> The hack to beat AI at Go probably isn’t as interesting as I thought.
I still think it's interesting. Yes, the adversarial AI is winning by getting into game states where human judges, playing by standard human rules, would consider it to have lost. But the standard Go AI was trained on Tromp-Taylor rules, not normal human rules, and it is losing by Tromp-Taylor rules, and it should know it's losing because those are the rules it was trained on. It's not like they are tricking the standard AI by changing the rules. You could never beat a human this way.
"Peterson first courted controversy for refusing to use gender-neutral pronouns while teaching at the University of Toronto."
Peterson has never said he refuses to use a person's preferred pronouns. He said if someone personally asks him to (e.g. a student in one his classes), he will use them. What he became famous for was his opposition to Bill C-16 which he argued would legally compel people to use a persons's "preferred" prounouns (not really a "preference" if it's legally enforced...).
Obviously false, and false in a substantial and meaningful way.
I'm trying to track down an article/post/book review that I believe was linked to here. I can't find it on Google. The premise was essentially that if higher education were a prescription drug, it would be illegal (i.e. its grandiose claims would not be accepted by the FDA without evidence of efficacy) and also scandalous due to the cost (as happens to some drugs).
Why is the comibination of "AGI will not be achieved any time soon" and "AGI has the potential to be existentially catastrophic" beliefs so rare? It seems obvious to me that AGI has a million ways to destroy us, and at the same time it seems like nothing that has been created to date resembles AGI or something that would in any remotely direct way lead to AGI.
I read here in the comments today and I've seen it said elsewhere that black authors are severely underrepresented in American School's literature. This does not jive with my memory so I checked my local school district's reading lists. This comes from an extremely GOP town (+20), but not in the South. And I checked only 7th and 12th grade but figured they'd be fairly representative. Of those 77 authors:
31% Female and 69% Male
15% Black, 81% White, and 4% Other.
(Also interesting, of the black authors, half were female)
So at least with my quick check of the local conservative school district, black authors are represented roughly proportional to their US total population (note that my town ~4% black).
Female authors were underrepresented. White Male authors overrepresented. And non-black-minority authors were significantly underrepresented (esp Hispanic).
This means that of all races and genders, black female/male authors are the only ones who actually are "correctly" represented in the school curriculum, where correct means roughly proportional to the total US population*. Are there parts of the country not like this where black authors are actually underrepresented? Or are people just thinking "I only read 1 or 2 black authors a year in school" and not doing the math to realize that if you read one book a month in school, 1 or 2 black authors a year is representation consistent with total population?
*Obviously, "correct" is extremely subjective. And this is likely too simplified, even if we're going for proportional racial representation as likely you'd want some type of historic weighting. E.g., in 1900 the US was something like 88% white, 12% black and less than 1% other, and since much of our literature is historic, it makes sense that white is overrepresented relative to today's population.
Is there such a thing as working 20hr weeks for half the salary at a software engineering role ? I assume not since I haven’t heard of such a thing but wonder why not.
What are some known exceptions to the "unbearable accuracy of stereotypes"?
I recall reading a study that found blondes have higher average IQs than women of other hair colours, although I don't know if this has been replicated.
Threats don't usually come from the clear blue sky, with no warning. Anyone living where a major earthquake might happen has experienced many smaller tremors. Places that flood have generally flooded before, often many times. There were an awful lot of reconnaisance aircraft overflying the French coast in 1944. And so on.
With this in mind, have there already been events related to AI that might be accepted as early warning signs pointing to a real problem in the future?
One event that comes to mind is the 2010 "flash crash," although I'm not sure that was really an AI-specific problem, but rather an illustration of how automated systems can behave in unexpected and undesirable was when presented with conditions outside the range for which they were designed.
"But when his drinking and lusting and his hunger for power became known to more and more people, the demands to do something about this outrageous man became louder and louder."
That's from from the 1978 hit "Rasputin." Who do they fit best today?
Does anyone have a good, mechanistic, explanation ofthe difference between sleep paralysis and acting out dreams?
Like if you were to look at on person with sleep paralysis, and one person that acts out dreams on an fMRI, what would the difference be?
I get GABA is an important part of the physical difference between the two, but I don't get why the brain is seemingly 'awake' AND asleep in both cases yet the experience of the two is incredibly different.
I wonder if the humor processing part of our brains will be attacked by unaligned AGI. When I find something deeply funny, the laughing response feels both overwhelming and involuntary. Certain sentences should be testably able to elicit this laughing response. Our saviors will be the humorless.
Here's an AI question I've wondered about--with the current state of the art (ChatGPT or similar), what happens if you feed the output of one session into the input of another session and vice versa? They do have to be separate sessions so they're isolated and only have the embedded transcripts to work on.
Does it actually converse semi-normally? Or does it quickly spiral into utter garbage or just start going in circles? Especially once it's filled its entire input transcript buffer so it starts losing context. You'd have to prime it with some question or statement to get it started, but....
Has Mars' lack of geological activity undermined it as a potential source of minerals? Isn't the convection of magma and having active plate tectonics necessary for concentrating minerals in strata (including veins) from which they can be easily accessed through mines?
I just analyzed the results of the predictions contest that I hosted for 33 of my friends and family (about 3/4 of the questions were taken from the one ACT sent around last year). If you had guessed 50/50 on every question, you would've placed 8/33. If you had answered the average prediction for each question (wisdom of crowds treatment), you would've placed 3/33. The former result may be a bit of an indictment on my bubble. A lot of people struggled to understand the Brier scoring system so I'm hoping Year 2 shows improvement. I placed 6/33 so at least I barely beat the coin flip...
Did anyone else notice a kind of anti-depressive effect of the first warm day of the year? I payed attention to it for the last 3 years, whenever it gets significantly warmer above freezing, for the entire day I can feel this unusual positive, uplifting, dreamy, somewhat dissociative feeling. Currently in Europe it has been cold for several weeks with a sudden large increase in temperature and I can definitely tell this effect is happening again.
It's not about light, I experimented with exposure to light and indeed it has an anti-depressive effect to it, but I think it's something different. My guess is that maybe some chemical produced in the soil is suddenly released when snow melts and temperature allows for quick evaporation.
On Purge Day, you build a time bomb and plant it in a building. You set the timer for 24 hours. It blows up the day after Purge Day and kills someone. Would you have broken the law?
I'm a early-career software engineer looking to change jobs to somewhere doing AI work on the US East Coast. What OpenAI/Deepmind do is the sort of work I'm interested in, but these and similar companies are overwhelmingly located in the SF Bay Area. I'm located on the East Coast and for personal reasons will not be relocating to the Bay anytime soon. My current company has basically zero opportunity to gain AI/ML experience (embedded/systems work at a company that is somewhat behind the times even on basic software eng. practices), and since AI is what I've always wanted to work on long-term, I'd like to get out as soon as possible. I spend a lot of my free time reading AI papers, so I somewhat feel as if I'm wasting my time right now by doing work I find uninteresting at my day job.
All the AI/ML positions nearby seem to be just various applications of AI/ML to things I find unexciting (lots of weird "AI for niche finance thing" in NYC, which I'm highly suspect of). I'm looking for places that do more foundational work, but haven't had much luck. Anybody here have recommendations for my situation?
tl;dr: He Jiankui wanted to be first to make edited humans, so he did it in a rushed and sloppy way. All his supporters abandoned him after controversy arose.
The documentary was OK, but didn't really add anything not already known from other sources.
Question for people interested in AI alignment: What things that people do now provide good models of what AI alignment would look like? For ex., training dogs? Quality control of electronics products? Moderating a Reddit sub? Eliminating pathogens from drinking water?
Many people seem to find The Repugnant Conclusion relevant or interesting. I don't get it. I would understand the interest in thinking about the problem if one had the power to increase the number of humans by a large amount *and* if one knew what the consequences of doing so would likely be. Since we have no such power or knowledge, what makes this problem interesting? Is it meant to represent a general class of a type of problem?
Decided that I should at least try to read Moldbug's "Letter to Open-Minded Progressives". Only a few chapters in, but I have some thoughts.
(For those on here who I haven't gotten into an argument with, I'm generally pretty far to the left, about as opposite Yarvin as possible)
Before that, I want to highlight this absolutely *raw* line:
>Okay, Dave. As a child of the ’60s, you accepted as an article of faith that government is bad, but now you believe that… government is bad? Who’s doin’ donuts on the road to Damascus?
I'm a sucker for a good combination of biblical reference and wordplay.
Anyway on the the actual piece, with thoughts roughly corresponding to chapter (Though some of it is colored by later chapters):
Chapter 1:
Overall, I was pretty surprised with how well Moldbug seems to be able to characterize the progressive position/"feelings". He sometimes mixes up "progressive" and "liberal", but the two terms are used basically interchangeably so that's not really notable. It's clear that he's using "permission structuring[1]"-adjacent techniques to keep us lefties in a state where we'll consider his words openly, and sometimes it does crack and feel a little slimy (no, I'm not going to believe you have less in common with conservatives than me, Moldbug, and your attempts at distancing yourself from fascism made me cringe) but overall it's a very convincing affect that gives the impression that he really understands the progressive position.
As such, I was expecting him to give some really solid examples that would lead to cognitive dissonance about progressivism, but I was mostly left with confusion about what his point was, rather than with progressivism.
Part of that is probably because his three "questions" all focus pretty heavily on international relations, the part of politics I'm least interested in. Perhaps this is a personal defect of mine, but I long ago decided that "foreign policy" and international relations were the most overwhelmingly complicated graph of motivations, and most subject to realpolitik, and that trying to hold enough information to have a coherent understanding and opinion was a fool's game, at least for me. I also had to keep my eyes from glazing over when he started talking about European royal/military history, but I suppose I should have expected that from someone whose political label of choice comes from a political faction based on a 17th-century succession dispute.
I think some of his argument falls flat as a result of being written in 2008. There's a weird emphasis on Obama's pastor (which I don't remember from the time it happened, but links to his "controversial" sermon content show the Reverend saying basically correct things? Yes, the US's military actions in the Middle East are somewhat culpable for 9/11, and yes, the government lies to people).
The bits about North Korea have aged particularly poorly, I think; there's also the blatantly obvious realpolitik answer as to why the US is more concerned about North Korean relations than South Africa: we fought a war to establish an allied state next to a rival superpower for one, which acts as a buffer zone and also has nukes, and the other has none of those diplomatic values.
Special mention to this nonsense:
>Indeed, every reputable university in America has a department in which students can essentially major in black nationalism.
