does imagining +singularity cause depression?
How many people feel despair in imagining a heaven (positive singularity) that they'll miss out on if they don't survive long enough? I don't think about it, but I already have plenty of reasons to like being alive.
Problems in Education
Post will be returning in Main, after a rewrite by the company's writing staff. Citations Galore.
Personal Evidence - Superstitions as Rational Beliefs
I'll start with a confession:
The evidence I have personally seen suggests haunted houses are, in fact, real, without given any particular credence to any particular explanation of what the haunting is. In particular, I own a house in which bizarre crap has happened since I first moved into it. Persistently. I've moved into another house, and have been making repairs in preparation to sell it; most recently, in a room with almost no furniture, in a space with absolutely no furniture, a key was dropped by myself. Four people searched the area for significant periods of time on three different occasions with no luck. I found it on the floor a week or two ago on top of something that wasn't there when it fell. Which is the straw that broke the camel's back in terms of my skepticism.
Other bizarre things that have happened include such things as my waking up to discover my recently-purchased bottle of key lime juice had been placed in the oven, and the oven turned on; the plastic bottle had just started to melt when I made the discovery. Another situation involved my sister, who one morning (while home alone) walked into the living room and discovered on a previously empty floor three sonograms of the previous occupant's baby. (There were -many- other things; I'm choosing for the purposes of this post the most unusual and least prone-to-outside-explanation occurrences. Night terrors, for example, are easily explained.)
Up until the last incident, the key, I was inclined to attribute the events to, say, sleepwalking and confirmation bias. At this point, I do not think the evidence really supports that conclusion anymore. My skepticism has been broken by personal experience; I'm not going to attribute anything to any -particular- explanation, but there is definitely something -not normal- about that house, whatever it may be; it has been the (nearly) sole repository of such experiences in my life. (The only other such experience was the day my grandfather (with whom I was extremely close) died, and given the mental turmoil I was experiencing, I'm disinclined to give that particular experience too much credit. For the curious, I was taking a shower, and the hot water repeatedly (3 times) turned off. As in, the knob was completely rotated to shut off the flow of hot water to the faucet.)
A key point of rationality is that evidence can in fact change your mind. Well, the evidence has changed my mind.
From a reader's perspective, this is all anecdotal evidence. So I don't expect to change anybody -else's- mind - indeed, you're probably making a mistake if you -do- change your mind, because out of millions of people, you -should- expect to see a few weird things being related by other people. The odds of somebody else relating an entirely factual series of anecdotes that suggest something unlikely are probably significantly higher than the odds of that unlikely thing being true. However, the odds of such things happening to you personally are considerably -lower- than the odds of hearing about the events from somebody else. Which all leads into a central conclusion: It's possible for the evidence to support one person believing something, while at the same time -not- supporting that anybody else believe that thing. If you win the lottery, that may be evidence for you believing you're living in a simulation or that some other mechanism "forced" the outcome - while at the same time the evidence doesn't suggest anything for somebody -else- winning the lottery.
I have a different purpose in mind: Making the claim that objectively irrational beliefs can, in fact, be subjectively rational. Prior to these experiences, I regarded the idea of a haunted house - I use the idea without prejudice for what "haunted" is or refers to - was that it was just superstitious people scaring themselves. At this point I'm forced by the evidence I've seen to conclude that there's something to the idea, even if it's not what people think it is. Maybe EMFs subtly messing with my brain (there is some weak evidence for the idea that electromagnetic fluctuations can induce metabolic changes in neurons - see http://jama.jamanetwork.com/article.aspx?articleid=645813 ), maybe something else.
If a pattern-recognition algorithm doesn't produce false positives, it's probably getting false negatives, and given that we can test false positives but do not know to test false negatives, pattern-recognition should favor false positives over false negatives. What does this have to do with anything? Well, it means superstitions aren't a product of a poor mind, only -untested- superstitions are. A good intelligence should develop superstitions. It should, when capable, discard them.
But it should only discard such superstitions as it has evidence to do so.
