Вы здесь

Новости LessWrong.com

Подписка на Лента Новости LessWrong.com Новости LessWrong.com
A community blog devoted to refining the art of rationality
Обновлено: 1 день 13 часов назад

Genetics: It sometimes skips a generation is a terrible explanation!

1 января, 2022 - 18:12
Published on January 1, 2022 3:12 PM GMT

Question: "How come your son Jimmie isn't colorblind when his father is"
Wrong answer: "It some times skips a generations, maybe his kids will inherent it"
Right Answer: "it's on the X chromosome, he got his fathers Y chromosome, so his kids will not inherent colorblindness from his side of the family!"

Question: "How come your daughter Susie isn't colorblind when her farther is"
Confused answer: "It some times skips a generations, maybe her kids will inherent it"
Right Answer: "The confused answer is less wrong than with the son, it's on the X chromosome, she got her fathers X chromosome, so half heir sons will be color blind"

Simple genetics, your Mother has XX your father has XY, let's use little x to denote colorblindness, you get a random chromosome from each parent.

Colorblind father -> sons have normal vision, daughters are carriers:
xY + XX - > 50% XY + 50% xX

Colorblind mother -> sons are colorblind, daughters are carriers
XY + xx - > 50% xY + 50% xX

Carrier mother -> half sons are colorblind, half of daughters are carriers
XY + xX - > 25% xY + 25% XY + 25% xX + 25% XX

Finally if colorblindness runs in both families:
xY + xX -> 25% xY + 25% XY + 25% xx + 25% xX

both parents colorblind -> all children colorblind

Question: "Have you ever wondered, how does the xX females know not to use the color blind gene?"
Answer: "They do not, half the relevant cells in their eyes will have a bad photo receptor... but the brain is pretty smart, it quickly learns to boost the signal from the other half of the cells"


$1000 USD prize - Circular Dependency of Counterfactuals

1 января, 2022 - 12:43
Published on January 1, 2022 9:43 AM GMT

I've previously argued that the concept of counterfactuals can only be understood from within the counterfactual perspective.

I will be awarding a $1000 prize for the best post that engages with this perspective. The winning entry may be one of the following:

a) A post that attempts to draw out the consequences of this principle for decision theory

b) A post that attempts to evaluate the arguments for and against adopting the principle that counterfactuals only make sense from within the counterfactual perspective

c) A review of relevant literature in philosophy or decision theory

Feel free to ask me for clarification about what would be on or off-topic. Probably the main thing I'd like to see is substantial engagement with this principle. If someone submits a high-quality post that only touches on this issue tangentially, but someone else submits an only okayish post that tries to deeply engage with this issue, then I would likely award it to the latter.

Why do I believe in this principle?

Roughly, my reasons are as follows:

  1. Rejecting David Lewis' Counterfactual Realism as absurd and therefore concluding that counterfactuals must be at least partially a human construction: either a) in the sense of them being an inevitable and essential part of how we make sense of the world by our very nature or b) in the sense of being a semi-arbitrary and contingent system that we've adopted in order to navigate the world
  2. Insofar as counterfactuals are inherently a part of how we interpret the world, the only way that we can understand them is to "look out through them", notice what we see, and attempt to characterise this as precisely as possible
  3. Insofar as counterfactuals are a somewhat arbitrary and contingent system constructed in order to navigate the world, the way that the system is justified is by imagining adopting various mental frameworks and noticing that a particular framework seems like it would be useful over a wide variety of circumstances. However, we've just invoked counterfactuals twice: a) by imagining adopting different mental frameworks b) by imagining different circumstances over which to evaluate these frameworks. (*).
  4. In either case, counterfactuals seem to be dependent upon themselves. Or at least, I find this argument persuasive.

Why do I believe this is important?

I've argued for the importance of agent meta-foundations before. Roughly, there seems to be a lot of confusion about what counterfactuals are and how to construct them. I believe that much of this confusion would be cleared up if we can sort out some of these foundational issues.

Why am I posting this bounty?

I believe in this idea, but:

  1. I haven't been able to dedicate nearly as much to time exploring this as I would like in between all of my other commitments
  2. Working on this approach just by myself is kind of lonely and extremely challenging (for example, it's hard to get good quality feedback)
  3. I suspect that more people would be persuaded that this was a fruitful approach if this principle was presented to them in a different light.

Fine print:

The submission should be posted LW/the alignment forum, although I'm open to private submissions assuming that they will be published in due course.

I'm currently planning to set the submission window to 3 months from the date of this post.

I'll award the prize assuming that there's at least one semi-decent submission (according to the standards of posts on Less Wrong). If this isn't the case, then I'll donate the money to an AI Safety organization instead. I'd be open to having this money be held in escrow.

I've written on this topic myself, so this probably biases me in some ways, but it's a small enough amount of money that's it's probably not worthwhile looking for external judges.

(*) Counterpoint: requiring counterfactuals to justify their own use isn't the same as counterfactuals only making sense from within themselves. Response: It's possible to engage in the appropriate symbol manipulation without a concept of counterfactuals, but we can't have a semantic understanding of what we're doing. We can't even describe this process without being to say things like "if given string of symbols s, do y". Similarly, counterfactuals aren't just justified by imagining the consequences of applying different mental over different circumstances, in this case, they are a system for performing well over a variety of circumstances.


Mailing Binax kits

1 января, 2022 - 09:17
Published on January 1, 2022 6:17 AM GMT

Is it safe with regard to test efficacy to mail Binax rapid covid test kits (i.e. to a friend) in the winter when the weather is below freezing? The official instructions say to store at above 2 degrees C, but they are on sale on Amazon. Going to visit a friend in another city, being able to mail them a test for their use would be easier logistically than having them pick one up. I've tried googling this and found nothing, maybe somebody here knows.


This Year I Tried To Teach Myself Math. How Did It Go?

31 декабря, 2021 - 20:55
Published on December 31, 2021 5:55 PM GMT

Cross-posted from my substack


This goes over a year of self teaching, the topics of math I was able to learn, how I view the best way to learn math and what my general view of math as a whole is with my limited current knowledge on the subject. At the end is a guide I made for anybody else that wishes to teach themselves math, no matter the skill level.

The Wild World of Online Learning:

When you first start out to do any project it's usually best to have instructions and a goal of what a ‘finished’ state looks like in mind. Most consumer bought appliances that require assembly come with a very hand holdy instruction sheet and all the parts required, the same can not be said for trying to learn an area of knowledge. I would argue a good chuck, >60% of the year,  was just spent on figuring out what I needed to learn and where the next logical step from that was. 

I will fully admit that GPT-3 and other AI advances got me interested in joining the industry. I will also fully admit that I immediately jumped into Coursera’s Intro to Machine Learning and crashed hard when I realized I wasn't getting anything out of it because I barely knew high school algebra and couldn't understand what I was looking at, because I wasn't interested in anything math related until now. Even reading about what partial derivatives were, there was too much scaffolding of previous calculus math needed to really set me up for understanding. The kicker was I could still pass the coding assignments and it made my “grade” look good, but that’s mostly because of how the course was set up. I complained about online learning because of it here, but I’m going to be a bit more dispassionate in this review.

This cycle repeated in a few more online courses. I would spend time going through the course, get by doing the assignments, start to get frustrated because I didn’t know why what I was learning was going to be important down the road, then eventually drop the course and start looking for another one that I felt could actually help me. It was about half way through the year that I decided to drop online learning due to the resentment of feeling like I wasted my time on the courses. The problem was I could do all the practice problems of the online courses fine, but I still felt like I had only islands of unconnected knowledge and I hadn't learned anything actually useful. It was like college all over again, just doing enough to get a grade then moving on. Many courses seem designed to give a myopic view of a topic, a few practice problems, and then moving onto another topic. I’m more okay with colleges doing it since you get a degree and usually classes build off each other, but knowing I wouldn't get any recognition and most courses dead stopped after it felt bitter. But I do have my share of the blame of course.

I didn’t know what to look for, and so what I found probably wasn't that good. I won’t say perfect learning materials exist, or that most online learning courses are not just a trade of money for a sense of accomplishment. But knowing what I was supposed to learn and would have helped me beforehand would have helped me filter courses much better. So what happens is what I’m calling the autodidact's paradox: you need to already have the knowledge about what you want to learn about to get the best learning resources for that subject. This is of course easily solved by mentors or professors giving guidance in the field, but I did this on my own and was struggling.

Even though I wouldn't rate my experience with online courses very high, it wasn't pointless to me because it helped me learn what I wanted to avoid in an online course. And the insular knowledge tidbits sometimes did come back and help make learning things easier in the long run. It’s just during the course that it felt pointless, and honestly a good book could probably cover the topics faster and more in depth. 

Knowing where my faults were from trying and failing many math and programming courses, and feeling burnt from online courses, I decided to go back to the basics and this time just to follow along with books. So the second half of the year, starting around July, I went on Kahn academy and started up high school algebra 1 again with the intent to really learn it this time.

Learning Math For Real.

This wasn't going to be like the first half of the year. I had a solid list of resources I could trust, I had a plan to move through them, and all I needed to do was do it. And I did.

I did all the way through Precalculus on Khan Academy. I thought the videos were good but took a while, so I only watched them if I was hard stuck on a section. Most of the time what I did was I would instantly go to the practice problems and continuously do them. Moving on when I got ¾ correct. It took about two solid months of three to four hours per day to get through it. Math just takes a good amount of time, especially learning and thinking through the problems. 

After that I decided to read How to Prove It by Velleman, Calculus Vol.1 by Apostol, and Probability by Degroot. Probability I got through the first 2 chapters or so until it started throwing out integration which went over my head at the time. My biggest math faux pas is I got through the first third of Proofs by Velleman before dropping it, not because it was bad but cause Calc and Prob also started with an intro to proofs along with set theory, and their introduction was good enough to work off that I went ahead. I will most likely go back to finish it… eventually… maybe. Apostol's Calculus book I loved and read all the way through, although it was much more rigorous than anything I had encountered thus far. 

The book's exercises being as rigorous as they are, I could only do a couple of the easier problems of the section at most. Some sections I couldn't do any, looking specifically at first and second order differential equations here. After reading for a while I realized I didn’t really like doing the problems in the book, mainly because they didn’t hold your hand as much.  I was thinking back on how the online problems usually were involved and offered greater explanation on why you got something wrong, and helped to retrace steps and correct it. One day I was looking through EDx again and realized something that made me embarrassed about not thinking of it before. Just do both: learn from two resources, use the book as a guide through the concepts and the area you wanted to learn, then do practice problems online where you can get greater feedback. Learning from two resources also allows coverage of concepts from two viewpoints and leads to better understanding. Learning from multiple resources I heard about in this post by LessWrong user TurnTrout, which has other useful tips.  So I found some equivalent online learning courses that I audited that I could go alongside with my math books, basically reinventing the college system of having readings, then lectures, then curated problems. Just from more diversified sources.

Apostol’s book covers what is typically called in the college programs Calc 1, Calc 2, and a bit of Linear Algebra. So from derivatives to integration and beyond, leaving nothing out using the axiom then proofs then exercises typical math book formula. Even though it was way over my head at the start, and some of it still is, I am way more comfortable now reading more advanced math texts then I was before. I also enjoy the little history lessons in each section and talks of how all the areas in the book tie together and what some of the areas are used for in future / different courses as well. Every complaint I had about online courses was solved by how good this book is.

How Does One Learn Math? 

Over the year this is something I asked myself more and more, how does one learn math? What does learning actually look like and what is a good way to learn? Learning math is deceptively simple, just read the section about it and do some practice problems, then you have “learned” that section. But that is a facile representation of the hard part of teaching yourself. How many problems do you have to do to feel like you learned something? How comfortable with the topic and the surrounding extensions to it do you have to be to feel comfortable in that subject as a whole? How much time can you afford to spend ensuring your knowledge in one area that could have been better spent just moving on and getting rough knowledge in another? There are always constraints and trade offs. We live in a finite world with finite time, and these questions bothered me a lot.

My idea of optimal math learning essentially boiled down to this over the year: breadth first exposure to math topics is better than the school like depth first exposure to math topics.

Taking an outside look at how schools seemed to try and teach mathematics it seemed like they would do a brief (sometimes feeling non-existent in my case) intro into a topic then drill as many practice problems as they can onto a student. The students presumably with many other classes will probably look up how to solve the problem quickly to save time, do all the assignments, take the test and then wait for the next topic to repeat the cycle. For non math majors this is a necessary evil we must do to get to what we actually want. For math majors maybe it's different, and they can see the connections between subjects we outsiders can not. How well this is done depends on school quality, but this is what I would call front-loading math skills. Doing many exercises to drill in precise instructions to a topic that will hopefully last long term. 

I think this way of learning is mostly a waste of time. Specifically because knowing that skills will decay when they are not used, if one doesn’t use many of the topics learned in their math courses then spending countless hours drilling problems is a waste of time that could be better allocated into learning how those problems arise and when to look for them. Schools seem to try to use a shotgun method of “we don’t know what this student will be doing in the future so we should try to get them all proficient in everything”, it has to be that way in our current system and that’s fine, it is what it is, most students I’ve talked to don’t know what they want to do in the future but have a blurry life outline that what they are majoring in now is correct. But if you choose to self learn with a clear end goal of a certain field there's benefits to be had by not over drilling exercises initially and putting them off till later.

My thinking came to be over the year that skills I will need will naturally be strengthened over time just due to the fact that I will be using them more, skills I don’t need will decay due to lack of use, but it will be useful to recognize at the very least what the problem is so I can look up how to solve it if need be. So what I do is when I read a section now I try to recognize where this math problem might arise, what it’s general structure looks like, and what related fields it might show up in. I do about 3-4 practice problems and I move on. I don’t get too deep into edge cases of the topic where most college students go, and I will fully admit it causes me to be less skilled. But I can move on faster and reinforce my prior learning while learning new material. As I go farther, what I need to know strengthens itself and what I don’t decays. Over time I get to the proficiency level of the college student who front-loaded all the problems, and over time they probably decay to my level using what's needed. But I will have covered more ground. 

So to me learning math is just problem recognition. I know I won’t remember every way of how to solve a problem, or be able to re-derive a solution easily. But if I can at least recognize what I am looking at I can find what I need to solve it. This is what I am aiming for as I go through these lower levels of math of Algebra, Calc, etc. That’s not to say math isn’t difficult when it is time to learn it; it does still take a lot of brain power. But it's the best way I’ve found to learn it so far.

The nice thing about problem recognition in learning math is that all previous topics learned get abstracted easily, that's why symbols are so nice. When I see the integration symbol I don’t have to go over every little aspect of step functions or least upper bound set theory in my head to eventually come to a re-derivation of the formula used for polynomials. I can just use the formula to solve the problem. If I had to go over the number line to the Cartesian plane to all the types of lines all just to do a simple derivative I would probably go insane. While many people initially complain about the alphabet in math problems, I’ve come to appreciate the abstraction so we don’t have to reinvent the wheel every problem.

Once I get more into my field of choice doing Artificial Intelligence research I will be forced by necessity to deal with edge cases and tough problems, but I look forward to that. I think anybody trying to do research in their field is dealing with really tough problems. While reading Apostol’s book I was thinking about how some people spend their life on one math formula that gets a paragraph dedicated to it if they are lucky. Only to be learned and abstracted by college students in about a week or two. I could, if I had different values, probably spend a good chunk or even my whole life diving into one specific topic out of the whole book like the mathematicians of the past. It’s a weird feeling knowing I get to read quickly what took lifetimes to accumulate.

My General View of Mathematics After One Year.

These are my rough thoughts on what I think mathematicians are doing, what they are trying to accomplish, and some thoughts on the problems I’ve seen them try to solve with a layman's view.

While reading Apostol’s I quickly picked up on a common pattern. The topic would introduce the simplest case of the problem with a single variable (if possible), then one with multiple variables, then as many variables as possible, maybe infinite. While I found math problems with the potential for so many variables interesting, I wondered how that would apply to the physical world which we live in. I have since seen linear equations using hundreds of dimensions of input but it did help me get an idea of how to categorize math into the areas of theory and practice.

In my current view, theorists see how far they can take their specific topic of math and contend with interesting and weird edge cases of the topic. If some of the proven theories seem useful for a project that could be made, that specific part of the theory is used by applied mathematicians to make the project work. Applied mathematicians working in our world usually seem to get different titles however; calling themselves physicists, finance managers, and computer scientists rather than just mathematicians. While I enjoy the outcomes of the applied mathematicians, some of the theorists work I see as incomprehensible. But I think that’s just my lack of knowledge showing.

I don’t know if I will ever venture down the road of mathematics far enough to understand why the Monster group is important in topology, or why Conway’s knot is important either. Graham’s Number I see as ridiculous, apparently one of the answers to his original problem could be as low as a single digit number, why have power towers on power towers then? I don’t understand spending so much time on such a problem, or why to try to solve Fermat's Last Theorem. In my current understanding there is no practical application to these things even if they were solved; they are just fun puzzles. I can relatively understand what a computer science theorist is trying to solve, even if I don’t understand it I can see its application. In pure math theory I can’t understand it nor see its application right now.

I think it's incredible what I’ve seen so far of how theory and application has come together with Calculus and Algebra to build the modern world. How the mathematicians of old were learning how to describe the physical world around them in precise language. As I go further into math I intuitively think the relationship between math and the physical world will get more blurry, not as clearly defined as tracking how things change or accumulate over time.  I’m excited to learn and battle my confusion about what is happening. Hopefully in another year's time I can call the person writing this clueless.