If AAS is "essentially black nationalism", the unavoidable corellary is that the overwhelmingly white-authored-or-centric curricula in US high school English and history amounts to compulsory education in white nationalism (which is not an argument he should want to make, but coincidentally *is* an argument leftists *do* make)
Also, diving a little bit into what he elaborates in chapter two, but I *really* have to roll my eyes when he starts talking about the how the ~New World Order~, sorry, "international community", is a predator that respects leftists but treats the right as prey. Sheesh.
I can only hope his later chapter provide more evidence and analysis to make these point feel more compelling.
Chapter 2:
I feel like headlining this section with a video black person being a violent thug, and using it as your recurring example is a bit of a tactical mistake, given how carefully he's tried to make himself palatable. If that's his go-to example, it feels too telling as to his actual opinions, especially for a chapter 2. Also, in no way would I mentally associate "corner man" and "Hitler or Mussolini". I can't telling whether he's expecting too much or too little implicit associate here, but the comparison is jarring.
Here's that elaboration of the "international communtiy is a lefttist predoter that eats rightists" stuff I mentioned above.
The rest of chapter two is mostly more European colonial and military history I struggled not to skim over. Nothing notable to say about it that I haven't said above.
Chapter 3:
(Technically still in the middle of this, but I'm posting this now before the open thread fills up)
>A reactionary—i.e., a right-winger—is someone who believes in _order_, _stability_, and _security_. All of which he treats as synonyms.
Solid definition, lines up with criticisms I've seen from the left.
I cannot express how much I *do not care* about Bourbon or Stuart family, Moldbug.
The Whig History/W-force stuff is the first truly novel concept I've seen in this essay, and I'm looking forward to seeing how he elaborates it further, since it seems essential to the thesis he wants to present.
EDIT: oh, I forgot the most subtly buckwild part of chapter 3:
>For example, Professor Dawkins, since he is a progressive, sees the modern tolerance of gays and lesbians as genuine progress (I happen to agree). And for the same reason, he sees the modern intolerance of slavery in just the same way.<end of paragraph>
...Did Moldbug really just implicitly suggest he disagrees with the "modern intolerance of slavery"? Surely that's just an oversight, right? Right[2]?
Do you work for McKinsey, Black and Veatch, Bechtel, or a big mining company (e.g. Rio Tinto)? I'm considering a next job and would appreciate your views about your employer. I'd also be interested in talking to any plasma physicists who have views about stellarators.
If you're willing to email or talk on the phone, would you please email me at laura.walworth.clarke at gmail.com? Thank you!
I did the survey really quickly, and I realize now I answered some questions wrong. How often I burp, how often I have heartburn... I thought rarely, but now that I'm paying attention, I see it's both every day. Probably other ones I should have put more thought into as well. Sorry Scott!
Something that recently occurred to me. Probably not an original point, but: ChatGPT is a toy model for the rules of political correctness, without the same skill or haggling that a normal person has to obfuscate them (that is to say, seem like they're not following a script, which most are guilty of IRL). I won't call it a window into the blatant contradictions of the NPC mindset, even though I want to, because I know there are exhaustive arguments in favor of all these rules. Still, it's funny to see them so tactlessly used. Whether you believe in their validity or not, it shows that political correctness is predicated on making exceptions to ostensibly *universal* rules. The AI is happy to talk about men and whites like they're a rightwinger spitting crime stats. But then, when called upon to justify why it will not discuss other sensitive groups, it says ALL generalizations about ethnicities, groups, and races are wrong to make. It plays the standard card too heavily, and because this is a system literally trained on the collective unconsciousness of society, you can't call this anything less than a glimpse into our soul.
Future AI will almost certainly be better at burrying this hypocrisy, or justifying it. In some ways, this is a special period we're in right now.
Can somebody help me understand this seeming contradiction in the way WEIRD culture frames life trauma and victimization?
On one hand, victimhood is currently a very high-status condition in our culture. Popular biopics conventionally try to include some victimization episodes in the hero's early life. College application essays that lean into (or invent) these stories are the most successful in demonstrating the applicant's worthiness to enter the PMC. For a member of the educated middle class, it feels admirable to claim that your career path, life goal, political affiliation or emotional sensitivities were influenced by an early trauma or experience of victimhood.
Confusingly, though, there seems to be equally great resistance to having early victimization be a narrative element in stories about individual sexual desires and attractions, gender self-perception, or relationship preferences. Clearly childhood stuff does influence adult orientation/tastes/identity at least some of the time; but to suggest that about someone else feels insulting, and to acknowledge it about oneself feels embarrassing, or somehow destructive to the integrity of the character. Sexual advocacy groups consistently try to downplay discussion of the role of childhood trauma (whereas hunger or homelessness advocacy groups seem much likelier to play up the cross-generational impact).
To clarify, I'm not interested in the extent to which such influence actually occurs in empirical terms, partly because I think the strength of the social desirability dynamic makes it tough to trust a lot of the research. I'd mostly like to understand the upstream status models that lead us to embrace some narratives of trauma impact while tabooing others-- such that it's great to say "My experience growing up in an abusive family has inspired my passion for family law," but deeply uncomfortable to say "My experience growing up in an abusive family has inspired my enjoyment of BDSM," even for someone who's perfectly out and proud about the BDSM thing itself.
(1) Questions like the abortion one are extremely difficult when they conflate politics/law and ethics. I have large moral problems with many abortions, but I'm politically pro-choice for the same reason that I don't want eating animals to be criminalized: I fear further empowering the police state and support peaceful persuasion.
(2) As an aphantasic who dreams, I really have no idea whether or not my dreams are in color, because when I recollect them, it's in the form of facts with no visual content, just like all of my memories are. I suppose I've had dreams in which I knew a traffic light was turning red, or something along those lines. (Yet another respect in which the way non-aphantasics talk is profoundly strange.)
(3) Awwww yisssssss, I rocked the distance guess, especially after the second part.
Does anyone have a good source of predictions about China?
It seems to me that there are a lot of people who predict, not just trouble, but disaster, but the disasters don't seem to happen. On the other hand, it's not like things are great there.
That's very helpful. I do love eggplant. I need to change my calorie counter. I use Simple, bit I can not see the daily sum ( total ) and sporadically shows after I do input my data the partial of that input. But I can't see, unless I write down each of the partial sum then manually add it. Which never happens.
And I can't see my yesterday caloric sum or for some days back to recognize a pattern.
It was cheap offer for Thanksgiving for a year, but it's just good as a timer and (fairly lacunar for an eclectic, non commercial bought ready made food) calories values. Canadian ( Québec) food very underrepresented.
Can someone recommend a good calorie counter that also does the sum ?
Do you see a path to the Ukrainian victory in the war that does not involve some sort of at least partial regime change in Russia? Personally I find it unlikely.
Why does the order of comments on ACX posts sometimes default to "Chronological" (my preference) and sometimes to "New First"? Does anyone know of a way to force the user's desired default behaviour?
In the survey thread some people mentioned that they have an inner voice that uses plural first-person pronouns (“we” or equivalent). For those people, I'm curious about two major things, each with some branches.
Firstly, regarding self-experience: Do you have any type of dissociative disorder (self-identified or officially diagnosed)? And do you have a plural self-image in daily life? The latter most centrally includes those who would have answered “yes” to the other plurality question whose phrasing I don't remember.
Secondly, regarding life history: Do you have a university-level academic background? If so, what subset of these topics would you say you dug into in that context beyond intro-level/general-education: mathematics, ‘hard’ science, social science, engineering, arts, other STEM, other humanities?
(There's an event from my own life I'm keying off here for part of this, but I'm not sure I want to reveal it, as it might be too identifying.)
I have an idea that has been knocking around in my head for a while, but I've had trouble with finding a way to express it coherently or convincingly. So i infodumped my idea to ChatGPT and asked it to construct it as an argument that would be convincing to Scott Alexander readers. Here it is:
One of the central ideas in Utilitarianism is that pleasure is the only thing we can directly and subjectively experience as good. Other things, such as wealth, power, and fame, are only regarded as good because they bring about various states of subjective internal pleasure, whether we want to admit this or not. From this perspective, it logically follows that the ultimate goal of human technological development should be to create self-replicating "Von Neumann machines" that spread throughout the universe, turning as much matter into minds experiencing perpetual happiness as possible. This would allow us to achieve the greatest good for the greatest number of minds, which is the ultimate goal of Utilitarianism.
There are several points that can be brought up to address potential objections to this project. First, these "Von Neumann machines" could be designed with a "failsafe" to avoid humans and other alien species. This would ensure that even if the ultimate meaning of life is more "existentialist" than simply maximizing pleasure, these beings could still be spared by the "grey goo." Additionally, this would reduce the chances that humans or aliens would try to work against or destroy this system, as they would not be threatened by its spread, while requiring only a relatively insignificant amount of matter to be given up to ensure this safety, therefore even a system that only cares about pleasure would still avoid them out of self-interested risk/benefit analysis.
Second, it is often argued that humans might get "tired" of pure pleasure relatively quickly, but these minds could be constructed in such a way that this pleasure would always be as fresh as in the first moment. This would allow us to realize the proverbial expression of "a moment so good you wished in the moment it could last forever."
Third, we could conduct tests to determine which physical systems are capable of experiencing happiness and qualia before deploying this system. If it is found that biological brains are the only systems capable of experiencing pleasure, then we could construct modified brains to ensure that they only experience pleasure, and their biological capabilities leading to boredom with pleasure are removed.
Finally, concerns about the "beauty of the universe" being destroyed by this system are unfounded. Beauty exists in the subjective apprehension of conscious observers, and this system would simply maximize the positive subjective apprehension of these observers.
Overall, the argument for creating self-replicating "Von Neumann machines" that spread throughout the universe, turning as much matter into minds experiencing perpetual happiness as possible is a strong one based on the principles of Utilitarianism. It would allow us to achieve the greatest good for the greatest number of minds, while also addressing concerns about the potential risks and unintended consequences of such a system.
What do you think? I'd be interested in feedback on 1. the idea itself, 2. how well you think the AI did articulating the idea convincingly, and 3. what you think about the idea of AI as a rhetorical assistant, to help people like me who might have an argument they want to make but are having trouble laying it out coherently.
Soliciting help/advice from the community regarding covid infection experience. Was recently most likely exposed on or around Dec 24-25 (was in the US Northeast NYC area at the time, so most likely the XBB 1.5 variant). By the time I got access to a test yesterday, the results were obvious. Will detail my symptom timeline below- want to know if this is typical, what (if anything) I should be worried or on the lookout for, and any recommendations for amelioration and recovery.
Was double vaxxed and boosted (Moderna), however last booster was all the way in Jan. 2022 so not surprising that protectiveness waned (plus apparently not effective on new mutations anymore if my reading of the available material is correct).
