collaboration would benefit from a snappy way to make jokes. For example, a funny Slack emoji to react to a comment.
• The effects of remote work on collaboration among information workers. An under looked at cause might be jokes.

In anticipation of a wave of cases due to the COVID-19 omicron variant, NYC requires private employers with 100 or more employees to mandate vaccination for in-person workers. Every worker must have their first vaccine does by December 27, 2021.

This is a sweeping mandate is the most ambitious to date. However, the next NYC Mayor takes over in January so it’s anyone’s guess if it will stay.

See also:

• OSHA mandate-or-test workplace vaccine rule pushes COVID-19 prevention onto employers (and is held up in court battles)
• COVID-19
• American values are incompatible with managing the pandemic

After looking into linting prose in Emacs, I found a way to roll your own prose linter setup using vale (an extensible prose linter), efm-langserver (a generic language server), and eglot (a language server mode for Emacs).

Configure efm-langserver

In /${HOME}/.config/efm-langserver/config.yaml add the following setup for linting with vale. (This was difficult to figure out because of the non-existent documentation, but I found someone else’s efm-langserver config for vale).

version: 2
root-markers:
  - .git/
log-file: /{SOME PATH FOR LOGS}/efm.log
log-level: 1
tools:
  vale: &vale-lint
    prefix: vale
    lint-command: 'vale --output line ${INPUT}'
    lint-stdin: false
    lint-ignore-exit-code: true
    lint-formats:
      - '%f:%l:%c:%m'

languages:
  org:
    - <<: *vale-lint

Configure Emacs

In your Emacs init.el, configure eglot to use efm-langserver when using org-mode. Vale has support for org-mode files as of v2.2.

(add-hook 'org-mode-hook #'eglot-ensure)
(add-to-list 'eglot-server-programs '((org-mode) "efm-langserver"))

Now when you open an org-mode file in Emacs you’ll see all your prose linting issues!

See also:

• Turn Emacs into a focused writing tool
• Using an LSP server with custom vale styles could be an interesting way to set up an organizational linter for engineers

When there is no formal structure in a group of people that interact over any significant amount of time, informal structures appear. Informal structures are communication channels through networks of friends in the group that share similar beliefs and traits that give rise to influence. This is the nature of elites—a small group of insiders that exhibits informal influence over a larger group because they know how decisions are made.

See also:

• This idea comes from my reading of The Tyranny of Structurelessness.
• The concept of elites seem to be explained by dyadic ties in network theory: the stronger the tie between two individuals, the more overlap they have in friendship circles. By definition it is exclusionary and the flow of information is more tightly bound to that group.
• In-group favoritism shows that elites (the in-group) will give preferential treatment to other members of the in-group.
• All bridges are weak ties, for elites that means they can have a wide reach if they also maintain a large number of acquaintances (such as a following on social media).

Emacs can be customized to create a beautiful, focused writing environment akin to Ulysses, iA Writer, Bear, or other minimalist markdown editors.

Reduce visual noise

• Use writeroom-mode to hide frames and headings in the default emacs chrome
• Compile emacs for the natural title bar in macOS. This further reduces visual noise and makes for a more minimalist look and feel.
• Use a theme like doom-plain-dark or doom-plain-light for fewer colors and visual cues that can be distracting.
• Change links styles to remove bold (it’s already underlined).
• Hide markup characters like ~/emphasis/ by enabling org-hide-emphasis-markers to reduce the number of sigils like brackets, parenthesis, tick marks, etc.

Use a different font for writing than for coding I personally prefer monospace fonts for writing because it pairs well with org-mode headings and gives it a typewriter look and feel.

I don’t like to use the same font for writing as I do for coding. Mainly because it needs to look good larger than what I code in (Cascadia Code) and having a separation between writing and coding makes it easier to concentrate (“I’m in writing mode now”).

Monospace font recommendations:

• iA Writer Duo (what I’m using)
• Colophon Space Mono (more rounded and spacious)

Other proportional fonts:

• Monotype Bembo (used by Edward Tufte, similar to Georgia or a Garamond to give it a journal-y feel)

Transparency You can add a little extra flair by adding transparency to the background. Eval the following elisp expression using M-:, (set-frame-parameter (selected-frame) 'alpha 80).

