Podcasts about Stochastic

Outline00:00 - Intro01:07 - Early steps02:47 - Why control?05:20 - The move to the US07:40 - The first journal paper13:30 - What is backstepping?17:08 - Grad school25:10 - Stochastic stabilization29:53 - The interest in PDEs43:24 - Navier-Stokes equations52:12 - Hyperbolic PDEs and traffic models57:51 - Predictors for long delays1:08:14 - Extremum seeking1:27:14 - Safe control1:36:30 - Interplay between machine learning and control1:42:28 - Back to the roots: robust adaptive control1:50:50 - On service1:55:54 - AdviceLinksMiroslav's site: https://flyingv.ucsd.edu/Tuning functions paper: https://tinyurl.com/yznv6r9rP. Kokotović: https://tinyurl.com/mwmbm9yhSeparation and swapping:  https://tinyurl.com/y4fre6t8Adaptive nonlinear stabilizers: https://tinyurl.com/4a9wmmvxKKK book: https://tinyurl.com/2kw2b4k6Stochastic nonlinear stabilization: https://tinyurl.com/4td3537aFollow-up with unknown covariance: https://tinyurl.com/4c4n7fd7Boundary state feedbacks for PIDEs: https://tinyurl.com/4e9y4tdrBoundary Control of PDEs: https://tinyurl.com/d8x38bmjStabilization of Navier–Stokes systems: https://tinyurl.com/4a8cbjemTraffic congestion control: https://tinyurl.com/525jphs5Delay compensation: https://tinyurl.com/5yz6uj9pNonlinear predictors for long delays: https://tinyurl.com/7wvce6vyStability of extremum seeking: https://tinyurl.com/mr5cvzd3Nash equilibrium seeking: https://tinyurl.com/yeywrysnInverse optimal safety filters: https://tinyurl.com/9dkrpvkkNeural operators for PDE control: https://tinyurl.com/5yynsp7vBode lecture: https://tinyurl.com/mp92cs9uCSM article: Support the showPodcast infoPodcast website: https://www.incontrolpodcast.com/Apple Podcasts: https://tinyurl.com/5n84j85jSpotify: https://tinyurl.com/4rwztj3cRSS: https://tinyurl.com/yc2fcv4yYoutube: https://tinyurl.com/bdbvhsj6Facebook: https://tinyurl.com/3z24yr43Twitter: https://twitter.com/IncontrolPInstagram: https://tinyurl.com/35cu4kr4Acknowledgments and sponsorsThis episode was supported by the National Centre of Competence in Research on «Dependable, ubiquitous automation» and the IFAC Activity fund. The podcast benefits from the help of an incredibly talented and passionate team. Special thanks to L. Seward, E. Cahard, F. Banis, F. Dörfler, J. Lygeros, ETH studio and mirrorlake . Music was composed by A New Element.