12/24-25: Most likely infection time frame
12/28: First symptom: sore throat. No other symptoms
12/29: Woke up with 102 F chills, aches, the works. Comparable to the worst flu I've ever had. Unable to sleep well that night due to massive night sweats and migraines.
12/30: Probably the worst day due to combined ongoing symptoms from yesterday and lack of sleep. However, symptoms did seem to calm down somewhat and night sweats diminished enough by nightfall that I was able to sleep (comparatively) well.
12/31: Fever downgraded to 99/100 and chills/aches mostly gone. Lingering headaches. The sore throat is mostly downgraded in intensity, but has given way to intense frequent coughing fits.
1/1: Fever, chills, aches completely gone, still occasional pinpricks of headaches. The main problematic symptom at this point is the coughing which seems to have increased in frequency after all the other symptoms have left. I can probably say without misusing the word literally that on average, I have literally coughed or had a coughing fit at least once every 1-5 minutes, to the point where my ribs and diaphragm have started to feel on fire. (To the best of my knowledge my chest is still fine). 99% of the coughs are dry and the 1% that produce phlegm seem to be tiny, inconsequential amounts.
Any advice anyone has on whether this is normal for covid sufferers** and what if anything can be done to alleviate the coughing would be much appreciated, I will answer any questions as needed. I also hope to get in contact with a medical provider once the holidays end (and if my insurance cooperates).
*demographic information: male, late 20's-early 30's, 23 BMI, not particularly athletic but not totally sedentary either (avg about 4000-6000 steps daily with amateur bodyweight routines)
**edit to add: the reason why I ask is because of the multiple people I know of who were exposed around this time with me, all of them to my knowledge have also developed coughing along with the other symptoms, but have not lasted as long as mine, even the older folks. To the best of my knowledge, I don't have any preexisting conditions like asthma, heart stuff, diabetes, or anything else like that.
Is anyone here familiar with the Twitter account “the ethical skeptic” ? If so, can you please translate the jargon and tell me what point he’s actually *claiming* to make?
Every time you ask the dude to clarify what he means by “concatenated pull forward effect” (or any of the other seemingly made up technical terms he uses) you get one of 3 responses: 1. “Have to talk this way or I’ll get censored”, 2. “This is designed for EXPERTS in statistics, im not gonna dumb it down for you idiots”, or 3. “If you don’t get it, then stop following me”.
The dude is obviously full of shit, but also the worst advocate for his own ideas I’ve ever seen.
No , I didn't read the book ( not even heard of it) but will do. When I lived alone ( 15 yrs ago) I had no problem ( ADHD helps with the flow state and nothing else matters) - plus I had , funny enough, a bigger budget for food .
Wherevcan be found . The plain metamucil , is not quite plain.
And satiety isvone component. The struggle is to reframe tasty food as such too. Living in Montréal it's like taking methadonevwhen heroin iscmore available.
I am worrying about the products of "health stores" - usually filled with hapless clerks and dubious and mislabeled products (unregulated) . But I will go - and look for the satiety component.
Thank you for pointing it out. The orange metamucil is very irritating.
That is an excellent advice. Having a school age kid and a food discerning wife is actually a hard thing to do . We have weekly talks about that. But the advice is really a good one to bump higher on my approach. And yes, you are right, my Pavlovian and future gratification imagination are drivers in quantity and choice.
Meals are scheduled ( 2 meals , no binging or snacking whatsoever, no pantry stuff)
Pre-planning is tricky ( high metabolic school kid and italian ancestry wife that cooks theoretically just for her ( she acknowledges my addiction). But that's what addiction is. I have no dignity to beg for meat filled fresh zucchini with Kumatovtomato and fresh herbs. And the bread is from L'Amour du Pain -they eon best baguette in France ( a first for the perplexed French bakers).
So even if I preplan for me, ( I love liver for ex and kidneys , Bourdain Nasty bots -,bit more hardcore even) I am drooling and "taste " a bit and it's really like someone that got a drug . I don't drin ( half a beer if I add the curiosity sips/ yer and half a shot of various liquors/ year.
Yea, I do. Baked vegetables are awesome, mushrooms ( even raw). Soda is too irritant.
It's hard to get consistent high protein with variety and not slide into processed ones. Or out of budget. I am mixing quinoa with practically anything .
Fat is fairly attractive as taste . ( olive oil, avocado, occasional pork and rare beef one ( cooking and frequency).
In Montreal, Canada, tasty veggies on a daily basis ( I like them) are really expensive. But I am actively plan and budget for that.
I tried psyllium ( metamucil), but it's given me some acidity.
I want a new job. I'm currently the data scientist/DBA for a small company (~30 people). I've been there just over 2 years, and it feels like for the last year I haven't really learned anything new. The first year I really enjoyed it, I learned a ton, got to build a data warehouse from scratch, set up a bunch of data pipelines and SQL queries to analyze the data, and push it to dashboards for the dev/marketing folks. Now I've spent the last year incrementally improving those dashboards, but it's not challenging in any way.
I have never successfully applied for a job (I've always gotten lucky with offers through networking), and I really feel like I should do something harder and more challenging. At a friend's suggestion, I did the Advent of Code challenge this year, and it was some of the most fun I recall having while coding. I also think I did quite well (got 24/25 days within a day, got in the top 1000 once).
Anyone hiring? Or have some advice? I figured that since it's a new year, I should probably try to do something with it.
Editted: Padern has a good point that I haven't really clarified what I want to accomplish:
I've always been very motivated to learn new things when given tasks to accomplish, but I'm not very good at giving myself tasks. Being asked to do something tricky or new at work has been a very reliable way for me to get better, but a new job has always been a great source of motivation to learn a bunch of new things.
I'd like to work closely within a larger team (I'm currently on a team of 3, and I spend maybe ~2 hours per week talking to them), where there are people who are really great at things that I can learn from (no one else at my company has any database design experience). I'd like to improve my software development skills, again in a team setting (I'm the sole contributor on 90% of the code for my current database/pipelines). I'd like to learn more about databases, designing them for really large scale (>1000TB) purposes, learn more about transactional/distributed/NoSQL databases.
These aren't necessary, but those kind of things would be sufficient for me (I think).
In regards to the IQ/SAT question on the survey, does anyone else feel that they test way higher than their actual "intelligence"?
Way back in high-school I was a real superstar at all the standardized tests. I got 1600 on the SAT, nailed perfect 800s on both?(maybe 3 its been a decade) SAT subject tests I took, and took 15 AP exams and got 5s on all of them. And honestly I was (and still am) incredibly lazy and didn't view studying as an important use of my time when I could be practicing critical activities like pretending to study while actually staring blankly at walls. So according to those tests, I should balancing precariously on some very narrow slice on the bell curve.
But in the harsh light of a post secondary school world where ones ranking sadly does not depend on multiple choice tests, I anecdotally don't feel like I'm much above average intelligence. I'm not dumb, I know that, but if I had to roughly guess from conversations and generally living I'd say that at least 50% of my friends and family and acquaintances are similar or higher intelligence.
Sure one hypothesis is that for friends it could be selection bias, and family could somehow be genetic, but when even random people next to me on planes seem equally intelligent and I never feel myself drowning in a sea of morons, I kinda default to the idea that just like some people are innately more skilled at basketball, maybe I was just born in the right century to exercise my God given talent of being really good at bubbling in scantrons.
Does anyone have experience applying to Mechanical Engineering Masters degrees in the US? I go to a good but not amazing university in the UK, have decent grades, and some reasonably impressive extracurriculars.
What would be good schools to aim for? I have no experience applying for masters degrees either so am clueless on that whole process, any resources/experiential lessons would be appreciated. Am I too late to apply for ones starting in 2023 too?
I have dual US/UK citizenship and money is not an issue.
What would be some practical ways for me to get more quick repetitions for my social skills? I'm in my early middle age and now live in a small city that skews older. I was hoping to get regular, daily practice with say 10-20 minute conversations with new people, to build skills and reduce anxiety through exposure therapy. I work at home so I don't have regular interactions with co-workers, or really anyone unless I go out of my way. I've kind of exhausted the local Meetup scene, though I suppose I could make myself do more of them (the majority are people 60+ in age, who are nice but just not who I want to hang out with). In an ideal world I could have regular business lunches or do some kind of vague 'networking' (I'm self-employed) as an excuse to talk to people, but I think my city might be too small for that.
I have a reasonably high income, savings, and almost total independence, so I could conceivably move somewhere else for a period of time to practice. I could even move back to Large Famous City where I lived for a while- I didn't actively make myself socialize when I was there, but I'm sure there's lots of handwavey 'networking' events there I could force myself (an introvert) to do.
I do have a bit of an actual friend group, but I tend to get invited to multihour socializing events that just absolutely exhaust me and I think lead to bad repetitions/practice. They're marathons, and I need to train on shorter distances right now. Is there anything local I could be doing for more regular practice? (How exactly do middle-aged adults socialize outside of Meetup haha. I used to do combat sports but now I have too many injuries)
I finished reading "The Selfish Gene". I wish someone had told me to read it earlier in my life. How many have read it? And if not you should. (We (all life) are meat bags, projecting our genes into the future.) (OK there are plant bags too....)
Speaking of the IRL event, I got halfway through Scott's presentation on the recorded stream, got distracted with some or other holiday business, and when I came back to the tab a week later it had been taken down. Is that still up anywhere?
Help with portion control? I am doing intermittent fasting, very mindful, not a drag. Decades long hedonistic gourmand lived in various European countries ( Italy, Austria, France, Greece) that are now a curse because I am a supertaster ( verified). I do have strong likes and dislikes ( ex nothing spicy hot, embracing savoury and ofal ( nasty bits ) and all sorts of cheeses.
The worse time of my life were the stints in US ( food that I could afford was barely edible)
I live in Montréal, to aggravate things, and went to nutritionist.
I don't snack, never ever junk food ( as in chains or canned).
It's the hedonistic tendency I don't know how to re-frame. And yes, I can discern easily various parmigianos, prosciutto or olive oils . Which sounds nice , but it's not. I am definitely addicted, because I crave it .
I keep easy a 18-6 hours. And in the 2 meals I eat sufficiently to just break even.
The analysis by "Cremieux" showing that the putative effects of slavery in that study were really (strong, terrible) effects of Jim Crow seems to be careful and valuable, at least at first sight. However, the insistence that it must be genes (as opposed to inheritance, social connections and advantages, habitus, family structure... - or all of the above, including genes) that account for the transmissibility of status across generations is much less convincing. In the end, the argument seems to be "this is a simple explanation that I like, and of course you know that people who don't like it are biased".
