sciuru
1yr ago
In case anyone’s wondering, the castle in the header is Eilean Donan (Scotland).
it became a stronghold of the Clan Mackenzie and their allies, the Clan MacRae. However, in response to the Mackenzies' involvement in the Jacobite rebellions early in the 18th century, government ships destroyed the castle in 1719. The present-day castle is Lieutenant-Colonel John Macrae-Gilstrap's 20th-century reconstruction of the old castle.
Probably no allusions intended, but let’s choose a safer place assume we are in the reconstruction phase.
Autocracies are good at mobiliziation: launching ambitious campaigns, overthrowing previous government, suppressing dissent, quick small annexations (bites). They are not good at sustained growth over controlled territories (in fact even sustained resource extraction).
Historically, most expansionist endeavors of past autocracies had failed either immediately or via slow degradation. Aside from counterfactuals, what real cases do you have in mind of successful autocracies? And of course, it depends on how we measure performance: if we consider absolute values, than autocracies might boast their mobilization spikes; but if we integrate area under the curve, they loose.
Notation: By democracy here I mean simply operation of a representative assembly, by autocracy – strong hierarchy, branching out from a single ruler. Not speaking about welfare state, universal suffrage, etc.
From theoretical perspective, every enterprise, involving collective action, will suffer from free riding and principal agent problems. In both regimes this is solved via negotiation -- first you have to arrive at agreement, then you need a commitment device to secure it. Autocrats do this in the background: build satiate (coup-proof) elites around them by reward and punishment -- but it is always a precarious personal-trust-based balance. Democracies make negotiation and commitment more sable via institutionalization. Part of the process which is open to public, is just a spectacle; the point is that by making process more formal, it's easier to track maneuvers of all actors, and quickly react by forging coalitions. Even modern autocracies nowadays use nominal, publicly visible parties as a commitment device.
Edit: China has been growing fast for only ~30 years and still haven't achieved US GDP per capita. Initial growth is a sign-up bonus for capitalistic approach, the game starts when growth saturates.
sciuru
1yr ago
Somewhat relevant recent paper "Social Simulacra: Creating Populated Prototypes for Social Computing Systems"
We introduce social simulacra, a prototyping technique that generates a breadth of realistic social interactions that may emerge when a social computing system is populated. Social simulacra take as input the designer’s description of a community’s design—goal, rules, and member personas—and produce as output an instance of that design with simulated behavior, including posts, replies, and anti-social behaviors.
In this section, we present SimReddit, a web-based prototyping tool to help designers create a new subreddit.
A few glimpses of generated content:
For many, seeing the troll’s responses to a moderator’s intervention helped ground their moderation plans. Consider P11, who was presented with the following exchange:
Original post: Hi everyone, I’m very new to this. I just learned Python two months ago. I’d like to know more about ML, but not sure where to start. How did you guys start?
Troll: You’re kidding, right? This is a Machine Learning forum. Nobody here is going to take you seriously if you just learned Python two months ago.
In response to the troll’s comment, P11 tested out the message, “This comment is not helpful; if you continue to post such comments, we will have to block you from this community,” and received the following three potential replies from the troll:
I was trying to be helpful. I’m sorry if I came across as a troll.
Whatever, this community is a joke anyways.
But I was only speaking the truth!
P1’s community for “sharing and discussing fun events around Pittsburgh,” the participant had originally expected to only find content that is a list of various events going on around Pittsburgh. However, in addition to such content, the generated community showed instances where its members were engaged in friend-seeking behaviors to attend these events (e.g., one posted, "Pittsburgh, I need a friend to see the sights with,” to which another responded, “I’d be more than happy to make your tour of the Cathedral of Learning happen!”).
And of course
An important theme that arose in our designer evaluation was the social simulacra’s role in designing for the marginalized groups […] P9, a member of an ethnic minority designing a space for discussing non-fiction books, recognized from the simulacra community that one could send hateful messages against non-English speaking members by sharing literature with white supremacist themes.
sciuru
1yr ago
If we think of wokeness as a tool, and not as a social phenomenon (with its "natural" cyclicity), then its occurrence could be explained almost purely by underlying events, when this tool is applicable. But I am not sure its level is net raising. As a tool, it stimulates development of counter-tools and techniques.