Concluding Thoughts:

While getting to the place I am required many small and sometimes large changes in my thinking and learning materials.  I feel much more comfortable in my ability to learn things and my current plan going forward for learning mathematics. I don’t feel like I am good at mathematics, I’m not sure I will ever feel that way, but I know I have gotten better and that’s a good feeling I am going to continue to pursue.  I will no doubt change my mind on what resources to use and get lost and confused while learning, but that's just part of the process. Thinking back to the beginning of the year, I certainly think I am in a better place than I was both in my knowledge of math, my position in life, and my perception of how I will fare in the future.

The study guide I’ve made and have been following is here.



31 декабря, 2021 - 20:48
Published on December 31, 2021 5:48 PM GMT

Untitled (Escalators) by Scott MutterI. Traversing the Landscape of Scientific Knowledge

Consider a landscape in which peaks represent scientific knowledge and valleys represent all that is false, irrational, or ill-conceived. Scientific progress depends on our ability, as individuals and as a scientific community, to (1) find new peaks and to (2) climb them. These two sub-goals require two fundamentally different modes of travel, what I will call wandering (i.e. random walk, or the explore in explore/exploit) and climbing (hill-climbing in algorithmic parlance). As will become clear momentarily, I have chosen these terms over the more conventional explore/exploit dichotomy because I have something in mind that is a little more expansive and multi-dimensional than is typically implied by that framing.

We begin by considering the wandering/climbing distinction at the level of an individual scientist and their research activities. Climbing is what Yanai and Lercher, and Francoise Jacob before them, call Day Science, “…the one you read about in the news, it is the one we learn about in school, the one captured by the phrase “hypothesis driven”. It’s epitomized by the women and men in white lab coats holding pipettes or looking intently at a computer screen. A day scientist is a hunter who has a clear picture of what she is pursuing.” Night science then is wandering, something that is made explicit in the following passage, “…it wanders blind. It hesitates, stumbles, recoils, sweats, wakes with a start. Doubting everything, it is forever trying to find itself, question itself, pull itself back together. Night science is a sort of workshop of the possible where what will become the building material of science is worked out”.

Nobel prize-winning biochemist Albert Szent-Györgi used different terminology to describe these same basic categories of scientific activity in a 1972 letter to Science.

In science the Apollonian tends to develop established lines to perfection, while the Dionysian rather relies on intuition and is more likely to open new, unexpected alleys for research…The future of mankind depends on the progress of science, and the progress of science depends on the support it can find. Support mostly takes the form of grants, and the present methods of distributing grants unduly favor the Apollonian. Applying for a grant begins with writing a project. The Apollonian clearly sees the future lines of his research and has no difficulty writing a clear project. Not so the Dionysian, who knows only the direction in which he wants to go out into the unknown; he has no idea what he is going to find there and how he is going to find it. Defining the unknown or writing down the subconscious is a contradiction in absurdum.

The ancient dichotomy of the Apollonian (order, logic, restraint, harmony) and the Dionysian (chaos, emotion, intuition, orgiastic revelry) provides the psychological bedrock upon which the higher-level dichotomy of climbing and wandering is built. This points us towards a crucial, perhaps the crucial, difference between our two modes of travel: climbing is about the quantifiable (the logic can be traced, the calculations can be checked, the data can be analyzed) while wandering is about the unquantifiable—ideas, intangible and ineffable, their value, causes, and consequences only knowable in hindsight, if at all (more on this later).

These two modes of individual scientific thought and activity also flow upward into contrasting epistemological orientations and organizational principles for the scientific community as a whole. To understand why this is so, let’s treat the climbing metaphor very literally by considering the first confirmed ascent of Mount Everest in 1953 by Edmund Hillary and Tenzing Norgay (George Mallory and Andrew Irvine might have reached the summit in 1924, but they died during the descent). The ascent was a massive operation, taking over 2 months from base to summit and requiring years of planning (great care was taken to select the most expert mountaineers), training (the principles of the crew trained in Wales before the ascent), and reconnaissance. The crew of over 200 climbers and porters was led by John Hunt, a British army colonel, and for all intents and purposes was a military operation.  

From this brief description, we can draw several lessons about the nature of climbing that apply whether the mountains are made of rocks or knowledge. Mountain climbing is a collaborative activity that requires some degree of hierarchical, top-down command. The ascent of truly challenging mountains requires the financial and administrative support of a large centralized institution (for Everest, it was the British military; for “knowledge mountains”, it is universities and funding bodies like the NSF and NIH). Climbing also requires special skills than can only be developed through extensive training (often provided or supported by those same large centralized institutions); following from this, some kind of vetting or credentialing process is needed to determine who are the best mountaineers. Lastly, mountain climbing is an inherently dangerous activity (the first pair chosen for ascent to the summit had to turn around due to equipment failure—climbing also requires equipment) and therefore is generally conservative in nature. This is not to say that climbing can’t be a highly risky endeavor, but that there is a general defensive posture which must be taken while ascending any significant peak.

To recap: climbing is collaborative, hierarchical, top-down, centralized, institutionalized, and conservative. It requires significant amounts of money, equipment, and training and therefore is professionalized, standardized, and homogenized (you don’t want any equipment surprises halfway up the mountain and you need to trust that your fellow climbers aren’t going to deviate from the plan and put everyone in danger). All of these qualities of climbing are interconnected and in some sense implied by the very nature of the activity.

Climbing leads to a specific form of scientific progress that has been variously characterized as incremental, “normal” science (in the Kuhnian sense), or “standing on the shoulders of giants in order to see further”. Climbing is the mode of science that aims towards known targets and clearly defined goals. The Manhattan project is a quintessential example of climbing science as I have described it here, and as such it shouldn’t be surprising that it was ultimately a military operation just like the first ascent of Mount Everest.

Wandering is the antithesis of climbing, its equal and opposite. It is those scientific endeavors that are individualistic, decentralized, egalitarian, unsystematized, unstandardized, and bottom-up. Wandering is revolutionary, paradigm-shifting science, the kind of science that finds entire mountain ranges of knowledge that were heretofore unimaginable (the unknown unknowns).

This metaphor has already gotten a little too unwieldy (a little?), but let’s add another metaphor into the mix in hopes that this will simplify our understanding of the first: climbing is religion and wandering is spirituality (much more on this in the final section).

II. Climbing and its Discontents

I hope that one thing is obvious at this stage—neither climbing nor wandering is inherently better or more useful than the other, and both are equally necessary for robust scientific progress. The challenge is to find the optimal balance between the two, a task that is complicated by the fact that the optimal balance is itself a moving target (e.g. physics may require more climbing than psychology at their current levels of advancement. What would be the equivalent of the Large Hadron Collider for psychology? We have no idea.). For most of scientific history, we suffered from a lack of climbing capacity—the scientific community was too distributed and unorganized, and there was not an enough emphasis on brick-by-brick experimentation and data collection. Now, I think it is fair to wonder if the pendulum has swung too far in the other direction, if what modern science suffers from is an excess of climbing.

I believe there are a few broad historical trends that in combination have served to tilt science too far towards the climbing end of the spectrum. To some degree, all of these trends are the result of scientific progress, and in this sense science has simply become a victim of its own success. This is seen perhaps most clearly in the inevitable consequences of knowledge accumulation, the so-called “burden of knowledge”. As knowledge grows, successive generations of scientists face a greater educational burden (i.e. there is more to learn in order to reach the frontier of a field). This growing burden necessitates longer educational periods and/or narrowing expertise, both of which reduce the capacities of individual thinkers. These compensatory moves compel a shift towards educational schemes/institutions which favor efficiency and standardization (i.e. there is a lot to learn so we need to learn it quickly and figure who knows it and who doesn’t) and towards a greater reliance on team science.

All of this is to say that we should naturally expect a shift towards climbing (team-oriented, institutionalized, efficient, standardized, professionalized, etc.) as a consequence of growing scientific knowledge. To put this point in metaphorical terms: when the “shoulders of giants” are low, it is conceivable that one individual can scale the shoulders by themselves, but teamwork and efficient standardized training are required when the giants truly become gigantic.

So the burden of knowledge creates a significant pressure towards climbing science, however on its own this pressure should not be enough to swing the pendulum too far in this direction; in fact, if the burden of knowledge were the only factor influencing the climbing/wandering balance then we would likely expect it to track optimally with the current state of scientific knowledge. We should look then for forces that have enhanced the underlying trend towards climbing beyond what is inevitable with a growing burden of knowledge.

Tyler Cowen has argued that the development of modern transportation, communication, and industrial technologies over historical timescales has enabled the bureaucracy needed to sustain the large powerful governments and corporations that we see today. Though Cowen doesn’t discuss this effect in the context of scientific organization, there is no reason to think that these same technological forces wouldn’t be acting here as well.

Interestingly, Cowen describes the phenomenon of “governmental overshooting”.

“…it appears that new technologies enabled the spread of a fascist intoxication with power, both among leaders and the general citizenry…Political and cultural institutions were not well equipped to handle the social implications of the new technologies of radio, electricity, and easy transportation. Those technologies made mass culture possible and in the realm of politics that mass culture translated into fascism. Only after bitter experience did fascist ideas become less popular and social and political norms subsequently evolved to protect electorates against the fascist temptation.”

This provides a good model for the notion that science has overshot the optimal balance between climbing and wandering. We could argue in the same manner that this technological climate enabled a kind of “intoxication” with centralized institutional power in science (culminating in the Manhattan project and the space race) that we have not yet learned to pushed back on by evolving new norms and values.

Dovetailing with this trend towards increasing centralization is the steadily growing importance of science and technology to national security; the confluence of these two forces has resulted in our current situation in which the entanglement of science and government is at all-time high. The corrupting influence of politics on science is well documented, but there is another more subtle issue that arises as the scientific community becomes intertwined with the governmental apparatus. In Seeing like a State, historian James Scott discusses how governments have sought over time to increase the legibility of their populace by imposing policies that make it easier quantify and keep track of people and property (e.g. social security numbers, the collecting of census data, the gridification of cities, standardized measurements, uniform languages, etc.). I think we can argue that a similar thing has happened as governments have increasingly become sponsors and organizers of science—there has been a mass shift towards greater legibility of all the inputs (money, equipment, administrative support) and outputs of research (publications, citations, patents, awards, etc.). Computers and the internet have been rocket fuel for legibility, easily allowing us to track nearly every aspect of the scientific enterprise.

I would argue that this intense “metrification” is responsible, either directly or indirectly, for much of what’s wrong with science. Much ink has been spilled on the nature of these issues so I won’t belabor the discussion, but suffice it to say that many of the metrics used to evaluate scientists and their research serve to incentivize conservatism and easily-quantified productivity over risk-taking and the development of new ideas (which, as previously noted, are by their very nature unquantifiable) (See Matt Clancy’s “Conservatism in Science”). In essence we have created a system where scientists are disincentivized from doing anything that won’t help them publish more research or bring in more grant money—performing replications, sharing data or ideas, mentoring undergraduates, improving their teaching, etc.

Some of these problems with metrics are due to the unsophisticated use of simple metrics and it should be possible to alleviate these problems with better metrics and greater awareness of their strengths and weaknesses. That being said, there are also some more fundamental issues that arise from the very nature of metrics and the incentive structures they create. It is easy to say that we need more complex metrics, but greater complexity comes with its own costs (e.g. more data might be needed to develop the new metric, more bureaucracy needed to collect that data, less intuitive metrics are less likely to be understood and correctly applied). Metrics, whether complex or simple, always suffer from the curse of Goodhart’s Law – “when a measure becomes a target, it ceases to be a good measure” (see the Eponymous Laws series). For example, citation counts may have been a good proxy for a scientist’s influence when we first started using them, but now that scientists are aware of their importance they have become a less effective measure because scientists are incentivized to game the citation counts however they can (publishing in hot research areas, rushing flashy results to publication without careful replication, developing superfluous new terms or metrics in hopes that they will catch on, self-citing). We might also mention here that citation norms themselves represent an implicit epistemological orientation that predisposes one towards incremental research that clearly builds off previous work (i.e. climbing) at the expense of truly novel ideas that might not have clear antecedents (I could provide a citation for this point, but I refuse to do so as an act of civil disobedience). Lastly, it is fair to wonder about the psychological cost of excessive metrification. I’m no expert here, but the awareness that your work is constantly being measured on a variety of dimensions probably isn’t conducive to creativity, job satisfaction, or mental health.

It might be helpful to make explicit something that has been lurking in the background of this section: the role of the internet in the scientific enterprise. The internet is, I think, usually seen as an unequivocal boon for science as it allows greater access to information and rapid communication between scientists (amongst many other benefits). My contrarianism runs deep, but not quite deep enough to deny that the internet has been, on balance, a good thing for science. But there is a balance, and I don’t think we have fully grappled with the costs of this mass shift in scientific thinking and doing. To put it in the terminology of this essay: the internet has supercharged our ability to climb but subtly (or not so subtly) harmed our ability to wander. In his song “Welcome to the Internet”, comedian-troubadour Bo Burnham sings, “Could I interest you in everything all of the time?”; I’m not the first to say it nor the last, but there is something incredibly unnatural about this everything-all-the-time state of affairs and it seems more than likely that it is warping our creativity and cultural dynamics in a problematic manner. Derek Thompson speculates on this theme in his recent article “American is Running on Fumes”:

The world is one big panopticon, and we don’t fully understand the implications of building a planetary marketplace of attention in which everything we do has an audience. Our work, our opinions, our milestones, and our subtle preferences are routinely submitted for public approval online. Maybe this makes culture more imitative. If you want to produce popular things, and you can easily tell from the internet what’s already popular, you’re simply more likely to produce more of that thing. This mimetic pressure is part of human nature. But perhaps the internet supercharges this trait and, in the process, makes people more hesitant about sharing ideas that aren’t already demonstrably pre-approved, which reduces novelty across many domains.

Lest I be mistaken, I am not advocating for the abandonment of metrics and centralized science (or the internet, as if that were even possible or desirable). This would amount to a return to the scientific anarchy of the pre-modern era, i.e. a science that is all wandering and no climbing. The challenge then, as I see it, is to find a way forward that allows us to overcome the bad incentives and constraints that make it difficult for scientists to freely wander while also maintaining the benefits of big institutional science.

III. Not All Who Wander are Lost

In his essay “We Don’t Know How to Fix Science”, José Luis Ricón argues that while we have many good ideas about how to reform science we need more meta-scientific experimentation to determine if these reforms will actually work. While I am certainly on board with more meta-scientific experimentation, I am dubious that this will “fix” science or even really improve it in any kind of significant way. Ricón is realistic about the difficulties of evaluating these kinds of experiments—the results may only become clear over the long term, a policy that has positive effects in a small experiment (e.g. shifting the length of a grant) might not scale—but remains optimistic that they will help us learn how to slowly improve science policy. My skepticism towards this whole philosophy comes in part from a striking passage in the essay:

Statistician Adrian Barnett reports having talked to Australian funding agencies asking them about using a lottery to allocate their funding. The reply didn’t involve, as one might have expected, a lack of belief in the effectiveness of lotteries. Rather, the answer he got was that “It would make it look like we [the agency] don’t know what we’re doing.”

Such is the nature of institutions—they will act to preserve the problem for which they are the solution (the Shirky principle). Producing evidence for new policies is one thing; implementation is another. The scientific ecosystem is so complex, and the interests so entrenched, that I am pessimistic about our ability to effect real top-down change. Moreover, I worry that even when we do effect structural change often times all we really accomplish is a shift to a new equilibrium that has different constraints and different bad incentives (goodharting never stops). Alexey Guzey hits the nail on the head in his recent essay “New Incentives will not Save Science”.

So then what will save science?

You. You are going to save science.

Musing on Marc Andreessen’s widely circulated 2021 essay “It’s Time to Build”, Tanner Greer quips that, “In the 21st century, the main question in American social life is not “how do we make that happen?” but “how do we get management to take our side?” (“On Cultures that Build”). 

Management will never take our side. Fuck management.

Do not try to change governmental institutions, universities, scientific journals, or the “scientific community”. Everyone is busy, no one wants more work to do, and no one gives a shit about what you have to say. There is no grand top-down scheme that will magically alleviate all of the problems in science. There is no quick fix and there never will be.

As I have argued, our current scientific epoch is characterized by an overabundance of this multi-dimensional constellation of qualities that I have termed climbing—goal-oriented, collaborative, institutionalized, centralized, professionalized, hierarchical, legible. The only way to remedy this excess is to lean radically in the opposite direction, that is to wander, and to do so fiercely. More than anything, this means an utter and complete rejection of anything that doesn’t start with the individual. The revolution will not trickle down from the top; it will bubble up from the bottom, from deep within the minds and hearts of each and every one of us. All of this obsession with meta-science and institutional design is just a cop-out. It’s easy—do an experiment, change a policy, build an institution, and voila!—no more publish or perish, no more endless grant writing, no more bad incentives, science enters a new golden age and we all live happily ever after.

There is no way around it: you must look in the mirror and ask yourself, moment-by-moment and day-by-day, if you are truly embodying the deepest values of science— creativity, courage, openness, honesty, tolerance, humility, collegiality, and the rest. As if this wasn’t hard enough on its own, there is an enemy in this endeavor that conspires to thwart us at every turn, what is variously known as The System, The Man, The Machine, or Moloch.