******Support the channel****** Patreon: https://www.patreon.com/thedissenter PayPal: paypal.me/thedissenter PayPal Subscription 1 Dollar: https://tinyurl.com/yb3acuuy PayPal Subscription 3 Dollars: https://tinyurl.com/ybn6bg9l PayPal Subscription 5 Dollars: https://tinyurl.com/ycmr9gpz PayPal Subscription 10 Dollars: https://tinyurl.com/y9r3fc9m PayPal Subscription 20 Dollars: https://tinyurl.com/y95uvkao   ******Follow me on****** Website: https://www.thedissenter.net/ The Dissenter Goodreads list: https://shorturl.at/7BMoB Facebook: https://www.facebook.com/thedissenteryt/ Twitter: https://x.com/TheDissenterYT   This show is sponsored by Enlites, Learning & Development done differently. Check the website here: http://enlites.com/   Dr. David Wolpert is a Professor at the Santa Fe Institute, external professor at the Complexity Science Hub in Vienna, adjunct professor at ASU, and research associate at the ICTP in Trieste. He has over 30,000 citations, with most of his papers in thermodynamics of computation, foundations of physics, dynamics of social organizations, machine learning, game theory, and distributed optimization / control.   In this episode, we focus on his paper, “The Past as a Stochastic Process”. We first talk about what a stochastic process is, how to study history, a stochastic process framework, and history itself as a stochastic process. We also discuss the jumps in the sociopolitical complexity of polities, narrative approaches in history, and predicting the future. -- A HUGE THANK YOU TO MY PATRONS/SUPPORTERS: PER HELGE LARSEN, JERRY MULLER, BERNARDO SEIXAS, ADAM KESSEL, MATTHEW WHITINGBIRD, ARNAUD WOLFF, TIM HOLLOSY, HENRIK AHLENIUS, FILIP FORS CONNOLLY, DAN DEMETRIOU, ROBERT WINDHAGER, RUI INACIO, ZOOP, MARCO NEVES, COLIN HOLBROOK, PHIL KAVANAGH, SAMUEL ANDREEFF, FRANCIS FORDE, TIAGO NUNES, FERGAL CUSSEN, HAL HERZOG, NUNO MACHADO, JONATHAN LEIBRANT, JOÃO LINHARES, STANTON T, SAMUEL CORREA, ERIK HAINES, MARK SMITH, JOÃO EIRA, TOM HUMMEL, SARDUS FRANCE, DAVID SLOAN WILSON, YACILA DEZA-ARAUJO, ROMAIN ROCH, DIEGO LONDOÑO CORREA, YANICK PUNTER, CHARLOTTE BLEASE, NICOLE BARBARO, ADAM HUNT, PAWEL OSTASZEWSKI, NELLEKE BAK, GUY MADISON, GARY G HELLMANN, SAIMA AFZAL, ADRIAN JAEGGI, PAULO TOLENTINO, JOÃO BARBOSA, JULIAN PRICE, EDWARD HALL, HEDIN BRØNNER, DOUGLAS FRY, FRANCA BORTOLOTTI, GABRIEL PONS CORTÈS, URSULA LITZCKE, SCOTT, ZACHARY FISH, TIM DUFFY, SUNNY SMITH, JON WISMAN, WILLIAM BUCKNER, PAUL-GEORGE ARNAUD, LUKE GLOWACKI, GEORGIOS THEOPHANOUS, CHRIS WILLIAMSON, PETER WOLOSZYN, DAVID WILLIAMS, DIOGO COSTA, ALEX CHAU, AMAURI MARTÍNEZ, CORALIE CHEVALLIER, BANGALORE ATHEISTS, LARRY D. LEE JR., OLD HERRINGBONE, MICHAEL BAILEY, DAN SPERBER, ROBERT GRESSIS, IGOR N, JEFF MCMAHAN, JAKE ZUEHL, BARNABAS RADICS, MARK CAMPBELL, TOMAS DAUBNER, LUKE NISSEN, KIMBERLY JOHNSON, JESSICA NOWICKI, LINDA BRANDIN, NIKLAS CARLSSON, GEORGE CHORIATIS, VALENTIN STEINMANN, PER KRAULIS, ALEXANDER HUBBARD, BR, MASOUD ALIMOHAMMADI, JONAS HERTNER, URSULA GOODENOUGH, DAVID PINSOF, SEAN NELSON, MIKE LAVIGNE, JOS KNECHT, ERIK ENGMAN, LUCY, MANVIR SINGH, PETRA WEIMANN, CAROLA FEEST, STARRY, MAURO JÚNIOR, 航 豊川, TONY BARRETT, BENJAMIN GELBART, NIKOLAI VISHNEVSKY, AND STEVEN GANGESTAD! A SPECIAL THANKS TO MY PRODUCERS, YZAR WEHBE, JIM FRANK, ŁUKASZ STAFINIAK, TOM VANEGDOM, BERNARD HUGUENEY, CURTIS DIXON, BENEDIKT MUELLER, THOMAS TRUMBLE, KATHRINE AND PATRICK TOBIN, JONCARLO MONTENEGRO, AL NICK ORTIZ, NICK GOLDEN, AND CHRISTINE GLASS! AND TO MY EXECUTIVE PRODUCERS, MATTHEW LAVENDER, SERGIU CODREANU, BOGDAN KANIVETS, ROSEY, AND GREGORY HASTINGS!