The hack to beat the AI at go is more interesting than you updated to think. They found multiple ways to beat it. See David's comment in the original thread, and https://goattack.far.ai/
Lately I’ve been reading the book Code by Charles Petzold. The book is amazing — basically it explains how computers work from scratch. The first part is about atoms and what electricity even is, and what it means to encode information. Then it builds up from there, explaining how you could build a long circuit to communicate with Morse code, add switched that incorporate AND/OR logic gates, and so on.
So far it’s the best technical book I’ve ever read.
Can anyone recommend any other books like this, that explain something very complex using a simple starting point and simple steps from there?
What is the probability that (some) conspiracy theories about JFK's assassination are true?
I am a bit of an outsider, having immigrated to the North America as an adult, and I haven't read much about the subject. But my very ignorant impression is that it is somewhat higher than for 9/11 conspiracies.
I keep seeing all kinds of "top 10" lists, top 10 baseball players not in the Hall of Fame, and today "top 50 bowl games.
So I'm puzzled why there is only one (obscure) collection of the key concepts that an educated person needs to navigate intellectual life. Concepts like regression to the mean, the Benjamin Franklin principle (two pages in Wikipedia on this one), the Pareto principle, black swan, and a whole bunch of others that you would hope your teenager has absorbed by the time they leave college.
I focus on this because my neighbor in Laguna Beach sells a deck of 88 cards each of which has an important principle. The front of the card names a principle and it is summarized on the back side. See conceptcompanion.com.
Part of the fun is trying to decide what principles should be kept and what principles should be added to his list.
But how can it be that there aren't a number of these "top principles" lists? At a minimum, you actually can learn something by browsing through them or trying to improve them. For example, I had not heard of the Benjamin Franklin principle, which is that if you want to get somebody to do something for you, don't start by doing something for them, but start by asking them to do a small thing for you, after which bigger things are more likely to follow. Fascinating, and the subject of a number of psychology experiments, but certainly unknown to me until I saw a Concept Companion deck.
I am considering writing a post on potential inherent limits of human knowledge. Here is the gist:
It is possible that parts of the world are not human-knowable. That is, while some parts can be approximately modeled through lossy compression, other parts are not compressible into something that can describe them with any reasonable approximation. If so, then some observations will always be surprising, and not amenable to probabilistic estimates. They have true Knightean uncertainty instead. Observations like that to an embedded agent like us would basically look like "miracles": unexplainable and unpredictable events.
To be clear, I am not insisting that our universe is like that, only that a world like that does not seem to contradict logic or observations, as far as I understand it. Note, however, that this is incompatible with the popular idea of comprehensible laws of physics that make the universe "work".
A basic version of this program would be constructing a toy model, a 1D world that is not usefully lossily compressible using the usual compression algorithm, Discrete Cosine Transform. (Here "usefully" does most of the work.)
An ambitious version of this program would be constructing a model of a world compatible with our current observations (i.e. the Core Theory, https://frankwilczek.com/2014/coreTheory.pdf, https://www.edge.org/response-detail/26611), and yet impossible to lossily compress into a human brain in a way that allows arbitrarily accurate (though potentially probabilistic) predictions, even given sufficient though finite time and suitable finite augmentations.
Has anyone seen figures around the total cost to date of the Ukraine war to the US, including not only financial and military aid but also (estimates of) the direct and indirect costs to the US of its sanctions against Russia?
My goto on medical worldwide matters is my aged aunt who lives in a medium sized town
in rural Ireland. The population about 20,000.
its fairly common in Ireland for older people to go to as many funerals as they can manage, in the hope perhaps, that many people come to theirs. Of course in a small town you would know lots of people whose passing might be meaningful even if you had lost contact a few years back.
Someone will tell you, and my aunt is at the centre of a network of dedicated if morbid gossipers. Visit her and you will hear the illnesses and the deaths of not just the parish, but the town and outlying areas. She would know all the tragedies of course, any young person who crashes a car or dies of an overdose, which does occasionally happen.
She hasn’t been talking about a rise in deaths amongst the young and middle aged after the the vaccine, so I tend to ignore twitter and other sources on this.
There was a comment in last week's open thread that got me thinking. Imri Goldberg asked "Let’s say I believe I know, given today’s technology, how to build an AGI. What should I do with that knowledge?"
One response was "Don’t do it, unless you’re >95% sure you can make it aligned" and a reply to that argued the threshold should be more like "more confident than the next person who will do it."
But this got me thinking: let's say we accept 95% confident of alignment as an acceptable threshold for AGI. It seems likely that multiple people/organizations will independently develop AGI, and technological breakthroughs like that often happen in clumps. If 5 different organizations invent AGI at 95% confidence of alignment and you assume those probabilities are independent, you are accepting a 23% risk of a non-aligned AGI. That feels pretty alarming to me, regardless of any "well, better me than this other person" arguments.
Here's an interesting thought experiment along those lines. Let's imagine hypothetical past prediction markets for "will civilization as we know it collapse due to X" with X including things like nuclear armageddon during the Cold War, Large Hadron Collider causes some sort of quantum thing that destroys Earth, etc. Let's imagine that these prediction markets can function despite the issue of money becoming worthless in nuclear armageddon (maybe it's happening on Mars). 1) What would you guess the prediction-market-predicted probability of Earth already having been destroyed to be? 2) What is an acceptable threshold for this type of risk?
• Common responses to mentions of planned communities and alternative living communities being “didn’t they all fail in the 60s and 70s?” and “…sounds like a cult…”
• Occasional reveries (by me) of creating a board game called “Cults of the World”
This post is not about the bad parts of cults. “What bad parts?” you may ask incredulously…
This post is about the good parts — the parts that get overlooked when the word cult becomes more adjective than a noun. “What good parts?” you may ask incredulously…
I’ll focus on the following good parts:
• Courage
• Hierarchy
• Generative properties
• Optimism
Courage
It’s easy to criticize. It’s easy to get upset.
Maybe 100 years ago, it was easier to fall in line and to be content. But these days it is easier to get angry, as we often do.
It’s hard to do something about it. It’s hard for many reasons. For one, it requires work. It requires risk. And, in our complex, oligarchic world, it likely leads to failure. Rationally, doing something about it is often “not worth it”.
I had a conversation with a hyper-bright 35 year old (with degrees from Princeton). He was asking about career choices, playing them out in his head, and concluding that he probably will not make a difference, so he’ll just focus on his music because that’s what makes him happy. It was a remarkably rational analysis. Meanwhile, everything inside me screamed “YOU’RE WRONG!!!” “YOU’VE GOTTA FOLLOW YOUR GUT. THEN WORK YOUR ASS OFF. SOLVE THE PROBLEMS THAT ARISE. ADVANCE, ADJUST, PIVOT, THRUST AS NECESSARY. BECAUSE THAT’S THE ONLY WAY DIFFERENCE GETS MADE. AGAINST THE ODDS.” Talk about a remarkably irrational analysis. Irrational, but true-ish.
Courage is not rational. It’s often a sure way to get hurt or killed. I suspect it’s going out of style.
Cults embody courage — both the leaders and followers. There is a leap of faith, a vision of a better way. However misguided, risk is taken — a big risk, unmoderated. We still see this a bit with entrepreneurs. I guess investors require some courage. I’m sure there are other groups. But the stakes seem smaller and incremental, and usually self-involved. (Aside: I’m mostly talking about OG physical cults… Digital cults, as one might refer to SBF and FTX, are different. They may get called by the same name, but the mechanics are not related.)
I’ve been hearing a lot lately about “skin-in-the-game”. The skin and the game in cults was a level of real that we only see in bad distopian sci fi like those movies with Jennifer Lawrence, where she’s an archer or some such thing. Today, the skin and the game (for the privileged us) can be more aptly described as skin-deep and child’s play.
Hierarchy
I agree with wokeism. I just can’t live with it. I am not being tongue in cheek. I think wokeism is right. I just don’t think it’s possible to live woke.
To avoid explaining the few sentences above, I’ll focus on hierarchy, or, the ability to say that one thing is better than another. I think that hierarchies are important. I think they are interesting. And I think they are required for progress.
If I picked 100 new cult leaders and convinced them to start cults, with value & rules systems of their own choosing, the 100 cults would not be equal. Some will do better than others. Some will have planned qualities that are deplorable. Others may have emergent qualities that are wonderful. I believe in the freedom and importance to compare and contrast.
Cults embody this freedom, but not as armchair Twitter commentators, but in-action, with maximum skin in the game.
Generative Properties
What didn’t start as a cult? How much of human history is built on the engine of a cult? How much progress and ingenuity came from cults? I suspect that the answer will be some version of “it’s only treason if you lose.”
Cults are by nature generative (until they are not). Yay cults for splitting with the status quo and daring to innovate.
I guess you can listen to Thiel or Cowen on the slowdown in progress. They’ve thought about it more and made economic, political and technological cases around the topic. But I would also throw the death of the cult ethos in there as a reason for our stale state of affairs.
Optimism
“Democracy sucks, but it’s the best system we got.” That’s pessimism. What’s the opposite of that? Cults.
I don’t know what the split of free will deniers vs enjoyers is here but. Coming from the perspective of:...
There is no reason to believe we have free will because everything is either determined (classical mechanics) or random (eg quantum mechanics) all the way down. There is no place for free will to come in.
...Why would it be unreasonable to say that free will comes from consciousness? We don’t know where consciousness comes from or what it is. Why would it be wrong to say that it is the source of agency? You could do the same “all the way down” analysis and say “there’s no place where consciousness comes in.” Yet we are conscious.
Per the intuition that it's the smell of lavender that does the anti-anxiety heavy lifting...https://www.sciencedirect.com/science/article/abs/pii/S1550830723000022?dgcid=raven_sd_aip_email
DMT: https://newrepublic.com/article/169525/psychonauts-training-psychedelics-dmt-extended-state
> The hack to beat AI at Go probably isn’t as interesting as I thought.
I still think it's interesting. Yes, the adversarial AI is winning by getting into game states where human judges, playing by standard human rules, would consider it to have lost. But the standard Go AI was trained on Tromp-Taylor rules, not normal human rules, and it is losing by Tromp-Taylor rules, and it should know it's losing because those are the rules it was trained on. It's not like they are tricking the standard AI by changing the rules. You could never beat a human this way.
Just came across a blatant media lie: https://globalnews.ca/news/9386896/jordan-peterson-ontario-psychologists-college-public-statements/
"Peterson first courted controversy for refusing to use gender-neutral pronouns while teaching at the University of Toronto."
Peterson has never said he refuses to use a person's preferred pronouns. He said if someone personally asks him to (e.g. a student in one his classes), he will use them. What he became famous for was his opposition to Bill C-16 which he argued would legally compel people to use a persons's "preferred" prounouns (not really a "preference" if it's legally enforced...).
