Dark, Gray, and Light Lore

The three kinds of lore and formulas for crafting them

Part 5/7 of the On Lore series.

Can lore can be crafted, and if so, how?

I think the answers are yes, and using one of three playbooks — dark, gray, and light.

You can craft lore from scratch, and it is easiest to do so if you’re up to no good. This is dark lore, and it points to how you might approach what we actually want: light lore, by way of something in-between which I call gray lore.

This is going to be a rougher set of notes than usual, since I’m not entirely satisfied with the account I have here so far. But I hope this rough initial scheme of dark/gray/light lore is useful for those of you following along on this series.

Dark Lore

Both internet-era works like Ryan Holiday’s Trust Me, I’m Lying, and the older work of thinkers like Jacques Ellul on classic propaganda, point to age-old traditions of creating lore from whole cloth, using devices such as pseudoevents.

Cultish subcultures of true believers can be created by planting the right kind of seeds, and they will then go on to generate vast dreamscapes of self-delusion on their own, congruent with the original suggestive seeds and the agendas driving them. This is the dark lore.

The formula that has emerged in the last decade seems to be:

1. Take a sufficiently complex and anxiety-inducing complex hyperobject (climate, vaccines, personality models, culture war, injustice, civilization, the meaning of life) where there is a vast amount of phenomenology and no obvious logical starting point, and many desirable conclusions are reachable so long as you start in the right places.

2. Offer a simple starter set of half-truths and hard-to-disprove lies that instantly alleviates anxieties and panders to the conceits and natural inclinations of a population with an interest in the hyperobject. A starter set that just happens to lie in the basin of attraction of conclusions that favor your interests, reachable by people “doing their own research.”

3. Offer an intuitive but not over-scripted pointer to a path to a satisfying view of the hyperobject with unending amounts of nerdy detail to get into along the way, along with a vague promise of decisive future agency (bitcoin solves this, just accept Jesus, MAGA).

4. Vaguely link this appreciative view to everyday ritual actions (consuming news, watching prices, prayer, protesting, stacking sats, being active on social media) that are stand-ins for promised future agency.

5. Provide a gathering space — today the internet is the default space — for the target population to self-organize and start accumulating lore, and rough gamified pathways along which they can reach predetermined conclusions by “doing their own research.”

That’s basically it. The formula for creating dark lore from scratch parallels that for creating a fandom around fiction. As in fiction you retain the ability to dictate what is and is not canon. As in fiction, you control the “ending” and everything else depends on the ending the landscape is herding people towards.

It just takes less work and resources because you’re drawing on reality itself to provide much of the source material, with crowdsourcing doing the rest.

The “tell” of dark lorecraft is that if you actually “do your own research” following the starting points and cues suggested, you’ll reach predictable conclusions useful to the lorecrafter — a conjuring trick that produces an illusion that resembles a popular cargo-cult view of how scientific discovery is supposed to work.

It works because complex systems have exploitable initial-condition-dependence dynamics. So if you pick the starting points and provide minimal cues, you can control the outcomes. The pathways from a given set of starting points may be unpredictable (fostering an illusion of autonomy in traversing them) but the outcomes are not. You could replace “do your own research” with “pick a card.”

Dark lore preferentially selects for the most desperate people, most traumatized by doubt and anxiety whirling about a hyperobject, and most eager to latch on to certainties. As with Nigerian Prince scam emails, only the most vulnerable marks get through the gatekeeping. The difference is, the vulnerable are not the dumbest (in fact they are often smarter than average in narrow ways), but the most desperate for ways to manage their psyches (which, if you recall, is the function of lore).

In the past decade or so, thousands around the world got good at running the dark lorecraft playbook. The worlds of lore they craft are typically based on transient false hopes that last just long enough for the dark lorecrafters to turn a profit and exit the scene before disillusionment sets in.

Distasteful though this formula is, dark lore does illustrate some of the necessary elements in the simplest possible configuration that can work to generate lore.

Gray Lore

In midwit company, if you ever mention Myers-Briggs, chances are, a midwit with some modest statistical ability (the sort of person who likes phrases like “correlation is not causation” or “that’s a false dichotomy”) will make some sort of superior remark about how it’s all superstition that does not replicate, and how you should use something like the Big 5 instead.