This week we talk about Luigi Mangione, VAW attacks, and mass shootings.We also discuss stochastic violence, terrorism, and Cybertrucks.Recommended Book: Some Desperate Glory by Emily TeshTranscriptThe terms “Lone Wolf,” “Lone Wolf Actor,” and “Lone Wolf Killer” are interchangeably used in many countries—though most commonly and prominently in the United States—to describe someone who commits a mass-killing or other mass-casualty event, but who is not part of an organization like a terrorist group or other criminal network like a gang.The term is hotly contested in the scholarly world, as it's applied loosely and inconsistently, and the definition varies somewhat by location, government, law enforcement entity investigating said killings, and the press reporting upon it. But in general, to be defined as a mass-casualty event or mass-killing, a collection of murders must occur in public—so it can't be a person killing their family at home, for instance—it must involve at least four victims—so someone killing or injuring three strangers in a public place will typically not be categorized in this way—and it must not occur as part of another crime, like a robbery gone wrong, or as part of a larger conflict between two rival gangs.Within this context of mass-killings and mass-casualty attacks, a lone wolf is someone who acts solo, the term originating with the concept of a wolf that has been separated from, or perhaps outcast from its pack.Someone who kills a bunch of people at the instruction of a terrorist organization like ISIS, then, would not be considered a lone wolf, even if they committed the act without any direct aid from that group; though this definition is wobbly even in that regard, as someone who takes inspiration from a group like ISIS, committing a mass-killing to support that group's cause, but not directly connected to the group, might be labeled a lone wolf, or not. And there's no hard-set rule as to which definition is correct.This was a somewhat common issue back in the late-20th century, when many so-called lone wolf terrorists were committing acts of violence in support of anarchist ends, but the anarchist groups from which they derived their inspiration, and in some cases with which they collaborated, were leaderless by nature—so it couldn't really be said that they were instructed to carry out these acts, they were just inspired by these fellow ideological travelers, and that made determining whether they acted on their own behest or not a tricky and perhaps impossible undertaking; a lot of it is semantics.Also confounding the simple categorization of such killers and attacks is the concept of stochastic terrorism, which is a type of violence that is almost always political or ideological in nature, as opposed to being revenge-driven or otherwise personal, and it's generally incited by someone with a public persona—a politician or other leader—who creates an environment in which violence is more likely to occur, that violence seemingly random, but on average directed in a specific direction.So a politician who says something like “Man, people from the opposing party really believe some horrible stuff, I wouldn't be surprised if something happened to them, considering how evil they are,” while at the same time stoking the flames of potential violence throughout the population by increasing animosity between political parties and maybe even religious groups, might be aiming to spark stochastic terror that would benefit them and their ambitions.By riling up their base in this way, by sowing the seeds for potential attacks against their perceived enemies, violence in their favor, aimed at those enemies, is more likely to happen, but in a way that's deniable for them—just a random act of ideological murder that they can denounce, despite arguably having asymmetrically instigated it.Is stochastic terror an example of planting seeds for violence that makes the resultant killings something more like directed attacks, and therefore not lone wolf in nature, then? Or are all lone wolves arguably inspired by something they've learned or experienced or been told, and thus arguably stochastic in nature—no direct guidance or instruction, but still inspired by someone or something, somewhere along the way?What I'd like to talk about today are three instances of recent supposedly lone wolf attacks, and why some experts are predicting we'll see more such attacks, especially but not exclusively in the US, in the coming years.—There were nearly 500 officially recognized mass-shootings in the US in 2024—and again, that means 4 or more people injured or killed in public, and not as part of another crime being committed.That's down from previous years, the preceding four of which have each had more than 600 mass shootings, and on average a little less than 10 people are killed in these shootings—though that figure is nudged upward by the largest of these mass killings, like one in Las Vegas in 2017 that saw 60 people killed and more than 800 wounded, many in the resulting stampede, by a 64-year-old seemingly lone wolf gunman who fired on an open-air music festival from the 32nd floor window of a nearby hotel.Gun homicides in the US are rampant beyond mass-killings: there were about 21,000 murders committed with guns in the country in 2021, alone—and notably, self-inflicted gun deaths, suicides using these weapons, eclipse that number, tallying more than 26,000 that same year.That means more than 50 people are killed by guns in the US every single day, and about 4 out of every 5 murders are committed using guns in the country; which makes sense, as guns are very effective at what they're meant to do, which is killing something, and there are a lot of guns in the US: about 120 of them per 100 people, as of 2018.And to be clear, that doesn't mean everyone owns a gun: that average is driven sky-high by the gun-enthusiasts who tend to buy a lot of the things, though gun ownership has continued to increase in scope in recent years, as political and economic uncertainty, especially in areas where perception of crime levels, if not always actual elevated crime levels, increases, tends to drive more widespread gun sales.Given all of that, it's maybe not a huge surprise that many apparent lone wolf attacks in the United States are committed using firearms; sometimes assault rifles, sometimes guns that have been augmented using bump-stocks or similar add-ons to make a normal gun into basically an assault rifle, and sometimes just using a pistol, which can be easily pocketed and carried around pretty much everywhere in this country.On December 4, 2024, the