The System desires to suppress all which cannot be measured, classified, controlled, manufactured, or predicted. It wants to sap you of your creativity, drain you of your joy, and turn you to cynicism (if you have found yourself thinking that all of this is idealistic bullshit then you are already lost). Those things which cannot be measured—ideas, inspiration, wonder, beauty, faith, hope, love—are grave threats to the System because they are, by their very nature, irrational, unpredictable, uncontrollable.

The Machine wants conformity, efficiency, and productivity above all else. It wants you to bend the knee, to sell your soul, to sacrifice all that you hold dear in the name of faster and faster and more and more. It wants you to forget that bad incentives can and should be resisted, that rules are meant to be broken, that at all times you have the right to say fuck this shit, I’m out.

“And this I believe: that the free, exploring mind of the individual human is the most valuable thing in the world. And this I would fight for: the freedom of the mind to take any direction it wishes, undirected. And this I must fight against: any idea, religion, or government which limits or destroys the individual. This is what I am and what I am about.” – John Steinbeck

The System wants you to think that scientific progress only comes from “the hard slog of large armies of individuals, each making—at best—a tiny step or two forward”. It whispers in your ear, “To think that you could be the next Darwin, the next Einstein, the next Newton—what arrogance! Please be realistic, think about your career! Just keep your nose to the grindstone and maybe one day you’ll be a real scientist, with a fancy PhD, a lab coat, and a published paper to your name.” The System wants you to think that all of the mountains of scientific knowledge have already been discovered (all that is left are known unknowns) and the only remaining task is to slowly schlep our way to the top.

I have to get you to drop modesty and say to yourself, “Yes, I would like to do first-class work.” Our society frowns on people who set out to do really good work. You're not supposed to; luck is supposed to descend on you and you do great things by chance. Well, that's a kind of dumb thing to say. I say, why shouldn't you set out to do something significant. You don't have to tell other people, but shouldn't you say to yourself, “Yes, I would like to do something significant.”
— Richard Hamming, “You and Your Research

So I say unto you: wander fiercely, discover the Dionysian, pursue the paradoxical, embrace the irrational, welcome the wasteful, analyze the absurd, nurture inefficiency, find belief that cannot be shaken, and cherish all that cannot be measured.

These are your marching orders, but let’s talk boot-on-the-ground tactics.

V. Reformation

Previously, I summarized the climbing/wandering distinction by saying that climbing is religion and wandering is spirituality. I know these are dirty words for most science folk, but I can’t think of a better way to characterize my argument. Religion is institutions and (rigid) hierarchy, honorifics and rituals, obedience and external displays of piety. Religion can be faked—are you really a holy man or do you just play one on TV? In contrast, spirituality pertains only to the individual and their personal quest for meaning and purpose, their personal struggle for virtue and transcendence. Religion is data and results—external and concrete, easily exaggerated or fabricated—whereas spirituality is ideas, intangible and immeasurable (you cannot fake a good idea or control it, at least not for long). This entire essay can thus be summarized as follows: science needs less religion and more spirituality.

It has been argued (by Niall Ferguson, Balaji Srinavasan, and no doubt many others) that the closest historical analogue of personal computing is the printing press, and that when trying to understand our current era we should look to the reformation instead of any intervening period. The reason our current era may be most similar to the reformation is that both the printing press and personal computing (along with the internet) were/are powerful forces for decentralization. In both cases, this is a significant reversal from the preceding era in which the technological climate (radio, TV, telephone) favored centralization and hierarchy (as discussed above). This should really worry us, Ferguson argues, because the reformation led to 130 years of escalating religious conflict, culminating in the Thirty Years War, one of the most destructive conflicts ever.

These same tensions are now playing out in the scientific domain. The pendulum has now swung towards decentralization and egalitarianism and those that would oppose these forces (academia, governmental institutions, corporations) are struggling to maintain their grasp on scientific power. Martin Luther launched a revolution that shook the world when he nailed his 95 theses to the door of All Saints’ church in Wittenberg; in these theses, he criticized the Church’s practice of selling indulgences and outlined a new spiritual philosophy that valued the primacy of the individual’s connection with God and rejected the Church as the only source of divine knowledge (he believed that if everyone could read the Bible it would create a “priesthood of all believers”). The internet has functioned as the door of All Saints’ Church; just as the church sold its soul for the indulgence money, we can now all see that science has also sold its soul (for grant money). Similarly, we need to develop a form of scientific spirituality, one that values the primacy of the individual and his connection, not with god, but with the core values and virtues of science (and yes I know that this means I am essentially comparing myself to Martin Luther which is just about the most grandiose and narcissistic thing ever, but so be it).

How can we develop this “scientific spirituality” and engrain it in the next generation of scientists? How can we launch the scientific reformation? I can imagine three paths forward. I’ve referenced Richard Hamming’s classic 1986 colloquium at Bell Labs “You and Your Research” a few times already, but let’s go back to the well one more time. At the end of the talk, there is a very interesting exchange during the Q&A session:

Question: The remarks about having courage, no one could argue with; but those of us who have gray hairs or who are well established don't have to worry too much. But what I sense among the young people these days is a real concern over the risk taking in a highly competitive environment. Do you have any words of wisdom on this?

Hamming: I'll quote Ed David more. Ed David was concerned about the general loss of nerve in our society. It does seem to me that we've gone through various periods. Coming out of the war, coming out of Los Alamos where we built the bomb, coming out of building the radars and so on, there came into the mathematics department, and the research area, a group of people with a lot of guts. They've just seen things done; they've just won a war which was fantastic. We had reasons for having courage and therefore we did a great deal. I can't arrange that situation to do it again. I cannot blame the present generation for not having it, but I agree with what you say; I just cannot attach blame to it. It doesn't seem to me they have the desire for greatness; they lack the courage to do it. But we had, because we were in a favorable circumstance to have it; we just came through a tremendously successful war. In the war we were looking very, very bad for a long while; it was a very desperate struggle as you well know. And our success, I think, gave us courage and self confidence; that's why you see, beginning in the late forties through the fifties, a tremendous productivity at the labs which was stimulated from the earlier times.

Perhaps WWIII will instill in us a new courage, a new sense of purpose that galvanizes the development of this spiritual ethos for scientists.

Peter Thiel has attributed the present state of scientific and technological stagnation to a lack of belief in secrets (i.e. the unknown unknowns). From Scott Alexander’s review of Thiel’s book Zero to One:

Past scientific discoveries came from a belief in secrets. Isaac Newton wondered why apples fell, thought “Maybe if I work really hard on this problem, I can discover something nobody has ever learned before”, and then set out to do it. Modern people aren’t just less likely to think this way. They’re actively discouraged from it by a culture which mocks stories like Newton’s as “the myth of the lone genius”. Nowadays people get told that if they think they’ve figured out something about gravity, they’re probably a crackpot. Instead, they should wait for very large government-funded programs full of well-credentialed people to make incremental advances.

As hinted at in that last sentence, Thiel attributes this lack of belief in secrets in part to incrementalism. From Zero to One:

From an early age, we are taught that the right way to do things is to proceed one very small step at a time, day by day, grade by grade. If you overachieve and end up learning something that’s not on the test, you won’t receive credit for it. But in exchange for doing exactly what’s asked of you (and for doing it just a bit better than your peers), you’ll get an A. This process extends all the way up through the tenure track, which is why academics usually chase large numbers of trivial publications instead of new frontiers.

To put this in my own terminology: too much climbing and not enough wandering. Thiel also points to the lack of a frontier as a culprit.

There are no blank spaces left on the map anymore…Today, explorers are found mostly in history books and children’s tales. Parents don’t expect their kids to become explorers any more than they expect them to become pirates or sultans. Perhaps there are a few dozen uncontacted tribes somewhere deep in the Amazon, and we know there remains one last earthly frontier in the depths of the oceans. But the unknown seems less accessible than ever.

This offers another source of hope as we finally seem on the cusp of opening the Martian frontier; perhaps the colonization of Mars will create a new zeal and sense of purpose for science. The development of a new scientific community on Mars will also offer a unique opportunity to start from scratch and rewrite the values, norms, and fundamental organization of science.

So we can imagine two developments that will spur us towards a scientific reformation: WWIII and the colonization of Mars. One of these options is not preferable for obvious reasons, but both of them have the disadvantage of being transient—the memory of WWIII will eventually fade and the martian frontier will eventually close. The third path forward is the most difficult, but it does have the potential to bring about a more lasting change. It all begins with a simple prescription, one that all of us can act on in this very moment.


Look within.

“I maintain that truth is a pathless land, and you cannot approach it by any path whatsoever, by any religion, by any sect. That is my point of view, and I adhere to that absolutely and unconditionally. Truth, being limitless, unconditioned, unapproachable by any path whatsoever, cannot be organized; nor should any organization be formed to lead or to coerce people along any particular path.

(Originally posted at Secretum Secretorum)


Counterexamples to some ELK proposals

31 декабря, 2021 - 20:05
Published on December 31, 2021 5:05 PM GMT

In this post I’ll describe some possible approaches to eliciting latent knowledge (ELK) not discussed in our report. These are basically restatements of proposals by Davidad, Rohin, Ramana, and John Maxwell. For each approach, I’ll present one or two counterexamples that I think would break it.

I think of these approaches as being based on two fundamental hopes:

  1. If our AI tampers with one sensor, there may be other sensors we could deploy, or experiments we could run, in order to notice that something fishy has happened. We could have an AI help us design those experiments/sensors.
  2. Actions that covertly tamper may be similar to actions that visibly tamper, e.g. they may exploit similar unknown facts about physics or they may involve common substeps. So we might be able to notice covert tampering by pointing out similar action sequences that lead to visible tampering.

While these hopes may hold in practice, in this post I’ll describe possible situations where they would fail. (I don't think anyone particularly expected these approaches to be counterexample-proof.)

I’m only claiming that these failures are possible. I think it is also interesting to try to generate the most probable failure cases and to actually estimate their probability, but that’s definitely not what I’m doing here. 

Strategy: Hold out sensors

(Proposal #1 here)

In addition to asking my model to predict one set of sensors, I can deploy a second “held out” set (e.g. more cameras or seismographs or microphones…). I don’t ask my model to predict those sensors, but I show them to a human before getting them to answer questions. Hopefully that sometimes lets them notice tampering that would have been invisible without the new sensors.

Worry: human simulator can use all the extra sensor readings too

I’m concerned that I will learn a better version of the human simulator which predicts the readings of all sensors and then outputs what a human would infer from the complete set. That task need not be much more computationally demanding than the original human simulator. 

Moreover, my AI may be able to deduce the presence of the additional sensors very cheaply. Perhaps it can notice the sensors, or it can learn about my past actions to get a hint about where I may have placed them. If this is possible, then “Predict the readings of all sensors” need not be much more complex than “Predict the readings of one sensor.”

An AI doesn’t have to be exactly correct about what sensors exist—when in doubt it can just predict what the sensor would say and imagine giving it to the human.

Strategy: Predict hypothetical sensors

(Proposal #2 here, also suggested with counterexample by Rohin in private communication)

Instead of installing a single sensor, I could randomly choose what sensors to deploy and how to install them. Then I could train a model Predict+ to map (description of sensor) → (sensor reading), and I could give a human the readings from many hypothetical sensors in order to give them the best chance of detecting tampering.

Worry 1: Predict+ may not be reasonable on non-existent sensors

To collect training data for Predict+, I actually install sensors and then use their readings as a target. In every situation, my AI is being asked to predict the readings from a sensor that I really did build.

If my AI is able to deduce the presence of sensors (as described in the last counterexample), then it’s unclear how Predict+(sensor description) will behave when we give it a description of a sensor that does not in fact exist. 

In particular, Predict+(sensor description) might behave in the following undesirable way: it predicts the future observations conditioned on its observations and on the fact that someone typed (sensor description) into the dataset. But it knows that no sensor with description (sensor description) exists. So if someone typed that description into the dataset, that just means something funny is going on—not that the dataset will magically contain readings from a non-existent sensor.

For a predictor which understands the world well it seems quite plausible that this is the most natural behavior. Otherwise it needs to reason about a physically impossible counterfactual to get the “intended” answer (which it never needed to do to answer any other questions). Moreover, it would also be undesirable if the predictor conditioned on such a sensor existing despite having strong evidence to the contrary (since its predictions may then be extremely strange).

(A more extreme version of the same worry applies to proposal #3 here. A predictor seems even more likely to behave badly when it is given a description of a sensor that the human obviously couldn’t have built, and it would definitely know that the sensor doesn’t actually exist.)

Worry 2: the human may not be able to tell what is going on even with all the sensors

For some kinds of tampering no possible sensors would reveal what had happened. For example, we can imagine that our AI has covertly built an army of tiny machines which deconstruct the objects in the room while replacing them with fake versions that look the same on the outside. Sufficiently sophisticated machinery could reproduce the results of almost any experiment we perform (e.g. if you broke an egg, the tiny machines could simulate the sound of an egg cracking and quickly create the yolk on demand).

More realistically, our AI could search for sensors and tamper with any that it finds. It’s not clear how it is supposed to answer questions about what a sensor “would have shown”—if the hypothetical sensor had actually existed, then it also would have been found and tampered with. In this scenario, it may be “correct” for Predict+ to report the tampered value.

(As usual, this is only a problem if the human overseer literally can’t notice anything suspicious about a sequence of actions which is in fact constructing an army of sophisticated tiny machines. We are putting the burden of proof on the builder to show that a given concern is impossible.)

Strategy: conduct new experiments to notice that something is wrong

(Part of the proposals here and here.)

In the previous section I described installing sensors before the AI acted so that we’d have a better chance of seeing what was going on in real time. But we could also imagine installing sensors or conducting experiments after arriving at a given state, in order to better understand that state.

For example, suppose I believe that the diamond is in the room and all sensors are functioning correctly. I could ask another AI to design an experiment which would show that I’m fundamentally misunderstanding the situation and the sensors aren’t functioning correctly (e.g. perhaps it would walk me through building an x-ray machine to show me that the room is actually empty).

Worry 1: there is no experiment that reveals the problem

Ultimately to implement an experiment we will execute some actions in the world and observe the results. These experiments won’t be successful if either:

  • The sensors used to record the results of an experiment are also tampered with, and there is no sequence of actions that “fix” those sensors. (As in the “delusion box” introduced here and discussed as a counterexample for an ELK-like problem here.)
  • The actuators for carrying out an experiment have been tampered with. Perhaps our AI “proposes an experiment,” but actually the human sees something totally different (or is incapacitated).

You could hope to avoid this problem by using the same sensors and actuators your AI uses to interact with the world; then any form of tampering which undermines those sensors/actuators seems like it must also undermine the AI’s ability to act in the world.

But it seems like we could still have a problem:

  • Your AI may not continue acting using the same actuators or sensors, or at least not only the same actuators, with which it was originally constructed. It may have constructed tools in the environment, other autonomous processes, or other copies of itself.
  • Even if your AI is continuing to use the same sensors, it could tamper with sensors in a way that preserves their usefulness while forcing them to look innocuous to the human (e.g. by putting information in higher-order bits of a benign-looking video).
  • In some situations an AI would be motivated to tamper with sensors even if doing so reduced its own ability to act in the world.
Worry 2: an experiment can itself perform tampering

If we let our AI propose experiments to “reveal” tampering, it may instead propose experiments that do tampering from scratch. That tampering could then provide convincing evidence that something is fishy. If this can be done in any state, then it is useless for telling whether something fishy has actually happened.

We could try to avoid this by restricting to actions that are simple enough that a human understands how they work and can be confident that they won’t tamper. But it then becomes even more unclear whether there is any experiment that would identify problems (since we are effectively pitting a human experimenter against a very powerful manipulator).

Alternatively, we could hope that “reporting existing tampering” involves shorter or simpler action sequences than tampering from scratch. Then we could identify a state as fishy if it’s possible to find a short sequences of actions that look fishy.

But unfortunately the fishiest states could end up looking the safest. For example, suppose that if an AI spends 10 * N + 100 timesteps tampering with the sensors, it takes N steps to “undo” its efforts and reveal that something has gone wrong. A simple tampering attempt takes 200 timesteps to execute and 10 timesteps to undo, which sounds good. But if the AI spent 1100 timesteps tampering, it would take 100 timesteps to undo its work, which is less than the amount it would take to tamper from scratch. So this extremely-tampered-with state would end up looking less fishy than a normal state.

Strategy: detect covert tampering by analogy to visible tampering

(Part of the proposal here.)

As we’ve just discussed, some forms of sensor tampering might also destroy your ability to ever figure out whether tampering has occurred (by disabling actuators or sensors). But it’s still plausible that on the way to tampering we will pass through states in which something fishy is clearly about to happen. If we interrupt these action sequences in the middle, maybe we can do an experiment that shows that something has gone wrong.

Put differently: maybe action sequences that result in covert tampering have the same initial steps as action sequences that result in visible tampering.

More generally: we could look for any recognizable similarity between action sequences that result in covert tampering and action sequences that result in nearly-identical yet visible tampering.

