Since the 1930s seed companies keep inbred lines of corn, not sold to farmers (e.g. AAbb & aaBB), that when crossed make genetically identical F1 hybrids, which is what is sold to farmers (e.g. all AaBb).
The F1 hybrids have "hybrid vigor" like higher yield etc., but if you save and replant it then the F2 hybrids are genetically variable (e.g. some Aabb, aabb, etc.) and aren't reliably vigorous—hybrid vigor doesn't "breed true". Farmers want reliable vigor, and buy new F1 seed every year, and seed companies profit from repeat business.
There are two main theories for how hybrid vigor works. Under the "dominance" theory, good dominant genes from each parent cover up bad recessive genes from the other (e.g. A dominant over a, B dominant over b). Under the "overdominance" theory, hybrids are vigorous because they have heterozygote advantage at many loci (e.g. Aa fitter than either AA or aa, Bb fitter than either BB or bb).
Under the first theory, you could also get the same vigor just by a selection program, but that *would* breed true—if seed companies haven't done so, it's not because of scientific obstacles, it's strictly business. Under the second theory, you couldn't select for reliable true-breeding vigor anyway. Of course seed companies prefer the second theory (I'm not sure if scientists are 100% on which is true).
Nowadays of course seed companies also try to patent their plants, or make GMO seeds that can't reproduce, etc., anything to keep the farmers reliant on buying new seed. But hybrid corn was the start.
"As of today (November 23, 2020), there are 15,039 positive cases of COVID-19 and 631 total confirmed deaths on the Navajo Reservation. This marks a spike of over 2,200 cases in 12 days on Navajo. Numbers are also on the rise on the Hopi Reservation, and both tribal governments have placed their communities in lockdown again. Our relief effort is the primary relief effort providing assistance on both reservations. We have already deployed roughly $6 million in direct relief in both communities, which has us nearing depletion of existing GoFundMe donations. We need a second wave of donations to help us meet the second wave--tsunami, really--of COVID that is inundating our communities.
To date, the Relief Fund has raised over $6 million, and has spent almost all of that on bringing vital resources to Navajo and Hopi communities to help shield them from COVID. The team and a legion of hundreds of Navajo and Hopi volunteers all over the two nations has provided food and water to over 46,000 households (each roughly averaging 4 persons per household, and amounting to over 186,000 people served—which is more than the combined population of the Navajo and Hopi Reservations) in 335 distributions in over 530 Navajo communities and Hopi Villages.
The Relief Fund currently seeks to raise enough funding to carry Navajo and Hopi families through this second wave, through the cold and flu season, and to the end of the COVID pandemic in these communities. We estimate needing $6.5 million in order to meet that goal, so we have now adjusted our GoFundMe campaign goal to $13 million total.
By donating, you will help the most heavily impacted community in the USA. For $100, you will help a family of four by providing them with 2 weeks’ worth of food and Personal Protective Equipment (PPE). For as little as $10, you can help feed a family of 4 for an entire day. Donate today to the Navajo & Hopi Families COVID-19 Relief Fund."
I can't tell yet how much the Roe V Wade leak is affecting people in general.
For my part, because I live in Texas, a purple state, I plan on donating a lot of time and money to the Democratic party to help them win in November, even though it might be a longshot, because of the apparent overturn of Roe.
I'm still curious how seriously people take the leak, though. Given the leak, what odds do you give Roe V Wade being overturned this semester? Might the leak be misleading?
I'm hoping I can help change Texas politics and put it back in a more Libertarian direction, instead of the Trumpist direction it has gone recently. I'd join the Republican party if I thought I could make them more moderate, because all I want is a moderate outcome. But my experience with the parties in Texas is that the Republicans are organized and wouldn't like me, whereas the Texas Democrats are entirely disorganized and clueless and maybe my organizational skills could help them.
Not saying we can win, but I'm ready to fight. I encourage others in Texas who believe in a woman's right to choose to join the fight.
What is the best source for getting a good understanding of what Trump was like as a businessman? Is there some biography out there that gives a good overview of how he worked? I thought of getting Art of the Deal but it's unclear to me how much of it would just be self-aggrandizement.
For people who are into genealogy and family trees, what is your preferred format, or alternatively what website do you think has the best format? Is Ancestry food for instance or are their better ways to display family trees?
A while ago, our project won an ACX grant to try to pursue a potentially far-reaching idea of spellchecking genomes with genetic editing. We are looking for a genetic editing expert to help in an advisory capacity, or potentially even join the core team.
Hey so I was wondering what people here think about the following: Is it "normal" to have different political leanings based on one's location?
I will explain my situation: I have lived in three countries (Germany, Canada and the US). In Germany, I tend to favour more right-leaning parties like the FDP, CDU/CSU and some others. Sometimes I am even sympathetic to the AFD (which is considered far-right). On most political issues in Germany, I tend to take the more right-leaning view (e.g. Nuclear power, "Energiewende", Refugees/Immigration etc.). In the US and Canada I consistently support the left-leaning parties and view on (almost) all issues.
I tend to explain (or justify) this by arguing from an utilitarian perspective: it makes more sense for Germany (and other Western European countries) to be more conservative, since the population density is so high and the geopolitcal neigbourhood (e.g. being close to the Middle East, Russia/Ukraine and Africa) makes it more prudent from an utilitarian perspective to be more cautious, ergo conservative. Thus I think that Western European countries should support policies of gradual improvement instead of being revolutionary (e.g. in terms of energy, developing Nuclear instead of renewables, and not letting in too many immigrants so as not make the social systems collapse).
For the US and Canada, with their many resources and low population density, it makes more sense to be more left-wing: letting in many immigrants, being socially liberal, and building lots of infrastructure (I guess having low population density means having a higher carbon footprint, which is "conservative". I do support more environmentalism in the US and Canada, especially since they have so much land space they could support more renewable energy like solar, wind and hydro). From a utilitarian global perspective I think it also makes more sense to support left-wing policies for the US and Canada (particularly in regards to immigration - because the US and Canada have so much space, they should take in much more immigrants from Africa and the Middle East than Europe). For Western Europe I don't think it's clear that from a global utilitarian perspective it's better to have more left-wing governments (e.g. the refugee crisis destabilised the EU and led to populism in many countries, and the EU doesn't have enough area to develop renewable energy to the extent the US and Canada can.)
So anyway what I am asking/wondering is if my argumentation makes sense from an utilitarian perspective, or if I am being an (European) egoist, since I will admit that generally being conservative means being egoistical (I guess you could disagree, but I think that statement makes sense from and utilitarian perspective).?
On AI alignment: recursive self-improvement seems like a prerequisite for the type of AGI that would become an existential threat. To what extent have alignment researchers looked into ways to prevent AGI from trying to self-improve? Has there been research into e.g. utility functions that disallow self-improvement?
Has anyone really looked into the literature on the relationship between mathematical ability and musical aptitude? Bearing in mind how consistently this correlation occurs, and the fact that ridiculously large fraction of the people I know in math heavy majors, were either in a band or compose their own music, I'm confident the correlation is real.
A lot of the literature seem to (incorrectly, in my opinion) infer that learning how to play music on an instrument directly improves mathematical ability, but can this really be right? I have four competing hypotheses and wondered if anyone could verify which, if any, are most likely to be correct:
1. Learning to play music somehow ,directly or indirectly, improves mathematical ability.
2.Kids who get involved in music are more likely to have smart, educated, middle-class parents, and by some combination of nature and nurture have an advantage in math.
3. People who are better at math often have superior general intelligence, thus making them more likely to be able to excel in multiple domains, like music and math. While people with intelligence closer to avg general intelligence could not maintain both, so they must choose one.
4. (And for my wackiest theory)It's not really intelligence alone driving the correlation, it is the combination of intelligence and autism-with the latter being arguably more important. It is generally accepted that STEM majors are, on average, more autistic than the general population. So my theory goes that the need for order and pattern recognition abilities that are often characteristic of autism would lend itself well musical pursuits.
Hi, on a previous thread (or maybe on the subreddit) someone had mentioned an app for setting goals and reflecting on them. It had a paid version where you could talk to a human and a name that dropped the first letter of a common word (think "concile" from "reconcile" but clever). Can anyone remember what it was?
Why does Twitter have such high operating expenses? What do they need all those programmers for? They've made only a couple of changes to the site (all for the worse) in 15 years.
What are the programmers spending all their time doing?
What do people's office climates look like with the lifting pandemic restrictions?
My office used to encourage remote work (both explicitly, as in "if you can work from home, please work from home," and also implicitly through mask requirements and closing the dining room). It recently switched to encouraging a return to the office (in particular, masks are optional, dining room is open again) -- though many people have concluded that actually working from home is great, and are still doing a lot of it (there's no strong pressure to show up at the office unless you need to work with a physical object there).
One development I was expecting to see was "if you're feeling under the weather, please wear a mask," but the official policy is instead "if you're feeling under the weather, please stay home". This isn't terrible as policies go, but it means that "masks optional" actually means "masks off" -- there's just no room for "I'm feeling sick enough to wear a mask, but not sick enough to stay home." I'm not yet decided how I feel about this, does anyone have a coherent opinion?
