Friday, April 28, 2017

Basilisk Incident: Dumb or Smart?

On the face it's really fucking dumb.
Humans can't even get good birthday gifts for conspecifics. How are they supposed to know the desires of a superintelligence defined as being unable to communicate with them? Doing tests instead of trusting your first guess is kind of the point of science. Can't do tests on future entities.

On the other hand...

Yudkowsky and co. seem highly convinced that the basilisk meme has nonzero effectiveness and produces donations to MIRI (etc) up to the level of self-destructiveness. It is implausible that Yudkowsky didn't know about the Streisand effect, as the term was coined five years before the basilisk theory.

A truly deviant machiavelli who benefits from donations to AI research would then try to maximize the amplitude of the Streisand effect by maximizing the amplitude of the attempted suppression, on the assumption there's a positive correlation.

Yudkowsky reacted with maximum plausible emotions and repression only restricted by diminishing returns.

So either he's a true defector, or he's really, really, really dumb. Also, plz into emotional continence.


More dumb:

For it to be possible to defect on me, I have to define 'me' as including sensations I do not perceive, namely the sensations of future simulations of me. Or, alternatively, I do feel those sensations, meaning it's not acausal, it's just interaction across spacelike separations, such as time travel. Because that wouldn't break the universe or anything.

Yudkowsky accepts that causal decision theory concludes you should defect in the prisoner's dilemma. In other words Yudkowsky could have discovered that conclusion is untrue rather than trying to invent a whole new theory which incidentally creates the apparent possibility of basilisks.

"Since there was no upside to being exposed to Roko's Basilisk, its probability of being true was irrelevant."
Xeno's paradox was a brilliant dig at the idea that Greek philosophy understood phsyics, motion in particular. Equally, the basilisk shivs Yudkowsky's decision theory. But there's no upside to knowing Yudkowsky's theory has holes in it, now is there?

Classical decision theory already resists blackmail, it simply requires the theory investigator to not stop when they find an emotionally valent conclusion, but to continue until the logic stabilizes.


Yudkowsky's sequences are pretty okay. I want to know whether applying logic consistently really is that hard or if Yudkowsky isn't even genuinely trying. Also, plz into emotional continence.

Friday, March 31, 2017

Democratic Decay

Democracy punishes the responsible and rewards the irresponsible, the responsible die and the irresponsible thrive, and the population degenerates until it is so irresponsible infrastructure starts suffering obvious, unmistakeable failures.

If properly husbanded, the decay can be slow enough to last centuries. However, the decay is monotonic and inevitable as long as Democracy, the Demon Prince, holds power.

Tuesday, March 21, 2017

Folk Epistemology: Granularity

Example from here, via.
The last point is not a trivial one, since although our country is only about 13 percent black, according to a 2001 Gallup survey most people thought the figure was 33 percent, with the average non-white putting it at 40 percent.104 This was roughly confirmed by the GSS respondents in 2000, who also believed that nearly 18 percent of Americans were Jewish, a figure more than eight times too large.105 A very recent 2012 survey found that Americans believe Protestants outnumber Jews in this country by only 2.5 to 1, when the actual ratio is ten times greater.
From the perspective of the average mind, there's often nothing between 0 percent and 33 percent. The other options are 66 percent and 100 percent.
There are more than zero blacks in America, but it's not a majority, so that's 33%.

Essentially this is all confabulation. They have little to no idea what statistics mean, and trying to get quantities out of their fuzzy magnitude estimates is always going to be a pointless exercise in inaccuracy.

The average mind has broad categories. "None," "almost none," "a few," "some," "a bunch," "a whole bunch," and so on, based purely on the emotional experience of what they see in front of them. And why would they bother being more precise? They don't need to take a census, they just need to 'bond' with other average minds by having similar experiences. And maybe sometimes avoid places with "a whole bunch" of blacks.