Intuitively you might be optimistic about this if e.g. a would-be tamperer needs to somehow specify what they want to show up on the camera after it’s tampered with. If they just specified something slightly different then the action sequence would still be almost the same yet it would be easy to tell that the camera was tampered with (e.g. because it’s showing something physically impossible). Even if a human doesn’t understand how the covert tampering worked, they could say that something bad has happened because the action sequence is similar to one that causes overt tampering.

Worry: action sequences may be brittle

There may be sequences of action that covertly tamper with the sensors, but for which any small change leads to no tampering at all.

As an extreme example, we can imagine an AI whose first action is to commit to the hash of its next 50,000 actions. Perhaps it chooses a value x, and then installs a device that will cut power to its motors unless the hash of the next 50k actions is equal to x. Perhaps it “should have been” possible to change action 34,279 in order to carry out visible tampering rather than covert tampering—but instead what will happen is that the device will trigger and incapacitate the AI, so that no tampering at all occurs.

More realistically you might imagine an environment where an AI needs to prepare for chaotic influences in the environment—if it changes the action in timestep 17 it could have subtle effects later that change what actions in timestep 140 are feasible. In this case, even if the effects of tampering are determined by actions at late timesteps, you still need to take corresponding actions at early timesteps in order to prepare.

A broader way of viewing this concern is that covert and visible tampering may require “similar” actions in an abstract sense that is meaningful to an AI, which nevertheless correspond to very different sequences of low-level actions and which look quite different to a human.

If there are any kinds of tampering that are brittle in this way, then we won’t be able to recognize them (and so if we select against tampering we may be left with such actions).


D&D.Sci Holiday Special: How the Grinch Pessimized Christmas

31 декабря, 2021 - 19:23
Published on December 31, 2021 4:23 PM GMT

STORY (skippable)

Every Who down in Whoville enjoys Christmas Day

But you are the Grinch!  And whatever they say

You HATE Christmas Day, with those terrible lights

(They shine into your eyes, and they shine much too bright)

You HATE Christmas Day, with those huge awful trees

(You think you're allergic...they're making you sneeze)

But you HATE most of all all those Christmasy toys

That bring Christmas morning that Christmasy noise

(Oh god, all the noise!  All the noise, noise, NOISE, NOISE!)

(If there's one thing you hate, it's the noise, noise, NOISE, NOISE!!!!!!)

And tomorrow you know all those noises will start

As the presents bring joy to each Who child's heart

And they'll show all this joy (as Who children do)

By making loud noises and bothering you!

But as you are leaving earplugs by your bed

And preparing the pillows to cover your head

You realize there might be a thing you can do!

You can't steal the presents (they'd know it was you)

But for fifty-three years you've been hearing this riot

And sometimes it's louder!  And sometimes more quiet!

For the more a Who child enjoys all their toys

The more they will make all that terrible noise

But what if - you think - they were feeling less glee?

Yes, that is the way, it's so clear now you see!

For it's long been tradition (among all the Whos)

That when you give presents, each child must get two

And it's also tradition there (don't ask me why)

To distribute them randomly, closing your eyes

So at night, when the Whos sleep and dream of their joy

You'll creep in and rearrange all of those toys

And the very next morn, when the Who children wake,

They'll run down the stairs and their presents they'll take

But then, why, instead of the gifts that they're hoping

You'll give each one gifts that will make them start moping!

A doll for the Who who thinks dolls are for babies!

Puppies for the Whos who think they carry rabies!

And then, why at last, why on Christmas this year

You'll sleep soothed by the sound of those Who children's tears!

But the whisper, so soft, from your treacherous heart:

You could bring them more joy, and no more stand apart...

  • Each Who child has been given two presents.  Based on long Who tradition, you believe these presents have historically been allocated among the children randomly.
  • You must allocate two presents to each child, and may not assign any child two of the same present, but other than that you may distribute the gifts however you wish.
  • Your goal is to minimize the amount of noise produced (by all children combined).  You have a dataset of past presents received and noise made to help with this.
  • (You could instead/also submit a solution where you invert the goal and try to maximize the noise.  But why would you want to do that?)

The ten current Who children are:

NameAgeGenderAndy Sue Who12MBetty Drew Who11FSally Sue Who11FPhoebe Drew Who9FFreddie Lou Who8MEddie Sue Who8MCindy Drew Who6FMary Lou Who6FOllie Lou Who5MJohnny Drew Who4M

You have the following 20 toys to allocate among them (remember, exactly two toys to each child, and no child may get two of the same toy, or you'll give yourself away!):

  • Four Blum-Bloopers
  • Four Fum-Foozlers
  • Two Gah-Ginkas
  • Three Sloo-Slonkers
  • Three Trum-Troopas
  • Four Who-Whonkers

An answer key and leaderboard will be posted on Monday the 10th (my holiday schedule permitting).  

As usual, working together is allowed and encouraged, but for the sake of those who wish to work alone please spoiler-tag (type '>!' at the start of a line) any comments with information on the dataset.

Thank you to abstractapplic for feedback on a draft of this!  (For clarification, abstractapplic has no inside information on the dataset and can still play the scenario).


The Other Earth (Chapter 2)

31 декабря, 2021 - 19:20
Published on December 31, 2021 4:20 PM GMT

In this post, Chapter 2 of the sequence, I share the fable of The Other Earth from Olaf Stapledon’s Star Maker. It is abridged and lightly edited. I have also added a brief introduction titled “Learning to Travel” to provide context found elsewhere in Star Maker. It is within the fable of The Other Earth that Stapledon engages in a beautiful world modeling exercise that is relevant to our own journey.


Learning to Travel

It was the discovery of the psychical power of “projecting my mind” or of allowing my subjective viewpoint to expand and travel beyond earth from which I began my travels. And, as I would learn later, this was a skill that had to be exercised and studied before I could experience a wider range of the diversity of other minds inhabiting the cosmos. It turns out though that in the beginning of my journey I could only understand or communicate with minds that were more similar in constitution to my own. It was near the beginning of my journey that I found The Other Earth.

My consciousness had floated out into the greater cosmos. It floated out of my home solar system, out of the Great Milky Way, out into the great darkness between stars. This was something I had experienced already while honing my mind expanding power. However, this time was different. This time as I, I as my mind, floated out in the darkness, I spotted a planetary system not that dissimilar to my own. The sun seemed larger and darker, showing its age, but there, a planet was encircling it at an apparently approximate optimal distance, because I saw what could only be described as oceans and green continents scattered across this alien planet.

Being so drawn to the planet as I was, I decided to explore it closer to the ground:

On the Other Earth

As I slowly descended toward the surface of the little planet, I found myself searching for a land which promised to be like my home planet. But not sooner did I realize what I was doing than I reminded myself that conditions here would be entirely different from terrestrial conditions, and that it was very unlikely that I should find intelligent beings at all. If such beings existed, they would probably be quite incomprehensible to me. Perhaps they would be huge spiders or creeping jellies. How could I hope to ever make contact with such monsters?

I glided with wingless flight over the surface of the planet, through glades, across tracts of fractured rock, along the banks of streams. Presently I came to a wide region covered by neat, parallel rows of fern-like plants, bearing masses of nuts on the lower surfaces of their leaves. It was almost impossible to believe that this vegetable regimentation had not been intelligently planned. Or could it after all be merely a natural phenomenon not known on my own planet?

Such was my surprise that my power of locomotion, always subject to emotional interference, now began to fail me. I reeled in the air like a drunk man. Pulling myself together, I staggered on over the ranked crops toward a rather large object which lay some distance from me beside a strip of bare ground. Presently, to my amazement, my stupefaction, this object revealed itself as a plow. It was rather a queer instrument, but there was no mistaking the shape of the blade, which was rusty, and obviously made of iron. There were two iron handles, and chains for attachment to a beast of burden. It was difficult to believe that I was many light-years distant from home. Looking round, I saw an unmistakable cart track, and a bit of dirty ragged cloth hanging on a bush. Yet overhead was the unearthly sky, full noon with stars.

I followed a lane through a little wood of queer bushes, whose large fat drooping leaves had cherry-like fruits along their edges. Suddenly, round a bend in the lane, I came upon a man. Or so at first he seemed to my astounded and star-weary sight. I should not have been so surprised by the strangely human character of this creature had I at this early stage understood the forces that controlled my adventure. Influences which I shall later describe doomed me to discover first such worlds as were most akin to my own. Meanwhile the reader may well conceive my amazement at this strange encounter. I had always supposed that man was a unique being. An inconceivable complex conjunction of circumstances had produced him, and it was not to be supposed that such conditions would be repeated anywhere in the universe. Yet here, on the very first globe to be explored, was an obvious peasant. Approaching him, I saw that he was not quite so like terrestrial man as he seemed at a distance; but he was a man for all that. Had God, then, peopled the whole universe with our kind? Did he perhaps in very truth make us in his image? It was incredible. To ask such questions proved that I had lost my mental balance.

As I was a mere disembodied viewpoint, I was able to observe without being observed. I floated about him as he strode along the lane. He was an erect biped and in general plan definitely human. I had no means of judging his height, but he must have been approximately of normal terrestrial stature, or at least not smaller than a pigmy and not taller than a giant. He was of slender build. His legs were almost like a bird's, and enclosed in rough narrow trousers. Above the waist he was naked, displaying a disproportionately large thorax, shaggy with greenish hair. He had two short but powerful arms, and huge shoulder muscles. His skin was dark and ruddy, and dusted plentifully with bright green down.

All his contours were uncouth, for the details of muscles, sinews and joints were very plainly different from our own. His neck was curiously long and supple. His head I can best describe by saying that most of the brain-pan, covered with a green thatch, seemed to have slipped backwards and downwards over the nape. His two very human eyes peered from under the eaves of hair. An oddly projecting, almost spout-like mouth made him look as though he were whistling. Between the eyes, and rather above them, were a pair of green equine nostrils which were constantly in motion. The bridge of the nose was represented by an elevation in the thatch, reaching from the nostrils backwards over the top of the head. There were no visible ears. I discovered later that the auditory organs opened into the nostrils. Clearly although evolution on this Earth-like planet must have taken a course on the whole surprisingly like that which had produced my own kind, there must also have been divergencies.

The stranger wore not only boots but gloves, seemingly of tough leather. His boots were extremely short. I was to discover later that the feet of this race, the "Other Men" as I called them, were rather like the feet of an ostrich or a camel. The instep consisted of three great toes grown together. In place of the heel there was an additional broad, stumpy toe. The hands were without palms. Each was a bunch of three gristly fingers and a thumb.

When I had studied this agriculturalist for a while, I began to be strangely oppressed by his complete unawareness, of myself. It was not until I had followed my companion to his home, and had spent many days in that little circular stone house with roof of mudded wicker, that I discovered the power of entering into his mind, of seeing through his eyes, sensing through all his sense organs, perceiving the world just as he perceived it, and following much of his thought and his emotional life. Not till very much later, when I had passively "inhabited" many individuals of the race, did I discover how to make my presence known, and even to converse inwardly with my host.

I spent on the Other Earth many "other years," wandering from mind to mind and country to country, but I did not gain any clear understanding of the psychology of the Other Men and the significance of their history till I had encountered one of their philosophers, an aging but still vigorous man whose eccentric and unpalatable views had prevented him from attaining eminence. Most of my hosts, when they became aware of my presence within them, regarded me as an evil spirit or as a divine messenger. The more sophisticated, however, assumed that I was a mere disease, a symptom of insanity in themselves. They therefore promptly applied to the local "Mental Sanitation Officer." After I had spent according to the local calendar, a year or so of bitter loneliness among minds who refused to treat me as a human being, I had the good fortune to come under the philosopher's notice. One of my hosts, who complained of suffering from "voices," and visions of "another world," appealed to the old man for help. Bvalltu, for such approximately was the philosopher's name, the "ll" being pronounced more or less as in Welsh, Bvalltu effected a "cure" by merely inviting me to accept the hospitality of his own mind, where, he said, he would very gladly entertain me. It was with extravagant joy that I made contact at last with a being who recognized in me a human personality.

A Busy World

So many important characteristics of this world-society need to be described that I cannot spend much time on the more obvious features of the planet and its race. Civilization had reached a stage of growth much like that which was familiar to me. I was constantly surprised by the blend of similarity and difference. Traveling over the planet I found that cultivation had spread over most of the suitable areas, and that industrialism was already far advanced in many countries. On the prairies huge flocks of mammal-like creatures grazed and scampered. Larger mammals, or quasi-mammals, were farmed on all the best pasture. I say "quasi-mammal" because, though these creatures were viviparous, they did not suckle. The chewed cud, chemically treated in the maternal belly, was spat into the offspring's mouth as a jet of pre-digested fluid. It was thus also that human mothers fed their young.

My first visit to the metropolis of one of the great empires of the Other Earth was an outstanding experience. Everything was at once so strange and so familiar. There were streets and many-windowed stores and offices. In this old city the streets were narrow, and so congested was the motor traffic that pedestrians were accommodated on special elevated tracks slung beside the first-story windows and across the streets.

Here, then, was a host of persons who, in spite of their oddity, were as essentially human as Londoners. They went about their private affairs with complete assurance, ignorant that a spectator from another world found them one and all grotesque, with their lack of forehead, their great elevated quivering nostrils, their startlingly human eyes, their spout-like mouths. There they were, alive and busy, shopping, staring, talking. Children dragged at their mother's hands. Old men with white facial hair bowed over walking-sticks. Young men eyed young women. The prosperous were easily to be distinguished from the unfortunate by their newer and richer clothes, their confident and sometime arrogant carriage.

How can I describe in a few pages the distinctive character of a whole teeming and storied world, so different from my own, yet so similar? Here, as on my own planet, infants were being born every hour. Here, as there, they clamored for food, and very soon for companionship. They discovered what pain was, and what fear, and what loneliness, and love were. They grew up, molded by the harsh or kindly pressure of their fellow, to be either well nurtured, generous, sound, or mentally crippled, bitter, unwittingly vindictive. One and all they desperately craved the bliss of true community; and very few, fewer here, perhaps, than in my own world, found more than the vanishing flavor of it. They howled with the pack and hounded with the pack. Starved both physically and mentally, they brawled over the quarry and tore one another to pieces, mad with hunger, physical or mental. Sometimes some of them paused and asked what it was all for; and there followed a battle of words, but no clear answer. Suddenly they were old and finished. Then, the span from birth to death being an imperceptible instant of cosmical time, they vanished.

I must begin by speaking of the biological equipment of the Other Men. Their animal nature was at bottom much like ours. They responded with anger, fear, hate, tenderness, curiosity, and so on, much as we respond. In sensory equipment they were not unlike ourselves, save that in vision they were less sensitive to color and more to form than is common with us. The violent colors of the Other Earth appeared to me through the eyes of its natives very subdued. In hearing also they were rather ill-equipped. Though their auditory organs were as sensitive as ours to faint sounds, they were poor discriminators. Music, such as we know, never developed in this world.

In compensation, scent and taste developed amazingly. These beings tasted not only with their mouths, but with then-moist black hands and with their feet. They were thus afforded an extraordinarily rich and intimate experience of their planet. Tastes of metals and woods, of sour and sweet earths, of the many rocks, and of the innumerably shy or bold flavors of plants crushed beneath the bare running feet, made up of a whole world unknown to terrestrial man.

The genitals also were equipped with taste organs. There were several distinctive male and female patterns of chemical characteristics, each powerfully attractive to the opposite sex. These were savored faintly by contact of hands or feet with any part of the body, and with exquisite intensity in copulation.

This surprising richness of gustatory experience made it very difficult for me to enter fully into the thoughts of the Other Men. Taste played as important a part in their imagery and conception as sight in our own. Many ideas which terrestrial man has reached by way of sight, and which even in their most abstract form still bear traces of their visual origin, the Other Men conceived in terms of taste. For example, our "brilliant," as applied to persons or ideas, they would translate by a word whose literal meaning was "tasty." For "lucid" they would use a term which in primitive times was employed by hunters to signify an easily runnable taste-trail. To have "religious illumination" was to "taste the meadows of heaven." Many of our non-visual concepts also were rendered by means of taste. "Complexity" was "many flavored," a word applied originally to the confusion of tastes round a drinking pool frequented by many kinds of beasts. "Incompatibility" was derived from a word meaning the disgust which certain human types felt for one another on account of their flavors.

Difference of race, which in our world are chiefly conceived in terms of bodily appearance, were for the Other Men almost entirely differences of taste and smell. And as the races of the Other Men were much less sharply localized than our own races, the strife between groups whose flavors were repugnant to one another played a great part in history. Each race tended to believe that its own flavor was characteristic of all the finer mental qualities, was indeed an absolutely reliable label of spiritual worth. In former ages the gustatory and olfactory differences had, no doubt, been true signs of racial differences; but in modern times, and in the more developed lands, there had been great changes. Not only had the races ceased to be clearly localized, but also industrial civilization had produced a crop of genetic changes which rendered the old racial distinctions meaningless. The ancient flavors, however, though they had by now no racial significance at all, and indeed members of one family might have mutually repugnant flavors, continued to have the traditional emotional effects. In each country some particular flavor was considered the true hallmark of the race of that country, and all other flavors were despised, if not actually condemned.