Now, the skeptical reader might ask what odds I place on each of these events occurring. My answer is as follows: Each event was highly unlikely in itself, explainable as an independent event only by positing pretty unlikely circumstances (what odds would I place on me or my housemate sleepwalking multiple times when neither of us have any history of such behavior, and such behavior has entirely ceased since leaving that house? Keep in mind that neither I nor my sister were initially inclined to regard such events as even needing explanation; it's only been until the most recent episode that I've decided the evidence suggests anything at all, so the possible explanation that the sleepwalking was a product of disturbance at the first few unusual events seems unlikely). Further evidence has rendered each event less likely as an independent phenomenon - since moving to a different house, the occurrences have ceased. When returning to the house, occurrences resume within its context. My control, while hardly blind, is controlling. But meaningfully, the same evidence doesn't mean the same thing if it is coming from somebody else; out of millions of people, I would expect such things to occur. I simply cannot expect them to occur -to me-. (And I wasn't the only one who found the house to be... off. There's a sense of not-quite-rightness to one basement room which I cannot explain without resorting to Lovecraftian cliches about alien geometries. The house was burgled several times; the only room that was left completely untouched, even when the copper piping was stolen (and subsequently the water meter - I got a waterfall in my basement!), was that room, which is conveniently where I left a thousand or so dollars worth of building materials for a project I hadn't finished yet.)
Evidence is personal. The odds of something happening are not equal to the odds of that something happening to you. Therefore, while we should not be surprised if miracles (that is, really unlikely and contextually significant events) occur, it is still legitimate to be surprised when they occur to us individually. The qualitative rationality of an individual belief is not equal to the qualitative rationality of the same belief on a social scale; individuals get different evidence than society, even when the same evidence is apparently present both for the individual and the community.
And just as it is a mistake for people to judge the beliefs of others based on the community standard of evidence, rather than the individual standard of evidence, it is likewise a mistake for an individual to judge society based on the individual standard of evidence, rather than the community. Just as it is possible for the individual to rationally believe something that society should not rationally believe as a whole, it is possible for society to rationally reject something the individual has overwhelming personal evidence for.
Aumann was, in short, wrong, because Aumann Updating is based on the belief that two individuals -can- share evidence. Evidence is incompletely transferable.
(Note: Anthropic reasoning can potentially remedy this at least to some extent for -past- experiences; reproducible and continuing experiences somewhat less.)
The meaning of "existence": Lessons from infinity
[Crossposted; Based on Can infinite quantities exist? A philosophical approach (downvoted)
The topic is the concept of existence, not why there's something rather than nothing—not the fact of existence—but the bare concept brings its own austere delights. Philosophical problems arise from our conflicting intuitions, but “existence” is a primitive element of thought because our intuitions of it are so robust and reliable. Of course, we disagree about whether certain particulars (such as Moses) have existed and even about whether some general kinds (such as the real numbers) exist, but disputes don’t concern the concept of existence itself. If Moses’s existence poses any conceptual problem, it concerns what counts as being him, not what counts as existence. Adult readers never seriously maintain that fictitious characters exist; they disagree about whether a given character is fictitious; even the question of the existential status of numbers is a question about numbers rather than about existence. As will be seen, sometimes philosophers wrongly construe these disputes as being about existence.
When essay 19.0 asked “Can infinite quantities exist?” existence’s meaning wasn't in play—infinity’s was. Existence is well-suited for the role as a primitive concept in philosophy because it is so unproblematic, but it’s unproblematic nature can be thought of as a kind of problem, in that we want to know why this concept is uniquely unproblematic. We would at least like to be able to say something more about it than merely that it’s primitive, but in philosophy, we acquire knowledge by solving problems and existence fails to provide any but the unhelpful problem of its being unproblematic. The problem of infinity provides, in the end, some purchase on the concept of existence, which concept I assumed in dealing with infinity.
In one argument against actual infinity, I proposed as conceptually possible that separate things might be distinguishable only concerning their being separate things. Then, if we assume that infinite sets can exist, the implication is the contradiction that an infinite set and its successor—when still another point pops into existence—are the same set because you can’t distinguish them. (In technical terms, the only information that could distinguish the set and its successor, given that their members are brutely distinguishable, is their cardinality, which is the same—countably infinite—for each set.)
What’s interesting here is the role of existence, which imposes an additional constraint on concepts besides the internal consistency imposed by the mathematics of sets. Whereas we are unable to distinguish existing points, we are able—in a manner of speaking—to distinguish points that exist from those that don’t exist. While no proper subsets are possible for existing brutely distinguishable points, the distinction within the abstract set of points between “those” that exist and “those” that don’t exist allows us to extend the successor set by moving the boundary, resulting in contradiction.