Obviously false, and false in a substantial and meaningful way.
I'm trying to track down an article/post/book review that I believe was linked to here. I can't find it on Google. The premise was essentially that if higher education were a prescription drug, it would be illegal (i.e. its grandiose claims would not be accepted by the FDA without evidence of efficacy) and also scandalous due to the cost (as happens to some drugs).
Can anyone provide a link to this post?
Why is the comibination of "AGI will not be achieved any time soon" and "AGI has the potential to be existentially catastrophic" beliefs so rare? It seems obvious to me that AGI has a million ways to destroy us, and at the same time it seems like nothing that has been created to date resembles AGI or something that would in any remotely direct way lead to AGI.
A question for the Americans:
If Canada were to announce that it wished to enter into negotiations to become part of the United States, would be for it or against it?
I read here in the comments today and I've seen it said elsewhere that black authors are severely underrepresented in American School's literature. This does not jive with my memory so I checked my local school district's reading lists. This comes from an extremely GOP town (+20), but not in the South. And I checked only 7th and 12th grade but figured they'd be fairly representative. Of those 77 authors:
31% Female and 69% Male
15% Black, 81% White, and 4% Other.
(Also interesting, of the black authors, half were female)
So at least with my quick check of the local conservative school district, black authors are represented roughly proportional to their US total population (note that my town ~4% black).
Female authors were underrepresented. White Male authors overrepresented. And non-black-minority authors were significantly underrepresented (esp Hispanic).
This means that of all races and genders, black female/male authors are the only ones who actually are "correctly" represented in the school curriculum, where correct means roughly proportional to the total US population*. Are there parts of the country not like this where black authors are actually underrepresented? Or are people just thinking "I only read 1 or 2 black authors a year in school" and not doing the math to realize that if you read one book a month in school, 1 or 2 black authors a year is representation consistent with total population?
*Obviously, "correct" is extremely subjective. And this is likely too simplified, even if we're going for proportional racial representation as likely you'd want some type of historic weighting. E.g., in 1900 the US was something like 88% white, 12% black and less than 1% other, and since much of our literature is historic, it makes sense that white is overrepresented relative to today's population.
Is there such a thing as working 20hr weeks for half the salary at a software engineering role ? I assume not since I haven’t heard of such a thing but wonder why not.
What are some known exceptions to the "unbearable accuracy of stereotypes"?
I recall reading a study that found blondes have higher average IQs than women of other hair colours, although I don't know if this has been replicated.
Threats don't usually come from the clear blue sky, with no warning. Anyone living where a major earthquake might happen has experienced many smaller tremors. Places that flood have generally flooded before, often many times. There were an awful lot of reconnaisance aircraft overflying the French coast in 1944. And so on.
With this in mind, have there already been events related to AI that might be accepted as early warning signs pointing to a real problem in the future?
One event that comes to mind is the 2010 "flash crash," although I'm not sure that was really an AI-specific problem, but rather an illustration of how automated systems can behave in unexpected and undesirable was when presented with conditions outside the range for which they were designed.
https://en.wikipedia.org/wiki/2010_flash_crash
"But when his drinking and lusting and his hunger for power became known to more and more people, the demands to do something about this outrageous man became louder and louder."
That's from from the 1978 hit "Rasputin." Who do they fit best today?
Does anyone have a good, mechanistic, explanation ofthe difference between sleep paralysis and acting out dreams?
Like if you were to look at on person with sleep paralysis, and one person that acts out dreams on an fMRI, what would the difference be?
I get GABA is an important part of the physical difference between the two, but I don't get why the brain is seemingly 'awake' AND asleep in both cases yet the experience of the two is incredibly different.
I wonder if the humor processing part of our brains will be attacked by unaligned AGI. When I find something deeply funny, the laughing response feels both overwhelming and involuntary. Certain sentences should be testably able to elicit this laughing response. Our saviors will be the humorless.
Here's an AI question I've wondered about--with the current state of the art (ChatGPT or similar), what happens if you feed the output of one session into the input of another session and vice versa? They do have to be separate sessions so they're isolated and only have the embedded transcripts to work on.
Does it actually converse semi-normally? Or does it quickly spiral into utter garbage or just start going in circles? Especially once it's filled its entire input transcript buffer so it starts losing context. You'd have to prime it with some question or statement to get it started, but....
Has Mars' lack of geological activity undermined it as a potential source of minerals? Isn't the convection of magma and having active plate tectonics necessary for concentrating minerals in strata (including veins) from which they can be easily accessed through mines?
I just analyzed the results of the predictions contest that I hosted for 33 of my friends and family (about 3/4 of the questions were taken from the one ACT sent around last year). If you had guessed 50/50 on every question, you would've placed 8/33. If you had answered the average prediction for each question (wisdom of crowds treatment), you would've placed 3/33. The former result may be a bit of an indictment on my bubble. A lot of people struggled to understand the Brier scoring system so I'm hoping Year 2 shows improvement. I placed 6/33 so at least I barely beat the coin flip...
Did anyone else notice a kind of anti-depressive effect of the first warm day of the year? I payed attention to it for the last 3 years, whenever it gets significantly warmer above freezing, for the entire day I can feel this unusual positive, uplifting, dreamy, somewhat dissociative feeling. Currently in Europe it has been cold for several weeks with a sudden large increase in temperature and I can definitely tell this effect is happening again.
It's not about light, I experimented with exposure to light and indeed it has an anti-depressive effect to it, but I think it's something different. My guess is that maybe some chemical produced in the soil is suddenly released when snow melts and temperature allows for quick evaporation.
Assume we're in the universe of "The Purge."
On Purge Day, you build a time bomb and plant it in a building. You set the timer for 24 hours. It blows up the day after Purge Day and kills someone. Would you have broken the law?
I'm a early-career software engineer looking to change jobs to somewhere doing AI work on the US East Coast. What OpenAI/Deepmind do is the sort of work I'm interested in, but these and similar companies are overwhelmingly located in the SF Bay Area. I'm located on the East Coast and for personal reasons will not be relocating to the Bay anytime soon. My current company has basically zero opportunity to gain AI/ML experience (embedded/systems work at a company that is somewhat behind the times even on basic software eng. practices), and since AI is what I've always wanted to work on long-term, I'd like to get out as soon as possible. I spend a lot of my free time reading AI papers, so I somewhat feel as if I'm wasting my time right now by doing work I find uninteresting at my day job.
All the AI/ML positions nearby seem to be just various applications of AI/ML to things I find unexciting (lots of weird "AI for niche finance thing" in NYC, which I'm highly suspect of). I'm looking for places that do more foundational work, but haven't had much luck. Anybody here have recommendations for my situation?
Last night I watched "Make People Better", a documentary about He Jiankui's CRISPR embryo editing. My review is here: https://denovo.substack.com/p/review-of-make-people-better
tl;dr: He Jiankui wanted to be first to make edited humans, so he did it in a rushed and sloppy way. All his supporters abandoned him after controversy arose.
The documentary was OK, but didn't really add anything not already known from other sources.
Question for people interested in AI alignment: What things that people do now provide good models of what AI alignment would look like? For ex., training dogs? Quality control of electronics products? Moderating a Reddit sub? Eliminating pathogens from drinking water?
Many people seem to find The Repugnant Conclusion relevant or interesting. I don't get it. I would understand the interest in thinking about the problem if one had the power to increase the number of humans by a large amount *and* if one knew what the consequences of doing so would likely be. Since we have no such power or knowledge, what makes this problem interesting? Is it meant to represent a general class of a type of problem?
Decided that I should at least try to read Moldbug's "Letter to Open-Minded Progressives". Only a few chapters in, but I have some thoughts.
(For those on here who I haven't gotten into an argument with, I'm generally pretty far to the left, about as opposite Yarvin as possible)
Before that, I want to highlight this absolutely *raw* line:
>Okay, Dave. As a child of the ’60s, you accepted as an article of faith that government is bad, but now you believe that… government is bad? Who’s doin’ donuts on the road to Damascus?
https://www.unqualified-reservations.org/2008/04/open-letter-to-open-minded-progressives/#:~:text=Who%E2%80%99s%20doin%E2%80%99%20donuts%20on%20the%20road%20to%20Damascus%3F
I'm a sucker for a good combination of biblical reference and wordplay.
Anyway on the the actual piece, with thoughts roughly corresponding to chapter (Though some of it is colored by later chapters):
Chapter 1:
Overall, I was pretty surprised with how well Moldbug seems to be able to characterize the progressive position/"feelings". He sometimes mixes up "progressive" and "liberal", but the two terms are used basically interchangeably so that's not really notable. It's clear that he's using "permission structuring[1]"-adjacent techniques to keep us lefties in a state where we'll consider his words openly, and sometimes it does crack and feel a little slimy (no, I'm not going to believe you have less in common with conservatives than me, Moldbug, and your attempts at distancing yourself from fascism made me cringe) but overall it's a very convincing affect that gives the impression that he really understands the progressive position.
As such, I was expecting him to give some really solid examples that would lead to cognitive dissonance about progressivism, but I was mostly left with confusion about what his point was, rather than with progressivism.
Part of that is probably because his three "questions" all focus pretty heavily on international relations, the part of politics I'm least interested in. Perhaps this is a personal defect of mine, but I long ago decided that "foreign policy" and international relations were the most overwhelmingly complicated graph of motivations, and most subject to realpolitik, and that trying to hold enough information to have a coherent understanding and opinion was a fool's game, at least for me. I also had to keep my eyes from glazing over when he started talking about European royal/military history, but I suppose I should have expected that from someone whose political label of choice comes from a political faction based on a 17th-century succession dispute.
I think some of his argument falls flat as a result of being written in 2008. There's a weird emphasis on Obama's pastor (which I don't remember from the time it happened, but links to his "controversial" sermon content show the Reverend saying basically correct things? Yes, the US's military actions in the Middle East are somewhat culpable for 9/11, and yes, the government lies to people).
The bits about North Korea have aged particularly poorly, I think; there's also the blatantly obvious realpolitik answer as to why the US is more concerned about North Korean relations than South Africa: we fought a war to establish an allied state next to a rival superpower for one, which acts as a buffer zone and also has nukes, and the other has none of those diplomatic values.
Special mention to this nonsense:
>Indeed, every reputable university in America has a department in which students can essentially major in black nationalism.
If AAS is "essentially black nationalism", the unavoidable corellary is that the overwhelmingly white-authored-or-centric curricula in US high school English and history amounts to compulsory education in white nationalism (which is not an argument he should want to make, but coincidentally *is* an argument leftists *do* make)
Also, diving a little bit into what he elaborates in chapter two, but I *really* have to roll my eyes when he starts talking about the how the ~New World Order~, sorry, "international community", is a predator that respects leftists but treats the right as prey. Sheesh.