CEO of UnitedHealthcare, which is part of the largest health insurance company in the United States, UnitedHealth Group, Brian Thompson, was gunned down in front of the Midtown Manhattan Hilton Hotel.The alleged killer, who was later identified by law enforcement officials as Luigi Mangione, was captured on nearby CCTV cameras, was wearing a hoodie and an expensive backpack while shooting Thompson, and used a pistol with a suppressor—a silencer—to shoot him multiple times, the bullet casings left behind inscribed with the words Delay, Deny, and Depose; terms that have been associated with the US health insurance industry for legal tactics they lean on in order to pocket more money, allegedly at the expense of their customers who have their claims denied or long-term delayed, in some cases leaving them without the care they require, and in some cases leaving them in crippling debt following a necessary medical procedure that the insurance company says they won't pay for.The response to killings of any kind, even in a gun-happy country like the US, tends to be fairly grim and sad; the endless mutterings of “thoughts and prayers” by politicians and other public figures has become so common and toothless as to be near-satire at this point, but generally the tone is antagonistic toward whomever committed the killing, before then swinging toward calls for more security and policing if you're on the political right, and more gun regulation if you're on the political left. And that's generally where we leave things until the next headlines-capturing shooting; and we typically, unfortunately, don't have long to wait.Thompson's murder, though, was almost immediately met with celebration across the political spectrum; working class folks, Democrats and Republicans and everyone in between and on the furthest political extremes basically muttering about how it serves him right, before realizing everyone else was muttering the same thing, and that led to outright enthusiasm, especially online, and even calls for more of the same across the social media landscape—many normal people doing the politician and ideologue thing by basically posting their hopes that someone will knock off other CEOs as well, seemingly aiming to spark more stochastic violence in their favored direction.The wealthy and especially the CEO class were horrified at this response, perhaps understandably, and there was pushback from mainstream journalistic and political entities across the board, with lots of tut-tutting and finger-wagging at anyone who dared celebrate what looked to be the cold-blooded murder of another human being.But the nature of American healthcare and especially health insurance being what it is—massively imperfect at least, and by some assessments borderline abusive or even outright evil—this was seen by many as just desserts for someone who himself had committed millions of dollars worth of fraud and gotten away with it, and who was running UnitedHealthcare in such a way that it denies more claims than any of its peers, which in turn has allowed itself to massively enrich itself and its shareholders at the expense of its customers.There were many cries of “serves him right,” then, alongside some requests that other CEOs be next; many of these requests couched in memes and jokes, but also seemingly earnest.The nature of the alleged killer, who was eventually shown to be a good-looking young man of privilege who had maybe suffered under the auspices of the American healthcare system, due to chronic ailments and an insurance system that didn't even serve someone like him, who grew up with substantial advantages, further fanned those flames, and as of the day I'm recording this he's in custody, has pleaded not guilty, and is facing eleven state and four federal charges, including first-degree murder and a terrorism charge, the former of which could lead to the death penalty.Just shy of a month later, in the early morning hours of January 1, 2025, a new year's celebration on the well-know and well-traversed, and on that night, incredibly crowded Bourbon Street in New Orleans was attacked by a man in a large pickup truck, who plowed the vehicle into a crowd of revelers, driving at high-speed across three blocks that were partitioned-off for the celebration.The driver was apparently trying to hit as many people as possible, and then, after crashing into a utility vehicle, he stepped out of the truck and started firing a gun into the crowd.Police fired back at him, but he was wearing body armor, and two of them were injured before they managed to kill him, recovering an assault rifle and a semi-automatic pistol from his body. They also found a pair of explosive devices in coolers he had planted around the area before the attack, and further investigation led to the discovery of more bomb-making materials where he was staying in New Orleans.At least 35 people were injured and 14 people were killed in the attack, alongside the killer, who was later identified as 42-year-old Shamsud-Din Jabbar: an American-born Army veteran and Texas resident who had apparently been recently radicalized, possibly by online content posted by ISIS, and who had posted videos pledging his allegiance to the group mere hours before he drove into the crowd, an ISIS flag adorning the vehicle.More guns in this attack, then, but much of the damage was caused by the truck, and similar so-called “vehicle as a weapon,” or VAW attacks have been committed around the world in recent years, raising concerns especially in places where firearms are harder to come by, though also at large, open-air events where vehicles might cause more deaths and injuries in a short period of time than even an assault rifle, as seemed to be the case here.This attacker seemed to be self-radicalized, based on testimony from his friends and family, who were shocked at the change in his personality and expressed beliefs. The FBI has said they're pretty confident he acted alone, though they're looking into recent trips he took to Egypt and Canada, in case he met up with someone from ISIS or a similar group, while traveling.And apparently while he initially planned to kill his family—he's had several divorces that led to financial problems, due to many child support payments that exceeded his means—he didn't believe killing his family would have provoked enough of a response to spark a “war between the believers and the disbelievers.” Jabbar was brought up Muslim but left the faith for years, before apparently adopting a more intense and violent reinterpretation of it just recently, and that seemingly helped him justify and perhaps