There's "several times" more Protestants than Jews. (Assuming they even share the surveyor's idea of a "Protestant.") Three is several times, isn't it? If it was 25 times, there would be "almost no" Jews. There's clearly "a whole bunch" of Jews in some places, so that can't be right, can it?

More than half of America can barely grasp what numbers mean when it comes to balancing their chequebook. (Do folk still do that? Did they ever, really?) Why do you think they're going to take the effort to not be 'ignorant' of the correct demographic-statistical ratios?

It's not like they actually can't into numbers at all. Carpentry is not a high-IQ profession. It's just work, an experience the average surveyor could never imagine. However, asking for work from folk who have already been to work and would rather be having dinner is not going to end as the surveyor expects.

This is however a problem when politicians consult the public on things like airplane crashes and terrorism. Emotion drives out reason, even if they were into the work necessary, and you'll get two broad factions: those who say planes "never" crash and those who say 33% of frequent fliers will die in a plane crash. The solution being, predictably enough, don't do democracy.

Thursday, March 16, 2017

Long Yet Still Very Short History of Sophism

Early hominids* evolved to kill other hominids in conflicts by having bigger groups. Especially after evolving throwing and activating Lanchester's Square Law, (PDF warning) having a bigger troupe was crucial. This required bigger brains, which incidentally makes you smarter.

*(Yes I'm starting that far back.)
For completeness I'll also mention that secondarily, good killer instincts evolved and solidified.

Having become smart and deadly, hominids become not only ecologically successful, but ecologically dominant. The only real selective threat was other hominids. This resulted in a troupe size arms race. However, this had an unexpected result: troupes became so big that most members would spend the bulk of their time interacting not with other troupes or the environment, but other ingroup members.

Thus hominids started to form three castes; the warriors, for dealing with outgroup conspecifics; the shamans or priests, for dealing with the nonverbal, nonstrategic environment; and the merchants, who dealt primarily with internal exchanges.

Surface area being what it is compared to volume, merchants are the most numerous caste. For them, the most important resource-acquisition skill did not involve contact with reality, but opposingly with fooling other ingroup members into thinking defection was cooperation, and of course of detecting and defeating the others' efforts to do same.

Result: the arms race stopped being group size, and become liar-liar competition. Humans now practice lying instinctively. Humans have exquisitely sophisticated lying assistance adaptations, such as the consciousness/subconsciousness compartmentalization. However, necessarily, this makes you smarter, and as a result of these two chained smartness arms races, human brain size has increased blisteringly fast over evolutionary history.

Perhaps the pelvic limits on brain size also helped shift the war from group size to liars. Brain spent on having a bigger troupe doesn't help as well for lying and detecting lies, and ditto neurons spent on killing instinct, so there's a tradeoff there.

Result: lying is natural. Agriculture thus inherently creates an environmental niche for a guild of liars. Expert craftsmen at exploiting lies for mind control. This guild will have its own subculture, techniques and tricks, and I call it Sophism, of which the earliest clear record is Protagoras. They are a perversion of the shaman caste: they use their grasp of reality, but are treating conspecifics like a natural problem to be manipulated, instead of as peers like the merchant caste do. Instead of shamans of the forest, they are shamans of the human.

For completeness I'll mention that super-Dunbar sized groups necessarily require anthropology; the science of being human; the shaman type very similar to Sophism.

Sophism has a problem. There's responsible folk, who will have stuff, and irresponsible folk, who are easy to fool, but little overlap. Irresponsible folk irresponsibly waste their windfalls and return to equilibrium. Responsible folk who hit hard times have a tendency to turtle up and survive them.

Humans have strong egalitarian instincts from back before money evolved. With no way to store wealth and troupe size being critical to survival in war, every individual counted and hoarding was pure waste. Similarly, because a strongman is a defector, and defection harms the whole group, anyone who made themselves out to be too dominant was quite rationally torn down by a coalition.

Sophism, the apotheosis of natural lying instincts, can appeal to natural egalitarianism to propose an equal-sounding form of government: democracy. In theory everyone is in charge, but we still get the benefits of division of labour, by making someone else do the actual ruling.