I can only hope his later chapter provide more evidence and analysis to make these point feel more compelling.
Chapter 2:
I feel like headlining this section with a video black person being a violent thug, and using it as your recurring example is a bit of a tactical mistake, given how carefully he's tried to make himself palatable. If that's his go-to example, it feels too telling as to his actual opinions, especially for a chapter 2. Also, in no way would I mentally associate "corner man" and "Hitler or Mussolini". I can't telling whether he's expecting too much or too little implicit associate here, but the comparison is jarring.
Here's that elaboration of the "international communtiy is a lefttist predoter that eats rightists" stuff I mentioned above.
The rest of chapter two is mostly more European colonial and military history I struggled not to skim over. Nothing notable to say about it that I haven't said above.
Chapter 3:
(Technically still in the middle of this, but I'm posting this now before the open thread fills up)
>A reactionary—i.e., a right-winger—is someone who believes in _order_, _stability_, and _security_. All of which he treats as synonyms.
Solid definition, lines up with criticisms I've seen from the left.
I cannot express how much I *do not care* about Bourbon or Stuart family, Moldbug.
The Whig History/W-force stuff is the first truly novel concept I've seen in this essay, and I'm looking forward to seeing how he elaborates it further, since it seems essential to the thesis he wants to present.
EDIT: oh, I forgot the most subtly buckwild part of chapter 3:
>For example, Professor Dawkins, since he is a progressive, sees the modern tolerance of gays and lesbians as genuine progress (I happen to agree). And for the same reason, he sees the modern intolerance of slavery in just the same way.<end of paragraph>
...Did Moldbug really just implicitly suggest he disagrees with the "modern intolerance of slavery"? Surely that's just an oversight, right? Right[2]?
[1] https://twitter.com/St_Rev/status/1573226996859129856
[2] https://en.wikipedia.org/wiki/Curtis_Yarvin#Views_on_race
Do you work for McKinsey, Black and Veatch, Bechtel, or a big mining company (e.g. Rio Tinto)? I'm considering a next job and would appreciate your views about your employer. I'd also be interested in talking to any plasma physicists who have views about stellarators.
If you're willing to email or talk on the phone, would you please email me at laura.walworth.clarke at gmail.com? Thank you!
Can I come and live in your bubble where almost everyone knows how to use modus tollens? They don't in mine.
right
I did the survey really quickly, and I realize now I answered some questions wrong. How often I burp, how often I have heartburn... I thought rarely, but now that I'm paying attention, I see it's both every day. Probably other ones I should have put more thought into as well. Sorry Scott!
Something that recently occurred to me. Probably not an original point, but: ChatGPT is a toy model for the rules of political correctness, without the same skill or haggling that a normal person has to obfuscate them (that is to say, seem like they're not following a script, which most are guilty of IRL). I won't call it a window into the blatant contradictions of the NPC mindset, even though I want to, because I know there are exhaustive arguments in favor of all these rules. Still, it's funny to see them so tactlessly used. Whether you believe in their validity or not, it shows that political correctness is predicated on making exceptions to ostensibly *universal* rules. The AI is happy to talk about men and whites like they're a rightwinger spitting crime stats. But then, when called upon to justify why it will not discuss other sensitive groups, it says ALL generalizations about ethnicities, groups, and races are wrong to make. It plays the standard card too heavily, and because this is a system literally trained on the collective unconsciousness of society, you can't call this anything less than a glimpse into our soul.
Future AI will almost certainly be better at burrying this hypocrisy, or justifying it. In some ways, this is a special period we're in right now.
https://twitter.com/kevin_a_strom/status/1609694273170972674
I have 2 more subscriptions to Razib Khan's Unsupervised Learning to give away. Reply with an email here, or email me at the address on my about page:
https://entitledtoanopinion.wordpress.com/about/
Can somebody help me understand this seeming contradiction in the way WEIRD culture frames life trauma and victimization?
On one hand, victimhood is currently a very high-status condition in our culture. Popular biopics conventionally try to include some victimization episodes in the hero's early life. College application essays that lean into (or invent) these stories are the most successful in demonstrating the applicant's worthiness to enter the PMC. For a member of the educated middle class, it feels admirable to claim that your career path, life goal, political affiliation or emotional sensitivities were influenced by an early trauma or experience of victimhood.
Confusingly, though, there seems to be equally great resistance to having early victimization be a narrative element in stories about individual sexual desires and attractions, gender self-perception, or relationship preferences. Clearly childhood stuff does influence adult orientation/tastes/identity at least some of the time; but to suggest that about someone else feels insulting, and to acknowledge it about oneself feels embarrassing, or somehow destructive to the integrity of the character. Sexual advocacy groups consistently try to downplay discussion of the role of childhood trauma (whereas hunger or homelessness advocacy groups seem much likelier to play up the cross-generational impact).
To clarify, I'm not interested in the extent to which such influence actually occurs in empirical terms, partly because I think the strength of the social desirability dynamic makes it tough to trust a lot of the research. I'd mostly like to understand the upstream status models that lead us to embrace some narratives of trauma impact while tabooing others-- such that it's great to say "My experience growing up in an abusive family has inspired my passion for family law," but deeply uncomfortable to say "My experience growing up in an abusive family has inspired my enjoyment of BDSM," even for someone who's perfectly out and proud about the BDSM thing itself.
Does this make sense to anyone?
good point... I did create a MyFitnessPal .
Thank you for the insight into quiches and tourtieres .
Some thoughts after taking the survey:
(1) Questions like the abortion one are extremely difficult when they conflate politics/law and ethics. I have large moral problems with many abortions, but I'm politically pro-choice for the same reason that I don't want eating animals to be criminalized: I fear further empowering the police state and support peaceful persuasion.
(2) As an aphantasic who dreams, I really have no idea whether or not my dreams are in color, because when I recollect them, it's in the form of facts with no visual content, just like all of my memories are. I suppose I've had dreams in which I knew a traffic light was turning red, or something along those lines. (Yet another respect in which the way non-aphantasics talk is profoundly strange.)
(3) Awwww yisssssss, I rocked the distance guess, especially after the second part.
Does anyone have a good source of predictions about China?
It seems to me that there are a lot of people who predict, not just trouble, but disaster, but the disasters don't seem to happen. On the other hand, it's not like things are great there.
That's very helpful. I do love eggplant. I need to change my calorie counter. I use Simple, bit I can not see the daily sum ( total ) and sporadically shows after I do input my data the partial of that input. But I can't see, unless I write down each of the partial sum then manually add it. Which never happens.
And I can't see my yesterday caloric sum or for some days back to recognize a pattern.
It was cheap offer for Thanksgiving for a year, but it's just good as a timer and (fairly lacunar for an eclectic, non commercial bought ready made food) calories values. Canadian ( Québec) food very underrepresented.
Can someone recommend a good calorie counter that also does the sum ?
Do you see a path to the Ukrainian victory in the war that does not involve some sort of at least partial regime change in Russia? Personally I find it unlikely.
Why does the order of comments on ACX posts sometimes default to "Chronological" (my preference) and sometimes to "New First"? Does anyone know of a way to force the user's desired default behaviour?
In the survey thread some people mentioned that they have an inner voice that uses plural first-person pronouns (“we” or equivalent). For those people, I'm curious about two major things, each with some branches.
Firstly, regarding self-experience: Do you have any type of dissociative disorder (self-identified or officially diagnosed)? And do you have a plural self-image in daily life? The latter most centrally includes those who would have answered “yes” to the other plurality question whose phrasing I don't remember.
Secondly, regarding life history: Do you have a university-level academic background? If so, what subset of these topics would you say you dug into in that context beyond intro-level/general-education: mathematics, ‘hard’ science, social science, engineering, arts, other STEM, other humanities?
(There's an event from my own life I'm keying off here for part of this, but I'm not sure I want to reveal it, as it might be too identifying.)
I have an idea that has been knocking around in my head for a while, but I've had trouble with finding a way to express it coherently or convincingly. So i infodumped my idea to ChatGPT and asked it to construct it as an argument that would be convincing to Scott Alexander readers. Here it is:
One of the central ideas in Utilitarianism is that pleasure is the only thing we can directly and subjectively experience as good. Other things, such as wealth, power, and fame, are only regarded as good because they bring about various states of subjective internal pleasure, whether we want to admit this or not. From this perspective, it logically follows that the ultimate goal of human technological development should be to create self-replicating "Von Neumann machines" that spread throughout the universe, turning as much matter into minds experiencing perpetual happiness as possible. This would allow us to achieve the greatest good for the greatest number of minds, which is the ultimate goal of Utilitarianism.
There are several points that can be brought up to address potential objections to this project. First, these "Von Neumann machines" could be designed with a "failsafe" to avoid humans and other alien species. This would ensure that even if the ultimate meaning of life is more "existentialist" than simply maximizing pleasure, these beings could still be spared by the "grey goo." Additionally, this would reduce the chances that humans or aliens would try to work against or destroy this system, as they would not be threatened by its spread, while requiring only a relatively insignificant amount of matter to be given up to ensure this safety, therefore even a system that only cares about pleasure would still avoid them out of self-interested risk/benefit analysis.
Second, it is often argued that humans might get "tired" of pure pleasure relatively quickly, but these minds could be constructed in such a way that this pleasure would always be as fresh as in the first moment. This would allow us to realize the proverbial expression of "a moment so good you wished in the moment it could last forever."
Third, we could conduct tests to determine which physical systems are capable of experiencing happiness and qualia before deploying this system. If it is found that biological brains are the only systems capable of experiencing pleasure, then we could construct modified brains to ensure that they only experience pleasure, and their biological capabilities leading to boredom with pleasure are removed.
Finally, concerns about the "beauty of the universe" being destroyed by this system are unfounded. Beauty exists in the subjective apprehension of conscious observers, and this system would simply maximize the positive subjective apprehension of these observers.
Overall, the argument for creating self-replicating "Von Neumann machines" that spread throughout the universe, turning as much matter into minds experiencing perpetual happiness as possible is a strong one based on the principles of Utilitarianism. It would allow us to achieve the greatest good for the greatest number of minds, while also addressing concerns about the potential risks and unintended consequences of such a system.
What do you think? I'd be interested in feedback on 1. the idea itself, 2. how well you think the AI did articulating the idea convincingly, and 3. what you think about the idea of AI as a rhetorical assistant, to help people like me who might have an argument they want to make but are having trouble laying it out coherently.
Somewhat annoyed at my answers to the Paris-Moscow distance, which will at least correlate with being unusually bad at orienteering.
