A magician never reveals his secrets.

The secret behind nearly every magic trick ever performed is available at your local library. Magicial secrets stay secret because they're inconsequential. Unless you are a magician or aspire to become one, you have better things to learn than magic tricks. If magic tricks did anything that mattered then they wouldn't be magic tricks. They'd be technology.

Magicians don't need a conspiracy to keep our tricks secret. It takes work to learn how to do magic. Friction and inertia are sufficient to keep out the riffraff.

This is true of more important subjects too, like computer security. Though zero-day exploits themselves are precious secrets, "how to find" zero-days is public knowledge. And since zero-day exploits have a limited shelf-life it's "how to find" zero-days that matters.

Three may keep a secret, if two of them are dead.

―Benjamin Franklin

Organizations leak like a sponge. Organizations can keep passwords secret most of the time only because a good authentication system is easy to reset. If you're even the slightest bit concerned that your passwords have been stolen then you can re-randomize them. Similarly, an intelligence agency maintains its stockpile of zero-day exploits by constantly replenishing them. To an organization, "preserving secrecy" really means "restoring secrecy". Techniques can't be kept secret because they change too infrequently to restore secrecy after they get stolen.

In practice, organizations face the opposite problem: not enough knowledge is widely-known. Training people is so hard that the limiting factor of an organization's size is how many skilled employees it can hire. The bigger your organization gets the more it'll suffer a regression to the mean. Scaling a company is an exercise in dumbing down your employees' jobs to counteract the regression to the mean.

Large organizations can neither keep knowledge secret nor spread it around. In other words, a dependence on smart people of any kind inhibits the growth of an organization. An organization can scale to the extent it makes its employees'—and especially its customers'—intelligence unnecessary.

SCP-055 is a "self-keeping secret" or "anti-meme".

internal document, SCP Foundation

The largest organizations are precisely those that make knowledge the most obsolete. The public school system is, by headcount, among the largest organizations in modern civilization. It must therefore, by necessity, minimize the need for students to learn anything hard[1].

Most adults are employed by large companies. Most adults buy most of our products from large companies. Small businesses are dying out[2]. Modern civilization is increasingly dominated by large organizations. These organizations don't just shape our society. They are our society. We are our jobs. We are the products we use. We are the media we consume. We are our communities.

Our most popular activities are those that scale the best. Those that scale the best are those that require the least thinking, the least skill, the least specialized knowledge, the least individuality. If you want to measure your individuality, ask yourself this: of all the things you do, how much of it is so hard your friends and coworkers literally can't do it.


  1. By "hard" I mean "conceptual". Schools can effectively force students to learn by rote. However, as a coercive institution, any school with mandatory attendance is definitionally incapable of forcing students to productively misbehave or otherwise exercise critical thinking. (Except to oppose the institution itself.) ↩︎

  2. Small operations that concentrate a lot of talent in a tiny number of employees are doing well. These companies will continue to constitute an insignificant fraction of total employment. ↩︎

New Comment
9 comments, sorted by Click to highlight new comments since:

There's a piece of knowledge available in certain altered states (meditative and entheogen, possibly some manic states) that behaves a bit like scp-055. Right down to the fact that people find it easier to describe in terms of negatives. Repeated exposure allows you to bring a bit more back with you but it seems to make people susceptible to bad epistemics (i.e. most such people wind up woo). I think this negative payload isn't directly bad epistemics but something that collides with people's badly grounded ontology/metaphysics.

It's not quite moral realism, but it does relate to things actually being important/precious in a way that, while in the state, we are concerned that our normal self doesn't seem to understand.

One phenomenological signature of the thing is that it feels 'too big' for normal cognition. Like you need a higher than normal branch factor on your thought process to be able to hold its disparate parts at once.

It *isn't* unity consciousness, though that's nearby in mind-space or has some overlap.

It *isn't* 'no-self' (itself a bad translation of not-self and endlessly confusing for spiritual seekers.)