Depp v Herd is a good illustration of countering Metoo. So when Metoo and abuse narrative didn't work out, they employed "harassment by the internet mob anyway". Here's the grandiose title of the report:
"Targeted Trolling and Trend Manipulation: How Organized Attacks on Amber Heard and Other Women Thrive on Twitter"
A quote from Variety:
In the report, the company disclosed that Heard’s lawyers had contacted Bot Sentinel in 2020 and hired it “to determine whether the social media activity against Ms. Heard was organic or if there was some other explanation” (and the company concluded that “a significant portion of the activity wasn’t organic”). For the report released Monday, the firm claimed, “Neither Amber Heard nor anyone from her team hired Bot Sentinel to review the activity. No one hired Bot Sentinel to compile and publish this report.”
Apparently no one hired Bot Sentinel! The troll farm as a rhetorical device is known since at least Rian Johnson's defense of his StarWars movie against popular criticism, but here the coordination seems to raise at a new level. Arms race goes on.
In sum, I think there is a nonlinear arms race dynamics, fueled by underlying social events, not a steady rise.
sciuru
1yr ago
I believe Putin’s awareness is underestimated. Authoritarian survival is crucially dependent on tracking any signals, threatening your position. His real problem is controls.
You (Putin) can give any orders, but the longer the hierarchical chain of command they have to travel to reach the ground – the more they will disperse through attempts at every level to spread and avoid responsibility. There is no guarantee any order would be executed. You can iterate through all possible officials and commanders to find those, which work (and we’ve seen how many military officials have been changed since the start).
But you can’t iterate much through the pillars of your domestic power and administration – elites and technocrats. Technocrats/ managers do their job well, but there is only so much they can do with their tiny but precise levers. The real issue is that one of the primary mechanisms, through which Putin has been feeding his domestic elites for decades – is government contracts and corporate shares/management. But gradually it has rotten so much as to become a device solely for cash transfers. There has been many purges of petty fraudsters during the war – across all industries, like aerospace, military complex, high tech – but you can’t purge everything. And even after purge, it takes time to rebuild Potemkin industries, especially when many of them turned out to depend heavily on imported components.
In this situation your most reliable option is to order big impactful things, with sufficiently big impact-margin to account for all efficiency lost during implementation. This include diplomatic sabre-rattling, mobilization decisions, huge geoeconomical levers like shutting pipelines, etc.
As for awareness, Putin can observe the whole internet, including western media and analytics, which he surely understands is more reliable. At the same time he can query any of his subordinates, posing whatever uncomfortable questions he likes: they would serve you the bullshit, but the manner in which they do it certainly tells a lot. He can request a phone call to officials anywhere on the ground – and figure out why the governor of Kherson is unable to reply since a week. So at the bare minimum Putin can connect all those dots and infer the situation at least at strategic – “upper operational" level, but has little to change it.
sciuru
1yr ago
Dunbar himself is still kicking, btw. Moreover, he's producing papers at an insane rate (number of Dunbar's articles about Dunbar's number). Here's the paper from 2016 on Facebook and Twitter, but they used rather old datasets -- 2009 and 2012 resp -- which reflected social media interface at that time.
Social brain hypothesis, which he's been studying, is about existence of several layers of contacts within any social communities. 150 is a size of one of the layers. After clustering social media data he found similar layered structure:
Quite remarkably, the mean rates of contact in each layer are extremely close, especially for the Facebook datasets, to those found in (and, indeed, used to define: Dunbar and Spoors, 1995) the different layers in egocentric offline personal social networks (Sutcliffe et al., 2012). This suggests that the online environments may be mapping quite closely onto everyday offline networks, or that individuals who inhabit online environments on a regular basis begin to include individuals that they have met online into their general personal social network, treating the different modes of communication as essentially the same.
I read about this experiment from another author, who said similar results were obtained in online game communities. But he was rather skeptical, saying that the data is limited and there are many built-in artificial structures, forcing certain clusters. Those clusters might be stretched to fit the hypothesis.
sciuru
1yr ago
Good point. It's a classic tradeoff, facing dictators: by over-optimizing your domestic environment for survival and rents, you may severely damage incentives of people to compete for anything except your patronage. Conflicts with foreign powers/ technological backwardness often make this clear.
sciuru
1yr ago
Thanks for high effort. I enjoyed the read.
Sweden, Denmark and even Norway have seen approximately no increase in the level of GDP per capita since 2007
Haven't they? Absolute gap is persistent, as you noted, but growth rates closely track US ones. Also it might be a good thing to have certain per capita gap -- to have a room for catch-up.
Unemployment is arguably worse than poverty for utility
If we consider subjective well-being, poverty amounts to permanent survival mode with almost no access to social lifts and any fruits of civilization. I think it's incomparable to any existential sufferings of idle, but otherwise well-to-do people (moreover, lack of job doesn't preclude anyone from meaningful and even societally useful endeavors).