If finitude is a condition for existence, we’ve learned something new about the concept of existence. Its meaning is imbued with finitude, with definite quantity. Everything that exists does so in some definite quantity. Existence is that property of conceptual referents such that they necessarily exist in some definite quantity.
Existence is primitive because almost everyone knows the term and can apply it to the extent they understand what they’re applying it to. The alternative to primitive existence is primitive sensation, as when Descartes derived his existence from his “thinking.” But sensationalism is incoherent; “experiences” inherently lacking in properties (“ineffable”) are conceived as having properties (“qualia”). So, the heirs of extreme logical empiricism, from Rudolf Carnap to David Lewis, have challenged existence’s primitiveness. Carnap defined existence by the place of concepts in a fruitful theory. Lewis applies this positivist maxim to find that all possible worlds exist. Lewis isn’t impelled by an independent theory of logical existence, such as a Platonic theory that posits actually realized idealizations. Rather, the usefulness of possible worlds in logic requires their acceptance, according to Lewis, because that’s all that we mean by “exists.” Lewis is driven by this theory of existence to require infinitely many existing possible worlds, which disqualifies it on other grounds. But the grounds aren’t separate. When you don’t apply the constraints of existence because you deny their intuitive force, you lose just that constraint imposing finitude. The incoherence of sensationalism and of actual infinities argues for a metaphysics upholding the primacy of common-sense existence.
Can infinite quantities exist? A philosophical approach
Initially attracted to Less Wrong by Eliezer Yudkowsky's intellectual boldness in his "infinite-sets atheism," I've waited patiently to discover its rationale. Sometimes it's said that our "intuitions" speak for infinity or against, but how could one, in a Kahneman-appropriate manner, arrive at intuitions about whether the cosmos is infinite? Intuitions about infinite sets might arise from an analysis of the concept of actually realized infinities. This is a distinctively philosophical form of analysis and one somewhat alien to Less Wrong, but it may be the only way to gain purchase on this neglected question. I'm by no means certain of my reasoning; I certainly don't think I've settled the issue. But for reasons I discuss in this skeletal argument, the conceptual—as opposed to the scientific or mathematical—analysis of "actually realized infinities" has been largely avoided, and I hope to help begin a necessary discussion.
1. The actuality of infinity is a paramount metaphysical issue.
2. The principle of the identity of indistinguishables applies to physics and to sets, not to everything conceivable.
3. Arguments against actually existing infinite sets.
A. Argument based on brute distinguishability.
B. Argument based on probability as limiting relative frequency.
4. The nonexistence of actually realized infinite sets and the principle of the identity of indistinguishable sets together imply the Gold model of the cosmos.
Rationality by Other Means
Political violence is a terrible thing - but, sometimes, not quite as terrible as the alternative. I recently commented that a post focusing on such things might be worthwhile, and since the voting has been positive, here we are.
[Link] 12 Myths about Hunger
Copy and pasted from here.For those interested, there's an book that expands on the article.
EDIT: people seem to like the first few myths, but they get rather political as the article goes on.
I couldn't find much on hunger on GiveWell's site other than these three articles.
The Advanced Civilization Wiki's page on Food probably covers the good aspects of this article and expands on them.
Why so much hunger?
What can we do about it?
To answer these questions we must unlearn much of what we have been taught.
Only by freeing ourselves from the grip of widely held myths can we grasp the roots of hunger and see what we can do to end it.
Myth 1
Not Enough Food to Go Around
Reality: Abundance, not scarcity, best describes the world's food supply. Enough wheat, rice and other grains are produced to provide every human being with 3,500 calories a day. That doesn't even count many other commonly eaten foods - vegetables, beans, nuts, root crops, fruits, grass-fed meats, and fish. Enough food is available to provide at least 4.3 pounds of food per person a day worldwide: two and half pounds of grain, beans and nuts, about a pound of fruits and vegetables, and nearly another pound of meat, milk and eggs-enough to make most people fat! The problem is that many people are too poor to buy readily available food. Even most "hungry countries" have enough food for all their people right now. Many are net exporters of food and other agricultural products.
Myth 2
Nature's to Blame for Famine
Reality: It's too easy to blame nature. Human-made forces are making people increasingly vulnerable to nature's vagaries. Food is always available for those who can afford it—starvation during hard times hits only the poorest. Millions live on the brink of disaster in south Asia, Africa and elsewhere, because they are deprived of land by a powerful few, trapped in the unremitting grip of debt, or miserably paid. Natural events rarely explain deaths; they are simply the final push over the brink. Human institutions and policies determine who eats and who starves during hard times. Likewise, in America many homeless die from the cold every winter, yet ultimate responsibility doesn't lie with the weather. The real culprits are an economy that fails to offer everyone opportunities, and a society that places economic efficiency over compassion.