In this world, as in our own, nearly all the chief means of production, nearly all the land, mines, factories, railways, ships, were controlled for private profit by a small minority of the population. These privileged individuals were able to force the masses to work for them on pain of starvation. The tragic farce inherent in such a system was already approaching. The owners directed the energy of the workers increasingly toward the production of more means of production rather than to the fulfillment of the needs of individual life. For machinery might bring profit to the owners; bread would not. With increasing competition of machine with machine, profits declined, and therefore wages, and therefore effective demand for goods. Marketless products were destroyed, though bellies were unfed and backs unclad. Unemployment, disorder, and stem repression increased as the economic system disintegrated. A familiar story!

As conditions deteriorated, and the movements of charity and state-charity became less and less able to cope with the increasing mass of unemployment and destitution, the new pariah-race became more and more psychologically useful to the hate-needs of the scared, but still powerful, prosperous. The theory was spread that these wretched beings were the result of secret systematic race-pollution by riff-raff immigrants, and that they deserved no consideration whatever. They were therefore allowed only the basest forms of employment and the harshest conditions of work. When unemployment had become a serious social problem, practically the whole pariah stock was workless and destitute. It was of course easily believed that unemployment, far from being due to the decline of capitalism, was due to the worthless-ness of the pariahs.

At the time of my visit the working class had become tainted through and through by the pariah stock, and there was a vigorous movement afoot amongst the wealthy and the official classes to institute slavery for pariahs and half-pariahs, so that these might be openly treated as the cattle which in fact they were. In view of the danger of continued race-pollution, some politicians urged wholesale slaughter of the pariahs, or, at the least, universal sterilization. Others pointed out that, as a supply of cheap labor was necessary to society, it would be wiser merely to keep their numbers down by working them to an early death in occupations which those of "pure race" would never accept. This, at any rate, should be done in times of prosperity, but in times of decline, the excess population could be allowed to starve, or might be used up in the physiological laboratories.

The persons who first dared to suggest this policy were scourged by the whips of generous popular indignation. But their policy was in fact adopted; not explicitly but by tacit consent, and in the absence of any more constructive plan.

The first time I was taken through the poorest quarter of the city I was surprised to see that, though there were large areas of slum property far more squalid than anything in England, there were also many great clean blocks of tenements worthy of Vienna. These were surrounded by gardens, which were crowded with wretched tents and shanties. The grass was worn away, the bushes damaged, the flowers trampled. Everywhere men, women and children, all filthy and ragged were idling.

I learned that these noble buildings had been erected before the world-economic crises (familiar phrase!) by a millionaire who had made his money in trading an opium-like drug. He presented the building to the City Council, and was gathered to heaven by the way of the peerage. The more deserving and less unsavory poor were duly housed; but care was taken to fix the rent high enough to exclude the pariah-race. Then came the crisis. One by one tenants failed to pay their rent, and were ejected. Within a year the buildings were almost empty.

There followed a very curious sequence of events, and one which, as I was to discover, was characteristic of this strange world. Respectable public opinion, though vindictive toward the unemployed, was passionately tender toward the sick. In falling ill, a man acquired a special sanctity, and exercised a claim over all healthy persons. Thus no sooner did any of the wretched campers succumb to a serious disease than he was carried off to be cared for by all the resources of medical science. The desperate paupers soon discovered how things stood, and did all in their power to fall sick. So successful were they, that the hospitals were soon filled. The empty tenements were therefore hastily fitted out to receive the increasing flood of patients.

Observing these and other farcical events, I was reminded of my own race. But though the Other Men were in many was so like us, I suspected increasingly that some factor still hidden from me doomed them to a frustration which my own nobler species need never fear. Psychological mechanisms which in our case are tempered with common sense or moral sense stood out in this world in flagrant excess. Yet it was not true that Other Man was less intelligent or less moral than man of my own species. In abstract thought and in practical invention he was at least our equal. Many of his most recent advances in physics and astronomy had passed beyond our present attainment. I noticed, however, that psychology was even more chaotic than with us, and that social thought was strangely perverted.

In radio and television, for instance, the Other Men were technically far ahead of us, but the use to which they put their astounding inventions was disastrous. In civilized countries everyone but the pariahs carried a pocket receiving set. As the Other Men had no music, this may seem odd; but since they lacked newspapers, radio was the only means by which the man in the street could learn the lottery and sporting results which were his staple mental diet. The place of music, moreover, was taken by taste- and smell-themes, which were translated into patterns of ethereal undulation, transmitted by all the great national stations, and restored to their original form in the pocket receivers and taste-batteries of the population. These instruments afforded intricate stimuli to the taste organs and scent organs of the hand. Such was the power of this kind of entertainment that both men and women were nearly always seen with one hand in a pocket. A special wavelength had been allotted to the soothing of infants.

A sexual receiving set had been put upon the market, and programs were broadcast for it in many countries, but not in all. This extraordinary invention was a combination of radio — touch, taste, odor, and sound. It worked not through the sense organs, but direct stimulation of the appropriate brain-centers. The recipient wore a specially constructed skullcap, which transmitted to him from a remote studio the embraces of some delectable and responsive woman, as they were then actually being experienced by a male “love-broadcaster” or as electromagnetically recorded on a steel tape on some earlier occasion.

The principle of radio-brain-stimulation was soon developed much further. Programs of all the most luscious or piquant experiences were broadcast in all countries, and could be picked up by simple receivers that were within the means of all save the pariahs. Thus even the laborer and the factory hand could have the pleasures of a banquet without expense and subsequent repletion, the delights of proficient dancing without the trouble of learning the art, the thrills of motor-racing without danger. In an ice-bound northern home he could bask on tropical beaches, and in the tropics indulge in winter sports. Governments soon discovered that the new invention gave them a cheap and effective kind of power over their subjects. Slum-conditions could be tolerated if there was an unfailing supply of illusory luxury. Reforms distasteful to the authorities could be shelved if they could be represented as inimical to the national radio-system. Strikes and riots could often be broken by the mere threat to close down the broadcasting studios, or alternatively by flooding the ether at a critical moment with some saccharine novelty.

During my last years on the Other Earth a system was invented by which a man could retire to bed for life and spend all his time receiving radio programs. His nourishment and all his bodily functions were attended to by doctors and nurses attached to the Broadcasting Authority. In place of exercise he received periodic massage. Participation in this scheme was at first an expensive luxury, but its inventors hoped to make it at no distant date available to all. It was even expected that in time medical and menial attendants would cease to be necessary. A vast system of automatic food-production, and distribution of liquid pabulum by means of pipes leading to the mouths of the recumbent subjects, would be complemented by an intricate sewage system. Electric massage could be applied at will by pressing a button. Medical supervision would be displaced by an automatic endocrine-compensation system. This would enable the condition of the patient’s blood to regulate itself automatically by tapping from the communal drug-pipes whatever chemicals were needed for correct psychological balance.

Even in the case of broadcasting itself the human element would no longer be needed, for all possible experiences would have been already recorded from the most exquisite living examples. These would be continuously broadcast in a great number of alternative programs. A few technicians and organizers might still be needed to superintend the system; but, properly distributed, their work would entail for each member of the World Broadcasting Authority’s staff no more than a few hours of interesting activity each week.

Children, if future generations were required, would be produced ectogenetically. The World Director of Broadcasting would be requested to submit psychological and physiological specifications of the ideal “listening breed.” Infants produced in accordance with this pattern would then be educated by special radio programs to prepare them for adult radio life. They would never leave their cots, save to pass by stages to the full-sized beds of maturity. At the latter end of life, if medical science did not succeed in circumventing senility and death, the individual would at least be able to secure a painless end by pressing an appropriate button.

Now at this time economic confusion had been driving the great commercial empires of the Other Earth into more and more desperate competition for markets. These economic rivalries had combined with ancient tribal passions of fear and hate and pride to bring about an interminable series of war scares each of which threatened universal Armageddon.

Trouble was successfully created between one of the older commercial empires and a certain state which had only recently adopted mechanical civilization, but was already a Great Power, and a Power in desperate need of markets. Radio, which formerly had been the main force making cosmopolitanism, became suddenly in each country the main stimulus to nationalism. Morning, noon and night, every civilized people was assured that enemies, whose flavor was of course subhuman and foul, were plotting its destruction. Armament scares, spy stories, accounts of the barbarous and sadistic behavior of neighboring peoples, created in every country such uncritical suspicion and hate that war became inevitable.

A dispute arose over the control of a frontier province. During those critical days, Bvalltu and I happened to be in a large provincial town. I shall never forget how the populace plunged into almost maniacal hate. All thought of human brotherhood, and even of personal safety, was swept away by a savage blood-lust. Panic-stricken governments began projecting long-range rocket bombs at their dangerous neighbors. Within a few weeks several of the capitals of the Other Earth had been destroyed from the air. Each people now began straining every nerve to do more hurt than it received.

Of the horrors of this war, of the destruction of city after city, of the panic-stricken, starving hosts that swarmed into the open country, looting and killing, of the starvation and disease, of the disintegration of the social services, of the emergence of ruthless military dictatorships, of the steady or catastrophic decay of culture and of all decency and gentleness in personal relations, of this there is no need to speak in detail.

Instead, I shall try to account for the finality of the disaster which overtook the Other Men. My own human kind, in similar circumstances, would never, surely, have allowed itself to be so completely overwhelmed. No doubt, we ourselves are faced with the possibility of a scarcely less destructive war; but, whatever the agony that awaits us, we shall almost certainly recover. Foolish we may be, but we always manage to avoid falling into the abyss of downright madness. At the last moment sanity falteringly reasserts itself. Not so with the Other Men.

Prospects Of The Race

Increasingly I suspected that this race, in spite of all its triumphs, was now living on the great ideas of its past, mouthing concepts that it no longer had the sensibility to understand, paying verbal homage to ideals which it could no long sincerely will, and behaving within a system of institutions many of which could only be worked successfully by minds of a slightly finer temper. These institutions, I suspected, must have been created by a race endowed not only with much greater intelligence, but with a much stronger and more comprehensive capacity for community than was now possible on the Other Earth. They seemed to be based on the assumption that men were on the whole kindly, reasonable, and self-disciplined.

I had often questioned Bvalltu on this subject, but he had always turned my questions aside. It will be remembered that, though I had access to all his thoughts so long as he did not positively wish to withhold them, he could always, if he made a special effort, think privately. I had long suspected that he was keeping something from me, when at last he told me the strange and tragic facts.

It was a few days after the bombardment of the metropolis of his country. Through Bvalltu's eyes and the goggles of his gas-mask I saw the results of that bombardment. We had missed the horror itself, but had attempted to return to the city to play some part in the rescue work. Little could be done. So great was the heat that still radiated from the city's incandescent heart, that we could not penetrate beyond the first suburb. Even there the street's were obliterated, choked with fallen buildings. Human bodies, crushed and charred, projected here and there from masses of tumbled masonry. Most of the population was hidden under the ruins. In the open spaces many lay gassed. Salvage parties impotently wandered. Between the smoke-clouds the Other Sun occasionally appeared, and even a daytime star.

After clambering among the ruins for some time, seeking vainly to give help, Bvalltu sat down. The devastation round about us seemed to "loosen his tongue," if I may use such a phrase to express a sudden frankness in his thinking toward myself. I had said something to the effect that a future age would look back on all this madness and destruction with amazement. He sighed through his gas-mask, and said, "My unhappy race has probably now doomed itself irrevocably." I expostulated; for though ours was about the fortieth city to be destroyed, there would surely some day be a recovery, and the race would at last pass through this crisis and go forward from strength to strength. Bvalltu then told me of the strange matters which, he said, he had often intended to tell me, but somehow he had always shunned doing so. Though many scientists and students of the contemporary world-society had some vague suspicion of the truth, it was clearly known only to himself and a few others.

The species, he said, was apparently subject to strange and long-drawn-out fluctuations of nature, fluctuations which lasted for some twenty thousand years. All races in all climates seemed to manifest this vast rhythm of the spirit, and to suffer it simultaneously. Its cause was unknown. Though it seemed to be due to an influence affecting the whole planet at once, perhaps it actually radiated from a single starting point, but spread rapidly into all lands. Very recently an advanced scientist had suggested that it might be due to variations in the intensity of "cosmic rays." Geological evidence had established that such a fluctuation of cosmical radiation did occur, caused perhaps by variations in a neighboring cluster of young stars. It was still doubtful whether the psychological rhythm and the astronomical rhythm coincided, but many facts pointed to the conclusion that when the rays were more violent the human spirit declined.

Bvalltu was not convinced by this story. On the whole he inclined to the opinion that the rhythmical waxing and waning of human mentality was due to causes nearer home. Whatever the true explanation, it was almost certain that a high degree of civilization had been attained many times in the past, and that some potent influence had over and over again damped down the mental vigor of the human race. In the troughs of these vast waves Other Man sank to a state of mental and spiritual dullness more abject than anything which my own race had ever known since it awoke from the subhuman. But at the wave's crest man's intellectual power, moral integrity, and spiritual insight seem to have risen to a pitch that we should regard as superhuman.

Again and again the race would emerge from savagery, and pass through barbarian culture into a phase of worldwide brilliance and sensibility. Whole populations would conceive simultaneously an ever-increasing capacity for generosity, self-knowledge, self-discipline, for passionate and penetrating thought and uncontaminated religious feeling.

Consequently within a few centuries the whole world would blossom with free and happy societies. Average human beings would attain an unprecedented clarity of mind, and by massed action do away with all grave social injustices and private cruelties. Subsequent generations, inherently sound, and blessed with a favorable environment, would create a world-wide Utopia of awakened beings.

Presently a general loosening of fiber would set in. The golden age would be followed by a silver age. Living on the achievements of the past, the leaders of thought would lose themselves in a jungle of subtlety, or fall exhausted into mere slovenliness. At the same time moral sensibility would decline. Men would become on the whole less sincere, less self-searching, less sensitive to the needs of others, in fact less capable of community. Social machinery, which had worked well so long as citizens attained a certain level of humanity, would be dislocated by injustice and corruption. Tyrants and tyrannical oligarchies would set about destroying liberty. Hate-mad submerged classes would give them good excuse. Little by little, though the material benefits of civilization might smolder on for centuries, the flame of the spirit would die down into a mere flicker in a few isolated individuals. Then would come sheer barbarism, followed by the trough of almost sub-human savagery.

On the whole there seemed to have been a higher achievement on the more recent crests of the wave than on those of the "geological" past. So at least some anthropologists persuaded themselves. It was confidently believed that the present apex of civilization was the most brilliant of all, that its best was as yet to come, and that by means of its unique scientific knowledge it would discover how to preserve the mentality of the race from a recurrence of deterioration.

The present condition of the species was certainly exceptional. In no earlier recorded cycle had science and mechanization advanced to such lengths. So far as could be inferred from the fragmentary relics of the previous cycle, mechanical invention had never passed beyond the crude machinery known in our own mid-nineteenth century. The still earlier cycles, it was believed, stagnated at even earlier stages in their industrial revolutions.

Now though it was generally assumed in intellectual circles that the best was yet to be, Bavalltu and his friends were convinced that the crest of the wave had already occurred many centuries ago. To most men, of course, the decade before the war had seemed better and more civilized than any earlier age. In their view civilization and mechanization were almost identical, and never before had there been such a triumph of mechanization. The benefits of a scientific civilization were obvious. For the fortunate class there was more comfort, better health, increased stature, a prolongation of youth, and a system of technical knowledge so vast and intricate that no man could know more than its outline or some tiny corner of its detail. Moreover, increased communications had brought all the peoples into contact. Logical idiosyncrasies were fading out before the radio, the cinema, and the gramophone. In comparison with these hopeful signs it was easily overlooked that the human constitution, though strengthened by improved conditions, was intrinsically less stable than formerly. Certain disintegrative diseases were slowly but surely increasing. In particular diseases of the nervous system were becoming more common and more pernicious. Cynics used to say that the mental hospitals would soon outnumber even the churches. But the cynics were only jesters. It was almost universally agreed that, in spite of wars and economic troubles and social upheavals, all was now well, and the future would be better.

The truth, said Bvalltu, was almost certainly otherwise. There was, as I had suspected, unmistakable evidence that the average of intelligence and of moral integrity throughout the world had declined; and they would probably continue to do so. Already the race was living on its past. All the great seminal ideas of the modern world had been conceived centuries ago. Since then, world-changing applications of these ideas had been made; but none of these sensational inventions had depended on the extreme kind of penetrating the whole course of thought in an earlier age. Recently there had been, Bvalltu admitted, a spate of revolutionary scientific discoveries and theories, but not one of them, he said, contained any really novel principle. They were are recombinations of familiar principles. Scientific method, invented some centuries ago, was so fertile a technique that it might well continue to yield rich fruit for centuries to come even in the hands of workers incapable of any high degree of originality.

But it was not in the field of science so much as in moral and practical activity that the deterioration of mental caliber was most evident. I myself, with Bvalltu's aid, had learnt to appreciate to some extent the literature of that amazing period, many centuries earlier, when every country seemed to blossom with art, philosophy and religion; when people after people had changed its whole social and political order so as to secure a measure of freedom and prosperity to all men; when state after state had courageously disarmed, risking destruction but reaping peace and prosperity; when police forces were disbanded, prisons turned into libraries or colleges; when weapons and even locks and keys came to be known only as museum pieces; when the four great established priesthoods of the world had exposed their own mysteries, given their wealth to the poor, and led the triumphant campaign for community; or had taken to agriculture, handicrafts, teaching, as befitted humble supporters of the new priest-less, faithless, Godless religion of world-wide community and inarticulate worship. After some five hundred years locks and keys, weapons and doctrines, began to return. The golden age left behind it only a lovely and incredible tradition, and a set of principles which, though now sadly misconceived, were still the best influences in a distraught world.