even inspired these acts.This has been called a lone wolf attack, then, but it was apparently heavily influenced by ISIS ideology, despite Jabbar possibly never having been in contact with anyone from that group.Just a handful of hours later, that same morning, at 8:49 January 1, 2025, a Tesla Cybertruck that was parked outside the front lobby of the Trump International Hotel in Las Vegas exploded—its occupant apparently having died from a self-inflicted gunshot wound to the head right before a bunch of fireworks and gas canisters placed in the trunk were detonated.That occupant was Matthew Alan Livelsberger, who was also American-born, and like Jabbar, had been in the US military, though Livelsberger was an active-duty Special Forces soldier from Colorado who was on leave at the time.The blast didn't kill anyone, and while it hurt a few bystanders, no one was seriously injured. But the intention, according to two letters recovered from his phone by the FBI, was apparently to make a political statement related to alleged clandestine US military operations, and advanced technologies the US and China allegedly secretly possess; though he was also apparently in the midst of a serious mental health crisis, including significant PTSD episodes and what might have been paranoid delusions.The vehicle also contained an assault rifle and two pistols, though none of these weapons were used, as while Livelsberger was seemingly intent on escaping across the Mexican border following the attack, based on what he said in those aforementioned letters, he seemingly decided to kill himself instead—which may support the assertion that this was primarily, if not exclusively, a mental health crisis issue.Livelsberger also apparently had family issues, due in large part to his support of president-elect Trump and his family's opposition to that support, and he was apparently suffering from untreated depression, that lack of treatment possibly the result of stigma toward such things within the military, which sometimes results in people not getting treatment that they might benefit from, because they worry doing so will see them sidelined by their superiors.A manifesto penned by Livelsberger that was sent to a retired Army Intelligence officer claims that he was being monitored by the military because of his knowledge of war crimes and those aforementioned military advanced technologies, and that he didn't intend to self-harm, the divulgence of which has led to some conspiracy theories about this not having been a suicide.That said, this attack is being investigated as potential terrorism, and while it was initially being explored as part of a larger wave of such actions, since that attack in New Orleans happened just hours earlier, and both attackers used the same online car rental service to procure the vehicles they were driving, investigators have since indicated they don't believe these attacks were connected.Interestingly, Livelsberger's letters also criticized income inequality, though with a politically conservative bent, basically saying that the country had become too liberal and effeminate, and that Trump, Elon Musk, and Robert Kennedy Jr needed to take control and make the US more masculine so that it could compete against entities like China, Russia, and Iran.Experts on ideological violence and political fracturing have warned that we may see more lone wolf and lone wolf-esque violence in a more polarized society, in which people are less likely to consider those on the opposite side of the aisle to be people they disagree with, and more likely to think of them as bad or evil or even subhuman, which makes violence more thinkable.That's not ideal, as these sorts of attacks are difficult to prevent, their solo nature meaning there's no network to track and pluck apart, nothing to infiltrate and fewer easily accessible data points to aggregate and in which to recognize a pattern. Lone wolf attackers tend to cause less damage than groups can, then, but they're often almost invisible, to the organizations that hope to stop them, anyway, right up till the moment they start killing and injury people.We're also entering an era in which trust in authority has degraded substantially, new technologies have made the research, hardware procurement, and implementation of such attacks a lot more attainable to more people, which means folks suffering from different sorts of psychological or physical torments, or those who simply have strong opinions and a lot of perceived enemies, are more likely to be able to act on that confusion or those hatreds, in some cases at a moment's notice, and in many cases without anyone beyond their immediate friends and family recognizing that something might be up.We may be entering a period of heightened threat, then, in the US especially, because of the number and wide distribution of highly effective weapons throughout the population, and because of the period of political polarization and animosity we seem to be wading through, but also throughout the rest of the world, to some degree at least, because of those same political and ideological factors, and because of how big and weapon-like vehicles have become, and how relatively easy it is to get one's hands on information that allows for the construction of things like bombs and the technologies required to 3D-print and otherwise manufacture deadly implements of all shapes and sizes.Show Noteshttps://en.wikipedia.org/wiki/Luigi_Mangionehttps://www.vox.com/politics/390438/luigi-mangione-healthcare-shooting-ghost-gunhttps://en.wikipedia.org/wiki/Killing_of_Brian_Thompsonhttps://en.wikipedia.org/wiki/2017_Las_Vegas_shootinghttps://www.bbc.com/news/world-us-canada-41488081https://en.wikipedia.org/wiki/2025_New_Orleans_truck_attackhttps://www.bbc.com/news/articles/c205ek63433ohttps://www.nytimes.com/2025/01/03/us/new-orleans-victims-truck-attack.htmlhttps://en.wikipedia.org/wiki/Trump_International_Hotel_Las_Vegas_Tesla_Cybertruck_explosionhttps://www.nytimes.com/2025/01/04/us/matthew-livelsberger-las-vegas-cybertruck.htmlhttps://www.nytimes.com/2025/01/05/us/new-orleans-attack-travel.htmlhttps://www.nytimes.com/2025/01/04/us/new-orleans-attack-shamsud-din-jabbar-isis.htmlhttps://scholarlycommons.law.northwestern.edu/cgi/viewcontent.cgi?article=1257&context=nulr_online&preview_mode=1&z=1519320539https://en.wikipedia.org/wiki/Stochastic_terrorismhttps://en.wikipedia.org/wiki/Lone_wolf_attackhttps://www.ojp.gov/ncjrs/virtual-library/abstracts/lone-wolf-terrorism-americahttps://journals.sagepub.com/doi/10.1177/1088767917736797 This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