The 1950s, 60s, 70s, 80s and 90s had very distinctive styles in terms of fashion, clothing, graphic design, music, food, pretty much everything.
Then sometime around 1999, we got stuck. There was no distinctive 2000s style that looks different to a 2010s style, and the 2020s aren't looking great either unless you count pandemic-induced slobbiness as a fashion trend. Basically, 2022 looks like 2002 except with smartphones.
What gives? Possible explanations:
1. Something about the internet soaking up all the innovation leaving meatspace stuck in an eternal 1999
2. The 1950s-1990s period was a weird abberation, a manifestation of the "generation gap" between the boomers and their parents.
3. This is all an illusion and actually 2022 looks as unlike 2012 as 1992 looked unlike 1982, and I'm just not sensitive enough to notice it.
Question about Nato expansion, "not one inch to the East" etc.
I've seen people on Russia's side of this argument (eg Jeff Sachs) state that promises were broken, and I've seen the West's response which I've tried to describe below, but I've never found someone engage with or try refute that response.
What am I getting wrong/missing?
What happened -
1. During discussions re reunification of Germany, Russia were given verbal assurances that Nato would not expand eastwards within reunified Germany. The phrase "not one inch to the east" was used.
2. But in the end the actual treaty agreed allowed for some troops in what was East Germany.
3. No assurances were ever discussed or given about other former Soviet states joining Nato.
4. Russia itself discussed joining Nato and were told they would need to apply like everyone else and so didn't (my guess is they wouldn't have been admitted if they did apply)
West's view -
A. if it's not in the treaty it doesn't count, democracies change governments all the time, new governments can't be bound by verbal assurances of their predecessors
B. This is especially so if the treaty that was signed already wasn't in line with those verbal assurances
C. No assurances re former soviet states were given.
D. The idea that the states that were actually part of the USSR and then became sovereign not being allowed to join Nato is strange - it's essentially like saying, we promise not to let you join our club even if you want to?
Is there a way to view the Substack archive in the order the posts were published (or just any way remotely more convenient than the standard option)? I am just about caught up on SSC, and scrolling through the whole ACT archive to start from the beginning seems a little ineffective.
My view of politics in the United States is that the country is fairly evenly divided into two groups: Republicans and Democrats. Each group considers the other group to be a threat to the nation, and their most important political goal is to stop the other group from succeeding.
Now in my mind this is a recipe for failure of the United States. I believe the two-party system is the true threat to the nation.
I have been thinking about this a lot and devising a possible solution. The sad irony is that given our political stalemate, the solution is impossible to implement. But if you are willing to imagine the best possible electoral system for the United States (not just the puny changes currently possible), please read this proposal and share your thoughts with me.
“ If he wants to become an electrician, that's wonderful! The only problem is... I have no idea how to help him do that! As a university-attender myself, I know advice and procedures on how to do that. And while it was kind of expected of me, I have my own questions about the value of it both on its own, and compared to the costs. But when he asks me how to become an electrician, well, I just don't know!”
Since at DSL I am banned at my own request till June I can’t respond there soon, I suppose I’ll and I ask others to pass on the info to DSL, thanks if you do:
@no one special,
The aptitude tests to be an apprentice electrician in most counties locals is similar to the test to be an apprentice plumber but typically the IBEW has additional tests for English language skills while the UA stresses arithmetic more, regardless being able to do multiple choice arithmetic problems very quickly will be the core of the tests, your son should throughly practice that, there will also be tests for “spatial relations” and “mechanical aptitude” and a little practice on those is good, but arithmetic is the most important.
For higher paying more urban counties a high school diploma is a requirement, a GED isn’t an acceptable substitute, so complete high school, or test out, but do not drop out. More rural locals will accept a GED, but a drug test will be required instead.
Some locals will just accept apprentices based on their scores on the tests but most will interview as well, at one Union local the question was “Who do you know?” but for others it was “Do you have a reliable car?”, and “Are you used to working outside lifting heavy things?”.
Christopher Mims's 'Keywords" column in this weekend's WSJ creeps me out.
It's about Apple building its own modems. That isn't the creepy part. But the phrasing Mims uses in describing the technology for "always-on smart glasses and augmented reality, more wearables with their own independent connection to cellular networks" and a "computer-generated reality atop the real world" that gets me.
A new, improved version of reality will be overlaid on the natural world. How clever.
Isn't it bad enough that perceptions of public reality are systemically and professionally manufactured by those with the deepest pockets? Now some want to escape biological consciousness altogether. Alcohol, designer drugs and ridiculously video-gaming aren't enough escape.
Isn't putting on a headset in a gravity-free recliner with some kind of lens projecting 'augmented reality' into your eyes a bit narcissistic? I mean, there could be a cute-looking girl next door who's equally desperate.
There have been several comments here in the last few weeks explaining why nuclear war is unlikely to literally kill everyone.
To what extent is this because of arms controls treaties? What fraction of the human population would be killed if a full-scale nuclear war broke out today, compared to if a full-scale nuclear war broke out in the late 1980s when nuclear stockpiles were at their largest?
If these two numbers are significantly different from each other, then nuclear arms control treaties need to be celebrated a lot more.
Is this true? I know a large number of EA supporters are also AI researchers, but it's unclear that EA itself has any impact. The only argument I can see is that maybe OpenAI not releasing GPT-2's weights was inspired by AI safety ideology, but it's unclear that that action itself was hugely influential. Anyone know what this tweet could be referring to specifically?
It's crazy to me how the lab-leak theory has not permeated 'normie' circles. No one knows who Peter Daszek is, that Fauci has been in office for almost 40 years, or that the lead bioethicist of the NIH is his wife.
As a self sanity test: In a more sane world these people are on trial/investigation, no?
I have literally no medical research background, but seems like many here do. The question is directed to people who identify as having some. Would you say it is true that there is a stagnation in obesity and obesity prevention research, or is it actually progressing about as fast as other (which? comparable how?) comparable research fields? If it is indeed stagnating, what would you identify as the chief bottleneck?
This is part of a new series of posts I have planned on weird (potential) EA cause areas that are either completely new or that have been discussed before but are under-appreciated.
I'd like to see somebody review the book "The Shrinking of America" by Bernie Zilbergeld. Here's the best paragraph in the book. According to Arthur Janov, whose Primal Therapy created such a stir in the early 1970s, his methods produce "a tensionless, defense-free life in which one is completely his own self and experiences deep feeling and internal unity....People become themselves and stay themselves." Clients become more intelligent; are better coordinated; enjoy sex more; work better but do not overwork; lose their depressions, phobias, anxieties, as well as their compulsions to take drugs and alcohol, to overeat, and to smoke; and they are "never moody." There are special benefits for women: increased breast size for flat-chested women and the disappearance of premenstrual cramps and irregular periods. In a way, says Janov, "the post-Primal person is a new kind of human being," one who "is truly in control of his life." (Before readers rush off to sign up for Primal Therapy, I should mention that I am unaware of any independent research demonstrating the validity of these claims.)
Question: could consciousness have evolved as a solution to the black-box problem?
One of the challenges with AI is that we often don't know why the AI is spitting out certain results, and the lack of knowledge can cause trouble. My favourite example is Google Flu Trends. The model seemed really good at predicting the flu based on what people searched for, but it turns out it had over-fitted for spurious seasonal search terms, like "high school basketball." Its predictive capabilities therefore seemed impressive at first but then totally flunked.
Consciousness didn't necessarily evolve for anything useful (it could be a byproduct of other useful stuff). But if it did evolve for something useful, solving the black-box problem is one possibility. When brains evolved algorithms to find food or predators in sense data, perhaps they evolved the ability to attach qualia to those interpretations to better enable a top-down analysis of where those interpretations came from. It might make it easier to grasp what inputs are spurious bullshit.
I've noticed a number of "rationalist" type folks taking Dominic Cummings's recent writings on nuclear deterrence very seriously; I'm thinking of these pieces...
...in which Cummings tends to claim in strong and unequivocal language that (his words) "In the Cold War America based its nuclear strategy on an intellectual framework that was false. It defined standards of ‘rationality’ then concluded the Soviets would not use nuclear weapons in many scenarios. There was a governing tautology: rational leaders would be deterred otherwise they would be irrational. Given this tautology, more vulnerability improves ‘stability’ (e.g submarine launched weapons), while better defence is ‘destabilising’ (e.g missile defence)."
I've looked into Cummings's main source for this line of thought and found it to be inaccurate in its read of Soviet Cold War strategy, so I've been sharing this in some rationalist spaces (this comment is adapted from an earlier one on Zvi Mowshowitz's blog). Consider this a warning about taking Cummings too seriously on these matters.
I don't disagree with Cummings about everything, and I certainly share his concerns about the disturbingly high overall likelihood of great power nuclear conflict, but be careful with his stuff--he is not a well-informed source on nuclear deterrence. He often says things about how the USSR was willing to push the button at the drop of a hat, and this turns out to be not at all accurate from what I can tell. He also suggests at times that constructing missile defense is a useful policy in great power nuclear standoffs; this is badly mistaken for some pretty basic reasons.