Shortly before I left the Other Earth a geologist discovered a fossil diagram of a very complicated radio set. It appeared to be a lithographic plate which had been made some ten million years earlier. The highly developed society which produced it had left no other trace. This find was a shock to the intelligent world; but the comforting view was spread abroad that some non-human and less hardy species had long ago attained a brief flicker of civilization. It was agreed that man, once he had reached such a height of culture, would never have fallen from it.

In Bvalltu's view man had climbed approximately to the same height time after time, only to be undone by some hidden consequence of his own achievement.

When Bvalltu propounded this theory, among the ruins of his native city, I suggested that some time, if not this time, man would successfully pass this critical point in his career. Bvalltu then spoke of another matter which seemed to indicate that we were witnessing the final act of this long-drawn-out and repetitive drama. It was known to scientists that, owing to the weak gravitational hold of their world, the atmosphere, already scant, was steadily decreasing. Sooner or later humanity would have to face the problem of stopping this constant leakage of precious oxygen. Hitherto life had successfully adaptive itself to the progressive rarefaction of atmosphere, but the human physique had already reached the limit of adaptability in this respect. If the loss were not soon checked, the race would inevitably decline. The only hope was that some means to deal with the atmospheric problem would be discovered before the onset of the next age of barbarism. There had only been a slight possibility that this would be achieved. This slender hope the war had destroyed by setting the clock of scientific research back for a century just at the time when human nature itself was deteriorating and might never again be able to tackle so difficult a problem.

The thought of the disaster which almost certainly lay in wait for the Other Men threw me into a horror of doubt about the universe in which such a thing could happen. That a whole world of intelligent beings could be destroyed was not an unfamiliar idea to me; but there is a great difference between an abstract possibility and a concrete and inescapable danger. On my native planet, whenever I had been dismayed by the suffering and the futility of individuals, I had taken comfort in the thought that at least the massed effect of all our blind striving must be the slow but glorious awakening of the human spirit. This hope, this certainty, had been the one sure consolation. But now I saw that there was no guarantee of any such triumph. It seemed that the universe, or the maker of the universe, must be indifferent to the fate of worlds. That there should be endless struggle and suffering and waste must of course be accepted; and gladly, for these were the very soil in which the spirit grew. But that all struggle should be finally, absolutely vain, that a whole world of sensitive spirits fail and die, must be sheer evil. In my horror it seemed to me that Hate must be the Star Maker.

Not so to Bavalltu. "Even if the powers destroy us," he said, "who are we, to condemn them? As well might a fleeting word judge the speaker that forms it. Perhaps they use us for their own high ends, use our strength and our weakness, our joy and our pain, in some theme inconceivable to us, and excellent." But I protested, "What theme could justify such waste, such futility? And how can we help judging; and how otherwise can we judge than by the light of our own hearts, by which we judge ourselves? It would be base to praise the Star Maker, knowing that he was too insensitive to care about the fate of his worlds." Bvalltu was silent in his mind for a moment. Then he looked up, searching among the smoke-clouds for a daytime star. And then he said to me in his mind, "If he saved all the worlds, but tormented just one man, would you forgive him? Or if he was a little harsh only to one stupid child? What has our pain to do with it, our failure? Star Maker! It is a good word, though we can have no notion of its meaning. Oh, Star Maker, even if you destroy me, I must praise you. Even if you torture my dearest. Even if you torment and waste all your lovely worlds, the little figments of your imagination, yet I must praise you. For if you do so, it must be right. In me it would be wrong, but in you it must be right."

He looked down once more upon the ruined city, then continued, "And if after all there is no Star Maker, if the great company of galaxies leapt into being on their own accord, and even if this little nasty world of ours is the only habitation of the spirit anywhere among the stars, and this world doomed, even so, even so, I must praise. But if there is no Star Maker, what can it be that I praise? I do not know. I will call it only the sharp tang and savor of existence. But to call it this is to say little."


Reflections on The Other Earth


I use the space here for this extended fable to help us all in exercising our world modeling skills. This is done deliberately by Stapledon, and we can easily imagine the continued relevance of imagining a world in which the human species loses control of its machines and its hybrid superintelligences. I will attempt to highlight throughout this sequence, including through literature and lyrics, the grave concern artists, writers, and thinkers have expressed, for at least the past 200+ years of humanity’s tenuous relationship with its machines and its organizations.

In The Other Earth, we find a world, less lucky than ours. One that succumbed to similar trends that we see present in our own world, and then, eventually, to civilization and global collapse. In this world, we found that despite the growth in mechanical  and scientific civilization leading to great progress in wealth and comfort, great progress was also made in efficiency in killing, general destruction, and in the degradation of the human spirit. In this world, these negative consequences led to self destruction of the mechanical creators at the hands of their own inventions. This seemed to manifest for the other men, some innate inability, over and over, to rise above their very human condition.

As noted earlier in this sequence, it may be the case that the flaws of our own human nature are unavoidably bad in their collective, or herd mentality. It may be a feature of this sort of group or hive mind rather than a bug. It does seem that our collective minds do have a similar flavor to our individual minds, but in great excess in every direction. (They howled with the pack and hounded with the pack.) A form of collective minds can be found not only for groups of humans and animals, but also in organizations and modern hybrid superintelligences.

That is, if we think of collective minds as being constituted by smaller minds that function in some semi autonomous and semi coordinated fashion that contributes to the processing of the broader collection of minds, then each of these collections will be defined in part by the nature of the minds that constitute its decisionmaking or mindedness functioning, and in part on the nature of those kinds of minds and the specific ways in which they are connected and communicate with other minds within the collective.

Thus the behavior of the collective mind, as with any sufficiently complex decision making system, is determined, in large part, by the individuals agents that constitute that system and their interactions.

This is the major point of the fable: The human condition imposes a strained relationship between humanity and its machines. As machines play an increasing role in shaping the production process of culture and economy, humans put at risk their very humanity. That is, there may be some point of mechanization that is detrimental to the human spirit, to human civilization, to human flourishing.

And as our narrator reminds us:

That a whole world of intelligent beings could be destroyed was not an unfamiliar idea to me; but there is a great difference between an abstract possibility and a concrete and inescapable danger.


This has been Chapter 2 of the sequence. In Chapter 3 we turn to examine Huxley’s argument that over-population and over-organization have given rise to Our Brave New World for which we have had to contend with since the latter half of the 20th Century.


Omicron Post #11

31 декабря, 2021 - 18:50
Published on December 31, 2021 3:50 PM GMT

The traditional action for New Year’s Eve is to go to a party, stay up until midnight and watch the bell drop, and the traditional song is Auld Lang Syne.

I’ve stopped doing any of these things.

My traditional action is now to watch a bunch of great college football games that are the cumulation of the year’s season, and maybe have a few friends over.

As for songs, I have two. First, to look back on the year and the ways one is stuck, and think about and motivate what one might change, Anastasia’s Same Old Story. Then, to remind that life is awesome and my life is awesome, Jewel’s This Way.

It’s important to have rituals and traditions that speak to you.

It’s also important to contain multitudes.

The darkness is coming. So is the light.

Life is good. All signs point towards the golden scenario of a quick peak under reduced severity.

Happy new year!

Around the World

South Africa, the UK and Denmark are no longer uniquely informative places, so I’m combining all the areas into one section now.

There are a few cities where we may have peaked, but otherwise outside of South Africa? Not so much.

Cases still growing in the UK, but perhaps no longer growing in London.

Here’s a UK hospital update, from the comments via MoveMyCat:

I work at a large teaching hospital in the UK, but outside London. I’ve been isolating for a while (very mild if any symptoms, but was undergoing regular testing at work). Prior to my isolation, there were rising numbers, but they were still using respiratory beds for ‘Incidental Covid’ types. Anecdotal observation supported the idea that staff isolation and the need to isolate from vulnerable Covid-negatives was more of a challenge at that point than any influx of very unwell Covid pneumonitis types.

I’ll aim to update when I’m next in work.

I’m back in work now. There’s a small uptick in actual respiratory admissions but most of the increased Covid capacity the hospital has put together is being filled by geriatrics patients who happen to have Covid. A friend who works at a nearby hospital reports that she has about a quarter of her unit in self isolation for one reason or other.

It’s actually remarkable how few people are coming in with fevers and breathlessness.

If we do have access to the numbers on any of that, please do share, for now I only know of such numbers in San Francisco.

Germany is behind other countries in terms of spread and has stalled for time, but it won’t be that much time, and it will change very little. That last week in gray isn’t fully reported yet, so this is the same old story as everywhere else.

In hindsight, that first Boston wastewater jump was probably Delta. The one now, not so much:

If we use the black line, the average is now over four times the previous maximum level. If we use the recent data points as an indication, there’s at least one more doubling after that.

An update from San Francisco, where the ‘with Covid’ and ‘for Covid’ distinction is actually being measured (and I’m happy that Dr. Fauci has ‘broken the seal’ and is talking about the distinction in earnest.)

This seems roughly right to me. If you have a highly vaccinated population and solid hospital capacity, even with this rapid a storm, we should mostly be able to muddle through, things can only add to 100% (and presumably cap out at more like a total of 80%-90%). If the peak is as high in rural areas with less beds and less vaccinations, I’d be a lot more nervous.

Note that San Francisco’s positive rate is somewhat lower than the national average, so it’s probably missing a slightly lower percentage of its cases than other places.

21% of NYPD is out sick.

21% of the NYPD was out sick Thursday in what outgoing Police Commissioner Dermot Shea said was the highest level of sick employees he had ever seen. 17% were out a day earlier.

“When you look at March 2020 we peaked at 19.8%, we hit 21% today,” said Shea during a briefing on security for New Year’s Eve festivities in Times Square.

30% of EMS and 17% FDNY were on medical leave.

That can’t go on for that long, because math.

Forgot to include it in the weekly post, but the daily positive rate in the USA is another good measure.

I still don’t know how to properly translate a 16% positive rate into an expected percentage of cases being missed, but this will likely be the best way to tell now whether things have actually peaked or not. We previously peaked at around 23% on this chart early in the pandemic, with 50%+ rates in New York City, on vastly less testing.

Vaccine Effectiveness

A commenter on the last post notes that in Canada, while measured vaccine effectiveness in preventing infections dropped to zero or below (I continue to think this is a rural/urban effect in large part), against hospitalization it’s still strong. Graham thinks the essential story is that vaccine protection against severe outcomes is holding but doesn’t believe that Omicron is much less virulent than Delta (to continue to note that opinions on this differ.)

Using Alberta, Canada numbers. The current rate of those hospitalized who were vaccinated is 3.68/100k, and among the unvaccinated it is 20.17/100k. On December 4 it was vaccinated 3.84/100k and unvaccinated 26.69/100k. Since the rise of Omicron, the rate of hospitalizations has actually gone down in both populations, but the vaccinated now have a slightly higher relative rate of hospitalization than they did pre-Omicron. (Vaccinated rate was 14.4% that of the unvaccinated rate 3 weeks ago, and now it is closer to 18.2%.)

Two shots of J&J in South Africa gave 85% protection against hospitalization. The first shot alone gave 63%, so the second shot was an additional 50% from there.

T-Cells from vaccination still holding up well. Looking good.

It’s still a pandemic of the unvaccinated when it comes to hospitalization and death.

Test Sensitivity

The FDA uses the magic weasel words and claims rapid tests ‘may have reduced sensitivity’ to Omicron. Given the nature of what ‘may’ be true, such statements are always technically correct, and also always the ‘may’ mostly gets translated as ‘do’ or at least as ‘probably.’

I translate it that way too, because that’s the claim they’re effectively making. If you don’t have a very good reputation for your words having meaning, I’m not going to let you pull this transparent a weaseling that doesn’t deserve the dignify of being called a motte-and-bailey.

Abbot flat out denies the FDA’s claim of potential lowered test sensitivity, says their tests are as effective against Omicron as they were against previous variants.

This is at least some actual preliminary data suggesting that reduced sensitivity, but it has enough caveats and low enough sample size I’m not updating much on it.

Michael Mina, as designated fire-backer, fires back.

He links to this report.

Are you pissed, Mina? Yeah.

Mina’s prior seems right to me, but that interpretation doesn’t mean everything is fine. This still means that a rapid test is less effectively sensitive for knowing if you’re infectious. Which is what they’re for, so there’s still a problem whether or not the test is technically less sensitive. What matters is the practical utility of the test, and what it tells you.

I do agree that giving out vague ‘may be less effective’ warnings based on preliminary data is a good way to completely freak out a lot of people in an unhelpful way, without much in the way of upside.


Mouse model finds less pneumonia risk from Omicron.


What are we doing about Omicron?

One can see this as the ‘less action’ camp versus the ‘more action’ camp, and there’s something to that. But it’s also important to notice that different actions make sense in different circumstances. The CDC reduced isolation requirements because conditions got worse, and it was necessary. That doesn’t contradict taking more precautions in other ways, nor does it require having an attitude of moving on.

Lack of government restrictions is good, but doesn’t mean business as usual, some places won’t be shrugging this off.

Three or four weeks with the promise of light after that seems highly survivable. That’s an accelerated timeline and it could easily take longer, but the restaurants should be fine not too long from now, if people use their heads.

The thing is, if you have already had and recovered from Omicron, restaurants are fully open for your business. About one in ten Londoners had the virus at one time this past week. All those people, or least the ones that are aware of it, can then fully exit their lockdowns. A month later, half the population can be partying like it’s 2019, on top of the ones who already didn’t much care, even if things haven’t fully died down yet.

The future’s so bright we may, in that future, require shades.

What if?

It’s a good question. Some of the confusion is that ‘collapse society’ is not a well-formed concept here. There are different levels of sacred value, and things we are and aren’t prepared to accept, as things get more dire, and trade-offs are being made. If the situation were different and worse, it would require different trade-offs.

Letting the original strain rage freely would not have collapsed society. It would have killed a lot of people and scarred pretty much everyone, but most of us would have survived and moved on. Forcing everyone with Omicron into extended isolation would shut down a lot of things over the next few weeks (with little upside to compensate) and if this included hospital staff it likely kills more people rather than less people. But we can survive those kinds of disruptions, it would be quite bad but the unthinkable now becomes the thinkable tomorrow.

But maybe that’s wrong? If we’d suffered a 4% death rate in the first half of 2020, what happens? What about if the supply chain problems hadn’t turned around?

The other perspective is that perhaps we really are a society on the brink of collapse, that can no longer be resilient in the face of threats, and we’re now dodging civilizational collapse on a rolling basis until at some point we fail to dodge it. In this model, the close calls we’ve had recently were actually remarkably close and could easily have gone in other ways.

In this case, I think we know the answer, because there was a lot of uncertainty about severity, and Bruno’s scenario was very much on the table until a week or two ago. It seems clear we were pretty committed to this course of action either way, and we’d have had to accept the consequences.

What about if we go to the next hypothetical? What if Omicron was more infectious by default and had also rendered the vaccines completely ineffective on all levels? At that point, I do think we have a real crisis where all the choices seem impossible, because we really can’t actually contain this thing under those conditions, but letting it rip would have unthinkable consequences and we’d be forced to try to slow things down, even though it would be a mistake – we wouldn’t be able to do enough slowing to have that much marginal effect.

At that point, the correct reaction really would be a ‘protect the vulnerable as best we can’ style reaction, while actively letting things go as quickly as possible otherwise to give that strategy a chance. I don’t think we’re capable of implementing that. But it’s important to note that we don’t know, and won’t until we are forced to run the experiment. The wartime generals are available if things get bad enough for the phone to ring.

Threads and Posts

29 December thread from Bob Wachter about what might happen over the next two months. It lays out the same scenario I’ve been increasingly expecting as the puzzle pieces have fallen into place, matching my non-logistical priors quite well.

This seems like wishful thinking with regard to testing capacity, and also Paxlovid availability in the short-to-medium term (it’s also explicitly labeled wishful thinking, which is fully allowed.) My guess is London won’t peak until this coming week, but it’s impossible to tell because at such high levels you’re in the Fog of War and the official numbers aren’t helpful. Medicine and logistics are different magisteria.

But the rest is most of it and doing most of the work, and it checks out, so long as one reads February as the end of the month.

The current shortage in hospitals, WaPo reports (no paywall), is not equipment or beds but health care workers, because so many are sick and everyone is burnt out and stretched to the limit already after two years. Short term we’ve lowered isolation requirements, but there’s little we can do. Long term, the obvious solution is to expand supply by lowering the regulatory barriers that create artificial scarcity.

Kai looks at the question of how much of increased transmissibility is evasion, versus being more infectious (I’d add versus there being a shorter generation time, as well).

It’s worth noting that such official reports are often rather far behind, and a ‘no clear epidemiological demonstration’ type of statement from such a report means little.

Preprint here.

Comparing households infected with the Omicron to Delta VOC, we found an 1.17 (95\%-CI: 0.99-1.38) times higher SAR for unvaccinated, 2.61 times (95\%-CI: 2.34-2.90) higher for fully vaccinated and 3.66 (95\%-CI: 2.65-5.05) times higher for booster-vaccinated individuals, demonstrating strong evidence of immune evasiveness of the Omicron VOC. Our findings confirm that the rapid spread of the Omicron VOC primarily can be ascribed to the immune evasiveness rather than an inherent increase in the basic transmissibility.

