Links

AI:

• “Decentralized Reinforcement Learning: Global Decision-Making via Local Economic Transactions”, Chang et al 2020 (blog; toy problems but interesting)

• “Progressive Generation of Long Text”, Tan et al 2020 (GPT-3 users have been experimenting with various kinds of expansion/summarization for GPT-3, like bullet ↔︎ email, to avoid the context window limit; models finetuned on summaries to do expansion might be another way around it)

• “Possible takeaways from the coronavirus pandemic for slow AI takeoff” (corona was x-risk on easy mode: a risk (global influenza pandemic) warned of for many decades in advance, in highly specific detail, by respected & high-status people like Bill Gates, which was easy to understand with well-known historical precedents, fitting into standard human conceptions of risk, which could be planned & prepared for effectively at small expense, and whose absolute progress human by human could be recorded in real-time happening rather slowly over almost half a year while highly effective yet cheap countermeasures like travel bans & contact-tracing & hand-made masks could—and in some places did!—halt it. Yet, most of the world failed badly this test; and many entities like the CDC or FDA in the USA perversely exacerbated it, interpreted it through an identity politics lenses in willful denial of reality, obstructed responses to preserve their fief or eek out trivial economic benefits, prioritized maintaining the status quo & respectability, lied to the public “don’t worry, it can’t happen! go back to sleep” when there was still time to do something, and so on. If the worst-case AI x-risk happened, it would be hard for every reason that corona was easy. When we speak of “fast takeoffs”, I increasingly think we should clarify that apparently, a “fast takeoff” in terms of human coordination means any takeoff faster than ‘several decades’ will get inside our decision loops. Don’t count on our institutions to save anyone: they can’t even save themselves.)

• “This Pony Does Not Exist”, Arfafax

• Matters Of Scale:

  • “Are We In An AI Hardware Overhang?”

  • “Direct Fit to Nature: An Evolutionary Perspective on Biological and Artificial Neural Networks”, Hasson et al 2020 (perhaps NNs ‘just’ interpolate… but in what high-dimensional abstract space biased towards simple programs? See also “Neural Networks, Manifolds, and Topology”)

  • “Breaking the Quadratic Attention Bottleneck in Transformers like GPT-3?” (discussion)

Genetics:

• Everything Is Heritable:

  • “The Contribution of Cognitive and Noncognitive Skills to Intergenerational Social Mobility”, McGue et al 2020

Statistics/Meta-Science:

• “How Are the Mighty Fallen: Rejected Classic Articles by Leading Economists”, Gans & Shepherd1994

• “Lights and Shadows”, Bartosz Ciechanowski (explorable on lighting, color, geometry)

Politics/religion:

• “Stereotype accuracy: One of the largest and most replicable effects in all of social psychology”, Jussim et al 2016

• “Thrones Wreathed in Shadow: Tacitus and the Psychology of Authoritarianism”, Rehman2020: “History Will Judge the Complicit: Why have Republican leaders abandoned their principles in support of an immoral and dangerous president?”, Applebaum2020; “The American Press Is Destroying Itself: A flurry of newsroom revolts has transformed the American press”/“The Left is Now the Right”, Matt Taibbi2020 (Singal)

• the Uighur Genocide & Concentration Camps¹ (a decade of links):

  • “China cuts Uighur births with IUDs, abortion, sterilization” (report; commentary); “China’s Government Has Ordered a Million Citizens to Occupy Uighur Homes. Here’s What They Think They’re Doing.”; “‘This is mass rape’: China slammed over program that ‘appoints’ men to sleep with Uighur women”

  • “How China Tracked Detainees and Their Families”; “China’s Operating Manuals for Mass Internment and Arrest by Algorithm”; “‘Absolutely No Mercy’: Leaked Files Expose How China Organized Mass Detentions of Muslims”; “China has turned Xinjiang into a police state like no other”; “China Is Detaining Muslims in Vast Numbers. The Goal: ‘Transformation.’”; “A Week in Xinjiang’s Absolute Surveillance State”; “‘We Are Afraid to Even Look for Them’: Enforced Disappearances in the Wake of Xinjiang’s Protests”

  • “China convicts Uighurs in sham trials at Xinjiang camps”

  • “Her Uighur Parents Were Model Chinese Citizens. It Didn’t Matter.”; “A model’s video gives a rare glimpse inside internment”; “Beijing Squeezes Exiles in U.S. by Detaining Family Back Home”

  • “One Month, 500,000 Face Scans: How China Is Using AI to Profile a Minority” (one thinks of Qiao et al 2018 and wonders what research has not been published & what their real purpose is); “China’s Algorithms of Repression: Reverse Engineering a Xinjiang Police Mass Surveillance App” (similar to North Korea); “China Is Forcing Tourists to Install Text-Stealing Malware at its Border”; “The Panopticon Is Already Here”; “Revealed: Massive Chinese Police Database—Millions of Leaked Police Files Detail Suffocating Surveillance of China’s Uighur Minority”

  • “China Secretly Built A Vast New Infrastructure To Imprison Muslims”; “Blanked Out Spots On China’s Maps Helped Us Uncover Xinjiang’s Camps”

  • “China Is Collecting DNA From Tens of Millions of Men and Boys, Using U.S. Equipment”; “The Xinjiang Procedure”; “Independent Tribunal into Forced Organ Harvesting from Prisoners of Conscience in China”; “Analysis of official deceased organ donation data casts doubt on the credibility of China’s organ transplant reform”, Robertson et al 2019; 13 tons of exported Uighur hair seized

  • “I am an Uighur who faced China’s concentration camps. This is my story.”