We talk about the Cybertruck psyop with Insurgent Notes, Race Traitor, and Hard Crackers vets John Garvey and Zhana Kurti.In the fun half, available to subscribers at http://patreon.com/theantifada, we read the prophecies of alien intervention and free energy by the Balkan Nostradamus Baba Vanga, and hear Sean's interpretation of Charli XCX's Brat.Song: Kendrick Lamar - wacced out murals

久々の選手回、JCL TEAM UKYO所属の石橋学選手（@manabu_1484）のキャリア前半の話、確率的なレースの組み立て方、アシスト、Strava表彰などなどについて高田くん（@offcourseyass）話を聞きました。 お便りはTwitterのハッシュタグ #sxsradio か、番組のTwitterダイレクトメールまで。番組をサポートする投げ銭はPatreonページ でお願いします。 サポートグッズも販売しています！Tシャツやパーカー、トートバッグはUT Me!から。マグカップ・iPhoneケース・クージーなどはsuzuriより購買できます Show Notes 上野みなみ氏の話 元シマノレーシングの湊諒氏 青森山田高等学校 黒枝士揮・山本元喜 徳田鍛造・中里仁 2015ジロ出場 伝染性単核球症 日本で一番登る人（Strava） Special Thanks to 小泉享亮 様

How far can following your passions take you in science? To find out, Dr. Charles Liu and co-host Allen Liu welcome the multi-faceted Leah Voytovich, a software development engineer for Project Kuiper who planned on going to med school but instead ended up working on space lasers for Amazon's satellite internet constellation. As always, though, we start off with the day's joyfully cool cosmic thing: the use of deuterium levels in the search for exoplanets with advanced civilizations. That's because signs of deuterium depletion can indicate that a planet has developed nuclear fission power.  After that, Leah tells us a little about Amazon's Project Kuiper. She explains the wide variety of challenges to keeping the satellites, which are in low earth orbit, in communication with the ground. The project she's working on now is using infrared space lasers to communicate more securely within the satellite constellation.  Chuck brings up the impact that these massive groups of satellites like Starlink are having on astronomy, from satellite streaks on dark sky surveys to image-destroying flares. Leah said there are people working on reducing Project Kuiper's impacts across multiple areas of concern. For our first question. Yasmin asks, “Will our satellites all be controlled by AI someday? If so, when? And is this a good idea?”  Leah speculates that there may be some projects already using Machine Learning here on Earth, but that it will be further down the line before there's AI actually present in satellites themselves. Leah doesn't see a problem with the use of ML tools for science, as long as there are protections in place and that humans are still writing the software. After our break, Leah tells us about what she does outside her work – literally – from climbing Mount St. Helens in winter using an ice ax and crampons, to skiing, to running marathons – 3 so far! Then we're back to AI, ML, and the differences between the two. Charles talks about how important AI has beome as a tool for astronomers and other scientists. Leah explains the “Chinese Room Argument” and why it's a good analogy for machine learning, Allen shares the description of AI as a “stochastic parrot.”  Our next question comes from Jonathan, who asks, “With so many problems facing the world like poverty, war and the climate crisis here on Earth, why should we spend money and resources on space exploration and astronomy?” Leah talks about how Project Kuiper is intended to serve underserved communities by providing internet access to people who don't already have it. That in turn increases global communication and connectivity and can then help address those other problems. And of course, to launch those satellites, we need rockets. She also explains how valuable internet access is to communities– something she knows a little about. She and Martin Leet co-founded Maji, a nonprofit in Uganda that uses solar power to make clean water easily accessible for refugee communities. Leah, a former EMT, explains how Maji also provides first aide medical training and agricultural training to the communities. The story of how Maji came to be is remarkable – we'll leave it to Leah to share that with you in the episode. We end with Leah recounting what it was like to be in Mission Control for the launch of the first Project Kuiper satellites. She was the representative for her team in the Mission Control room, so she was the one who actually got to say “Go” for her service during the final pre-launch Go/No Go review.  If you'd like to know more about Maji, you can check out the nonprofit's website here. We hope you enjoy this episode of The LIUniverse, and, if you do, please support us on Patreon. Credits for Images Used in this Episode: Illustration of Kepler-138 exoplanets– NASA, ESA, Leah Hustak (STScI) Deuterium, aka hydrogen-2 (²H or D) – Center for Deuterium Depletion An axonometric view displays various Earth orbits, illustrating space debris and active satellites – WikiCommons/Pablo Carlos Budassi The history of the universe since the Big Bang – NASA Satellite streaks in an astronomical image – University of Washington/ Abhilash Biswas, Kilando Chambers, and Ashley Santos Space-based lasers – IEEE Spectrum Illustration of NASA's Laser Communications Relay Demonstration (LCRD) communicating with the I.S.S. over laser links – NASA's Goddard Space Flight Center Illustration of machine learning and artificial intelligence – Creative Commons / mikemacmarketing Computer simulation of dendrites of pyramidal neurons – Hermann Cuntz/ PLoS Computational Biology, Vol. 6(8) August 2010. Earth illuminated time-lapse video if ISS over Earth – NASA Martin Leet and Leah Voytovich – University of Pennsylvania The Maji water tank for the Olua I community – University of Pennsylvania/Martin Leet

In this lecture, we prepare for the final exam and give a brief review of all topics from the course. Students are encouraged to bring their own questions so that the focus of the class is on the topics that students feel they need the most help with.

In this lecture, we wrap up the course content in IEE 475. We first do a quick overview of the four variance reduction techniques (VRT's) covered in Unit K. That is, we cover: common random numbers (CRN's), antithetic variates (AV's), importance sampling, and control variates. We then  remember some general comments about the goal of modeling and commonalities seen across simulation platforms (as well as the different types of simulation platforms in general).

In this lecture, we review four different Variance Reduction Techniques (VRT's). Namely, we discuss common random numbers (CRNs), control variates, antithetic variates (AVs), and importance sampling. Each one of these is a different approach to reducing the variance in the estimation of relative or absolute performance of a simulation model. Variance reduction is an alternative way to increase the power of a simulation that is hopefully less costly than increasing the number of replications.