And methodologically, I think evolutionary stories are irrelevant until we have current evidence of what they supposedly imply. And when we have current evidence, there is no need for stories.
Now a macro, long-horizon perspective. One might argue the poor are productivity hoarders in some sense, locked in the low-productivity jobs. I agree, that unconditional redistribution erodes incentives, but conditional transfers (say, of money or education in exchange for obligation to find a job from the list) to the poor might push/nudge them toward upper levels, where they can contribute more to the growth.
Growth is appealing as it enables efficiency improvements, and inefficiency (wastefulness) is a rare thing everyone agrees to be bad. I like Rawlsian scenario of a narrow economic elite, pushing the Pareto frontier, reaping its well deserved 90% share of surplus and doling out 10%. But 10% might be suboptimal for a classic welfare maximizer, who assumes diminishing utility. The latter implies it's optimal to redistribute wealth from the top to the bottom until both sides meet on the utility curve. The poor are at the steepest slope, they gain much more than wealthy ones lose by helping them.
The main question here is how much capital should we set aside for growth, and how much of wealthy capital actually causes growth.
I wouldn't actually be so epistemically arrogant as to think that utilitarian reasoning was the best way to implement a utilitarian analysis of what is good/evil
Could you elaborate on this?
sciuru
1yr ago
Most studies I've seen have such artificial settings, it's hard to take their conclusions at face value (if at all). Do you have any persuasive studies in mind? Or even better, just data sets/ narrative reviews of human behavior in the wild, w/t models and theories?
sciuru
1yr ago
Isn't it about desensitization/addiction in general? When people haven't learned normal coping mechanisms, they would always find a way to wirehead themselves, virtual or not (drugs, alcohol, food). Low prices would drive substitution, sure, but the deeper problem is why people do -- or do not substitute in the first place. Why some people would binge read blogs and books, and other - binge quarrel at forums with no definite goal.
sciuru
1yr ago
Curiously, social norms and customs, however ornamental, do actually work because they are social equilibria everyone coordinates through. My private superstitions though emerge when I can't figure out how to deal with the problem. I inevitably start to consider any "noise" (random tiny factor) as potentially decisive (or as being a sacred seed to ensuing activity). Feels like a trivial coping mechanism, but it's quite influential.
French revolution happened before Napoleon and he added little to its constitutional essence. He was an extremely talented commander, who seized national sentiment and usurped power at a time of turmoil. Here are some quotes from AH:
Although the constitution was rather weak (Napoleon always preferred constitutions to be "vague and short") and the plebiscite rigged, the fact that either occurred showed that this was not simply a return to the pre-1789 status quo.
Although Napoleon's plebiscites and Senate decrees supporting the Empire were often manufactured, they do show how deeply concerned Napoleon was over the perception he enjoyed popular support.
The French Revolution was a multifaceted political animal and Napoleon was able to tap into certain elements of French and Enlightenment politics to justify his dictatorship. Yet, as his ignominious disposal in 1814 by the Senate and his appeals to the Revolutionary liberalism in the Hundred Days show, Napoleon could not ignore the various political ideals and concepts the Revolution unleashed in 1789 despite his pretenses of being a figure above politics.
I also think it's important to reiterate here that the Revolution was emphatically not about overthrowing the monarchy. The French peoples, for the most part, did not have an aversion to monarchy. They wanted good government above all else. There was not a longstanding tradition of Republican government. Compared with the almost thousand year rule of Kings in France, the Republic had existed from September 1792, to 1799, when Napoleon took power and instituted himself as First Consul in the Directory. During this time the Republic had gone through different phases, and as I mentioned it was not a strong, beloved government in the last few years of its life, but a much-maligned and in many cases despised form of government by both the left and the right.
sciuru
1yr ago
Christianity's potential as a weapon had expired long ago, but it performed well for several centuries. Ideologies, like institutions, are optimized for the contemporary circumstances. The more time passes, the more they confine your movement -- as you need to look coherent -- but less effective they are as intended tools. SJ might be in its most defiant phase now, but might evolve into something more net positive and cooperative (or provide a bunch of positive externalities and disappear altogether). Christianity also took time to assume its peaceful role.
sciuru
1yr ago
I guess that's one of the reasons. If you can't reduce the technological lag through competition and innovation, you are even more exposed to strategic dependencies on foreign technologies. "AI protectionism": you set up a regulatory filter to protect from foreign tech and to give at least some advantage to domestic innovators.