See also:

• Using vale with emacs for prose linting further improves the Emacs writing environment
• This is a good example of improving the ‘hand feel’ for writing on a computer

The Tyranny of Structurelessness by Jo Freeman.

Some of the downsides of informal structures in an unstructured group are:

• Decision making: people will listen to others because they like them, not because it is a good idea or significant
• Accountability and responsibility: the informal group has no formal responsibility to the wider group. Their influence is not based on what they do and so they can not be compelled to act in the best interest of the entire group.

Read The Tyranny of Structurelessness.

When rules are implicit, only those who know how decisions are made can participate. The only way one can know how decisions are made are to spend enough time cultivating friendships with the elites that establish the informal structures of the group. Therefore, informal structures are exclusionary (one must be a member of the elite) and have other negative consequences.

Formal structures counteract these issues by making decision making explicit. Everyone now has the opportunity to participate because it’s more clear how to do that.

See also:

• This idea comes from The Tyranny of Structurelessness
• Unstructured groups form the basis for elites

An elite individual is an oxymoron. The nature of an elite is membership to a group that exhibits informal influence or control over a larger group. This is often misused in politics “candidate x is an elite”. Regardless of how well known an individual is they can never be an elite.

From The Tyranny of Structurelessness.

A commonly cited reason for Web3 is that centralized systems (government, banking) are broken and decentralized systems are the way forward (blockchains, cryptocurrencies). The problem is that fully decentralized systems have all sorts of issues in the same way that, for example, truly free markets do—exploitation, theft, perverse incentives, and so on. As such, engineering solutions can only go so far and are doomed to rediscover the need for hierarchical regulation to make it work.

Read 100 years of whatever this will be from apenwarr.

A problem with Twitter’s advertising business is that it’s product is used for text-based information (tied to the physical world like work or current events). Users are more engaged and combative on Twitter, making it a poor fit for direct response advertising (users are probably more annoyed by it too).

Contrast that to Instagram, images and short videos that are highly produced to show your best self. Users are taking a break from something else to look at it and the aspirational content makes it a great fit for direct response advertising.

From Twitter Has a New CEO; What About a New Business Model? by Stratechery.

A commonly cited success story of NFTs is CryptoPunks. They were 10,000 uniquely generated pixel art avatars released on the blockchain for free. They’ve since become a hugely sought after item with market prices greater than $7MM.

I didn’t understand what’s going on until hearing Kevin Rose and Aftab Hossain talk about collecting them—these are Star Wars action figure collectors sharing their passion for buying the digital equivalent of a Rocket Firing Boba Fett.

It’s far more endearing to think of NFT enthusiasts as Star Wars action figure collectors.

See also:

• CryptoPunks prove that NFTs provide digital scarcity because there is a market (buyers and sellers).
• Web3 as weaponized Star Wars action figure collectors is an interesting dystopian angle.
• Bitcoin and Ethereum use the same amount of electricity as the country of Austria whereas those tiny Star Wars action figures were a one time plastic. The difference in relative pollution curves might match the difference in price ($150,000 for the most rare Star Wars toy, $7MM for the most rare CryptoPunk).

Zettelkasten notes become more useful the better links you make between them. That’s because you want to find connections to things you did not expect, not what you already know (otherwise you would have gone directly there to begin with).

To improve links, every note should have links to a more specific idea, a more general idea, and a related topic.

A good test of the quality of the links between notes is to see how often you search directly for a note vs follow a backlink.

Read Make Your Notes Work for You: the Secret Sauce of Zettelkasten.

See also:

• In reflections on writing 500 notes, writing notes as definitions of terms is also not useful and it’s better to write as if you were making claims.
• This works for personal notes, but it is not enough for multi-party content—hyperlinks as a foundation for distributed knowledge fails because it lacks transclusion.

I came across the idea of a hyperfine village from Lisa Hardy. It’s a novel way to organize your ideas into a metaphorical “village” so you can more easily recall them in context later. Rather than search or rote memory to recall an idea later, you can go for a stroll in your village.

Hyperfine village might include:

• The Cabin: journal, wardrobe, home-related things.
• The Lab: experiments, work-in-progress
• The Museum: finished works for archive
• The Office: task lists and project management
• The Library: books read or reading
• The Garden: notes and ideas
• The Planetarium: visions of the future
• The Gazebo: community and people

This can be extended or changed as needed similar to how you might use a “mind palace” to assist with memory.