Myth 3
Too Many People
Reality: Birth rates are falling rapidly worldwide as remaining regions of the Third World begin the demographic transition—when birth rates drop in response to an earlier decline in death rates. Although rapid population growth remains a serious concern in many countries, nowhere does population density explain hunger. For every Bangladesh, a densely populated and hungry country, we find a Nigeria, Brazil or Bolivia, where abundant food resources coexist with hunger. Costa Rica, with only half of Honduras' cropped acres per person, boasts a life expectancy—one indicator of nutrition —11 years longer than that of Honduras and close to that of developed countries. Rapid population growth is not the root cause of hunger. Like hunger itself, it results from underlying inequities that deprive people, especially poor women, of economic opportunity and security. Rapid population growth and hunger are endemic to societies where land ownership, jobs, education, health care, and old age security are beyond the reach of most people. Those Third World societies with dramatically successful early and rapid reductions of population growth rates-China, Sri Lanka, Colombia, Cuba and the Indian state of Kerala-prove that the lives of the poor, especially poor women, must improve before they can choose to have fewer children.
Myth 4
The Environment vs. More Food?
Reality: We should be alarmed that an environmental crisis is undercutting our food-production resources, but a tradeoff between our environment and the world's need for food is not inevitable. Efforts to feed the hungry are not causing the environmental crisis. Large corporations are mainly responsible for deforestation-creating and profiting from developed-country consumer demand for tropical hardwoods and exotic or out-of-season food items. Most pesticides used in the Third World are applied to export crops, playing little role in feeding the hungry, while in the U.S. they are used to give a blemish-free cosmetic appearance to produce, with no improvement in nutritional value.
Alternatives exist now and many more are possible. The success of organic farmers in the U.S. gives a glimpse of the possibilities. Cuba's recent success in overcoming a food crisis through self-reliance and sustainable, virtually pesticide-free agriculture is another good example. Indeed, environmentally sound agricultural alternatives can be more productive than environmentally destructive ones.
Myth 5
The Green Revolution is the Answer
Reality: The production advances of the Green Revolution are no myth. Thanks to the new seeds, million of tons more grain a year are being harvested. But focusing narrowly on increasing production cannot alleviate hunger because it fails to alter the tightly concentrated distribution of economic power that determines who can buy the additional food. That's why in several of the biggest Green Revolution successes—India, Mexico, and the Philippines—grain production and in some cases, exports, have climbed, while hunger has persisted and the long-term productive capacity of the soil is degraded. Now we must fight the prospect of a 'New Green Revolution' based on biotechnology, which threatens to further accentuate inequality.
Myth 6
We Need Large Farms
Reality: Large landowners who control most of the best land often leave much of it idle. Unjust farming systems leave farmland in the hands of the most inefficient producers. By contrast, small farmers typically achieve at least four to five times greater output per acre, in part because they work their land more intensively and use integrated, and often more sustainable, production systems. Without secure tenure, the many millions of tenant farmers in the Third World have little incentive to invest in land improvements, to rotate crops, or to leave land fallow for the sake of long-term soil fertility. Future food production is undermined. On the other hand, redistribution of land can favor production. Comprehensive land reform has markedly increased production in countries as diverse as Japan, Zimbabwe, and Taiwan. A World Bank study of northeast Brazil estimates that redistributing farmland into smaller holdings would raise output an astonishing 80 percent.
Myth 7
The Free Market Can End Hunger
Reality: Unfortunately, such a "market-is-good, government-is-bad" formula can never help address the causes of hunger. Such a dogmatic stance misleads us that a society can opt for one or the other, when in fact every economy on earth combines the market and government in allocating resources and distributing goods. The market's marvelous efficiencies can only work to eliminate hunger, however, when purchasing power is widely dispersed.
So all those who believe in the usefulness of the market and the necessity of ending hunger must concentrate on promoting not the market, but the consumers! In this task, government has a vital role to play in countering the tendency toward economic concentration, through genuine tax, credit, and land reforms to disperse buying power toward the poor. Recent trends toward privatization and de-regulation are most definitely not the answer.