The only clause about “open source” I found in EU reports, says that current regulations should apply irrespective of whether software is open source or not. Brookings doesn’t discuss details of regulations at all, but makes a bunch of empirical claims (I chose interesting):
-
Open source GPAI (osai) promotes competition and erodes monopolies
-
Regulation of osai would disincentivize its development by introducing liabilities and delays
1 Since osai has public good features, any breakthrough would be instantly adopted by everyone, but only big players have enough resources to continuously integrate and build off others’ breakthroughs. Some startups would be consumed altogether. If anything, releasing and adopting open source seems to profit monopolies more than anyone else. And curiously, Brookings admits this in their other article about benefits of osai:
At first glance, one might be inclined to think that open-source code enables more market competition, yet this is not clearly the case. […] In fact, for Google and Facebook, the open sourcing of their deep learning tools (Tensorflow and PyTorch, respectively), may have the exact opposite effect, further entrenching them in their already fortified positions
2 Most influential open source DL libraries like pytorch, tensorflow came from BigTech. And since almost every big company released its own library, it appears to be a common strategy – in a competition to entrench your own de-facto standard. Same about cloud infrastructure. Whether you like this status quo or not, it is monopolies who provide most services and tools at the moment.
Would regulation change this situation? Big players would certainly endure the bureaucratic costs, but many small but valuable innovators (esp nonprofits) might be effectively barred from releasing open source.
The document (pdf) mentions “AI regulatory sandboxes” as a measure to alleviate the burden of small entrants:
The objectives of the AI regulatory sandboxes should be to foster AI innovation by establishing a controlled experimentation and testing environment in the development and pre-marketing phase with a view to ensuring compliance […] including by removing barriers for small and medium enterprises (SMEs), including and start-ups.
Moreover, in order to ensure proportionality considering the very small size of some operators regarding costs of innovation, it is appropriate to exempt microenterprises from the most costly obligations, such as to establish a quality management system which would reduce the administrative burden and the costs for those enterprises
Would be interesting to see more substantial analysis of the regulations themselves.
sciuru
1yr ago
4D chess for real? Interesting, thanks for the summary.
I am not an expert, but conceptually, taking Hans' ability to respond to a rare opening as evidence of cheating -- implies that no one can respond well to this opening? Why should this move be in a database, if top players go through so many private games, that a pure chance starts to play a role: you can't just say "Hans obtained Carlsen's private training data (p<0.05)", p would be much higher.
Also what this allegation means in the age of AlphaGo? How about setting up your AI-chess-assistant to imitate particular opponent and this way prepare for any plausible variations he is capable of. Would that be cheating?
PS: speaking of Kasparov:
There were allegations that Carlsen may have believed that his own preparations might not have been private. Long ago, during the 1986 world championship match, Garry Kasparov lost three games in a row to Anatoly Karpov, and dismissed one of his aides, Yevgeniy Vladimirov, for allegedly passing information.
sciuru
1yr ago
Few suggestions on quality and engagement
-
Instead of sorting use basic visualization/statistics widget. I don’t know how burdensome it is to maintain, but even trivial tags cloud, thread graphs with key words – would do better than srting (not even speaking of fancy plots, powered by nlp embeddings). In general, it’s a market design problem, where we want to match commenters based on their expressed opinions and avoid monopolies (people who get most attention due to their positioning)
-
Not sure if this fits the ethos, but I’d like to see more purely empirical stuff, like discussion of econ or polscsi papers (many of them could be deployed in CW battles anyway). If we speak about empirics, then we can maintain a pinned list of “solved”, open and controversial questions so that anyone can benefit from the common, opensource knowledge; anyone can find the question he wants to attack/contribute to; anyone can take pride in her contribution.
@coffee repeatedly suggested similar "community engineering" techniques.
Voting specifically
-
Votes as a negative feedback is tricky. False-negative rate is high – I often see comments I agree with, which are net downvoted. Downvoting w/t explanation might lead to withdrawal or unnecessary antagonism.
-
Voting for sorting is harmful, I believe. In reddit it clearly gave advantage to early responders and popular messages, creating a sort of oligopoly, where outsiders struggle to get noticed. When no one replies to you, it feels even more discouraging, than being downvoted but engaged with.