Soliciting help/advice from the community regarding covid infection experience. Was recently most likely exposed on or around Dec 24-25 (was in the US Northeast NYC area at the time, so most likely the XBB 1.5 variant). By the time I got access to a test yesterday, the results were obvious. Will detail my symptom timeline below- want to know if this is typical, what (if anything) I should be worried or on the lookout for, and any recommendations for amelioration and recovery.
Was double vaxxed and boosted (Moderna), however last booster was all the way in Jan. 2022 so not surprising that protectiveness waned (plus apparently not effective on new mutations anymore if my reading of the available material is correct).
12/24-25: Most likely infection time frame
12/28: First symptom: sore throat. No other symptoms
12/29: Woke up with 102 F chills, aches, the works. Comparable to the worst flu I've ever had. Unable to sleep well that night due to massive night sweats and migraines.
12/30: Probably the worst day due to combined ongoing symptoms from yesterday and lack of sleep. However, symptoms did seem to calm down somewhat and night sweats diminished enough by nightfall that I was able to sleep (comparatively) well.
12/31: Fever downgraded to 99/100 and chills/aches mostly gone. Lingering headaches. The sore throat is mostly downgraded in intensity, but has given way to intense frequent coughing fits.
1/1: Fever, chills, aches completely gone, still occasional pinpricks of headaches. The main problematic symptom at this point is the coughing which seems to have increased in frequency after all the other symptoms have left. I can probably say without misusing the word literally that on average, I have literally coughed or had a coughing fit at least once every 1-5 minutes, to the point where my ribs and diaphragm have started to feel on fire. (To the best of my knowledge my chest is still fine). 99% of the coughs are dry and the 1% that produce phlegm seem to be tiny, inconsequential amounts.
Any advice anyone has on whether this is normal for covid sufferers** and what if anything can be done to alleviate the coughing would be much appreciated, I will answer any questions as needed. I also hope to get in contact with a medical provider once the holidays end (and if my insurance cooperates).
*demographic information: male, late 20's-early 30's, 23 BMI, not particularly athletic but not totally sedentary either (avg about 4000-6000 steps daily with amateur bodyweight routines)
**edit to add: the reason why I ask is because of the multiple people I know of who were exposed around this time with me, all of them to my knowledge have also developed coughing along with the other symptoms, but have not lasted as long as mine, even the older folks. To the best of my knowledge, I don't have any preexisting conditions like asthma, heart stuff, diabetes, or anything else like that.
Is anyone here familiar with the Twitter account “the ethical skeptic” ? If so, can you please translate the jargon and tell me what point he’s actually *claiming* to make?
Every time you ask the dude to clarify what he means by “concatenated pull forward effect” (or any of the other seemingly made up technical terms he uses) you get one of 3 responses: 1. “Have to talk this way or I’ll get censored”, 2. “This is designed for EXPERTS in statistics, im not gonna dumb it down for you idiots”, or 3. “If you don’t get it, then stop following me”.
The dude is obviously full of shit, but also the worst advocate for his own ideas I’ve ever seen.
No , I didn't read the book ( not even heard of it) but will do. When I lived alone ( 15 yrs ago) I had no problem ( ADHD helps with the flow state and nothing else matters) - plus I had , funny enough, a bigger budget for food .
Wherevcan be found . The plain metamucil , is not quite plain.
And satiety isvone component. The struggle is to reframe tasty food as such too. Living in Montréal it's like taking methadonevwhen heroin iscmore available.
I am worrying about the products of "health stores" - usually filled with hapless clerks and dubious and mislabeled products (unregulated) . But I will go - and look for the satiety component.
Thank you for pointing it out. The orange metamucil is very irritating.
That is an excellent advice. Having a school age kid and a food discerning wife is actually a hard thing to do . We have weekly talks about that. But the advice is really a good one to bump higher on my approach. And yes, you are right, my Pavlovian and future gratification imagination are drivers in quantity and choice.
Meals are scheduled ( 2 meals , no binging or snacking whatsoever, no pantry stuff)
Pre-planning is tricky ( high metabolic school kid and italian ancestry wife that cooks theoretically just for her ( she acknowledges my addiction). But that's what addiction is. I have no dignity to beg for meat filled fresh zucchini with Kumatovtomato and fresh herbs. And the bread is from L'Amour du Pain -they eon best baguette in France ( a first for the perplexed French bakers).
So even if I preplan for me, ( I love liver for ex and kidneys , Bourdain Nasty bots -,bit more hardcore even) I am drooling and "taste " a bit and it's really like someone that got a drug . I don't drin ( half a beer if I add the curiosity sips/ yer and half a shot of various liquors/ year.
Is there any chance that heating oil will make a comeback in America, and the use of natural gas will decline?
Yea, I do. Baked vegetables are awesome, mushrooms ( even raw). Soda is too irritant.
It's hard to get consistent high protein with variety and not slide into processed ones. Or out of budget. I am mixing quinoa with practically anything .
Fat is fairly attractive as taste . ( olive oil, avocado, occasional pork and rare beef one ( cooking and frequency).
In Montreal, Canada, tasty veggies on a daily basis ( I like them) are really expensive. But I am actively plan and budget for that.
I tried psyllium ( metamucil), but it's given me some acidity.
Good advice.
I want a new job. I'm currently the data scientist/DBA for a small company (~30 people). I've been there just over 2 years, and it feels like for the last year I haven't really learned anything new. The first year I really enjoyed it, I learned a ton, got to build a data warehouse from scratch, set up a bunch of data pipelines and SQL queries to analyze the data, and push it to dashboards for the dev/marketing folks. Now I've spent the last year incrementally improving those dashboards, but it's not challenging in any way.
I have never successfully applied for a job (I've always gotten lucky with offers through networking), and I really feel like I should do something harder and more challenging. At a friend's suggestion, I did the Advent of Code challenge this year, and it was some of the most fun I recall having while coding. I also think I did quite well (got 24/25 days within a day, got in the top 1000 once).
Anyone hiring? Or have some advice? I figured that since it's a new year, I should probably try to do something with it.
Editted: Padern has a good point that I haven't really clarified what I want to accomplish:
I've always been very motivated to learn new things when given tasks to accomplish, but I'm not very good at giving myself tasks. Being asked to do something tricky or new at work has been a very reliable way for me to get better, but a new job has always been a great source of motivation to learn a bunch of new things.
I'd like to work closely within a larger team (I'm currently on a team of 3, and I spend maybe ~2 hours per week talking to them), where there are people who are really great at things that I can learn from (no one else at my company has any database design experience). I'd like to improve my software development skills, again in a team setting (I'm the sole contributor on 90% of the code for my current database/pipelines). I'd like to learn more about databases, designing them for really large scale (>1000TB) purposes, learn more about transactional/distributed/NoSQL databases.
These aren't necessary, but those kind of things would be sufficient for me (I think).
In regards to the IQ/SAT question on the survey, does anyone else feel that they test way higher than their actual "intelligence"?
Way back in high-school I was a real superstar at all the standardized tests. I got 1600 on the SAT, nailed perfect 800s on both?(maybe 3 its been a decade) SAT subject tests I took, and took 15 AP exams and got 5s on all of them. And honestly I was (and still am) incredibly lazy and didn't view studying as an important use of my time when I could be practicing critical activities like pretending to study while actually staring blankly at walls. So according to those tests, I should balancing precariously on some very narrow slice on the bell curve.
But in the harsh light of a post secondary school world where ones ranking sadly does not depend on multiple choice tests, I anecdotally don't feel like I'm much above average intelligence. I'm not dumb, I know that, but if I had to roughly guess from conversations and generally living I'd say that at least 50% of my friends and family and acquaintances are similar or higher intelligence.
Sure one hypothesis is that for friends it could be selection bias, and family could somehow be genetic, but when even random people next to me on planes seem equally intelligent and I never feel myself drowning in a sea of morons, I kinda default to the idea that just like some people are innately more skilled at basketball, maybe I was just born in the right century to exercise my God given talent of being really good at bubbling in scantrons.
Does anyone have experience applying to Mechanical Engineering Masters degrees in the US? I go to a good but not amazing university in the UK, have decent grades, and some reasonably impressive extracurriculars.
What would be good schools to aim for? I have no experience applying for masters degrees either so am clueless on that whole process, any resources/experiential lessons would be appreciated. Am I too late to apply for ones starting in 2023 too?
I have dual US/UK citizenship and money is not an issue.
What would be some practical ways for me to get more quick repetitions for my social skills? I'm in my early middle age and now live in a small city that skews older. I was hoping to get regular, daily practice with say 10-20 minute conversations with new people, to build skills and reduce anxiety through exposure therapy. I work at home so I don't have regular interactions with co-workers, or really anyone unless I go out of my way. I've kind of exhausted the local Meetup scene, though I suppose I could make myself do more of them (the majority are people 60+ in age, who are nice but just not who I want to hang out with). In an ideal world I could have regular business lunches or do some kind of vague 'networking' (I'm self-employed) as an excuse to talk to people, but I think my city might be too small for that.
I have a reasonably high income, savings, and almost total independence, so I could conceivably move somewhere else for a period of time to practice. I could even move back to Large Famous City where I lived for a while- I didn't actively make myself socialize when I was there, but I'm sure there's lots of handwavey 'networking' events there I could force myself (an introvert) to do.
I do have a bit of an actual friend group, but I tend to get invited to multihour socializing events that just absolutely exhaust me and I think lead to bad repetitions/practice. They're marathons, and I need to train on shorter distances right now. Is there anything local I could be doing for more regular practice? (How exactly do middle-aged adults socialize outside of Meetup haha. I used to do combat sports but now I have too many injuries)
I finished reading "The Selfish Gene". I wish someone had told me to read it earlier in my life. How many have read it? And if not you should. (We (all life) are meat bags, projecting our genes into the future.) (OK there are plant bags too....)
Speaking of the IRL event, I got halfway through Scott's presentation on the recorded stream, got distracted with some or other holiday business, and when I came back to the tab a week later it had been taken down. Is that still up anywhere?
Help with portion control? I am doing intermittent fasting, very mindful, not a drag. Decades long hedonistic gourmand lived in various European countries ( Italy, Austria, France, Greece) that are now a curse because I am a supertaster ( verified). I do have strong likes and dislikes ( ex nothing spicy hot, embracing savoury and ofal ( nasty bits ) and all sorts of cheeses.
The worse time of my life were the stints in US ( food that I could afford was barely edible)
I live in Montréal, to aggravate things, and went to nutritionist.
I don't snack, never ever junk food ( as in chains or canned).
It's the hedonistic tendency I don't know how to re-frame. And yes, I can discern easily various parmigianos, prosciutto or olive oils . Which sounds nice , but it's not. I am definitely addicted, because I crave it .