Of course the wealthy know that it's two wolves and a sheep deciding what to have for dinner. Though in reality the wealthy are the wolves, thus two sheep and a wolf, and in reality the sheep are monkeys, so it's a wolf versus a monkey troupe. Before agriculture, troupe size mattered. After, army sized still mattered a whole lot even though armour disabled Lanchester's Square Law, and so the wealthy wolves tended to surrender to Sophist sophistry. Not to mention it soothes their own egalitarian guilt.

(And thus democracy is an envy/spite engine.)

Thus Sophists establish democracy, to establish a siphon between rich and poor, and skim some off the top. It's not like the irresponsible poor are going to notice there's some missing. Plus hey, free stuff.

Sophism first appears clearly in the historical record in Greece. From there it spread around, reaching most of the world within a couple centuries. Folk contaminated by Sophism lost their kings or suffered similar problems.

When Rome fell, we entered a 'dark' age, where Sophism was lost. However, Islam kept (and subsquently suffered from, when Al Ghazali Sophisted them to death) enough texts for the Sophism virus to re-appear in Europe around 1100. Whereupon demogogues started reappearing, you got St. Francis, Luther etc. causing chaos, and ultimately...they lost their kings.

What is Progressivism? It is our particular strain of Sophism. Given a Hajnal folk who believe in Christianity, the easy lies are predictable. These lies get laid down as culture, and then new niches for lies appear on top of them, and thus Progressivism is largely predictable.


For now, I'll be brief about mentioning Sophism resistance, also known as Philosophy. Sophism is parasitism, so sooner or later the population adapts to resist the parasite. Lies are unstable form of predation, as they require the prey to trust the predator. No matter how well adapted, the predator cannot make the prey trust them.

Philosophy's root is the art of checking your facts. Generalized due diligence.


Democracy is socialist redistribution of power. Power is often used to acquire wealth. Democracy thus implies socialism of wealth.

Or more generally, democracy is in itself the defeat and surrender of wealth. The responsible are punished, and the irresponsible rewarded. The responsible etiolate, while the irresponsible multiply. As per Plato et al, the electorate degenerates, moving leftward, becoming ever more vulnerable to Sophist manipulation.

Full socialism is communism. Full communism is full equality. We are only equal in death. 


As part of Sophism resistance, the responsible right starts noticing the irresponsible left voters are a problem, especially as responsibility declines so much that infrastructure starts to break down. For now, they're not realizing that responsibility is a spectrum. There's always a left half of the bell curve, who will suffer envy and independently be more vulnerable to Sophism. Even Massachusetts votes right 40% of the time. Only DC is pure, and that's probably due to filtering.

The Left is not the enemy. Even if the electorate was subject to drastic action and radically shifted right, there will still be a left half. The mind control parasites are the enemy.

More accurately, egalitarian hubris is the enemy. Envy is the enemy. Pride is the enemy. Lies and the lying instincts are the enemy. These ideas may be familiar...

Tuesday, March 14, 2017

Cthulu's Leftism Summary

As long as you accept that leftism == irresponsibility, it's simple. Democracy punishes the responsible. You get less of what you tax, more of what you subsidize. So, as per Plato et al, democracy degenerates the population, making them increasingly irresponsible. Hence, further left, in a positive feedback loop.

Cthulu always drifts left, says Moldbug. But here at casa AI, we insist on proof.

This is a highlight of a key section of a near-future post.

Progressivism is a parasite culture. It exists to use mind control via lies to siphon off resources for the minimum of effort. But it has a problem: the irresponsible folk, who are easy to fool, are poor. Responsible folk, who are hard to fool because they into due diligence, have all the stuff. Progressives must find some way to empower the irresponsible to take the responsible folk's stuff.

Hence, democracy. Let everyone, responsible or irresponsible alike, have one equal vote. The irresponsible naturally envy the wealth of the responsible, and it's easy to partition the electorate such that the bottom half siphons off the top half. Progressives simply skim some off the current - it's not like the irresponsible are going to run the numbers and find the accounts don't line up. (The responsible have already lost, so if they run the numbers it's easy to make them look like sore losers.)