--
After Zvi M talked up Cummings's views on nukes, I became curious about the Keith Payne book “The Fallacies of Cold War Deterrence” which is often cited by Cummings in support of the view that nuclear deterrence is much less stable than we think and rests on stupid assumptions about adversary psychology. So I gave the book a read and checked out some of the references.
The book makes some conceptual points that I think are useful as caveats to the usual theory of deterrence. The point is that no deterrence regime will work against every conceivable opponent. You need to take the specific goals and psychology of the opponent you face into account, rather than just assuming they are rational actors and leaving it at that. I don’t think these points count as objections to the prevailing school of thought, exactly, unless you turn it into a straw man version of itself (no, I don’t think Schelling et al believed that nuclear deterrence would’ve definitely worked against Hitler–but it is still an important point to note that sometimes a Hitler does come around). That part of the book was clarifying.
But my main goal was to clarify whether the empirical evidence Payne cites supported his contention about the disconnect between US and Soviet war plans, which he states thus (p 25-26):
“Only recently, courtesy of greater access to past Soviet decision-making practices, has it become virtually inarguable that the Soviet leadership never accepted the West’s definition of rationality with regard to nuclear weapons, and that Soviet expectations of US behavior, largely derived from the dogma of Marxist-Leninist ideology, appear to have been a significant factor in Soviet nuclear war planning.”
He refers several times to Soviet plans that “called for very heavy and very early nuclear and chemical strikes throughout Western Europe in the event of war.” (There is a grain of truth to this, but I think you’ll agree with me that it misrepresents his sources in an important way.)
The meatiest English-language source he cites is Heuser, “Warsaw pact military doctrines in the 1970s and 1980s: Findings in the East German archives,” which I hadn’t read before. It is a treasure trove, but it doesn’t say what Payne says it says. Let me quote the abstract, because the difference with Payne’s interpretation of the facts is night and day:
“Paradoxically, while the USSR was deploying more usable and survivable nuclear weapons (the SS‐20), it was developing a strategy which attempted to win a limited war in Europe with conventional weapons only. Pact records do show planning for preemptive nuclear strikes in response to observations of NATO preparations for nuclear launches. Great care was taken not to proceed to a nuclearization of the conflict unless the enemy was about to do so. These planning documents also reveal that the Pact was not expecting to launch all the nuclear weapons at its disposal.”
Essentially, Heuser’s read on the evidence is that for much of the Cold War the Soviets expected NATO to use nuclear weapons 5-6 days into a large conventional war (which was indeed NATO doctrine) and intended to use their nuclear weapons *in counterforce attacks* to destroy NATO nuclear weapons before this could be done. This was because they lacked the ability to gain early warning of NATO intent to launch. To quote Heuser again:
” Consequently, with fewer nuclear weapons at its disposal until the early 1970s than the United States had, the Soviet Union initially had to adopt a nuclear strategy of a massive, simultaneous strike of all forces available to it. Because of the limitations of Soviet abilities to monitor NATO activities, this would have had to be a preemptive strike. Only in the late 1970s did the USSR gradually acquire the technical means required for its preferred strategy, launch-on-warning: this meant Warsaw Pact nuclear release at the moment it became clear that NATO had launched or was about to launch nuclear weapons against the Soviet Union [18].”
By the late 70s, when the Soviets had the technology and capabilities to adopt a more American-style strategy, they did so, contrary to Payne’s claims. For example:
PAYNE:
“NATO policy sought to maintain a lid on the potential for nuclear escalation via sophisticated intra-war deterrence concepts of rational wartime bargaining and limited nuclear war. These US deterrence concepts mistakenly assumed a similarly minded, “rational” opponent, and thus were wholly incompatible with Soviet war plans.”
HEUSER:
“The Warsaw Pact exercise plans, however, confirm the careful analysis of contrary evidence by Notra Trulock, who argues that the Soviet Union did plan also for limited nuclear strikes [87]. Warsaw Pact planning assumed in the late 1970s and early 1980s that there might be several nuclear exchanges, but there was obviously the hope that followon exchanges might be avoidable.”
I went into this willing to update in Payne’s direction, and I think some of his theoretical insights are sound, but I think it became clear from my fact-checking that he misrepresents the empirical evidence about Soviet planning quite badly. The truth is much more complicated than his take, and in some ways contradicts it. I certainly don’t see any evidence that the Soviets were “irrational” by Schelling’s lights. (Castro may have been that irrational, but on the other hand he may have urged a launch knowing that the Soviets wouldn’t agree… kind of like how the Poles lately keep asking NATO to do things against Russia that they know the US will veto, as a sort of brinksmanship virtue signaling.)
I conclude that Cummings is too credulous, and is willing to assert with great confidence claims from a semi-popular book without checking the references to confirm its key claims.
--
To dismiss his claims about missile defense is much more straightforward; I've been surprised to see some smart people convinced by them, which must just mean that these people haven't looked into the issue in any detail. The classic argument against missile defense is that it only incentivizes the construction of more warheads by the enemy. MIRVing (putting multiple warheads on one's offensive missiles) makes it cheap and easy to overwhelm any missile defense system by launching more warheads than the defense system has interceptors. If you build interceptors to destroy your enemy's warheads, they will simply (for much cheaper) build delivery systems for an equal number of additional warheads and you've gotten less than nowhere on the whole. I don’t see any way around this calculus unless the technology changes drastically.
The calculus is different against N Korea for the near future, since they can’t afford very many missiles and lack MIRV technology. That said, MIRV is a technology that took the US and the Soviets about 10 years to develop following their invention of intercontinental missiles. North Korea is working now to develop this technology ( https://en.wikipedia.org/wiki/Hwasong-17 ), so it's unlikely that missile defense will provide a feasible check against them for much more than a decade longer.
How do I choose a school for my children? My partner and I are open to public school, private school, homeschooling - but how do we evaluate our options? I'm looking for book recommendations, essay recommendations, general advice. I'm aware of Great Schools, but it seems like their metric boils down to state test scores and SAT test scores. Should I believe these metrics capture academic rigor? What's more, social and moral development are of comparable import to me - how do I evaluate a school's ability to foster - or at least not impede - those?
I'm a maths and physics tutor looking for some more students. I can teach to anyone below undergrad level, and for undergrad stuff I can teach: mechanics, QM, ODE/PDEs, chaos theory, stat mech, probability theory, logic, QFT, quantum computing basics, ML, RL and some deep learning stuff. Message me if you want my services, my rates are negotiable.
Hi all! I'm a Master's student and for my dissertation I'm doing a qualitative study on expert forecasters. If you're reading this, interested participating, and are in the top percentile on either Metaculus or GJO, or a Superforecaster, could you leave a comment or contact me via the mail connected to this account or through Twitter @meraxion? Thanks! If you want more information, read on;
The particular interview methodology I'll be using is called Applied Cognitive Task Analysis (ACTA), which is a mostly structured series of (3) interview techniques used to figure out how experts in a field view their task, and what cues and strategies they look out for that they think novices won't know about. The methodology is often used as a first step for further construction of training manuals and things like that. The 3 techniques are; the Task Diagram, the Knowledge Audit, and the Simulation Interview. The Task Diagram asks the expert to lay out the (up to six) high-level steps of their task. The Knowledge Audit digs into the cues and strategies experts use for understanding aspects of their task (and making decisions in task situations). The Simulation Interview looks at the expert's actions in situ while doing the task, asking questions about actions, situation assessment, critical cues, and potential errors, at each step (as identified in the Task Diagram).
The product of the is the Task Diagram, and a Cognitive Demands Table outlining what came up during the interview(s) as the important/difficult/demanding parts of the task.
Interviews would be conducted in a month's time (June, maybe late May).
If you think this doesn't sound too daft, please don't hesitate to contact me using one of the methods I linked to above, and/or sharing with others whom you think might be interested!
Can anyone recommend a (free, online) introduction to statistical methods in climatology.(especially statistical methods in paleoclimatology) which doesn't assume domain knowledge, but also doesn't sacrifice mathematical rigour? What is the CS229 of climatology?
how do you manage the acid base balance in your body? What supplements do you take? How do you increase that HCO3 in your blood (i don’t intend to produce an alkalosis)? I am certain that taking Magnesiumcitrate is not enough (as it is converted to bicarbonate in the kidneys, but it doesn’t mean that it necessarily increases serum HCO3). What else do you do to manage it? don’t call it “bogus science” because clearly it’s not. Drop some knowledge here, please.
I literally *cannot imagine* enduring that kind of sleep deprivation. I feel utterly awful and am largely incapable of any productive work the next day if I get less than around 7 hours of sleep. And I find it basically impossible to nap during the day, no matter how exhausted I am.
Are there any parents here who had a similar lack of previous ability to cope with sleep deprivation? Are there any good studies about what parents can do to get their children to sleep through the night as soon as possible?
(I've read claims that bed-sharing helps--it just seems implausible to me that mothers in traditional high-fertility societies were enduring modern levels of new-parent sleep deprivation throughout their reproductive years.)