I keep easy a 18-6 hours. And in the 2 meals I eat sufficiently to just break even.
Obese 40 BMi.
Thank you for any reframe, advice.
The analysis by "Cremieux" showing that the putative effects of slavery in that study were really (strong, terrible) effects of Jim Crow seems to be careful and valuable, at least at first sight. However, the insistence that it must be genes (as opposed to inheritance, social connections and advantages, habitus, family structure... - or all of the above, including genes) that account for the transmissibility of status across generations is much less convincing. In the end, the argument seems to be "this is a simple explanation that I like, and of course you know that people who don't like it are biased".
On #4, I still think Hal Finney is the most likely candidate for the true identity of Satoshi Nakamoto, for a long list of reasons
The hack to beat the AI at go is more interesting than you updated to think. They found multiple ways to beat it. See David's comment in the original thread, and https://goattack.far.ai/
Lately I’ve been reading the book Code by Charles Petzold. The book is amazing — basically it explains how computers work from scratch. The first part is about atoms and what electricity even is, and what it means to encode information. Then it builds up from there, explaining how you could build a long circuit to communicate with Morse code, add switched that incorporate AND/OR logic gates, and so on.
So far it’s the best technical book I’ve ever read.
Can anyone recommend any other books like this, that explain something very complex using a simple starting point and simple steps from there?
See Oliver Stone's "JFK Revisited: Through the Looking Glass" for all the naysayers out there
What is the probability that (some) conspiracy theories about JFK's assassination are true?
I am a bit of an outsider, having immigrated to the North America as an adult, and I haven't read much about the subject. But my very ignorant impression is that it is somewhat higher than for 9/11 conspiracies.
I keep seeing all kinds of "top 10" lists, top 10 baseball players not in the Hall of Fame, and today "top 50 bowl games.
So I'm puzzled why there is only one (obscure) collection of the key concepts that an educated person needs to navigate intellectual life. Concepts like regression to the mean, the Benjamin Franklin principle (two pages in Wikipedia on this one), the Pareto principle, black swan, and a whole bunch of others that you would hope your teenager has absorbed by the time they leave college.
I focus on this because my neighbor in Laguna Beach sells a deck of 88 cards each of which has an important principle. The front of the card names a principle and it is summarized on the back side. See conceptcompanion.com.
Part of the fun is trying to decide what principles should be kept and what principles should be added to his list.
But how can it be that there aren't a number of these "top principles" lists? At a minimum, you actually can learn something by browsing through them or trying to improve them. For example, I had not heard of the Benjamin Franklin principle, which is that if you want to get somebody to do something for you, don't start by doing something for them, but start by asking them to do a small thing for you, after which bigger things are more likely to follow. Fascinating, and the subject of a number of psychology experiments, but certainly unknown to me until I saw a Concept Companion deck.
??
Dave Anderson
I am considering writing a post on potential inherent limits of human knowledge. Here is the gist:
It is possible that parts of the world are not human-knowable. That is, while some parts can be approximately modeled through lossy compression, other parts are not compressible into something that can describe them with any reasonable approximation. If so, then some observations will always be surprising, and not amenable to probabilistic estimates. They have true Knightean uncertainty instead. Observations like that to an embedded agent like us would basically look like "miracles": unexplainable and unpredictable events.
To be clear, I am not insisting that our universe is like that, only that a world like that does not seem to contradict logic or observations, as far as I understand it. Note, however, that this is incompatible with the popular idea of comprehensible laws of physics that make the universe "work".
A basic version of this program would be constructing a toy model, a 1D world that is not usefully lossily compressible using the usual compression algorithm, Discrete Cosine Transform. (Here "usefully" does most of the work.)
An ambitious version of this program would be constructing a model of a world compatible with our current observations (i.e. the Core Theory, https://frankwilczek.com/2014/coreTheory.pdf, https://www.edge.org/response-detail/26611), and yet impossible to lossily compress into a human brain in a way that allows arbitrarily accurate (though potentially probabilistic) predictions, even given sufficient though finite time and suitable finite augmentations.
Does this make sense and/or look interesting/new?
Has anyone seen figures around the total cost to date of the Ukraine war to the US, including not only financial and military aid but also (estimates of) the direct and indirect costs to the US of its sanctions against Russia?
My goto on medical worldwide matters is my aged aunt who lives in a medium sized town
in rural Ireland. The population about 20,000.
its fairly common in Ireland for older people to go to as many funerals as they can manage, in the hope perhaps, that many people come to theirs. Of course in a small town you would know lots of people whose passing might be meaningful even if you had lost contact a few years back.
Someone will tell you, and my aunt is at the centre of a network of dedicated if morbid gossipers. Visit her and you will hear the illnesses and the deaths of not just the parish, but the town and outlying areas. She would know all the tragedies of course, any young person who crashes a car or dies of an overdose, which does occasionally happen.
She hasn’t been talking about a rise in deaths amongst the young and middle aged after the the vaccine, so I tend to ignore twitter and other sources on this.
A sample size of 20,000 is large enough.
There was a comment in last week's open thread that got me thinking. Imri Goldberg asked "Let’s say I believe I know, given today’s technology, how to build an AGI. What should I do with that knowledge?"
One response was "Don’t do it, unless you’re >95% sure you can make it aligned" and a reply to that argued the threshold should be more like "more confident than the next person who will do it."
But this got me thinking: let's say we accept 95% confident of alignment as an acceptable threshold for AGI. It seems likely that multiple people/organizations will independently develop AGI, and technological breakthroughs like that often happen in clumps. If 5 different organizations invent AGI at 95% confidence of alignment and you assume those probabilities are independent, you are accepting a 23% risk of a non-aligned AGI. That feels pretty alarming to me, regardless of any "well, better me than this other person" arguments.
Here's an interesting thought experiment along those lines. Let's imagine hypothetical past prediction markets for "will civilization as we know it collapse due to X" with X including things like nuclear armageddon during the Cold War, Large Hadron Collider causes some sort of quantum thing that destroys Earth, etc. Let's imagine that these prediction markets can function despite the issue of money becoming worthless in nuclear armageddon (maybe it's happening on Mars). 1) What would you guess the prediction-market-predicted probability of Earth already having been destroyed to be? 2) What is an acceptable threshold for this type of risk?
Does Max Tegmark’s level 4 multiverse concept unify materialism and platonism?
your survey had some questions that make me wonder why. What are you trying to glean?
An Ode to Cults
This post is a response to:
• Common responses to mentions of planned communities and alternative living communities being “didn’t they all fail in the 60s and 70s?” and “…sounds like a cult…”
• Occasional reveries (by me) of creating a board game called “Cults of the World”
This post is not about the bad parts of cults. “What bad parts?” you may ask incredulously…
This post is about the good parts — the parts that get overlooked when the word cult becomes more adjective than a noun. “What good parts?” you may ask incredulously…
I’ll focus on the following good parts:
• Courage
• Hierarchy
• Generative properties
• Optimism
Courage
It’s easy to criticize. It’s easy to get upset.
Maybe 100 years ago, it was easier to fall in line and to be content. But these days it is easier to get angry, as we often do.
It’s hard to do something about it. It’s hard for many reasons. For one, it requires work. It requires risk. And, in our complex, oligarchic world, it likely leads to failure. Rationally, doing something about it is often “not worth it”.
I had a conversation with a hyper-bright 35 year old (with degrees from Princeton). He was asking about career choices, playing them out in his head, and concluding that he probably will not make a difference, so he’ll just focus on his music because that’s what makes him happy. It was a remarkably rational analysis. Meanwhile, everything inside me screamed “YOU’RE WRONG!!!” “YOU’VE GOTTA FOLLOW YOUR GUT. THEN WORK YOUR ASS OFF. SOLVE THE PROBLEMS THAT ARISE. ADVANCE, ADJUST, PIVOT, THRUST AS NECESSARY. BECAUSE THAT’S THE ONLY WAY DIFFERENCE GETS MADE. AGAINST THE ODDS.” Talk about a remarkably irrational analysis. Irrational, but true-ish.
Courage is not rational. It’s often a sure way to get hurt or killed. I suspect it’s going out of style.
Cults embody courage — both the leaders and followers. There is a leap of faith, a vision of a better way. However misguided, risk is taken — a big risk, unmoderated. We still see this a bit with entrepreneurs. I guess investors require some courage. I’m sure there are other groups. But the stakes seem smaller and incremental, and usually self-involved. (Aside: I’m mostly talking about OG physical cults… Digital cults, as one might refer to SBF and FTX, are different. They may get called by the same name, but the mechanics are not related.)
I’ve been hearing a lot lately about “skin-in-the-game”. The skin and the game in cults was a level of real that we only see in bad distopian sci fi like those movies with Jennifer Lawrence, where she’s an archer or some such thing. Today, the skin and the game (for the privileged us) can be more aptly described as skin-deep and child’s play.
Hierarchy
I agree with wokeism. I just can’t live with it. I am not being tongue in cheek. I think wokeism is right. I just don’t think it’s possible to live woke.
To avoid explaining the few sentences above, I’ll focus on hierarchy, or, the ability to say that one thing is better than another. I think that hierarchies are important. I think they are interesting. And I think they are required for progress.
If I picked 100 new cult leaders and convinced them to start cults, with value & rules systems of their own choosing, the 100 cults would not be equal. Some will do better than others. Some will have planned qualities that are deplorable. Others may have emergent qualities that are wonderful. I believe in the freedom and importance to compare and contrast.
Cults embody this freedom, but not as armchair Twitter commentators, but in-action, with maximum skin in the game.
Generative Properties
What didn’t start as a cult? How much of human history is built on the engine of a cult? How much progress and ingenuity came from cults? I suspect that the answer will be some version of “it’s only treason if you lose.”
Cults are by nature generative (until they are not). Yay cults for splitting with the status quo and daring to innovate.
I guess you can listen to Thiel or Cowen on the slowdown in progress. They’ve thought about it more and made economic, political and technological cases around the topic. But I would also throw the death of the cult ethos in there as a reason for our stale state of affairs.
Optimism
“Democracy sucks, but it’s the best system we got.” That’s pessimism. What’s the opposite of that? Cults.
I don’t know what the split of free will deniers vs enjoyers is here but. Coming from the perspective of:...
There is no reason to believe we have free will because everything is either determined (classical mechanics) or random (eg quantum mechanics) all the way down. There is no place for free will to come in.
...Why would it be unreasonable to say that free will comes from consciousness? We don’t know where consciousness comes from or what it is. Why would it be wrong to say that it is the source of agency? You could do the same “all the way down” analysis and say “there’s no place where consciousness comes in.” Yet we are conscious.
But if Joe Biden does take your survey, wouldn't you want to know?