See also:

• This could be easily implemented in Zettelkasten or any heterarchical note taking system or tool for networked thought since they don’t impose a single organizational method.
• Locations have a builtin epistemic status—things in “The Lab” have permission to be less finished than say in “The Museum”.
• Ideas for making notes more inviting
• Digital gardening
• The Garden and the Stream
• Our digital lives are siloed

Zero-knowledge succinct non-interactive argument of knowledge (zk-SNARK) can verify that another party is in possession of information without the other party needing to reveal that information or leak parts of it. This is useful because it provides a way of doing private transactions in public.

In a distributed, trust-less system like a cryptocurrency, zk-SNARKs can be used to make private transactions that are verifiable and stored on the blockchain (Zcash already does this, but for the most part blockchains are anonymous, not private). That means privacy conscious parties can benefit from using a blockchain while maintaining privacy at the transaction level and decreasing the chances of being de-anonymized at the transaction history level.

What is this used for?

Extrapolating a bit, there are interesting things you can do using zk-SNARKs with blockchains. You could verify an account balance by checking if it has enough available funds without leaking the balance or the account. You could verify the possession of a certificate or license without revealing any information about the certificate itself (like checking an ID without having to look at it). You could pay taxes without revealing your assets. You could automate verification of things like lottery tickets or possessing a certain NFT.

See also:

• Web3

Epistemic status: low

1. Translate the check function into a polynomial equation (how do you translate information into a polynomial equation?)
2. Verifier picks a random point encrypted using homomorphic encryption
3. The prover computes the polynomial encrypting the values in the polynomial
4. The verifier checks equality in encrypted form zero knowledge is achieved by adding some random numbers to both sides

One downside to SNARKs is that they need a trusted setup which some consider a flaw.

See also:

• Trust models

Blockchains are transparent ledgers in that anyone with a copy of the blockchain can see every transaction. This is necessary due to the way blocks are validated. However, even though transactions are anonymous they are still public and can be used to associate wallet addresses with someone’s identity.

See also:

• Web3

This note does not have a description yet.

Smart contracts are not recognized as legal contracts and don’t offer the same protections. They are more like stored procedures in that they are scripts that execute on a blockchain.

Vitalek Buterin, founder of Ethereum regrets naming them smart contracts.

See also:

• NFTs are a subset of ownership

Smart contracts that run on a distributed ledger effectively need to work forever to be useful. Working forever requires maintenance and security—the longer a computer program runs, the more exposure it has to change and the opportunity to be exploited.

However, programming as an industry has little experience with this kind of operating model. It’s not surprising to see large scale exploits or errors. I think we will be surprised when maintenance burdens and tech debt cause issues for higher level abstractions (DeFi) or stifle development of the underlying protocols.

See also:

• An Engineer’s Hype-Free Observations on Web3 literary notes
• Smart contracts are not legal contracts
• Smart contracts and their accompanying blockchain need to exist long enough for the Lindy Effect to kick in
• Web3

NFTs lack the legal and regulatory protection of other forms of property. In the event of a dispute, there is no oversight. This fundamental principle of crypto projects (decentralization, no middle man, zero trust) limits what can be represented as a non-fungible token.

See also:

• An Engineer’s Hype-Free Observations on Web3 literary notes
• Smart contracts are not legal contracts and since NFTs are implemented as smart contracts this might have some bearing on the rights of ownership (but I’m not a lawyer!)
• Web3

Distributed autonomous organzations (DAOs) have more in common with limited liability corporations (LLCs) than C Corporations. There is near infinite variability possible in an LLC’s operating agreement similar to how DAOs can be grant DAOs, investment DAOs, protocol DAOs, and many more. C Corps on the other hand have converged on, more or less, a standard—Delaware C Corp and a well-known governance structure based on ownership of stock.

See also:

• Smart contracts are not legal contracts
• Continuous organizations don’t make sense for high growth startups
• Web3

The World Health Organization (WHO) classified B.1.1.529 (omicron), first detected in South Africa, as a COVID-19 variant of concern.

One model from the Financial Times suggests the omicron variant is 500% more transmissible than the delta variant.

See also:

• The COVID-19 Delta variant is 60% more transmissible than the Alpha variant

Read An Engineer’s Hype-Free Observations on Web3 (and its Possibilities) from PSL.

See also:

• Web3