Here’s Peter Hanson’s interpretation of the same study.

I do find this to be strong evidence that the transmission advantage is more heavily weighted towards immune evasion, which in turn suggests a lower generation time in order to make the whole picture work.

I’m updating in that direction based on the study, although the 17% inherent increase here (whether or not it’s ‘significant’) is still a substantial push on its own given we were already being pushed to the limit, I take the 17% seriously as far as it goes, and I agree with Hansen that in other contexts the edge could easily be far bigger.

Focus on vaccination still seems like the best approach. I agree that it’s less good at ‘stopping it from spreading’ but that’s no longer on the table. Western countries could not have contained Delta without vaccinations, so ‘the same NPIs that contained Delta’ are doubly not an option. We couldn’t do again what we did before even if it would work, and even if we did it again somehow anyway, they wouldn’t work.

This helps clarify what is happening and project things forward, but in the central case it changes things very little.

Other Omicron News

If you want the ‘oh no everything is terrible we are doomed’ perspective in relatively credible form, my attention was just drawn to Tatiana Prowell, who is good for an endless stream of terrible data points if that’s something you are interested in. Things definitely aren’t going to be fun in many places over January.

I didn’t find the ‘look for all the scariest possible data points’ data stream to be especially more or less scary than I expected it to be, so I didn’t update much, but it’s good to have it in one’s mix, and to know that some people are making predictions like this:

I don’t think that’s right, but if there are key supplies like prescriptions you need over January, yes, definitely get them now. Make sure you can weather some supply chain disruptions in case they happen, just in case. Better to play it safe, 10% is not 0%.

Omicron seems to have switched cell entry methods. Paper (pre-print).

Latest superspreader event case study, 21 of 33 boosted people infected. I’m primarily interested to look at generation times, since superspreader events are by definition abnormal and we’re not going to learn much new about infectiousness from outliers at this point.

If we assume that the exposure to SARS-CoV-2 was on the evening of the gathering, the incubation period was short, ranging from 2 to 6 days, with a mean incubation period of 3.24 days (95% CI 2.87- 3.60). Time to resolution of symptoms varied, and at the end of follow-up, five individuals still reported symptoms, while the rest reported symptoms lasting 1 to 9 days.

A ‘near miss’ with severe symptoms is disappointing, but in the most important sense this is still another 0 for 21.

We now know exactly what was up with the CDC nowcast. It was obvious nonsense. I hoped it was mostly harmless obvious nonsense, but perhaps not.

I’d also note that if you have the monoclonal antibodies, even if you think they’re only 27% to work, wouldn’t you use them anyway? I’m pretty sure they’d still pass the cost-benefit test, and next week they’re going to be even less likely to work.

Not Omicron so holding off on the full discussion but if you haven’t seen the thread discussing an attempt to publish a paper with the CDC, you should go ahead and read it now, in a ‘offered without comment’ kind of way, in the meantime.

Probability Updates

I was asked in a comment whether I am measuring my calibration. The answer is that I am not, and I consider that but a small part of prediction evaluation, but if someone else wants to do that (without me anchoring you on how to do that measurement) by all means go ahead.

Chance that Omicron has a 100% or bigger transmission advantage in practice versus Delta: 65% → 70%.

The new study says 161% in vaccinated people, 266% in the boosted, 17% in the unvaccinated. If you average that out, it’s higher than 100% in the populations we care about, but it’s somewhat close. Thus I’m creeping back up a bit.

Chance that Omicron is importantly (25%+ in the same person) less virulent than Delta: 80% → 85%.

Chance that Omicron is vastly (75%+ in the same person) less virulent than Delta: 20% → 20%.

Didn’t get much new evidence and thus not much additional clarity, but if there wasn’t at least the 25% level we’d expect to see trouble in ways we are not seeing, so we can get more confident there.

Chance we will be getting boosters modified for Omicron within 6 months of our previous booster shot: 25% → 20%.

Israel floated a fourth shot when it still mattered and pulled back. On reflection I see it as less likely that, in a world where Omicron is rapidly declining, we’ll tell people to get fourth shots this quickly, especially with severe disease protections holding.

Chance we are broadly looking at a future crisis situation with widely overwhelmed American hospitals, new large American lockdowns and things like that: 15% → 10%.

The math seems to indicate that the major city hospitals, under large vaccination rates with good capacity, will hold. If that’s true, even with the rural ones already in trouble now, I don’t see a way for this to be a crisis that makes this true, and everyone is moving towards a ‘life must go on’ position. I’m not ready to call this quite yet, panic can still be a thing, but Washington DC is being hit very hard and mostly not flinching. In another week we can likely mostly write this off, except in the scenarios where we suddenly can’t.

Generation time (serial interval) of Omicron is 3.5 days or less: 85% → 85%.

Didn’t learn much, the superspreader event wasn’t that helpful.


[Review] The Matrix Resurrections

31 декабря, 2021 - 14:10
Published on December 31, 2021 11:10 AM GMT

Do I recommend it?


Is it good?

I don't care.

The Waschowskis movies are always ambitions. The Wachowskis never phone it in. I get the sense they make every movie as if it's the last one they'll ever get to make. The Matrix Resurrections is imaginative. It does not hold back. I feel like Lana Wachowski had full creative control of The Matrix Resurrections (2021).

I heard the The Matrix Reloaded (2003) and The Matrix Revolutions (2003) were boring. Do I have to watch them first?

You should watch The Matrix (1999) before watching The Matrix Resurrections (2021).

It is not necessary to watch The Matrix Reloaded (2003) and The Matrix Revolutions (2003). You can figure out the important bits from context just by watching The Matrix Resurrections (2021). Another option is to just read the plot summaries of The Matrix Reloaded (2003) and The Matrix Revolutions (2003) on Wikipedia.

What about The Animatrix (2003)?

Animation is cheaper to produce than live-action. You should watch The Animatrix (2003) if you want to find out what would happen if the Wachowskis were totally unconstrained by budget.

I love The Matrix so much I can recite the whole thing from memory. How can I better appreciate such a cinematic masterpiece?

Check out this video about the music of The Matrix Trilogy.

Censorship Policy

Absolutely no spoilers. Not even if you use spoiler tags or you use ROT13. Not even "it's good" or "it's bad". Spoilers for The Matrix Resurrections (2021) are forbidden. Spoilers for anything else the Wachowskis have produced is allowed.


From Considerations to Probabilities

31 декабря, 2021 - 05:10
Published on December 31, 2021 2:10 AM GMT

The previous post started forecasting the UK hospital peak (based on information through Dec. 21, 2021). We generated several considerations and ultimately focused on the Omicron doubling time, the peak number of cases, the current number of Omicron cases, and the seasonality. In addition to a reference class forecast based on seasonality, we assumed the case peak would be roughly governed by hospital capacity and used the calculation:

DateOfPeak = Dec. 21
    + 10 days to reach case peak (2.4-day doubling time and 4.1 doublings)
    + 9 days (case peak to hospital peak)
    + 3 days (lag of 7-day average)
    = Jan. 12th

In this lecture we'll focus on going from this point estimate to a full probability distribution. This will involve two steps:

  1. Asking "what invalidating considerations could cause this forecast to be totally wrong"?
  2. Asking "which numerical quantities is my forecast most sensitive to, and how uncertain am I about them?"

The motivation for this is that most uncertainty is from either your entire estimate being structurally wrong (invalidating considertions), or from the specific numbers going into your estimate being inaccurate (numerical sensitivity). In many (most?) cases, the first form of uncertainty dominates, so it's good to check both.

We'll work through both steps, then combine them into a final uncertainty estimate. At the end I've also included a Q&A with Misha Yagudin on how this approach compares with his approach to forecasting.

Part 1: Invalidating Considerations

I did the brainstorming exercise of "If the previous estimate is totally off, why is that?" I recommend that you try this exercise as well before reading what I came up with.

(whitespace to avoid spoilers)




Okay, here's what I came up with:

  1. If the UK cases are capped by herd immunity rather than hospital strain (17+ million cases instead of 6.7 million)
  2. If the doubling time is actually 1.5 days (vs. 2.4 days), as suggested in some articles
  3. If the peak happens due to people self-adjusting their behavior to make $R$ barely less than $1$, leading to a very long "peak".

Let's see how much each of these could affect the answer.

Consideration 1: herd immunity. This would add at most 2 more doublings, or ~5 days, to the date of the peak.

Consideration 2: short doubling time. Since we assumed around 4 doublings before, this would subtract only ~4 days from the date of the peak.

Consideration 3: extended peak. We calculated before that hospital capacity would correspond to around 6 million confirmed cases/week. Herd immunity was around 17 million cases, so this would mean 3 weeks to reach herd immunity. But I now realize that this is confirmed cases, and undertesting is around a factor of 2. So I think this would only really add 1.5 weeks, or ~9 days, unless people adjust their behavior to stay significantly below hospital capacity. I'll add another 3 days of wiggle room (12 days total) in case the extended peak is at 75% of hospital capacity rather than 100% of capacity, or in case I underestimated the herd immunity threshold.

If I consider how subjectively surprised I would feel in each of the 3 worlds above, and turn that into probabilities, I get: 15% (herd immunity), 15% (short doubling time), 10% (extended peak).

Exercise. Do you agree with the above probabilities?

Brainstorming exercise. What other considerations am I missing?

Part 2: Numerical Sensitivity

Next I checked the numerical sensitivity of the mainline forecast. Our mainline forecast is based on several quantities:

  • The current number of UK Omicron cases, estimated at $N_0 = 200,000$
  • The total number of future Omicron cases, estimated at $N = 6,700,000$.
  • The Omicron doubling time, estimated at $t = 2.4$ days
  • The lag $\Delta_0$ between case peak and hospital peak, estimated at 9 days.
  • The lag $\Delta_1$ between single-day hospital peak and 7-day average hospital peak, estimated at 3 days.

Our formula for the number of days until the peak is then

$\log_2(N/2N_0) \cdot t + \Delta_0 + \Delta_1$

Let's assess the sensitivity of this formula to each consideration:

  • If $N$ or $N_0$ is off by a factor of $2$, then our answer changes by $2.4$ days.
  • If $t$ is $3.3$ instead of $2.4$, our answer changes by $3.7$ days.
  • If $\Delta_0$ or $\Delta_1$ is off by $1$, our answer changes by $1$ day.

To make this more quantitative I put it into table form, including my $70\%$ uncertainty intervals for each number:

Parameter Point estimate Range Effect on answer $N_0$ $0.2 \times 10^6$ $[0.15, 0.25] \times 10^6$ $[-0.8, +1.0]$ $N$ $6.7 \times 10^6$ $[5, 13] \times 10^6$ $[-1.0, +2.3]$ $t$ $2.4$ $[2.0, 3.3]$ $[-1.6, +3.7]$ $\Delta_0 + \Delta_1$ $12$ $[9, 14]$ $[-3, +2]$

Considering that probably not all errors will occur in the same direction, when I combine these errors together I subjectively end up with a 70% confidence interval of $[-3.6, +4.9]$ relative to the Jan. 12th point estimate. (I estimated these as e.g. $3.6 = \sqrt{0.8^2 + 1.0^2 + 1.6^2 + 9^2}$ based on the premise that variances add for independent quantities. I don't think this is a logically valid calculation but it gives a decent ballpark, and the final numbers also seemed reasonable to me.)

Misha: I generally got a sense that your ranges are a bit too narrow, e.g., for doubling time. Metaculus is super uncertain about R_0 (their 70% CI 5.2 to 11.9), and “average” doubling guesstimates should probably be pretty uncertain given conflicting info, the impact of the holidays, impact of public concern, and government action. [Followed by some additional comments on why $N_0$ and $N$ should have higher uncertainty.]

I asked Misha if he also thought my final uncertainty estimates (given in the next section) were too small. He said:

Misha: Nope, I think they are fine (because the additional 45% went to extreme outcomes).

Putting it Together

If we assume the mainline estimate is structurally correct and all errors are due to numerical sensitivity, then we end up with a 70% confidence interval of (rounded to whole numbers) Jan. 8th to Jan. 17th. That means there is a 15% chance of being earlier than Jan. 8th and of being later than Jan. 17th.

If we instead consider structural uncertainty, we get a 15% chance of +5 days (Jan. 17th), a 15% chance of -4 days (Jan. 8th), and a 10% chance of +12 days (Jan. 24th).

In reality, both forms of uncertainty are present. Overall, the uncertainty also skews a bit more towards later dates than earlier dates. If I subjectively combine this, I would put my overall forecast as follows:

  • Median of Jan. 13th
  • 10% chance of Jan. 24th or later
  • 25% chance of Jan. 18th or later
  • 25% chance of Jan. 7th or earlier

Exercise. Do you agree with this assessment?

Concluding Q&A

Since this is the first lecture that presents a fully integrated forecasting method, I asked Misha how close it matches his approach to forecasting.

Jacob: How closely does the method discussed in this and the previous lecture map onto your own approach to forecasting? I.e.: generate and prioritize considerations, reduce uncertainty, construct a mainline estimate (or multiple mainline estimates), consider numerical sensitivity + structural uncertainty.

Misha: Well, I do all of the things from time to time. I do not do this explicitly in a step-by-step way. It’s more of playing it by ear and attending to whatever feels most informative.

The core step, which is missing from your writeups, is getting less confused about what’s going on and assembling a world model. I usually start pretty cluelessly; for example, I was forecasting cultured meat progress last month. I spend a lot of time trying to understand how the processes might work, how to reference class might look like, and what technological limitations are.

Until I had some understanding (still limited), I wasn’t looking for considerations. But after building a world model, I developed ways to approach most questions (sometimes very structurally uncertain).

To me, the key insight in your writeup was to look at beds/herd immunity and doubling. Everything else seems more like technical details necessary for delivering a good forecast but not primarily to the process.

Jacob: Would you still consider [this lecture] good pedagogy for students?

Misha: I think it is a textbook example, looks good to me. I think it’s put a bit too much weight on legible steps and a bit less on "actual creative work." To be clear, these legible technical steps are important and worth having in front of you.

My takeaway is that the approach above is useful and valuable, but that it is also important to build a good world-model (especially when confronting a new domain). We'll hopefully have more to say about that in upcoming lectures.


Is Omicron less severe?

31 декабря, 2021 - 02:14
Published on December 30, 2021 11:14 PM GMT

Do we know this yet? It seems awfully unlikely that we couldn't have a pretty solid answer to this question by now, given how many Omicron cases there have been by now.


The Genetics of Space Amazons

31 декабря, 2021 - 01:14
Published on December 30, 2021 10:14 PM GMT

If you wanted to colonize other planets with meat suite humans, then females are the superior choice as they have lower nutritional needs. Here I imagine a fictional Amazon gene that could lead to a skewed sex ratio favoring females.

  • Human Females need less food than males
  • Human Females silence a random X chromosome because otherwise every protein from the X chromosome would be twice as abundant for XX females compared to as for XY males
  • If Humans could be engineered to have a female biased sex ratio, then they would have an advantage when traveling to new worlds because:
    • then they will need less food.
    • their space ships would require slightly less fuel as females tend to be lighter.
    • the could increase their population size faster as there are more gestating members of the species.

So this is a cool premise for an egalitarian feminist spacefaring society... or a harem anime :)


Imagine a novel X chromosome mutation that allows the X chromosome to silence the Y chromosome such that these XY mutants have a female phenotype and reproduction strategy, to make this less confusing we denote the mutated X chromosome as A, the Amazon chromosome

In this new society there are the following species with a male phenotype: XY, and the following with female phenotype XX, AX and AY (AA would also be female, but cannot exist as the males have no A's)

Try to stop up and guess the sex equilibrium ratio... Is it 1/4 of each one because there are now 4 sexes, is it 1/3 XY and 2/3 female genotypes?, or something completely different?

Amazon mating: The Next Generation

Let's try to "mate" these new sex genotypes and see what would be produced

XY and XX (the standard): XX has 100% chance of giving an X and XY has 50/50 for each, so the offspring will be 50% XY and 50% XX

XY + XX -> 50% XX and 50% XY

Let's try to mate XY with AX:

XY + AX ->25% XY, 25% AX, 25% XY and 25% AY

Finally XY with AY, if we assume that egg cells carrying a Y are nonviable, then AY amazons will always donate their A giving:

XY + AY -> 50% AX + 50X AY

We can put this in a table where the columns are the mating pairs and the rows are offspring probabilities. (For mathematical reasons explained later we add a Male on Male columns with has 0% chance of getting offspring)

XY + XY XX + XY AX + XY AY + XY XY 0 0.5 0.25 0 XX 0 0.5 0.25 0 AX 0 0 0.25 0.5 AY 0 0 0.25 0.5

Looking at this table, do you now have a new guess for the equilibrium sex ratio?