I’m three years into my sabbatical and <6 months before I get back into the market so could do with a little more cash. Thus I’m interested in tutelage. I’ve done pro-content-creation for three and a half years before the sabbatical (I’m 26, UK based) so can teach anyone much about the skill(s) of videography/cinematography/editing/photography. I’m also an adept meditator having achieved stage 7 shamatha (meditative quiescence, and there’s 10 stages), so could assist someone who wants to get up to speed with that.
If you’re interested I’ll send you my work (for the content), and fees I’ll do at £40ph but if you can’t afford it we can discuss a more suitable price for you.
I think such a marketing post is tolerated here, but if not let me know and I’ll take it down.
Skimming the hivemind blogpost on Lacan, i am more and more convinced that lacan is, as someone else remarked elsewhere, a rorschach test - everybody read what it wants to read in it. Who guarantees me that the post is right and she has really truly understood what is going on?
Beside, i found the comparison with sapir-whorf telling, given how whorf himself was a master of motte and bailey statements such as "the hopi do not have time because they speak hopi"
Did some meditating on the Zvi/Hanson/etc. posts on blackmail.
Here's a question that came out of it:
Does there exist rationally-hidden information (i.e. it would be bad for the person hiding it were it to get out) that is positive-sum to disclose, that is not and *should not be even in theory* evidence of illegal behaviour?
To put it another way, does there exist an act X for which a) revealing X is bad for the X-doer, b) revealing X is overall good for society, c) X shouldn't be banned?
Congratulations to Hidden Open Aphorism contest winners timujin, Elisabeth Piper and Gašo, and thanks for all the pithy aphorisms and thoughtful comments!
الكلاب تنبح والقافلة تسير" (The dogs bark, but the caravan moves on) so I'm proud to announce the inaugural ACX Open Political Punsters contest, in tribute to the great comedian Volodimir Zelenskyy.
Prizes (well kudos) for the punniest political puns, with extra points for originality and brevity.
Your starter for 10:
- Why do anarchists only drink herbal tea? Because all proper tea is theft! (makes me so Proudhon)
Contest closes 8pm UTC Monday 4 May the force be with you and your time starts... now!
When you tally up the ratings of the book review entries, it would be interesting to sort reviews not just by average rating, but also by variance. (Will we get to see average ratings for all or any of the reviews?)
Any mental health/psych types want to have a go at this one?
My wife: early 30s, hyper-conscientious, neurotic with history of intrusive thoughts. She's currently struggling with thinking back to a professional document she wrote and getting waves of anxiety about whether the information was absolutely correct and defensible in court (NB it is very unlikely to end up in court, and her job is not in law - the memo just had some legal implications). In general, her relentless searching for possible sources of fault and guilt, despite (or because of?) her life being very comfortable, makes her life hell.
She's considering which type of professional to talk to, and whether or not she should be regarding anti-anxiety meds as a last resort (due to side-effects etc.). She's seen a psychoanalyst once (ages ago) and was frustrated by their esoteric/lofty/impractical approach.
Any suggestions appreciated as to what type of professional to talk to and whether anti-anxiety meds are a slippery slope to another kind of hell.
Jaynes has convinced me that almost all hypothesis testing, as it is currently performed, is ill-founded.
He also clarified something that I've been coming to for a while: the goal of statistics is *always* prediction, and if your model can't predict anything, it's worthless. This has been my problem with string theory/SUSY, and it's come to the fore of my mind again because of the Lacan discussion. Who cares what your psychoanalysis theory is unless it can make predictions about behavior, so I can adjust my own behavior accordingly?
Speaking past one another because vomiting theories is good enough for a paycheck is such an indictment of modern science. Everyone know about the social science replication crisis, but the Null Hypothesis Significance Testing paradigm is fundamentally flawed, and we base all of hard science on it too! Particle Physics is recognized as being the most stringent applied statistical discipline, but 2 weeks ago a result (W boson mass) was published that is fundamentally incompatible with previous measurements, so one or both sets must be wrong. Even the most preeminent scientists are failing to be precise enough to weed out errors, and my fellow statisticians are happy to keep on teaching STAT 301 to engineers and scientists. What's gotta give?
I've learned a whole lot about real estate Mass Appraisal and gotten in touch with a bunch of really smart people with the relevant credentials, several of whom we've recruited onto our team. Apparently a whole bunch of new papers in the field were published just in the last 3 years, and it seems like there's a lot of room to move both the state of the art and the state of common practice forward.
What's become quickly apparent is that this project is mostly all about data. Luckily, we've been absolutely spoiled by the OpenPhilly dataset. They have more fine-grained data than you can shake a stick at, and the majority of our work is going to consist of cleaning and merging these data sets together, then using them to reproduce the best models from the last 3 years of papers. After that our plan is to publish a paper, publish the source code, and release all the data that's not encumbered by any licensing issues (which so far is basically all of it). We would also like to set up a public website to do "live ratio studies" where we will track ongoing sales and compare the values of those properties to what our model predicts, and take particular close attention to sales of vacant land. That way if our model sucks it will be super obvious and you can all make fun of us more easily.
I also did a general survey of the data landscape of the top 20 US cities. They're surprisingly good! Most cities nowadays have slick "open data" portals with all kinds of information just free to download. Philadelphia's is still looking like it's the absolute best, but quite a lot of the others are up there.
But the absolute hands-down WORST? Silicon Valley (San Jose / Santa Clara County), ironically:
Has Scott written an article explaining the replication crisis from the ground up? Alternatively, is there a good, well-written introduction to the crisis from another author geared toward an intelligent reader who is wholly new to the subject? I’d like to recommend something to a friend on this topic.
How would the gold standard EA charities compare with just giving money to poor people on the street? Not talking San Francisco mentally ill people, just regular people on the street selling wares and begging for cash (as it happens all the time in my country (see comment below)).
No charity bureaucracy, complete certainty the other person is getting the money, people usually seem obviously deserving, makes me feel good etc. Probably stacks up pretty well, no? What do you think?
Has anyone here used thyroid hormone (T3 or T4) for depression, or other issues besides low thyroid?
I know it hasn’t done great in studies, but I developed severe hyperthyroidism (Graves disease) last year and it essentially cured my acute depression and showed me I had been experiencing dysthymia and the like for years. Treating the hyperthyroidism made the depression come back in full force, which is when I made the connection.
My depression is now well-controlled through a combo of Wellbutrin and remaining slightly hyperthyroid (with my endocrinologist’s blessing)
Are there AI alignment groups / companies that are remote or remote-friendly for engineers? It seems like all the major organizations are still exclusively in-office (primarily Bay, although it seems like Conjecture is in London).
I came across this: "If you look at incomes today then you find that the income in the richest country in 1950 is very close to the average income of the average person in the world today ($14,570). Today the average person on the planet is as rich as the the average person in the richest country in 1950. And all those countries that have an income higher than the global average today are more prosperous than the US in 1950: Iran, Mexico, Bulgaria, … Have a look at that list."
I live in one of the third (second?) world countries mentioned and I can't help but think that this is the sort of thing that makes people distrust the upbeat narrative Our World In Data emphasizes. Does anyone else feel like this can't possibly be true in a "deeper" sense? My country is falling apart, panhandlers are endemic and basic functions of the state go unfulfilled systematically. I know how poor people live in this country and I just can't fathom their lot being anyhow comparable to that of poor Americans in the time most people instantly associate with prosperity. Am I being irrational about this? Does anyone think I have a point?
When genetics met capitalism in the 1930s corn belt! https://monthlyreview.org/1986/07/01/the-political-economy-of-hybrid-corn/
Since the 1930s seed companies keep inbred lines of corn, not sold to farmers (e.g. AAbb & aaBB), that when crossed make genetically identical F1 hybrids, which is what is sold to farmers (e.g. all AaBb).
The F1 hybrids have "hybrid vigor" like higher yield etc., but if you save and replant it then the F2 hybrids are genetically variable (e.g. some Aabb, aabb, etc.) and aren't reliably vigorous—hybrid vigor doesn't "breed true". Farmers want reliable vigor, and buy new F1 seed every year, and seed companies profit from repeat business.
There are two main theories for how hybrid vigor works. Under the "dominance" theory, good dominant genes from each parent cover up bad recessive genes from the other (e.g. A dominant over a, B dominant over b). Under the "overdominance" theory, hybrids are vigorous because they have heterozygote advantage at many loci (e.g. Aa fitter than either AA or aa, Bb fitter than either BB or bb).
Under the first theory, you could also get the same vigor just by a selection program, but that *would* breed true—if seed companies haven't done so, it's not because of scientific obstacles, it's strictly business. Under the second theory, you couldn't select for reliable true-breeding vigor anyway. Of course seed companies prefer the second theory (I'm not sure if scientists are 100% on which is true).
Nowadays of course seed companies also try to patent their plants, or make GMO seeds that can't reproduce, etc., anything to keep the farmers reliant on buying new seed. But hybrid corn was the start.