  • “Surviving the Crackdown in Xinjiang: As mass detentions and surveillance dominate the lives of China’s Uighurs and Kazakhs, a woman struggles to free herself”; “China Cannot Silence Me: Speaking of the crimes committed against my family and other Uighurs in Xinjiang has sparked a surprising reaction.”

  • “China cracks down on Inner Mongolian minority fighting for its mother tongue”

  • “The Strategic Consequences of Chinese Racism: A Strategic Asymmetry for the United States”, Anonymous & Anonymous2013 (on Chinese racism/Sinocentrism; see also The Scholar’s Stage)

  • “How China Spreads Its Propaganda Version of Life for Uighurs”

  • “One by One, My Friends Were Sent to the Camps”; “China Locks Up Xinjiang’s Uighur Businessmen; ‘In Their Eyes, We Are All Guilty’”

  • “‘Some are just psychopaths’: Chinese detective in exile reveals extent of torture against Uighurs”

  • In the Camps: China’s High-Tech Penal Colony, Byler2021

  • “Leaked Documents Detail Xi Jinping’s Extensive Role in Xinjiang Crackdown: Records of internal speeches released by U.K.-based panel show how China’s leader set the blueprint for vast campaign of forcible assimilation against Uighurs, other groups”

  • “How Beijing Influences the Influencers”; “The Big Business of Uighur Genocide Denial: A New Lines investigation reveals a network of charities funneling millions into left-wing platforms that take Beijing’s side on the genocide allegations—and they’re all connected to an American tech magnate”

  • “Uighur kids recall physical and mental torment at Chinese boarding schools in Xinjiang”

  • “The architecture of repression: unpacking Xinjiang’s governance”, ASPI 2021

• “How Many Atheists Are There?”, Gervais & Najle2017 (social desirability bias / preference falsification / chilling effects)

• “Ancient Egypt and the geological antiquity of man, 1847^–₁₆1863_163ya”, Gold2019 (how did Victorian Christian academia enforce Young Earth Creationist orthodoxy? Discouraging geologists from absolute dating, accusations of impiety, nitpicking, ad hominems, & in general avoiding empirical data in favor of textual hermeneutics & philosophizing)

Psychology/biology:

• “Depression’s Unholy Trinity: Dysregulated Stress, Immunity, and the Microbiome”, Cruz-Pereira et al 2019

• “Chimpanzee super strength and human skeletal muscle evolution”, O’Neill et al 2020

• “The Overfitted Brain: Dreams evolved to assist generalization”, Hoel2020 (dreams as exploration/regularization of deep models, not mere experience replay; seems consistent with evidence & other paradigms like Tononi’s SHY—dream regularization can help decide which synaptic weights to decay; see also “World Models: Can agents learn inside their own dreams?”)

Technology:

• “Potential for large-scale CO₂ removal via enhanced rock weathering with croplands”, Beerling et al 2020 (Nature; editorial; media; this enhanced weathering approach uses basalt for fertilizer, but Project Vesta prefers olivine; see also the 3-part review: “Negative emissions—Part 1: Research landscape and synthesis”, Minx et al 2018/“Negative emissions—Part 2: Costs, potentials and side effects”/“Negative emissions—Part 3: Innovation and upscaling”, and “We Need To Take CO₂ Out Of The Sky”)

Economics:

• “Modeling the Human Trajectory” (paper), Roodman2020 (Revisiting Hanson’s2000 stacked-exponentials growth model of human history with a simpler power law that fits well—human global population/GDP growth remains exponential or greater, with astronomically wide prediction intervals even within the near future implying radical uncertainty, with AI, given the scaling hypothesis, looking increasingly like the next source of exponential GDP/population growth as previous growth feeds back into technology & digitized minds can be combined/scaled efficiently in a way human brains never could, creating the next level in the organismal hierarchy & next paradigm of growth; see also “1960: The Year The Singularity Was Canceled”, “Hyperbolic Growth”, “Hyperbolic Takeoff”, “Is growth linear, not exponential?”.

  One oddity is an apparent collapse of superexponential growth to ‘merely’ exponential sometime in the 20^th century, perhaps post-1913, showing up clearly by the 1970s; if superexponential growth comes from wealth feeding back into technology & productivity, particularly humans, that suggests that humanity missed something around then—the combinatorial expansion of technological possibilities should’ve yielded some new paradigm around then to continue superexponential growth. What was supposed to happen 1913^–₅₇1970_56ya which would have kept growth going and yielded an economic singularity by now? What is the overhang?); “Combinatorial innovation and technological progress in the very long run”, Matt Clancy

• “‘Automation’ of Manufacturing in the Late Nineteenth Century: The Hand and Machine Labor Study”, Atack et al 2019 (more on automation-as-colonization-wave)

Philosophy:

• “Refrigerator safety study: Case study analysis”, Hunter2013 (on refrigerator death: was Bain et al 1958 immoral research because it used ~200 children to establish that magnetic seals were necessary given the surprising behavior of children trapped in a refrigerator, even if it appeared to cause only minor distress and helped cut the death rate from >1.5 deaths/million to <0.5/million?² Apparently some people actually believe that. See also: the ‘debate’ over doing coronavirus challenge trials.)

Film/TV

Live-action:

• American Psycho (review)

• Tokyo Drifter (review)

Animated:

• The Dragon Dentist (review)

• Rage of Bahamut (mediocre JRPG sekaikei anime; the worldbuilding is sketchy and internally incoherent—angels serve “Zeus”?—and far more care has been given to the poorly-integrated and flashy CG than to minor considerations like animating anything of interest, although even there I fear I overpraise the CG of things like its MacGuffin, Bahamut. I’m sure there must have been a reason I put this on my list but whatever it was, it failed.)

Music

MLP:

• “Crepuscularity” (Acoustic Brony {2011_15ya}) [instrumental/acoustic rock]