Myth 8
Free Trade is the Answer
Reality: The trade promotion formula has proven an abject failure at alleviating hunger. In most Third World countries exports have boomed while hunger has continued unabated or actually worsened. While soybean exports boomed in Brazil-to feed Japanese and European livestock-hunger spread from one-third to two-thirds of the population. Where the majority of people have been made too poor to buy the food grown on their own country's soil, those who control productive resources will, not surprisingly, orient their production to more lucrative markets abroad. Export crop production squeezes out basic food production. Pro-trade policies like NAFTA and GATT pit working people in different countries against each other in a 'race to the bottom,' where the basis of competition is who will work for less, without adequate health coverage or minimum environmental standards. Mexico and the U.S. are a case in point: since NAFTA we have had a net loss of 250,000 jobs here, while Mexico has lost 2 million, and hunger is on the rise in both countries.
Myth 9
Too Hungry to Fight for Their Rights
Reality: Bombarded with images of poor people as weak and hungry, we lose sight of the obvious: for those with few resources, mere survival requires tremendous effort. If the poor were truly passive, few of them could even survive. Around the world, from the Zapatistas in Chiapas, Mexico, to the farmers' movement in India, wherever people are suffering needlessly, movements for change are underway. People will feed themselves, if allowed to do so. It's not our job to 'set things right' for others. Our responsibility is to remove the obstacles in their paths, obstacles often created by large corporations and U.S. government, World Bank and IMF policies.
Myth 10
More U.S. Aid Will Help the Hungry
Reality: Most U.S. aid works directly against the hungry. Foreign aid can only reinforce, not change, the status quo. Where governments answer only to elites, our aid not only fails to reach hungry people, it shores up the very forces working against them. Our aid is used to impose free trade and free market policies, to promote exports at the expense of food production, and to provide the armaments that repressive governments use to stay in power. Even emergency, or humanitarian aid, which makes up only five percent of the total, often ends up enriching American grain companies while failing to reach the hungry, and it can dangerously undercut local food production in the recipient country. It would be better to use our foreign aid budget for unconditional debt relief, as it is the foreign debt burden that forces most Third World countries to cut back on basic health, education and anti-poverty
programs.
Myth 11
We Benefit From Their Poverty
Reality: The biggest threat to the well-being of the vast majority of Americans is not the advancement but the continued deprivation of the hungry. Low wages-both abroad and in inner cities at home-may mean cheaper bananas, shirts, computers and fast food for most Americans, but in other ways we pay heavily for hunger and poverty. Enforced poverty in the Third World jeopardizes U.S. jobs, wages and working conditions as corporations seek cheaper labor abroad. In a global economy, what American workers have achieved in employment, wage levels, and working conditions can be protected only when working people in every country are freed from economic desperation.
Here at home, policies like welfare reform throw more people into the job market than can be absorbed-at below minimum wage levels in the case of 'workfare'-which puts downward pressure on the wages of those on higher rungs of the employment ladder. The growing numbers of 'working poor' are those who have part- or full-time low wage jobs yet cannot afford adequate nutrition or housing for their families. Educating ourselves about the common interests most Americans share with the poor in the Third World and at home allows us to be compassionate without sliding into pity. In working to clear the way for the poor to free themselves from economic oppression, we free ourselves as well.
Myth 12
Curtail Freedom to End Hunger?
Reality: There is no theoretical or practical reason why freedom, taken to mean civil liberties, should be incompatible with ending hunger. Surveying the globe, we see no correlation between hunger and civil liberties. However, one narrow definition of freedom-the right to unlimited accumulation of wealth-producing property and the right to use that property however one sees fit-is in fundamental conflict with ending hunger. By contrast, a definition of freedom more consistent with our nation's dominant founding vision holds that economic security for all is the guarantor of our liberty. Such an understanding of freedom is essential to ending hunger.
12 Myths About Hunger based on World Hunger: 12 Myths, 2nd Edition, by Frances Moore Lappé, Joseph Collins and Peter Rosset, with Luis Esparza (fully revised and updated, Grove/Atlantic and Food First Books, Oct. 1998)
Institute for Food and Development Policy Backgrounder
Please don't vote because democracy is a local optimum
Related to: Voting is like donating thousands of dollars to charity, Does My Vote Matter?
And voting adds legitimacy to it.
Thank you.