If it’s a particularly diligent sort of midwit, they might gleefully share an article like Uncovering the Secret History of Myers-Briggs, “proving” that it is all a scam invented by a corporation.

If the listener is a different species of midwit, they might point to their own sources arguing that the Big 5 has its own methodological problems.

Myers-Briggs, the Big 5, and interestingly enough, anti-Myers-Briggs midwittery are all part of a regime of what I like to call gray lore. In this case the gray lore domain of personality psychology.

Much of what I do is traffic in gray lore, both as producer and consumer.

People who gravitate to gray lore are typically looking for new and improved methods to manage their own psyches better, but are typically not desperate for such methods. They are not being actively traumatized by lack of such methods. They are doing okay, and looking to do better.

Gray lore is lore that is usable enough to generate value, if handled with care. Typically, the mark of gray lore is that it attracts a much wider variety of people than dark lore. There are smart, midwit, and dumb people, there are imaginative and unimaginative people. There are liberals and conservatives. There are people of all races and classes. There are people of varied educational levels. What they lack is the existential desperation of those who seek out dark lore.

Because they already have varied working commitments to good-enough ways of managing their psyches (ie they are not blank slates, but come with their own initial conditions) participants in a tradition of gray lore tend to diverge rather than converge. They tend to land on a variety of mutually contradictory positions, none of which is obviously decisively superior, and all of which benefit from mutual disputation. They don’t all form a single cult or fandom with a single set of shared basic beliefs.

Gray lore also tends to be more persistent, and too big and divergent for any single dark lorecrafter to control outcomes. In gray lore, you can’t script the endings easily because people have starting points that actually work for them, which they are attached to, and which fall in the basins of attraction of competing equilibria in the landscape.

Gray lore obviously includes things like astrology, Myers-Briggs, the diet and exercise industry, productivity porn, management ideas, and so forth.

Less obviously, it also includes things with a veneer of statistical legitimacy, like the Big 5 personality test, large swathes of social psychology, survey research, and many genres of “technical analysis” in trading. Notably, gray lore rarely reaches for mathematics beyond basic college-level statistics, or takes root in domains that require going beyond, which makes it depth-limited.

Gray lore is marked by there being just enough of value in the tradition that a thoughtful person might be able to extract some utility out of it. Possibly by engaging with the tradition in an ironic or absurdist mode, in a sufficiently bracketed way, and perhaps cross-checking any ideas drawn from the gray lore using other methods.

Gray lore, in other words, is usable bullshit, where the risk of being exploited by dark lorecrafters is manageable, the signal/noise filter problem is tractable, the traditions of disputation are real (even if midwit-grade), and approaches to cross-checking are available to suit a variety of tastes.

Crucially, there is enough actual ground truth there to meaningfully “do your own research” to form your own judgments and arrive at conclusions that are not the same as everybody else’s, and not the result of a conjuring trick. This is one way you can tell it is gray rather than dark lore. Much of the time, in pursuing lines of exploration in gray lore, you’ll get nowhere particularly useful to anyone.

Maybe you don’t believe in astrology, but maybe you read your daily horoscope for fun and because sometimes you get an idea that makes sense for other reasons. It might effectively serve as a random-idea-generator for you (this wasn’t always the case: historically, astrology was closer to dark lore).

Maybe you don’t believe in Elliot Wave analysis, but you follow what fans of the method say about stock price movements because it allows you to look at those movements in a fresh way and break out of your own ruts.

Maybe you don’t believe the supposed “statistical significance” of the Big 5 test is any more meaningful than Myers-Briggs lack of it, given how it is grounded in muddy use of language, but you look at Big 5 papers that use it anyway, simply to get an unreliable starter view of the phenomenon being studied.

What redeems gray lore is not that it is intrinsically safe, but that there are ways for reasonable people to engage with it safely, and create a divergent, open thought space for themselves that’s friendly to actual insight and open-ended evolution. Gray lore usually doesn’t trap you and cut off contact with a broader universe of ideas the way dark lore does.