It's begging letter time! Anyone who can afford to make charitable donations and is looking for something to donate to, here's one:
Navajo & Hopi Families COVID-19 Relief Fund
GoFundMe link https://gofund.me/0656bab6
"As of today (November 23, 2020), there are 15,039 positive cases of COVID-19 and 631 total confirmed deaths on the Navajo Reservation. This marks a spike of over 2,200 cases in 12 days on Navajo. Numbers are also on the rise on the Hopi Reservation, and both tribal governments have placed their communities in lockdown again. Our relief effort is the primary relief effort providing assistance on both reservations. We have already deployed roughly $6 million in direct relief in both communities, which has us nearing depletion of existing GoFundMe donations. We need a second wave of donations to help us meet the second wave--tsunami, really--of COVID that is inundating our communities.
To date, the Relief Fund has raised over $6 million, and has spent almost all of that on bringing vital resources to Navajo and Hopi communities to help shield them from COVID. The team and a legion of hundreds of Navajo and Hopi volunteers all over the two nations has provided food and water to over 46,000 households (each roughly averaging 4 persons per household, and amounting to over 186,000 people served—which is more than the combined population of the Navajo and Hopi Reservations) in 335 distributions in over 530 Navajo communities and Hopi Villages.
The Relief Fund currently seeks to raise enough funding to carry Navajo and Hopi families through this second wave, through the cold and flu season, and to the end of the COVID pandemic in these communities. We estimate needing $6.5 million in order to meet that goal, so we have now adjusted our GoFundMe campaign goal to $13 million total.
By donating, you will help the most heavily impacted community in the USA. For $100, you will help a family of four by providing them with 2 weeks’ worth of food and Personal Protective Equipment (PPE). For as little as $10, you can help feed a family of 4 for an entire day. Donate today to the Navajo & Hopi Families COVID-19 Relief Fund."
Short video dated March 15th 2022: https://www.youtube.com/watch?v=6xQdn3Ffnf8
You can read further details over on the fundraiser page. I would ask you all to please consider giving if you can. Thank you!
I can't tell yet how much the Roe V Wade leak is affecting people in general.
For my part, because I live in Texas, a purple state, I plan on donating a lot of time and money to the Democratic party to help them win in November, even though it might be a longshot, because of the apparent overturn of Roe.
I'm still curious how seriously people take the leak, though. Given the leak, what odds do you give Roe V Wade being overturned this semester? Might the leak be misleading?
I'm hoping I can help change Texas politics and put it back in a more Libertarian direction, instead of the Trumpist direction it has gone recently. I'd join the Republican party if I thought I could make them more moderate, because all I want is a moderate outcome. But my experience with the parties in Texas is that the Republicans are organized and wouldn't like me, whereas the Texas Democrats are entirely disorganized and clueless and maybe my organizational skills could help them.
Not saying we can win, but I'm ready to fight. I encourage others in Texas who believe in a woman's right to choose to join the fight.
What is the best source for getting a good understanding of what Trump was like as a businessman? Is there some biography out there that gives a good overview of how he worked? I thought of getting Art of the Deal but it's unclear to me how much of it would just be self-aggrandizement.
For people who are into genealogy and family trees, what is your preferred format, or alternatively what website do you think has the best format? Is Ancestry food for instance or are their better ways to display family trees?
A while ago, our project won an ACX grant to try to pursue a potentially far-reaching idea of spellchecking genomes with genetic editing. We are looking for a genetic editing expert to help in an advisory capacity, or potentially even join the core team.
Please reach out to us at at spellcheckhealth@icloud.com if this sounds interest to you - thanks!
Hey so I was wondering what people here think about the following: Is it "normal" to have different political leanings based on one's location?
I will explain my situation: I have lived in three countries (Germany, Canada and the US). In Germany, I tend to favour more right-leaning parties like the FDP, CDU/CSU and some others. Sometimes I am even sympathetic to the AFD (which is considered far-right). On most political issues in Germany, I tend to take the more right-leaning view (e.g. Nuclear power, "Energiewende", Refugees/Immigration etc.). In the US and Canada I consistently support the left-leaning parties and view on (almost) all issues.
I tend to explain (or justify) this by arguing from an utilitarian perspective: it makes more sense for Germany (and other Western European countries) to be more conservative, since the population density is so high and the geopolitcal neigbourhood (e.g. being close to the Middle East, Russia/Ukraine and Africa) makes it more prudent from an utilitarian perspective to be more cautious, ergo conservative. Thus I think that Western European countries should support policies of gradual improvement instead of being revolutionary (e.g. in terms of energy, developing Nuclear instead of renewables, and not letting in too many immigrants so as not make the social systems collapse).
For the US and Canada, with their many resources and low population density, it makes more sense to be more left-wing: letting in many immigrants, being socially liberal, and building lots of infrastructure (I guess having low population density means having a higher carbon footprint, which is "conservative". I do support more environmentalism in the US and Canada, especially since they have so much land space they could support more renewable energy like solar, wind and hydro). From a utilitarian global perspective I think it also makes more sense to support left-wing policies for the US and Canada (particularly in regards to immigration - because the US and Canada have so much space, they should take in much more immigrants from Africa and the Middle East than Europe). For Western Europe I don't think it's clear that from a global utilitarian perspective it's better to have more left-wing governments (e.g. the refugee crisis destabilised the EU and led to populism in many countries, and the EU doesn't have enough area to develop renewable energy to the extent the US and Canada can.)
So anyway what I am asking/wondering is if my argumentation makes sense from an utilitarian perspective, or if I am being an (European) egoist, since I will admit that generally being conservative means being egoistical (I guess you could disagree, but I think that statement makes sense from and utilitarian perspective).?
On AI alignment: recursive self-improvement seems like a prerequisite for the type of AGI that would become an existential threat. To what extent have alignment researchers looked into ways to prevent AGI from trying to self-improve? Has there been research into e.g. utility functions that disallow self-improvement?
Has anyone really looked into the literature on the relationship between mathematical ability and musical aptitude? Bearing in mind how consistently this correlation occurs, and the fact that ridiculously large fraction of the people I know in math heavy majors, were either in a band or compose their own music, I'm confident the correlation is real.
A lot of the literature seem to (incorrectly, in my opinion) infer that learning how to play music on an instrument directly improves mathematical ability, but can this really be right? I have four competing hypotheses and wondered if anyone could verify which, if any, are most likely to be correct:
1. Learning to play music somehow ,directly or indirectly, improves mathematical ability.
2.Kids who get involved in music are more likely to have smart, educated, middle-class parents, and by some combination of nature and nurture have an advantage in math.
3. People who are better at math often have superior general intelligence, thus making them more likely to be able to excel in multiple domains, like music and math. While people with intelligence closer to avg general intelligence could not maintain both, so they must choose one.
4. (And for my wackiest theory)It's not really intelligence alone driving the correlation, it is the combination of intelligence and autism-with the latter being arguably more important. It is generally accepted that STEM majors are, on average, more autistic than the general population. So my theory goes that the need for order and pattern recognition abilities that are often characteristic of autism would lend itself well musical pursuits.
For the first time in the modern history of the Supreme Court, a judge's draft opinion has been leaked (on Roe v Wade).
https://www.politico.com/news/2022/05/02/supreme-court-abortion-draft-opinion-00029473
I'm *really* hoping this was just some rogue clerk. Either way, this leak is going to be a massive shitshow.
Hi, on a previous thread (or maybe on the subreddit) someone had mentioned an app for setting goals and reflecting on them. It had a paid version where you could talk to a human and a name that dropped the first letter of a common word (think "concile" from "reconcile" but clever). Can anyone remember what it was?
Why does Twitter have such high operating expenses? What do they need all those programmers for? They've made only a couple of changes to the site (all for the worse) in 15 years.
What are the programmers spending all their time doing?
Anyone who is self taught in AI/ML and landed a job in the field ? I’d love to hear about the resources/curriculums you used.
What do people's office climates look like with the lifting pandemic restrictions?
My office used to encourage remote work (both explicitly, as in "if you can work from home, please work from home," and also implicitly through mask requirements and closing the dining room). It recently switched to encouraging a return to the office (in particular, masks are optional, dining room is open again) -- though many people have concluded that actually working from home is great, and are still doing a lot of it (there's no strong pressure to show up at the office unless you need to work with a physical object there).
One development I was expecting to see was "if you're feeling under the weather, please wear a mask," but the official policy is instead "if you're feeling under the weather, please stay home". This isn't terrible as policies go, but it means that "masks optional" actually means "masks off" -- there's just no room for "I'm feeling sick enough to wear a mask, but not sick enough to stay home." I'm not yet decided how I feel about this, does anyone have a coherent opinion?
The 1950s, 60s, 70s, 80s and 90s had very distinctive styles in terms of fashion, clothing, graphic design, music, food, pretty much everything.
Then sometime around 1999, we got stuck. There was no distinctive 2000s style that looks different to a 2010s style, and the 2020s aren't looking great either unless you count pandemic-induced slobbiness as a fashion trend. Basically, 2022 looks like 2002 except with smartphones.
What gives? Possible explanations:
1. Something about the internet soaking up all the innovation leaving meatspace stuck in an eternal 1999
2. The 1950s-1990s period was a weird abberation, a manifestation of the "generation gap" between the boomers and their parents.
3. This is all an illusion and actually 2022 looks as unlike 2012 as 1992 looked unlike 1982, and I'm just not sensitive enough to notice it.