In this lecture, we start by reviewing approaches for absolute and relative performance estimation in stochastic simulation. This begins with a reminder of the use of confidence intervals for estimation of performance for a single simulation model. We then move to different ways to use confidence intervals on mean DIFFERENCES to compare two different simulation models. We then move to the ranking and selection problem for three or more different simulation models, which allows us to talk about analysis of variance (ANOVA) and post hoc tests (like the Tukey HSD or Fisher's LSD). After that review, we move on to introducing variance reduction techniques (VRTs) which reduce the size of confidence intervals by experimentally controlling/accounting for alternative sources of variance (and thus reducing the observed variance in response variables). We discuss Common Random Numbers (CRNs), which use a paired/blocked design to reduce the variance caused by different random-number streams. We start to discuss control variates (CVs), but that discussion will be picked up at the start of the next lecture.

In this lecture, we review what we have learned about one-sample confidence intervals (i.e., how to use them as graphical versions of one-sample t-tests) for absolute performance estimation in order to motivate the problem of relative performance estimation. We introduce two-sample confidence intervals (i.e., confidence intervals on DIFFERENCES based on different two-sample t-tests) that are tested against a null hypothesis of 0. This means covering confidence interval half widths for the paired-difference t-test, the equal-variance (pooled) t-test, and Welch's unequal variance t-test. Each of these different experimental conditions sets up a different standard error of the mean formula and formula for degrees of freedom that are used to define the actual confidence interval half widths (centered on the difference in sample means in the pairwise comparison of systems). We then generalize to the case of more than 2 systems, particularly for "ranking and selection (R&S)." This lets us review the multiple-comparisons problem (and Bonferroni correction) and how post hoc tests (after an ANOVA) are more statistically powerful ways to do comparisons.

In this lecture, we review estimating absolute performance from simulation, with focus on choosing the number of necessary replications of transient simulations of terminating systems. The lecture starts by overviewing point estimation, bias, and different types of point estimators. This includes an overview of quantile estimation and how to use quantile estimation to use simulations as null-hypothesis-prediction generators. We the introduce interval estimation with confidence intervals and prediction intervals. Confidence intervals, which are visualizations of t-tests, provide an alternative way to choose the number of required replications without doing a formal power analysis.

In this lecture, we start by further reviewing confidence intervals (where they come from and what they mean) and prediction intervals and then use them to motivate a simpler way to determine how many replications are needed in a simulation study (focusing first on transient simulations of terminating systems). We then shift our attention to steady-state simulations of non-terminating systems and the issue of initialization bias. We discuss different methods of "warming up" a steady-state simulation to reduce initialization bias and then merge that discussion with the prior discussion on how to choose the number of replications. In the next lecture, we'll finish up with a discussion of the method of "batch means" in steady-state simulations.

 In this lecture, we review statistical fundamentals – such as the origins of the t-test, the meaning of type-I and type-II error (and alternative terminology for both, such as false positive rate and false negative rate) and the connection to statistical power (sensitivity). We review the Receiver Operating Characteristic (ROC) curve and give a qualitative description of where it gets its shape in a hypothesis test. We close with a validation example (from Lecture H) where we use a power analysis on a one-sample t-test to help justify whether we have gathered enough data to trust that a simulation model is a good match for reality when it has a similar mean output performance to the real system.

In this lecture, we introduce the estimation of absolute performance measures in simulation – effectively shifting our focus from validating input models to validating and making inferences about simulation outputs. Most of this lecture is a review of statistics and reasons for the assumptions for various parametric and non-exact non-parametric methods. We also introduce a few more advanced statistical topics, such as non-parametric methods and special high-power tests for normality. We then switch to focusing on simulations and their outputs, starting with the definition of terminating and non-terminating systems as well as the related transient and steady-state simulations. We will pick up next time with discussing details related to performance measures (and methods) for transient simulations next time and steady-state simulations after that. Our goal was to discuss the difference between point estimation and interval estimation for simulation, but we will hold off to discuss that topic in the next lecture.

During this lecture slot, we start with slides from Lecture G3 (on goodness of fit) that were missed during the previous lecture due to timing. In particular, we review hypothesis testing fundamentals (type-I error, type-II error, statistical power, sensitivity, false positive rate, true negative rate, receiver operating characteristic, ROC, alpha, beta) and then go into examples of using Chi-squared and Kolmogorov–Smirnov tests for goodness of fit for arbitrary distributions. We also introduce Anderson–Darling (for flexibility and higher power) and Shapiro–Wilk (for high-powered normality testing).We close with where we originally intended to start – with definitions of testing, verification, validation, and calibration. We will pick up from here next time.

In this lecture, we (nearly) finish our coverage of Input Modeling, where the focus of this lecture is on parameter estimation and assessing goodness of fit. We review input modeling in general and then briefly review fundamentals of hypothesis testing. We discuss type-I error, p-values, type-II error, effect sizes, and statistical power. We discuss the dangers of using p-values at very large sample sizes (where small p-values are not meaningful) and at very small sample sizes (where large p-values are not meaningful). We give some examples of this applied to best-of-7 sports tournaments and voting. We then discuss different shape parameters (including location, scale, and rate), and then introduce summary statistics (sample mean and sample variance) and maximum likelihood estimation (MLE), with an example for a point estimate of the rate of an exponential. We introduce the chi-squared (lower power) and Kolmogorov–Smirnov (KS, high power) tests for goodness of fit, but we will go into them in more detail at the start of the next lecture.