It *does* seem related to our problems, both object and meta, with moloch and azathoth. At least by my read.

related

https://qualiacomputing.com/2017/12/20/the-universal-plot-part-i-consciousness-vs-pure-replicators/

[-]Yitz140

I find a similar phenomenon occurs with extreme depression. When I’m in that state, I literally cannot remember what it feels like to be happy, though I remember acting in ways consistent with happiness. Likewise, every single time I go into an extremely depressed state, it feels like the worst experience I’ve ever had, even if I know intellectually that it’s been worse before (ie not feeling suicidal, not screaming uncontrollably, etc., when I have before), which leads me to believe that my brain is somehow blocking the extent of the pain I’ve experienced from my memory. Once the experience is over, there is something about it that is inaccessible from my current perspective.

I can't remember pain, in much the same way. Perhaps extreme depression “counts as” mental pain enough to trigger this effect?

[-]nim40

I've socially observed that "many/most such people wind up woo" thing you describe. I have a speculation to elaborate on the mechanism of the payload's collision with peoples' metaphysics:

I notice that a payload I've encountered which sounds to me like what you're describing has a lot of "one-way" references to it throughout art and culture. When I consider the reference without the payload, it's kind of like a dangling pointer, and the mind tends to pick whichever of those possible meanings seems most appealing based on internal state. However, considering the reference and payload together makes it seem profoundly obvious that the payload is the best-fitting explanation for the reference, and the lack of an obvious best-fitting explanation before was due to the absence of that particular payload. I call the link "one-way" because the references are not useful for inferring the details of the payload, but the payload is useful for explaining the references. This contrasts against the usual "two-way" relationship between an observation and its explanation, where the observation usefully narrows the size of the relevant explanation-space as well as the explanation justifying the applicability of the observation. (or, the link between descriptions of the payload and the payload behaves like an NP problem, whereas the links between other descriptions and their described things-like-the-payload behave like P problems)

However, due to this strange behavior, I'm not actually convinced that this payload is "a piece of knowledge" in the sense that it's usually useful to describe things as pieces of knowledge. For things to qualify as pieces of knowledge, I think they need to be transmissible between thinkers and verifiable as having been accurately transmitted, and the payload as I know it meets neither criterion with any sort of reliability. I suspect that the most useful metaphor for the payload and its effects varies based on the reason one's trying to discuss it.

I like the one way two way distinction for pointers-referents. Another thing I can say about the thing is that it seems to point to our boundary/distinction thresholds bring too rigid or too...assumed across reference classes when that's not actually appropriate. Like abstractions in general being held too tightly and treated like territory even though we know intellectually that they are map.

[-]nim20

When attempting to parse that final sentence, I get "the map for which ordinary maps are the territory", and some filter in my head tells me that it sounds poetic enough that I should try to force the thought to do something other than just amuse me and then disappear.

While I think I see where you're coming from in describing those inferences from the thing, I can't really build on them like I would hope to in a proper conversation, because I tend to keep my understanding of it wrapped up in an e-prime inspired thought as a sort of defense against the quagmire of woo that it seems to commonly get embedded in.

This is a great idea. I should add an article on the subject to this series.

You kind of lost me at the end. Isn’t part of the appeal of magic tricks that even though they are easy to learn, they still take work to master, and even if you could do it, you don’t, but you DO enjoy watching someone else perform them?

I think a related, but somewhat opposite observation is: we have more and more niches for everyone. For example, I often run into people who watch TV shows, even TV shows that are similar to what I watch (scifi, fantasy), but we still have zero overlap. That just couldn’t happen even 10 years ago. It’s not that they can’t watch my shows, or I theirs. It’s just we don’t.

Also this: https://www.lesswrong.com/posts/XvN2QQpKTuEzgkZHY/being-the-pareto-best-in-the-world

Though zero-day exploits themselves are precious secrets, "how to find" zero-days is public knowledge.

I'll bite. How does one reliably find zero-days in a given system?