I would like to second Scott's recommendation of the Hivewired article on Lacan psychoanalysis.
https://hivewired.wordpress.com/2022/04/27/the-game-of-masks/
It is extremely well written....and really very mind bending.
Question about Nato expansion, "not one inch to the East" etc.
I've seen people on Russia's side of this argument (eg Jeff Sachs) state that promises were broken, and I've seen the West's response which I've tried to describe below, but I've never found someone engage with or try refute that response.
What am I getting wrong/missing?
What happened -
1. During discussions re reunification of Germany, Russia were given verbal assurances that Nato would not expand eastwards within reunified Germany. The phrase "not one inch to the east" was used.
2. But in the end the actual treaty agreed allowed for some troops in what was East Germany.
3. No assurances were ever discussed or given about other former Soviet states joining Nato.
4. Russia itself discussed joining Nato and were told they would need to apply like everyone else and so didn't (my guess is they wouldn't have been admitted if they did apply)
West's view -
A. if it's not in the treaty it doesn't count, democracies change governments all the time, new governments can't be bound by verbal assurances of their predecessors
B. This is especially so if the treaty that was signed already wasn't in line with those verbal assurances
C. No assurances re former soviet states were given.
D. The idea that the states that were actually part of the USSR and then became sovereign not being allowed to join Nato is strange - it's essentially like saying, we promise not to let you join our club even if you want to?
Is there a way to view the Substack archive in the order the posts were published (or just any way remotely more convenient than the standard option)? I am just about caught up on SSC, and scrolling through the whole ACT archive to start from the beginning seems a little ineffective.
Political Gridlock in the United States
https://fivethirtyeight.com/features/how-much-longer-can-this-era-of-political-gridlock-last/
My view of politics in the United States is that the country is fairly evenly divided into two groups: Republicans and Democrats. Each group considers the other group to be a threat to the nation, and their most important political goal is to stop the other group from succeeding.
Now in my mind this is a recipe for failure of the United States. I believe the two-party system is the true threat to the nation.
I have been thinking about this a lot and devising a possible solution. The sad irony is that given our political stalemate, the solution is impossible to implement. But if you are willing to imagine the best possible electoral system for the United States (not just the puny changes currently possible), please read this proposal and share your thoughts with me.
Bit.ly/RealElectors
At a DSL thread https://www.datasecretslox.com/index.php/topic,6391.0.html
@no one special posted:
“ If he wants to become an electrician, that's wonderful! The only problem is... I have no idea how to help him do that! As a university-attender myself, I know advice and procedures on how to do that. And while it was kind of expected of me, I have my own questions about the value of it both on its own, and compared to the costs. But when he asks me how to become an electrician, well, I just don't know!”
Since at DSL I am banned at my own request till June I can’t respond there soon, I suppose I’ll and I ask others to pass on the info to DSL, thanks if you do:
@no one special,
The aptitude tests to be an apprentice electrician in most counties locals is similar to the test to be an apprentice plumber but typically the IBEW has additional tests for English language skills while the UA stresses arithmetic more, regardless being able to do multiple choice arithmetic problems very quickly will be the core of the tests, your son should throughly practice that, there will also be tests for “spatial relations” and “mechanical aptitude” and a little practice on those is good, but arithmetic is the most important.
For higher paying more urban counties a high school diploma is a requirement, a GED isn’t an acceptable substitute, so complete high school, or test out, but do not drop out. More rural locals will accept a GED, but a drug test will be required instead.
Some locals will just accept apprentices based on their scores on the tests but most will interview as well, at one Union local the question was “Who do you know?” but for others it was “Do you have a reliable car?”, and “Are you used to working outside lifting heavy things?”.
-Plumber
Christopher Mims's 'Keywords" column in this weekend's WSJ creeps me out.
It's about Apple building its own modems. That isn't the creepy part. But the phrasing Mims uses in describing the technology for "always-on smart glasses and augmented reality, more wearables with their own independent connection to cellular networks" and a "computer-generated reality atop the real world" that gets me.
A new, improved version of reality will be overlaid on the natural world. How clever.
Isn't it bad enough that perceptions of public reality are systemically and professionally manufactured by those with the deepest pockets? Now some want to escape biological consciousness altogether. Alcohol, designer drugs and ridiculously video-gaming aren't enough escape.
Isn't putting on a headset in a gravity-free recliner with some kind of lens projecting 'augmented reality' into your eyes a bit narcissistic? I mean, there could be a cute-looking girl next door who's equally desperate.
There have been several comments here in the last few weeks explaining why nuclear war is unlikely to literally kill everyone.
To what extent is this because of arms controls treaties? What fraction of the human population would be killed if a full-scale nuclear war broke out today, compared to if a full-scale nuclear war broke out in the late 1980s when nuclear stockpiles were at their largest?
If these two numbers are significantly different from each other, then nuclear arms control treaties need to be celebrated a lot more.
https://twitter.com/MarkLutter/status/1520510178214383617
> EA has been hugely influential in shaping AI
Is this true? I know a large number of EA supporters are also AI researchers, but it's unclear that EA itself has any impact. The only argument I can see is that maybe OpenAI not releasing GPT-2's weights was inspired by AI safety ideology, but it's unclear that that action itself was hugely influential. Anyone know what this tweet could be referring to specifically?
It's crazy to me how the lab-leak theory has not permeated 'normie' circles. No one knows who Peter Daszek is, that Fauci has been in office for almost 40 years, or that the lead bioethicist of the NIH is his wife.
As a self sanity test: In a more sane world these people are on trial/investigation, no?
I welcome disagreements.
I have literally no medical research background, but seems like many here do. The question is directed to people who identify as having some. Would you say it is true that there is a stagnation in obesity and obesity prevention research, or is it actually progressing about as fast as other (which? comparable how?) comparable research fields? If it is indeed stagnating, what would you identify as the chief bottleneck?
I just published a new Substack post: "Weird cause area #1 - kidney stone pain"
https://moreisdifferent.substack.com/p/weird-cause-area-1-kidney-stone-pain?s=w
This is part of a new series of posts I have planned on weird (potential) EA cause areas that are either completely new or that have been discussed before but are under-appreciated.
I'd like to see somebody review the book "The Shrinking of America" by Bernie Zilbergeld. Here's the best paragraph in the book. According to Arthur Janov, whose Primal Therapy created such a stir in the early 1970s, his methods produce "a tensionless, defense-free life in which one is completely his own self and experiences deep feeling and internal unity....People become themselves and stay themselves." Clients become more intelligent; are better coordinated; enjoy sex more; work better but do not overwork; lose their depressions, phobias, anxieties, as well as their compulsions to take drugs and alcohol, to overeat, and to smoke; and they are "never moody." There are special benefits for women: increased breast size for flat-chested women and the disappearance of premenstrual cramps and irregular periods. In a way, says Janov, "the post-Primal person is a new kind of human being," one who "is truly in control of his life." (Before readers rush off to sign up for Primal Therapy, I should mention that I am unaware of any independent research demonstrating the validity of these claims.)
Humans are only 3.5% efficient at accomplishing useful work.
https://www.militantfuturist.com/the-extraordinary-inefficiency-of-humans/
Any opinions on who to vote for in the California Governor primaries, either from a strategic or idealistic voter's perspective?
On the Futurati Podcast we recently interviewed ACX grant winner Nathan P Young on prediction markets, policy, and futarchy:
https://www.youtube.com/watch?v=F4UfK9qfW3E
I also think our most recent episode with Rik on cybesecurity and the future of warfare is a banger too:
https://www.youtube.com/watch?v=0wtkTbGaB5s
Question: could consciousness have evolved as a solution to the black-box problem?
One of the challenges with AI is that we often don't know why the AI is spitting out certain results, and the lack of knowledge can cause trouble. My favourite example is Google Flu Trends. The model seemed really good at predicting the flu based on what people searched for, but it turns out it had over-fitted for spurious seasonal search terms, like "high school basketball." Its predictive capabilities therefore seemed impressive at first but then totally flunked.
(Article about the google flu trends: https://www.wired.com/2015/10/can-learn-epic-failure-google-flu-trends/ )
Consciousness didn't necessarily evolve for anything useful (it could be a byproduct of other useful stuff). But if it did evolve for something useful, solving the black-box problem is one possibility. When brains evolved algorithms to find food or predators in sense data, perhaps they evolved the ability to attach qualia to those interpretations to better enable a top-down analysis of where those interpretations came from. It might make it easier to grasp what inputs are spurious bullshit.
I've noticed a number of "rationalist" type folks taking Dominic Cummings's recent writings on nuclear deterrence very seriously; I'm thinking of these pieces...
https://dominiccummings.substack.com/p/people-ideas-machines-iii-more-on
https://dominiccummings.substack.com/p/people-ideas-machines-ii-catastrophic
...in which Cummings tends to claim in strong and unequivocal language that (his words) "In the Cold War America based its nuclear strategy on an intellectual framework that was false. It defined standards of ‘rationality’ then concluded the Soviets would not use nuclear weapons in many scenarios. There was a governing tautology: rational leaders would be deterred otherwise they would be irrational. Given this tautology, more vulnerability improves ‘stability’ (e.g submarine launched weapons), while better defence is ‘destabilising’ (e.g missile defence)."
