Explanation: The Ten Commandments of the Internet

Overview

In my last post, I shared the Ten Commandments of the Internet.

But because they're written with some intentionally vague language (so as not to immediately lose meaning if today's jargon isn't tomorrow's jargon), I thought they warranted a bit of explanation.

Terminology

Here are the terms used:

Explanation

I

Thou shalt assume all strangers to be knaves or false spirits, unless thou knowest them to be of true flesh and noble soul.

Unless you know who someone is online, assume it's either a troll or a bot. Operating on this assumption, until proven otherwise, will keep you safe from all the pitfalls that come with assuming others are earnest an well-intended, such as

• Simply wasting your own time
• Falling victim to concerted psychological attacks
• Being suckered into ruining your day over a pointless internet argument

II

Thou shalt not parley with knaves, nor bandy words with false spirits, nor witness utterances between knaves and false spirits.

Don't engage in discussion or disagreement with trolls or bots, or even bother to read discussions between any combination of troll and bot.

If you follow the first commandment, that means don't discuss or disagree with anyone, unless you know that they are a real, earnest person.

This doesn't preclude any engagement with trolls or bots, which would mean "don't engage with anyone you don't know" – which isn't tractable online. But it does mean you should ignore any attempts to engage in argument with anyone you don't know.

III

Thou shalt not act a knave, nor sow discord for mirth or malice.

Don't be a troll, whether it be for fun or bad intent. The internet is bad enough without you actively making it worse, so don't be part of the problem. This is kind of like "thou shalt not kill": it really doesn't need much explanation.

IV

Thou shalt take all tidings as falsehoods, until proven true and borne of a righteous source.

Assume any claim made online is false, until you (or someone else) proves it to be true, and traceable to a reliable source of information.

The obvious implication here is that, if it can't be traced to a reliable source of information, it should be treated as a lie – even if it's something that superficially appears true or mostly true, or even if it appears to have come from that source.

V

Thou shalt doubt all fonts of knowledge until they be shown to be steadfast and true.

Assume any source of information is unreliable, even one with a name that sounds trustworthy, until it's shown to be repeatedly reliable.

One accurate news story isn't enough to transitively decide that another is. A long history of reliable journalism, however, probably is.

In the past we treated written word as imbued with some sense of reliability, with the common justification that no scribe would have taken the time to write something unless they thought it to be true. While that has been false forever (anyone with enough money could pay a scribe to write their lies), we have passed an inflection point: Now that written word is cheaper than it's ever been, and will only get cheaper as AI lets us generate mass content on demand, we can't even use the justification that somebody bothered to write it at all.

Because of this, it's similarly trivial to create a false "news agency," (there are even games about this). Which means that we can't even use the analogous justification that somebody bothered to create an organization, therefore they must be reliable.

In other words, we must assume that any source of information, and any content shared through that source, is, at best, to be treated with a strong sense of doubt, until they prove themselves to be consistently reliable over time, and through common social consensus between reliable humans.

VI

Thou shalt not declare as truth, save what thou traces to a font of firm foundation.

Don't make any claims about truth or fact without yourself tracing it back to a reliable source of information. Don't share or repost anything without doing the same thing. Don't state opinions as fact. Don't posit anything unfounded and traced to a reliable source of information.

VII

Thou shalt not share utterances born of vapors and vanity, lest they lead others astray.

Don't yourself make any claims with unfounded positions, even as a joke. This doesn't necessarily mean you have to cite a source for everything you say, but an unfounded position is one that lacks basis in fact, evidence, and reason.

This is similar to the third commandment, inasmuch as it basically means "don't be part of the problem."

VIII

Thou shalt deem all fonts and utterances as crafted by maligned golems, unless it be made known they come from mortal hand of true flesh and noble soul.

Unless you know a well-intended human curates some source of information, or made the content you're seeing, assume it was made by an algorithm specifically designed to act against your best interests.

A growing majority of internet content being presented to us is gated through algorithms, whose design is to further someone else's agenda. Someone who does not care about you. That agenda might be to maximize revenue or engagement, but it might also be to sow social discord in another country, or foment malcontent in that society.

If you are not certain that the content comes from a reliable mechanism (hand-curated, latest story first, etc.), assume that someone else made a conscious decision whether to show the content to you, and further, assume that whoever made that decision does not have your best interests at heart.

The same goes for the content you see as well: unless you know it was written by a human hand, assume it was automatically generated by AI or an algorithm, and assume that the prompt used to generate it conveyed an intent not aligned with your best interests.

IX

Thou shalt not take the word of any golem, maligned or not, to be gospel, but seek witnesses and signs to confirm it.

Do not trust anything generated by an algorithm to be trustworthy, and seek to corroborate the information through other sources.

This is not just critical because a bad actor might be crafting content with bad intent, but also because even well-intended users of AI might accidentally generate information that isn't reliable.

X

That which is writ upon the net shall endure unto the end of days, and feed all golems borne there; post thou with reverence and fear.

Consider carefully what you say and how you say it, and treat everything you add to the internet as undying. Treat everything you add to the internet as training data for new algorithms and AI models. This isn't strictly true for all content, but if you operate on the assumption that it is, you are much less likely to be haunted by any content you added in the past (even misrepresented).