Thus, being responsible is selected against. In the new election cycle, the whole electorate has shifted to be more irresponsible - further left. However, envy is zero sum. The poor are now poorer, the rich are now poorer (except the top 0.01% or so) so it's eternally possible to partition the electorate so that the bottom half can siphon off the top half.

Badly husbanded, it collapses immediately, such as when Russia liquidated the kulaks. Properly husbanded, this siphoning can run for centuries. However, inevitably responsibility drops so low that basic infrastructure starts failing. Ten generations is a rule of thumb for the limit. The end comes at the next economic shock.

Friday, March 10, 2017

Sleeping Beauty is Easy

Some very clever arguments for making it much harder than it actually is.

The Monty Hall problem is easy if you simply write out the possibility tree and count the final nodes. If I get fancier than counting, it gets tricky. I got this one wrong too before I checked using the dumb counting method.
Though there's a trick depending on what I care about.

The options above the line are 50% each. All I can do is change the weightings of the possibilities below the line, which must also sum to 50%.
The brain gets confused because there's more Ts in the lower half. However, since they must sum to 50%, they're worth less. More on this at the bottom.

If you run the experiment once, then it is kind of hard to think about. Instead, run it a thousand times.
If I think the probability of tails is 1/2, I will report it half the time, and my probability will match the number of times I'm right.
If I report tails all the time, I will be right half the time. If I report heads all the time, I'll be right half the time.
If I think the probability of tails 2/3rds, I will report it more often in the bottom, but less often than in the top, and be right half the time. (1/9 + 1/9 + 4/9 = 6/9 = 2/3. 2/6 + 1/6 = 1/2.)

I dunno. Doesn't seem like I need a whole big book of calculus to solve the problem. I just have to put aside my pride long enough to do the dumb counting thing.

Self-location information? Conditionalization? Principle of indifference? Imagine rule? Impressive, clever-sounding things I don't need to know or care about. I just make a stupid picture and look at it.

The trick being my choice of perspective. If I want to be right, I can pick whatever I want, as I'll be right half the time. If instead we're talking about the number of reports the experimenter sees, then it's different. If the observer is writing down each report separately, and I want to generate as many 'matching' entries as possible, it favours reporting tails. Tails causes double the records, effectively doing the experiment twice on that branch.

Wednesday, March 1, 2017

Carving Reality, Not Butchering It

I've found the Less Wrong description to be profoundly disappointing, so let's do better.

When you talk about a group for a purpose, the label used for that group should be causally relevant to the purpose.

I am hungry. I have a stick, a carrot, a steak, and a mug. I can carve up this group in various ways.
"Let's eat something natural."
"Let's eat something fleshy."
"Let's eat something long and thin."

Technically, if I have the correct background knowledge, any of those would work. In this example, were I sharing my dinner with you, any of those, while a bit stilted, are perfectly understandable and we could continue a discussion and ultimately sit down and dine.

However, when doing something difficult, we need every advantage we can get. Apparent understandability is not good enough. If I want to eat something, I should start by considering an edible property. We can say, "Whites are outbred," but 'white' is not a relevant property, and leads to interminable arguments about whether slavs or jews are 'white.'

When investigating, it's entirely possible that nobody knows the relevant background information. If I say, "Long thin things are edible," even formally restricting it to a simple set like above, we may end up trying to eat a stick.

Hajnal Europeans are the set living behind the Hajnal line, and thus have a history of manorialism and Catholic social engineering against the clan and in favour of itself. These things have a causal relationship with outbred nuclear families, which has a causal relationship to democracy resistance and lack of corruption and so on. Their colour, like the shape of the fleshy thing above, is coincidental.

Politicians/sophists love doing this wrong, precisely because it leads to confusions which they can exploit to get you to serve their interests instead of your own.