I've looked into Cummings's main source for this line of thought and found it to be inaccurate in its read of Soviet Cold War strategy, so I've been sharing this in some rationalist spaces (this comment is adapted from an earlier one on Zvi Mowshowitz's blog). Consider this a warning about taking Cummings too seriously on these matters.
I don't disagree with Cummings about everything, and I certainly share his concerns about the disturbingly high overall likelihood of great power nuclear conflict, but be careful with his stuff--he is not a well-informed source on nuclear deterrence. He often says things about how the USSR was willing to push the button at the drop of a hat, and this turns out to be not at all accurate from what I can tell. He also suggests at times that constructing missile defense is a useful policy in great power nuclear standoffs; this is badly mistaken for some pretty basic reasons.
--
After Zvi M talked up Cummings's views on nukes, I became curious about the Keith Payne book “The Fallacies of Cold War Deterrence” which is often cited by Cummings in support of the view that nuclear deterrence is much less stable than we think and rests on stupid assumptions about adversary psychology. So I gave the book a read and checked out some of the references.
The book makes some conceptual points that I think are useful as caveats to the usual theory of deterrence. The point is that no deterrence regime will work against every conceivable opponent. You need to take the specific goals and psychology of the opponent you face into account, rather than just assuming they are rational actors and leaving it at that. I don’t think these points count as objections to the prevailing school of thought, exactly, unless you turn it into a straw man version of itself (no, I don’t think Schelling et al believed that nuclear deterrence would’ve definitely worked against Hitler–but it is still an important point to note that sometimes a Hitler does come around). That part of the book was clarifying.
But my main goal was to clarify whether the empirical evidence Payne cites supported his contention about the disconnect between US and Soviet war plans, which he states thus (p 25-26):
“Only recently, courtesy of greater access to past Soviet decision-making practices, has it become virtually inarguable that the Soviet leadership never accepted the West’s definition of rationality with regard to nuclear weapons, and that Soviet expectations of US behavior, largely derived from the dogma of Marxist-Leninist ideology, appear to have been a significant factor in Soviet nuclear war planning.”
He refers several times to Soviet plans that “called for very heavy and very early nuclear and chemical strikes throughout Western Europe in the event of war.” (There is a grain of truth to this, but I think you’ll agree with me that it misrepresents his sources in an important way.)
The meatiest English-language source he cites is Heuser, “Warsaw pact military doctrines in the 1970s and 1980s: Findings in the East German archives,” which I hadn’t read before. It is a treasure trove, but it doesn’t say what Payne says it says. Let me quote the abstract, because the difference with Payne’s interpretation of the facts is night and day:
“Paradoxically, while the USSR was deploying more usable and survivable nuclear weapons (the SS‐20), it was developing a strategy which attempted to win a limited war in Europe with conventional weapons only. Pact records do show planning for preemptive nuclear strikes in response to observations of NATO preparations for nuclear launches. Great care was taken not to proceed to a nuclearization of the conflict unless the enemy was about to do so. These planning documents also reveal that the Pact was not expecting to launch all the nuclear weapons at its disposal.”
Essentially, Heuser’s read on the evidence is that for much of the Cold War the Soviets expected NATO to use nuclear weapons 5-6 days into a large conventional war (which was indeed NATO doctrine) and intended to use their nuclear weapons *in counterforce attacks* to destroy NATO nuclear weapons before this could be done. This was because they lacked the ability to gain early warning of NATO intent to launch. To quote Heuser again:
” Consequently, with fewer nuclear weapons at its disposal until the early 1970s than the United States had, the Soviet Union initially had to adopt a nuclear strategy of a massive, simultaneous strike of all forces available to it. Because of the limitations of Soviet abilities to monitor NATO activities, this would have had to be a preemptive strike. Only in the late 1970s did the USSR gradually acquire the technical means required for its preferred strategy, launch-on-warning: this meant Warsaw Pact nuclear release at the moment it became clear that NATO had launched or was about to launch nuclear weapons against the Soviet Union [18].”
By the late 70s, when the Soviets had the technology and capabilities to adopt a more American-style strategy, they did so, contrary to Payne’s claims. For example:
PAYNE:
“NATO policy sought to maintain a lid on the potential for nuclear escalation via sophisticated intra-war deterrence concepts of rational wartime bargaining and limited nuclear war. These US deterrence concepts mistakenly assumed a similarly minded, “rational” opponent, and thus were wholly incompatible with Soviet war plans.”
HEUSER:
“The Warsaw Pact exercise plans, however, confirm the careful analysis of contrary evidence by Notra Trulock, who argues that the Soviet Union did plan also for limited nuclear strikes [87]. Warsaw Pact planning assumed in the late 1970s and early 1980s that there might be several nuclear exchanges, but there was obviously the hope that followon exchanges might be avoidable.”
I went into this willing to update in Payne’s direction, and I think some of his theoretical insights are sound, but I think it became clear from my fact-checking that he misrepresents the empirical evidence about Soviet planning quite badly. The truth is much more complicated than his take, and in some ways contradicts it. I certainly don’t see any evidence that the Soviets were “irrational” by Schelling’s lights. (Castro may have been that irrational, but on the other hand he may have urged a launch knowing that the Soviets wouldn’t agree… kind of like how the Poles lately keep asking NATO to do things against Russia that they know the US will veto, as a sort of brinksmanship virtue signaling.)
I conclude that Cummings is too credulous, and is willing to assert with great confidence claims from a semi-popular book without checking the references to confirm its key claims.
--
To dismiss his claims about missile defense is much more straightforward; I've been surprised to see some smart people convinced by them, which must just mean that these people haven't looked into the issue in any detail. The classic argument against missile defense is that it only incentivizes the construction of more warheads by the enemy. MIRVing (putting multiple warheads on one's offensive missiles) makes it cheap and easy to overwhelm any missile defense system by launching more warheads than the defense system has interceptors. If you build interceptors to destroy your enemy's warheads, they will simply (for much cheaper) build delivery systems for an equal number of additional warheads and you've gotten less than nowhere on the whole. I don’t see any way around this calculus unless the technology changes drastically.
The calculus is different against N Korea for the near future, since they can’t afford very many missiles and lack MIRV technology. That said, MIRV is a technology that took the US and the Soviets about 10 years to develop following their invention of intercontinental missiles. North Korea is working now to develop this technology ( https://en.wikipedia.org/wiki/Hwasong-17 ), so it's unlikely that missile defense will provide a feasible check against them for much more than a decade longer.
How do I choose a school for my children? My partner and I are open to public school, private school, homeschooling - but how do we evaluate our options? I'm looking for book recommendations, essay recommendations, general advice. I'm aware of Great Schools, but it seems like their metric boils down to state test scores and SAT test scores. Should I believe these metrics capture academic rigor? What's more, social and moral development are of comparable import to me - how do I evaluate a school's ability to foster - or at least not impede - those?
I'm a maths and physics tutor looking for some more students. I can teach to anyone below undergrad level, and for undergrad stuff I can teach: mechanics, QM, ODE/PDEs, chaos theory, stat mech, probability theory, logic, QFT, quantum computing basics, ML, RL and some deep learning stuff. Message me if you want my services, my rates are negotiable.
Hi all! I'm a Master's student and for my dissertation I'm doing a qualitative study on expert forecasters. If you're reading this, interested participating, and are in the top percentile on either Metaculus or GJO, or a Superforecaster, could you leave a comment or contact me via the mail connected to this account or through Twitter @meraxion? Thanks! If you want more information, read on;
The particular interview methodology I'll be using is called Applied Cognitive Task Analysis (ACTA), which is a mostly structured series of (3) interview techniques used to figure out how experts in a field view their task, and what cues and strategies they look out for that they think novices won't know about. The methodology is often used as a first step for further construction of training manuals and things like that. The 3 techniques are; the Task Diagram, the Knowledge Audit, and the Simulation Interview. The Task Diagram asks the expert to lay out the (up to six) high-level steps of their task. The Knowledge Audit digs into the cues and strategies experts use for understanding aspects of their task (and making decisions in task situations). The Simulation Interview looks at the expert's actions in situ while doing the task, asking questions about actions, situation assessment, critical cues, and potential errors, at each step (as identified in the Task Diagram).
The product of the is the Task Diagram, and a Cognitive Demands Table outlining what came up during the interview(s) as the important/difficult/demanding parts of the task.
Interviews would be conducted in a month's time (June, maybe late May).
If you think this doesn't sound too daft, please don't hesitate to contact me using one of the methods I linked to above, and/or sharing with others whom you think might be interested!
Can anyone recommend a (free, online) introduction to statistical methods in climatology.(especially statistical methods in paleoclimatology) which doesn't assume domain knowledge, but also doesn't sacrifice mathematical rigour? What is the CS229 of climatology?
how do you manage the acid base balance in your body? What supplements do you take? How do you increase that HCO3 in your blood (i don’t intend to produce an alkalosis)? I am certain that taking Magnesiumcitrate is not enough (as it is converted to bicarbonate in the kidneys, but it doesn’t mean that it necessarily increases serum HCO3). What else do you do to manage it? don’t call it “bogus science” because clearly it’s not. Drop some knowledge here, please.