#annoyedbymotivatedcognition
Hofstadter's Superrationality
Possibly the main and original inspiration for Yudkowsky's various musings on what advanced game theories should do (eg. cooperate in the Prisoner's Dilemma) is a set of essays penned by Douglas Hofstadter (of Godel, Escher, Bach) 1983. Unfortunately, they were not online and only available as part of a dead-tree collection. This is unfortunate. Fortunately the collection is available through the usual pirates as a scan, and I took the liberty of transcribing by hand the relevant essays with images, correcting errors, annotating with links, etc: http://www.gwern.net/docs/1985-hofstadter
The 3 essays:
- discuss the Prisoner's dilemma, the misfortune of defection, what sort of cooperative reasoning would maximize returns in a souped-up Prisoner's dilemma, and then offers a public contest
- then we learn the results of the contest, and a discussion of ecology and the tragedy of the commons
- finally, Hofstadter gives an extended parable about cooperation in the face of nuclear warfare; it is fortunate for us that it applies to most existential threats as well
I hope you find them educational. I am not 100% confident of the math transcriptions since the original ebook messed some of them up; if you find any apparent mistakes or typos, please leave comments.
A model of UDT with a halting oracle
This post requires some knowledge of mathematical logic and computability theory. The basic idea is due to Vladimir Nesov and me.
Let the universe be a computer program U that can make calls to a halting oracle. Let the agent be a subprogram A within U that can also make calls to the oracle. The source code of both A and U is available to A.
def U():
# Fill boxes, according to predicted action.
box1 = 1000
box2 = 1000000 if (A() == 1) else 0
# Compute reward, based on actual action.
return box2 if (A() == 1) else (box1 + box2)
A complete definition of U should also include the definition of A, so let's define it. We will use the halting oracle only as a provability oracle for some formal system S, e.g. Peano arithmetic. Here's the algorithm of A:
- Play chicken with the universe: if S proves that A()≠a for some action a, then return a.
- For every possible action a, find some utility value u such that S proves that A()=a ⇒ U()=u. If such a proof cannot be found for some a, break down and cry because the universe is unfair.
- Return the action that corresponds to the highest utility found on step 2.
Now we want to prove that the agent one-boxes, i.e. A()=1 and U()=1000000. That will follow from two lemmas.
Lemma 1: S proves that A()=1 ⇒ U()=1000000 and A()=2 ⇒ U()=1000. Proof: you can derive that from just the source code of U, without looking at A at all.
Lemma 2: S doesn't prove any other utility values for A()=1 or A()=2. Proof: assume, for example, that S proves that A()=1 ⇒ U()=42. But S also proves that A()=1 ⇒ U()=1000000, therefore S proves that A()≠1. According to the first step of the algorithm, A will play chicken with the universe and return 1, making S inconsistent unsound (thx Misha). So if S is sound, that can't happen.
We see that the agent defined above will do the right thing in Newcomb's problem. And the proof transfers easily to many other toy problems, like the symmetric Prisoner's Dilemma.
But why? What's the point of this result?
There's a big problem about formalizing UDT. If the agent chooses a certain action in a deterministic universe, then it's a true fact about the universe that choosing a different action would have caused Santa to appear. Moreover, if the universe is computable, then such silly logical counterfactuals are not just true but provable in any reasonable formal system. When we can't compare actual decisions with counterfactual ones, it's hard to define what it means for a decision to be "optimal".
For example, one previous formalization searched for formal proofs up to a specified length limit. Problem is, that limit is a magic constant in the code that can't be derived from the universe program alone. And if you try searching for proofs without a length limit, you might encounter a proof of a "silly" counterfactual which will make you stop early before finding the "serious" one. Then your decision based on that silly counterfactual can make it true by making its antecedent false... But the bigger problem is that we can't say exactly what makes a "silly" counterfactual different from a "serious" one.
In contrast, the new model with oracles has a nice notion of optimality, relative to the agent's formal system. The agent will always return whatever action is proved by the formal system to be optimal, if such an action exists. This notion of optimality matches our intuitions even though the universe is still perfectly deterministic and the agent is still embedded in it, because the oracle ensures that determinism is just out of the formal system's reach.
P.S. I became a SingInst research associate on Dec 1. They did not swear me to secrecy, and I hope this post shows that I'm still a fan of working in the open. I might just try to be a little more careful because I wouldn't want to discredit SingInst by making stupid math mistakes in public :-)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d)