Gray lore still exerts selection selection pressure though, driving different outcomes for those with and without the critical abilities to get value from it. Often the difference is not brains (we’re all one-sigma midwits here in gray lore culture), but whether or not you have the right few friends with whom to navigate gray lore. But it needs to be just a small group, not a massive crowd in which you can dissolve your individual identity.

Dark lore tends to put you in large stadium-sized crowds in the company of hundreds or thousands aggressively enforcing a thought monoculture upon each other, because aggregation in homogeneous crowds suits the interests of dark lorecrafters.

Gray lore tends to put you in the company of a handful of friends, with weak links beyond, with each small group forming a somewhat localized and customized instantiation of the larger body of lore. Instead of a lore monoculture, you get a polyglot lore culture.

In general lore is hazardous to navigate either alone, or in too-large groups.

I suspect the right group size for navigating lore is about a dozen individuals who trust each other far more than anyone outside the group. The default distrust even extends to charismatic authority figures whose work they may draw on. Gray lore is generally dominated by this group size (as an aside, the group-size distribution of a lore domain is probably a strong indicator of whether it is dark/gray/light).

Can you craft gray lore? Yes you can. The way you craft gray lore is by crafting a generative, internally consistent, learnable system of thought that takes some skill and time to master. This is why games are a great medium for distributing gray lore.

The learnable “system” at the core of gray lore is typically something anyone with a modest amount of intelligence and imagination can master. Therefore, unlike dark lore, it cannot be easily monopolized or autocratically controlled.

You could say dark lore typically features a rule-by-law epistemology, where crucial ways of knowing are handed down by one or more figures who are “above” the law and lay it down for others, while gray lore typically features a rule-of-law epistemology, where the only source of authority is the core system which all are subject to.

This is one reason I suspect Web3 has gravitated mainly to gray lore — the technology itself is a rule-of-law technology. It may evolve further to light lore, but that remains to be seen.

Also unlike dark lore, the systems are not hard to create. Many people create them, creating a degree of healthy competition of producers in a domain (think diet fads for example). I myself have peddled quite a few gray lore systems.

Within the system, there is skill involved in being at least internally more or less correct or consistent. Whether or not you believe in astrology or Myers-Briggs, there are ways to be more or less “correct” in both domains.

Whether or not you believe the gray lore of the Gervais Principle which launched my consulting and blogging careers, there are in fact correct and incorrect ways of applying the model.

What makes gray lore gray though is that there are still risks of self-delusion, and the potential for long-term upside often isn’t high enough to entirely make up for it. Getting value out of gray lore in a stand-alone way is therefore something of a midwit grind.

And no matter how much you grind, you’re not going to stumble on earth-shaking, world-changing insights and ideas. Only modest ones that might improve your own life somewhat.

Gray lore is not physics. Playing with Myers-Briggs or Big 5 is not going to lead to the invention of stable fusion reactors or inspire the creation of a billion dollar business. If you have other things going on, gray lore might play a supporting role, but that’s about it.

That brings us to light lore. Light lore is lore that can do those things.

Light Lore

Lore is light lore if it fosters a non-trivial probability of truly significant, non-accidental, high-impact developments, and this probability is higher for individuals who are exceptional in illegible and variable ways that cannot reliably be selected for.

But crucially these outcomes are not central to the lore itself, merely a measure (but not a metric) of the lightness of the lore.

In fact, outlier outcomes are viewed with a degree of suspicion, and dominance by them is resisted in the spirit of Goodhart’s Law (“when a measure becomes a metric, it ceases to be a good measure”). In curating light lore, you want to note the presence of such outcomes without making them an optimization or maximization target. Just as you might want a healthy population of blue whales in the ocean, but not necessarily an infestation of them.

In other words, light lore is lore that increases the probability of outlier outcomes driven by outlier people, but does not center either.

Light lore is lore that creates conditions conducive to epic outcomes, but as we saw last time, it doesn’t center epic outcomes. Instead, it centers the everyday behaviors that make them likelier up to a point, and naturally tends to regulate that point. The non-epic outcomes are just as important, just as krill are just as important as blue whales to the health of the oceans.

Silicon Valley is a good example. Whether or not you believe the output of Silicon Valley are good for the world, there is no question that it produces significant, non-accidental, high-impact developments, and that individuals who are exceptional in some way are more likely to be behind them. Perhaps not exceptional in a way you admire, or in a way you could manufacture with standardized testing and education, but exceptional nonetheless.