I'm interested in having a child at some point, but fear that I simply won't be able to cope with severe sleep deprivation.
E. g., according to this study, UK parents get less than 5 hours of sleep per night for the first year: https://www.independent.co.uk/life-style/health-and-families/baby-sleeping-child-new-parents-night-newborn-parenting-sleep-pattern-a8406536.html
I literally *cannot imagine* enduring that kind of sleep deprivation. I feel utterly awful and am largely incapable of any productive work the next day if I get less than around 7 hours of sleep. And I find it basically impossible to nap during the day, no matter how exhausted I am.
Are there any parents here who had a similar lack of previous ability to cope with sleep deprivation? Are there any good studies about what parents can do to get their children to sleep through the night as soon as possible?
(I've read claims that bed-sharing helps--it just seems implausible to me that mothers in traditional high-fertility societies were enduring modern levels of new-parent sleep deprivation throughout their reproductive years.)
I’m three years into my sabbatical and <6 months before I get back into the market so could do with a little more cash. Thus I’m interested in tutelage. I’ve done pro-content-creation for three and a half years before the sabbatical (I’m 26, UK based) so can teach anyone much about the skill(s) of videography/cinematography/editing/photography. I’m also an adept meditator having achieved stage 7 shamatha (meditative quiescence, and there’s 10 stages), so could assist someone who wants to get up to speed with that.
If you’re interested I’ll send you my work (for the content), and fees I’ll do at £40ph but if you can’t afford it we can discuss a more suitable price for you.
I think such a marketing post is tolerated here, but if not let me know and I’ll take it down.
Any questions or whatever. Throw em’ at me.
Skimming the hivemind blogpost on Lacan, i am more and more convinced that lacan is, as someone else remarked elsewhere, a rorschach test - everybody read what it wants to read in it. Who guarantees me that the post is right and she has really truly understood what is going on?
Beside, i found the comparison with sapir-whorf telling, given how whorf himself was a master of motte and bailey statements such as "the hopi do not have time because they speak hopi"
Did some meditating on the Zvi/Hanson/etc. posts on blackmail.
Here's a question that came out of it:
Does there exist rationally-hidden information (i.e. it would be bad for the person hiding it were it to get out) that is positive-sum to disclose, that is not and *should not be even in theory* evidence of illegal behaviour?
To put it another way, does there exist an act X for which a) revealing X is bad for the X-doer, b) revealing X is overall good for society, c) X shouldn't be banned?
Congratulations to Hidden Open Aphorism contest winners timujin, Elisabeth Piper and Gašo, and thanks for all the pithy aphorisms and thoughtful comments!
الكلاب تنبح والقافلة تسير" (The dogs bark, but the caravan moves on) so I'm proud to announce the inaugural ACX Open Political Punsters contest, in tribute to the great comedian Volodimir Zelenskyy.
Prizes (well kudos) for the punniest political puns, with extra points for originality and brevity.
Your starter for 10:
- Why do anarchists only drink herbal tea? Because all proper tea is theft! (makes me so Proudhon)
Contest closes 8pm UTC Monday 4 May the force be with you and your time starts... now!
In the schrodinger's cat Experiment why isn’t the cat the non-quantum observer?
Palladium mag posted a joke article on April 1st that they're constructing a new techno-religion, that is better suited to handle modern life.
Are there any real attempts in that direction, by anyone?
When you tally up the ratings of the book review entries, it would be interesting to sort reviews not just by average rating, but also by variance. (Will we get to see average ratings for all or any of the reviews?)
Any mental health/psych types want to have a go at this one?
My wife: early 30s, hyper-conscientious, neurotic with history of intrusive thoughts. She's currently struggling with thinking back to a professional document she wrote and getting waves of anxiety about whether the information was absolutely correct and defensible in court (NB it is very unlikely to end up in court, and her job is not in law - the memo just had some legal implications). In general, her relentless searching for possible sources of fault and guilt, despite (or because of?) her life being very comfortable, makes her life hell.
She's considering which type of professional to talk to, and whether or not she should be regarding anti-anxiety meds as a last resort (due to side-effects etc.). She's seen a psychoanalyst once (ages ago) and was frustrated by their esoteric/lofty/impractical approach.
Any suggestions appreciated as to what type of professional to talk to and whether anti-anxiety meds are a slippery slope to another kind of hell.
Jaynes has convinced me that almost all hypothesis testing, as it is currently performed, is ill-founded.
He also clarified something that I've been coming to for a while: the goal of statistics is *always* prediction, and if your model can't predict anything, it's worthless. This has been my problem with string theory/SUSY, and it's come to the fore of my mind again because of the Lacan discussion. Who cares what your psychoanalysis theory is unless it can make predictions about behavior, so I can adjust my own behavior accordingly?
Speaking past one another because vomiting theories is good enough for a paycheck is such an indictment of modern science. Everyone know about the social science replication crisis, but the Null Hypothesis Significance Testing paradigm is fundamentally flawed, and we base all of hard science on it too! Particle Physics is recognized as being the most stringent applied statistical discipline, but 2 weeks ago a result (W boson mass) was published that is fundamentally incompatible with previous measurements, so one or both sets must be wrong. Even the most preeminent scientists are failing to be precise enough to weed out errors, and my fellow statisticians are happy to keep on teaching STAT 301 to engineers and scientists. What's gotta give?
Update on our ACX grant project:
I've learned a whole lot about real estate Mass Appraisal and gotten in touch with a bunch of really smart people with the relevant credentials, several of whom we've recruited onto our team. Apparently a whole bunch of new papers in the field were published just in the last 3 years, and it seems like there's a lot of room to move both the state of the art and the state of common practice forward.
What's become quickly apparent is that this project is mostly all about data. Luckily, we've been absolutely spoiled by the OpenPhilly dataset. They have more fine-grained data than you can shake a stick at, and the majority of our work is going to consist of cleaning and merging these data sets together, then using them to reproduce the best models from the last 3 years of papers. After that our plan is to publish a paper, publish the source code, and release all the data that's not encumbered by any licensing issues (which so far is basically all of it). We would also like to set up a public website to do "live ratio studies" where we will track ongoing sales and compare the values of those properties to what our model predicts, and take particular close attention to sales of vacant land. That way if our model sucks it will be super obvious and you can all make fun of us more easily.
I also did a general survey of the data landscape of the top 20 US cities. They're surprisingly good! Most cities nowadays have slick "open data" portals with all kinds of information just free to download. Philadelphia's is still looking like it's the absolute best, but quite a lot of the others are up there.
But the absolute hands-down WORST? Silicon Valley (San Jose / Santa Clara County), ironically:
https://twitter.com/larsiusprime/status/1520608678495367168
If you want data from them they're going to charge you $47,000, which sticks out like a sore thumb compared to everybody else's policies.
I mean, at least they *have* the data. But I can't fathom any reasonable scenario under which it costs that much to retrieve ~328 MB of data.
Has Scott written an article explaining the replication crisis from the ground up? Alternatively, is there a good, well-written introduction to the crisis from another author geared toward an intelligent reader who is wholly new to the subject? I’d like to recommend something to a friend on this topic.
How would the gold standard EA charities compare with just giving money to poor people on the street? Not talking San Francisco mentally ill people, just regular people on the street selling wares and begging for cash (as it happens all the time in my country (see comment below)).
No charity bureaucracy, complete certainty the other person is getting the money, people usually seem obviously deserving, makes me feel good etc. Probably stacks up pretty well, no? What do you think?
Has anyone here used thyroid hormone (T3 or T4) for depression, or other issues besides low thyroid?
I know it hasn’t done great in studies, but I developed severe hyperthyroidism (Graves disease) last year and it essentially cured my acute depression and showed me I had been experiencing dysthymia and the like for years. Treating the hyperthyroidism made the depression come back in full force, which is when I made the connection.
My depression is now well-controlled through a combo of Wellbutrin and remaining slightly hyperthyroid (with my endocrinologist’s blessing)
Are there AI alignment groups / companies that are remote or remote-friendly for engineers? It seems like all the major organizations are still exclusively in-office (primarily Bay, although it seems like Conjecture is in London).
Reading this: https://ourworldindata.org/economic-growth-since-1950
I came across this: "If you look at incomes today then you find that the income in the richest country in 1950 is very close to the average income of the average person in the world today ($14,570). Today the average person on the planet is as rich as the the average person in the richest country in 1950. And all those countries that have an income higher than the global average today are more prosperous than the US in 1950: Iran, Mexico, Bulgaria, … Have a look at that list."
I live in one of the third (second?) world countries mentioned and I can't help but think that this is the sort of thing that makes people distrust the upbeat narrative Our World In Data emphasizes. Does anyone else feel like this can't possibly be true in a "deeper" sense? My country is falling apart, panhandlers are endemic and basic functions of the state go unfulfilled systematically. I know how poor people live in this country and I just can't fathom their lot being anyhow comparable to that of poor Americans in the time most people instantly associate with prosperity. Am I being irrational about this? Does anyone think I have a point?