Lets say the space ship starts with 2 XY males and 8 AX Amazons, let's calculate the expected number for each of the 4 genotypes after 1 generation (assuming each female mates twice). We can express the 10 astronauts as a vector where items of the vector corresponds to the index in the table above, so [nXY,nXX,nAX,nAY]T.mjx-chtml {display: inline-block; line-height: 0; text-indent: 0; text-align: left; text-transform: none; font-style: normal; font-weight: normal; font-size: 100%; font-size-adjust: none; letter-spacing: normal; word-wrap: normal; word-spacing: normal; white-space: nowrap; float: none; direction: ltr; max-width: none; max-height: none; min-width: 0; min-height: 0; border: 0; margin: 0; padding: 1px 0} .MJXc-display {display: block; text-align: center; margin: 1em 0; padding: 0} .mjx-chtml[tabindex]:focus, body :focus .mjx-chtml[tabindex] {display: inline-table} .mjx-full-width {text-align: center; display: table-cell!important; width: 10000em} .mjx-math {display: inline-block; border-collapse: separate; border-spacing: 0} .mjx-math * {display: inline-block; -webkit-box-sizing: content-box!important; -moz-box-sizing: content-box!important; box-sizing: content-box!important; text-align: left} .mjx-numerator {display: block; text-align: center} .mjx-denominator {display: block; text-align: center} .MJXc-stacked {height: 0; position: relative} .MJXc-stacked > * {position: absolute} .MJXc-bevelled > * {display: inline-block} .mjx-stack {display: inline-block} .mjx-op {display: block} .mjx-under {display: table-cell} .mjx-over {display: block} .mjx-over > * {padding-left: 0px!important; padding-right: 0px!important} .mjx-under > * {padding-left: 0px!important; padding-right: 0px!important} .mjx-stack > .mjx-sup {display: block} .mjx-stack > .mjx-sub {display: block} .mjx-prestack > .mjx-presup {display: block} .mjx-prestack > .mjx-presub {display: block} .mjx-delim-h > .mjx-char {display: inline-block} .mjx-surd {vertical-align: top} .mjx-surd + .mjx-box {display: inline-flex} .mjx-mphantom * {visibility: hidden} .mjx-merror {background-color: #FFFF88; color: #CC0000; border: 1px solid #CC0000; padding: 2px 3px; font-style: normal; font-size: 90%} .mjx-annotation-xml {line-height: normal} .mjx-menclose > svg {fill: none; stroke: currentColor; overflow: visible} .mjx-mtr {display: table-row} .mjx-mlabeledtr {display: table-row} .mjx-mtd {display: table-cell; text-align: center} .mjx-label {display: table-row} .mjx-box {display: inline-block} .mjx-block {display: block} .mjx-span {display: inline} .mjx-char {display: block; white-space: pre} .mjx-itable {display: inline-table; width: auto} .mjx-row {display: table-row} .mjx-cell {display: table-cell} .mjx-table {display: table; width: 100%} .mjx-line {display: block; height: 0} .mjx-strut {width: 0; padding-top: 1em} .mjx-vsize {width: 0} .MJXc-space1 {margin-left: .167em} .MJXc-space2 {margin-left: .222em} .MJXc-space3 {margin-left: .278em} .mjx-test.mjx-test-display {display: table!important} .mjx-test.mjx-test-inline {display: inline!important; margin-right: -1px} .mjx-test.mjx-test-default {display: block!important; clear: both} .mjx-ex-box {display: inline-block!important; position: absolute; overflow: hidden; min-height: 0; max-height: none; padding: 0; border: 0; margin: 0; width: 1px; height: 60ex} .mjx-test-inline .mjx-left-box {display: inline-block; width: 0; float: left} .mjx-test-inline .mjx-right-box {display: inline-block; width: 0; float: right} .mjx-test-display .mjx-right-box {display: table-cell!important; width: 10000em!important; min-width: 0; max-width: none; padding: 0; border: 0; margin: 0} .MJXc-TeX-unknown-R {font-family: monospace; font-style: normal; font-weight: normal} .MJXc-TeX-unknown-I {font-family: monospace; font-style: italic; font-weight: normal} .MJXc-TeX-unknown-B {font-family: monospace; font-style: normal; font-weight: bold} .MJXc-TeX-unknown-BI {font-family: monospace; font-style: italic; font-weight: bold} .MJXc-TeX-ams-R {font-family: MJXc-TeX-ams-R,MJXc-TeX-ams-Rw} .MJXc-TeX-cal-B {font-family: MJXc-TeX-cal-B,MJXc-TeX-cal-Bx,MJXc-TeX-cal-Bw} .MJXc-TeX-frak-R {font-family: MJXc-TeX-frak-R,MJXc-TeX-frak-Rw} .MJXc-TeX-frak-B {font-family: MJXc-TeX-frak-B,MJXc-TeX-frak-Bx,MJXc-TeX-frak-Bw} .MJXc-TeX-math-BI {font-family: MJXc-TeX-math-BI,MJXc-TeX-math-BIx,MJXc-TeX-math-BIw} .MJXc-TeX-sans-R {font-family: MJXc-TeX-sans-R,MJXc-TeX-sans-Rw} .MJXc-TeX-sans-B {font-family: MJXc-TeX-sans-B,MJXc-TeX-sans-Bx,MJXc-TeX-sans-Bw} .MJXc-TeX-sans-I {font-family: MJXc-TeX-sans-I,MJXc-TeX-sans-Ix,MJXc-TeX-sans-Iw} .MJXc-TeX-script-R {font-family: MJXc-TeX-script-R,MJXc-TeX-script-Rw} .MJXc-TeX-type-R {font-family: MJXc-TeX-type-R,MJXc-TeX-type-Rw} .MJXc-TeX-cal-R {font-family: MJXc-TeX-cal-R,MJXc-TeX-cal-Rw} .MJXc-TeX-main-B {font-family: MJXc-TeX-main-B,MJXc-TeX-main-Bx,MJXc-TeX-main-Bw} .MJXc-TeX-main-I {font-family: MJXc-TeX-main-I,MJXc-TeX-main-Ix,MJXc-TeX-main-Iw} .MJXc-TeX-main-R {font-family: MJXc-TeX-main-R,MJXc-TeX-main-Rw} .MJXc-TeX-math-I {font-family: MJXc-TeX-math-I,MJXc-TeX-math-Ix,MJXc-TeX-math-Iw} .MJXc-TeX-size1-R {font-family: MJXc-TeX-size1-R,MJXc-TeX-size1-Rw} .MJXc-TeX-size2-R {font-family: MJXc-TeX-size2-R,MJXc-TeX-size2-Rw} .MJXc-TeX-size3-R {font-family: MJXc-TeX-size3-R,MJXc-TeX-size3-Rw} .MJXc-TeX-size4-R {font-family: MJXc-TeX-size4-R,MJXc-TeX-size4-Rw} .MJXc-TeX-vec-R {font-family: MJXc-TeX-vec-R,MJXc-TeX-vec-Rw} .MJXc-TeX-vec-B {font-family: MJXc-TeX-vec-B,MJXc-TeX-vec-Bx,MJXc-TeX-vec-Bw} @font-face {font-family: MJXc-TeX-ams-R; src: local('MathJax_AMS'), local('MathJax_AMS-Regular')} @font-face {font-family: MJXc-TeX-ams-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_AMS-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_AMS-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_AMS-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-cal-B; src: local('MathJax_Caligraphic Bold'), local('MathJax_Caligraphic-Bold')} @font-face {font-family: MJXc-TeX-cal-Bx; src: local('MathJax_Caligraphic'); font-weight: bold} @font-face {font-family: MJXc-TeX-cal-Bw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Caligraphic-Bold.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Caligraphic-Bold.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Caligraphic-Bold.otf') format('opentype')} @font-face {font-family: MJXc-TeX-frak-R; src: local('MathJax_Fraktur'), local('MathJax_Fraktur-Regular')} @font-face {font-family: MJXc-TeX-frak-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Fraktur-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Fraktur-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Fraktur-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-frak-B; src: local('MathJax_Fraktur Bold'), local('MathJax_Fraktur-Bold')} @font-face {font-family: MJXc-TeX-frak-Bx; src: local('MathJax_Fraktur'); font-weight: bold} @font-face {font-family: MJXc-TeX-frak-Bw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Fraktur-Bold.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Fraktur-Bold.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Fraktur-Bold.otf') format('opentype')} @font-face {font-family: MJXc-TeX-math-BI; src: local('MathJax_Math BoldItalic'), local('MathJax_Math-BoldItalic')} @font-face {font-family: MJXc-TeX-math-BIx; src: local('MathJax_Math'); font-weight: bold; font-style: italic} @font-face {font-family: MJXc-TeX-math-BIw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Math-BoldItalic.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Math-BoldItalic.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Math-BoldItalic.otf') format('opentype')} @font-face {font-family: MJXc-TeX-sans-R; src: local('MathJax_SansSerif'), local('MathJax_SansSerif-Regular')} @font-face {font-family: MJXc-TeX-sans-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_SansSerif-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_SansSerif-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_SansSerif-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-sans-B; src: local('MathJax_SansSerif Bold'), local('MathJax_SansSerif-Bold')} @font-face {font-family: MJXc-TeX-sans-Bx; src: local('MathJax_SansSerif'); font-weight: bold} @font-face {font-family: MJXc-TeX-sans-Bw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_SansSerif-Bold.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_SansSerif-Bold.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_SansSerif-Bold.otf') format('opentype')} @font-face {font-family: MJXc-TeX-sans-I; src: local('MathJax_SansSerif Italic'), local('MathJax_SansSerif-Italic')} @font-face {font-family: MJXc-TeX-sans-Ix; src: local('MathJax_SansSerif'); font-style: italic} @font-face {font-family: MJXc-TeX-sans-Iw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_SansSerif-Italic.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_SansSerif-Italic.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_SansSerif-Italic.otf') format('opentype')} @font-face {font-family: MJXc-TeX-script-R; src: local('MathJax_Script'), local('MathJax_Script-Regular')} @font-face {font-family: MJXc-TeX-script-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Script-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Script-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Script-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-type-R; src: local('MathJax_Typewriter'), local('MathJax_Typewriter-Regular')} @font-face {font-family: MJXc-TeX-type-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Typewriter-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Typewriter-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Typewriter-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-cal-R; src: local('MathJax_Caligraphic'), local('MathJax_Caligraphic-Regular')} @font-face {font-family: MJXc-TeX-cal-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Caligraphic-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Caligraphic-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Caligraphic-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-main-B; src: local('MathJax_Main Bold'), local('MathJax_Main-Bold')} @font-face {font-family: MJXc-TeX-main-Bx; src: local('MathJax_Main'); font-weight: bold} @font-face {font-family: MJXc-TeX-main-Bw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Main-Bold.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Main-Bold.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Main-Bold.otf') format('opentype')} @font-face {font-family: MJXc-TeX-main-I; src: local('MathJax_Main Italic'), local('MathJax_Main-Italic')} @font-face {font-family: MJXc-TeX-main-Ix; src: local('MathJax_Main'); font-style: italic} @font-face {font-family: MJXc-TeX-main-Iw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Main-Italic.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Main-Italic.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Main-Italic.otf') format('opentype')} @font-face {font-family: MJXc-TeX-main-R; src: local('MathJax_Main'), local('MathJax_Main-Regular')} @font-face {font-family: MJXc-TeX-main-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Main-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Main-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Main-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-math-I; src: local('MathJax_Math Italic'), local('MathJax_Math-Italic')} @font-face {font-family: MJXc-TeX-math-Ix; src: local('MathJax_Math'); font-style: italic} @font-face {font-family: MJXc-TeX-math-Iw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Math-Italic.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Math-Italic.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Math-Italic.otf') format('opentype')} @font-face {font-family: MJXc-TeX-size1-R; src: local('MathJax_Size1'), local('MathJax_Size1-Regular')} @font-face {font-family: MJXc-TeX-size1-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Size1-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Size1-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Size1-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-size2-R; src: local('MathJax_Size2'), local('MathJax_Size2-Regular')} @font-face {font-family: MJXc-TeX-size2-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Size2-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Size2-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Size2-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-size3-R; src: local('MathJax_Size3'), local('MathJax_Size3-Regular')} @font-face {font-family: MJXc-TeX-size3-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Size3-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Size3-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Size3-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-size4-R; src: local('MathJax_Size4'), local('MathJax_Size4-Regular')} @font-face {font-family: MJXc-TeX-size4-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Size4-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Size4-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Size4-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-vec-R; src: local('MathJax_Vector'), local('MathJax_Vector-Regular')} @font-face {font-family: MJXc-TeX-vec-Rw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Vector-Regular.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Vector-Regular.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Vector-Regular.otf') format('opentype')} @font-face {font-family: MJXc-TeX-vec-B; src: local('MathJax_Vector Bold'), local('MathJax_Vector-Bold')} @font-face {font-family: MJXc-TeX-vec-Bx; src: local('MathJax_Vector'); font-weight: bold} @font-face {font-family: MJXc-TeX-vec-Bw; src /*1*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/eot/MathJax_Vector-Bold.eot'); src /*2*/: url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/woff/MathJax_Vector-Bold.woff') format('woff'), url('https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/fonts/HTML-CSS/TeX/otf/MathJax_Vector-Bold.otf') format('opentype')} like this [2,0,8,0]T

If we also consider the first row of the table as a vector, then we can multiply the two vectors as follows:


This is called the vector dot product, if we "keep sliding" the vector down the table, then it is called a matrix dot product, Thus to calculate the next generation genotype vector we simply need to slide [2,0,8,0]T trough the table (and multiply by the number of times the females mates, which is 2)

2×⎡⎢ ⎢ ⎢⎣00.50.25000.50.250000.250.5000.250.5⎤⎥ ⎥ ⎥⎦⎡⎢ ⎢ ⎢⎣2080⎤⎥ ⎥ ⎥⎦=⎡⎢ ⎢ ⎢⎣4444⎤⎥ ⎥ ⎥⎦

So after 1 generation there are now 16 people and a sex ratio of 1:3, and a 1:1:1:1 genotype distribution. One generation later we have:

2×⎡⎢ ⎢ ⎢⎣00.50.25000.50.250000.250.5000.250.5⎤⎥ ⎥ ⎥⎦⎡⎢ ⎢ ⎢⎣4444⎤⎥ ⎥ ⎥⎦=⎡⎢ ⎢ ⎢⎣6666⎤⎥ ⎥ ⎥⎦

Interesting The same distribution, thus if the Amazons decide to get only 43 kids per female they will have a stable population:

43×⎡⎢ ⎢ ⎢⎣00.50.25000.50.250000.250.5000.250.5⎤⎥ ⎥ ⎥⎦⎡⎢ ⎢ ⎢⎣6666⎤⎥ ⎥ ⎥⎦=⎡⎢ ⎢ ⎢⎣6666⎤⎥ ⎥ ⎥⎦

If you are a scifi author and want to write an Amazons in space story, then you can use this world building premise without needing to cite me :D

Optional: Other XY + AY offspring patterns

As mentioned above the rules for XY + AY are not clear. It all depends on how a female Sex cell handles lacking a X chromosome. Is the X chromosome vital for the viability of the cell? or is it only vital for cell proliferation? which happens after merging with a sperm cell. Under all circumstances YY are not viable as the X chromosome contains a lot of important genes.

Let's game out all possible combinations

If egg cells containing a Y chromosome are viable from AY females then two other mating patterns are possible depending on what happens if a Y sperm tries to fuse with a Y egg cells.

  1. Y eggs 'reject' Y sperm Then:

    • AY donates A -> 50% XY + 50% AX
    • AY donates Y -> 100% XY
    • Thus 75% XY + 25% AX
  2. Y eggs 'accept' Y sperm, but the cell fails to divide

    • AY donates A -> 50% XY + 50% AX
    • AY donates Y -> 50% XY + 50% egg fails to divide
    • Thus 2/3 XY + 1/3 AX


Personal Response to Omicron

30 декабря, 2021 - 23:40
Published on December 30, 2021 8:40 PM GMT

Throughout the pandemic, our household has been on the careful end. We started preparing in February 2020, isolated thoroughly from March to July, moderately over the summer, and then thoroughly again from October until all the adults in our house were vaccinated. Because I was immunocompromised, we had a newborn, and kids under five couldn't get vaccinated yet, however, our family's 2021 was still a lot like 2020. Then, in doing Christmas holidays with family I was very careful individually in the two weeks leading up to it and I was the one working on precautions and making sure we had enough tests: some of the people I really wanted to spend Christmas with are at elevated risk.

I wrote last week about how there are a range of paths we could take with Omicron as a society, but now that Christmas is over what am I going to do personally?

Well, what's the weather out there?

That's a lot of covid in the sewer, 5x the previous peak and still going up.

I think if it were critical that I or a housemate didn't get Omicron, it would be possible, but very difficult. We would need to go back to isolating the way we were in Fall 2020, treating vaccinated people as about as risky as unvaccinated people, pulling the kids out of school, and never going indoors anywhere. Given the likely effects of contracting covid as a child or boosted adult, this is not worth it for us or I suspect most people.

Instead, I expect that I, the people in my house, and pretty much everyone who doesn't take intense and careful effort to avoid it will catch Omicron at some point in the next ~month. It's not a good thing, but it is what it is. Afterward, people's immune systems will have had yet another covid exposure, and I expect cases to go low until next fall. So I'm not going to stress about it: I'll follow official guidance and mask regulations, cheerfully go along with precautions others need, and test+isolate when sick, but I'm not going to go above and beyond to attempt to reduce spread the way I did for earlier parts of the pandemic.

I'm thinking of Omicron as the first wave of the endemic phase of covid-19. While "it's just the flu" was completely wrong in 2020, in 2022 the situation has changed enough of that "what would I do in flu season" is now a good guide.

Comment via: facebook