And Silicon Valley is full of lore. How to come up with good product ideas, how to acquire customers, how to raise funding, how to make a good pitch deck, how to price a round, how to solve a two-sided market problem, how acquihires work, how to tell good VCs from bad VCs, how to tell good and bad startups apart, how Elon Musk thinks... the sheer quantity of lore in Silicon Valley is overwhelming. The world of Web3 pales in comparison, despite its more direct and deliberate focus on quality lore.

This lore creates the conditions where outlier people can produce outlier outcomes. These outcomes might later be narrativized as stand-alone epics, but that tendency is somewhat at odds with the inner currents of the lore itself, as we saw last time. In Silicon Valley for example, the dominance of FAANG is generally viewed as being unhealthy for the startup end of the ecosystem, and the epic stories of such individual outlier outcomes are generally downplayed when telling the story of Silicon Valley as a whole. The ecosystem depends on a sufficiently varied population of people at the intake, not a bunch of people all trying to be ersatz clones of Steve Jobs or Elon Musk.

A common failure mode of newcomers to Silicon Valley is in fact to overindex on the self-mythologizing epic narratives of the outlier successes, while being ignorant of “early stage” lore that actually gate-keeps the culture.

It is telling that outside Silicon Valley, the Steve-Jobs-impersonator Elizabeth Holmes managed to fool investors operating by an epic mental model of Silicon Valley based on FAANG founder archetypes. They still think of her as a product of Silicon Valley culture even though she raised the bulk of her funding from the old economy. Within the core of SV culture, she was largely rejected and failed to raise significant investment. The light lore of Silicon Valley worked internally, while the the epic mythos of it that prevails outside led outsiders astray.

Another good example is science (not Science!™ in the cargo-cult sense).

The popular (and often cargo-culted by dark lore) view of science as a domain of “rigorous” truths and “methodologies” that look like bureaucratic procedures is nowhere to be found in actual scientific institutions. Science as it is actually practiced is again full of lore. And again, there is a deliberate emphasis on the overall culture of science as opposed to (say) cults of personality around Nobel prize winners or “famous” labs. Again, naive entryist autodidacts operating with an external epic view of the domain often stumble (in physics, autodidacts always seem to overreach with epic theories that gesture at vaguely explaining everything rather than actually explaining any one tricky thing; they often think they just need someone to work out the math for them).

Can you craft light lore?

Yes. The trick is to encourage genuine variety in objectives on a sufficiently rich frontier, regardless of whether or not they serve your purposes as lorecrafter. When people are trying to actually do a variety of different things, there is a limit to how much value there is in sharing systems of thought (gray lore) or starting beliefs (dark lore).

But what does get shared tends to be significantly more valuable as a result, simply by virtue of being proven useful in a the context of a variety of individual goals, initial conditions, and idiosyncratic behavioral systems.

What makes dark lore dark is that it grows around the singular objectives established by the lorecrafters. You choose the path, but the start and end points are chosen for you.

What makes gray lore gray is the singular “system” established by lorecrafters. You choose the path, and the start points, but the vehicle is prescribed for you, which limits the range of end points you might reach.

What makes light lore light is that you have to choose your own objectives, and build your own systems for trying to get to them, but you get a rich supply of bits and pieces to build with, significantly improving your odds.

The job of the lorecrafters in a domain of light lore is to provide a sufficiently useful selection of bits and pieces, and build in systemic resistance to monocultures of either objectives or systems.

Light lore is ultimately just the byproduct of ordinary efforts to muddle through towards something you want, through trial-and-error, while being mindfully curious about what actually happens, and responding with a live intelligence by changing your beliefs, behaviors, or both. When hundreds of people do this, light lore begins to accumulate.

It is precisely because this is such an ordinary tendency that it takes a great deal of effort to curate and protect it. The charisma and drama of the extraordinary is all around us all the time, seeking to harness the ordinary to its own ends.

The job of the light lorecrafter is ultimately to get people to believe in themselves, rather than in something outside of themselves.

And it is hard because it is such an obvious thing to try to do.

Next: Part 6/7 — Mopium, Copium, Hopium