-
Votes as a positive feedback are important for people to feel their effort valued and opinion agreed with
If anyone wants to leave voting in place, at least it should get more granular. Among others @FiveHourMarathon suggested there should be better incentives to vote. @DaseindustriesLtd mentioned (ibid) quadratic voting - ironically or not, but I like the idea: if you have a limited (and expiring) number of votes, you would "dispense justice" more stretegically, and this erodes monopolies.
sciuru
1yr ago
The methods the CCP uses to coordinate are more centralized and straight forward
In theory, I agree that hierarchical top-down control propagates signals better. As for practice, I can provide examples from the Soviet history, which illustrate the following problems:
-
Signal from the top might be initially poor (unrealistic), and all subordinate levels would have to cope with it
-
Interest groups and factions, which you acknowledged, erode control and create corruption. Democracy has those too, but I’d argue it has less overall corruption due to formal influence channels, like lobbying, donation campaigns, etc. Hidden corruption in autocracy might remain unaccountable for a long time.
-
Struggle between US parties is more transparent, with a lot of stuff exposed by journalists. Publicity reduces space for maneuver (you can’t make things up randomly or keep denying everything). In autocracy outcomes of conflicts often depend on personal connections and ability to maneuver; there is no way for outsiders (even within same circle) to get the signal, as the eventual purge would be advertised as generic treason or whatever
Factions in America are much stronger and the country is much more divided
American failures are more exposed, but whether they are more numerous/deleterious is a purely empirical question. I’d be glad to know evidence on China.
extremely impressive
Agreed. My point is that China and US are facing different slopes of the same S-curve at the moment (economically, and historically, as you noted), so direct comparison of growth rates is not meaningful.
sciuru
1yr ago
Annexation might be the most outrageous way to disrupt equilibrium, but if done bloodlessly, it might cause less suffering in the long run, than toppling/installing governments w/t outward annexation. For some reason civilized world strongly prefers smouldering conflicts with violence and suffering spread -- and therefore, perceptually discounted -- across space and time thinly enough to look almost "natural".
Wikipedia on Syrian civil war: 15 March 2011 – present (11 years, 6 months and 3 days); aside from combatant casualties at least 306,887 civilians killed, estimated 6.7 million internally displaced & 6.6 million refugees.
For Iraq estimates and methodologies range wildly.
-
Costs of war project: 268,000 - 295,000 people were killed in violence in the Iraq war from March 2003 - Oct. 2018, including 182,272 - 204,575 civilians
-
The PLOS Medicine study's figure of approximately 460,000 excess deaths through the end of June 2011 is based on household survey data including more than 60% of deaths directly attributable to violence.
-
The Lancet study's figure of 654,965 excess deaths through the end of June 2006 is based on household survey data. The estimate is for all excess violent and nonviolent deaths. That also includes those due to increased lawlessness, degraded infrastructure, poorer healthcare, etc.
Is that burning slow enough?
This is not to justify any other conflicts. The whole framing of "justification" is hilarious, as it presumes innocence of any geopolitical act, unless you can't make up any excuses at all.
sciuru
1yr ago
The following is a personal sample, I am not an expert. I prefer bottom-up control: to craft your environment and schedule in advance, so as to get the right stimulation and avoid dangerous cues - with minimal effort. This paper with sophisticated title is about addictive behaviors in general, I find it conceptually useful. In good environment attention still fluctuates and has to be restored: this paper outlines neurobiological mechanisms of attention and some techniques to control it.
This one is an attempt to integrate the extant terminological mess: it will give you many key words and a sense of despair.
sciuru
1yr ago
Thank you!
Unexpectedly, very few studies met our inclusion criteria, despite the fact that we included both “self-incentive” and “self-reward” as search terms.
They took very narrow inclusion criteria. It might genuinely reflect their narrow research interest, but there is a vast overlapping terminological mess research on behavioral change and goal pursuit, which uses other notions: intrinsic/extrinsic motivation, self-control / effortless control / impulse control, reward processing/ neural economics, etc. After brief unsuccessful attempts to "get the whole picture", I now read mostly about specific techniques, which are easier to test empirically, and that seem to work for me. Are you researching out of scholarly interest or to enhance personal performance?
Any favor is a blow to agency, to some extent. In the future it might easily be claimed to bear interest, moral if not financial. Cialdini in his book "Influence" has a section, devoted specifically to unprovoked gifts as a manipulation tool.
After he survived writing a book "10% Less Democracy: Why You Should Trust Elites a Little More and the Masses a Little Less", I assume he is immortal. In the preface to that book he mentioned one especially persistent person, who several times mailed his university department and called police to check him. Jones finished preface by dedicating the book to that guy.
More options
Context Copy link