In this lecture, we continue discussing the choice of input models in stochastic simulation. Here, we pivot from talking about data collection to selection of the broad family of probabilistic distributions that may be a good fit for data. We start with an example where a histogram leads us to introduce additional input models into a flow chart. The rest of the lecture is about choosing models based on physical intuition and the shape of the sampled data (e.g., the shape of histograms). We close with a discussion of probability plots – Q-Q plots and P-P plots, as are used with "fat-pencil tests" – as a good tool for justifying the choice of a family for a certain data set. The next lecture will go over the actual estimation of the parameters for the chosen families and how to quantitatively assess goodness of fit.

In this lecture, we introduce the detailed process of input modeling. Input models are probabilistic models that introduce variation in simulation models of systems. Those input models must be chosen to match statistical distributions in data. Over this unit, we cover collection of data for this process, choice of probabilistic families to fit to these data, and then optimized parameter choice within those families and evaluation of fit with goodness of fit. In this lecture, we discuss issues related to data collection.

Bitcoin Stochastic BREAKOUT to $500,000 in play. BTC has the same pattern as the previous cycle. The breakout point indicates the local bottom at the root of MASSIVE Parabolic RALLY. Learn more about your ad choices. Visit megaphone.fm/adchoices

This is the episode that SHOULD have aired Sunday BlogTalk was having a breakdown.       Donald Trump was once again, the target of an assassination attempt near his golf course adjacent Mar-a-lago. The would be, hapless assailant was caught in a nano second. Of course, "Trump being Trump"--he had to whine about this near miss. Now, I don't condone violence. Period. That being said, Trump rapidly took advantage of this attempt, accusing democrats of 'inciting' violence. Frankly, the irony is too sharp to ignore. For 8 long years, Trump has routinely incited violence against minorities, democrats, and anyone he just didn't like--or anyone who failed to kiss his ring.This incitement used by Trump is called 'stochastic terrorrism." The January 6th insurrection was the most damning thus far--which Trump and his political lackeys clearly incited--and THEY have the gall to shriek foul! The attempted assailant by all reports thus far, looks like a crazy lone wolf and nothing more, and yet Trump's loyalists in Congress are accusing their democratic colleagues of 'stochastic terrorism.' The sense of white, Christian male entitlement is palpable in it's hypocrisy. I will discuss the true danger of Trump and his loyalists as they spread propagandistic lies about anyone they dislike.  BTW, it is NOT incitement when journalists, politicians, and citizens report the TRUTH about how Trump has consistently used the rhetoric of Adolph Hitler--though thinly veiled in 21st century colloquialisms.  Our second story deals with Project 2025's plans for a new monitoring function--namely the 'menstrual police,' as described by journalist Thom Hartmann. This is not satire. The plan is REAL.  We will also have a variety of parodies.  Come join me.  Jeanine

Author and columnist Wajahat Ali: Donald Trump's Repeated Stochastic Terrorism Now Endangers Haitian Immigrants in OhioStop the Money Pipeline Executive Director Alec Connon: 13-Week ‘Summer of Heat' Climate Protest Building Movement to End Fossil Fuel FinancingSinger/songwriter Billy Bragg Reflects on His 40 years of Political Activism and the Power of MusicBob Nixon's Under-reported News Summary• EU's top court rules Apple must pay back taxes• Project 2025 regulates the nation's public lands• Environmental groups challenge new EPA rulesVisit our website at BTLonline.org for more information, in-depth interviews, related links and transcripts and to sign up for our BTL Weekly Summary. New episodes every Wednesday at 12 noon ET, website updated Wednesdays after 4 p.m. ETProduced by Squeaky Wheel Productions: Scott Harris, Melinda Tuhus, Bob Nixon, Anna Manzo, Susan Bramhall, Jeff Yates and Mary Hunt. Theme music by Richard Hill and Mikata.

Jesse's grim prediction on Trump's assassin. Rome finding its footing on the world state. The continued rhetoric from the left for a decade. Stochastic terrorism from the left. Communist do the exact same thing every time they force a revolution. Democrats radicalize street animals to violence. See omnystudio.com/listener for privacy information.

This episode is presented by Create A Video – For people who have spent years decrying political rhetoric that they say inspires violence, the media and Democrats (but I repeat myself) seem unable to identify their own indulgences in "stochastic terrorism". Help Pete's for the Walk to End Alzheimer's by going here. Subscribe to the podcast at: https://ThePeteKalinerShow.com/  All the links to Pete's Prep are free: https://patreon.com/petekalinershow  Advertising inquiries: Pete@ThePeteKalinerShow.com  Get exclusive content here!: https://thepetekalinershow.com/See omnystudio.com/listener for privacy information.

最近気になった記事について話しました。 AIとはなんなのか？ https://www.technologyreview.com/2024/07/10/1094475/what-is-artificial-intelligence-ai-definitive-guide/ 確率論的オウム https://en.wikipedia.org/wiki/Stochastic_parrot フレーム問題 シンボルグラウンディング問題 (記号接地問題) クオリア AirbnbがPdMをなくした話 https://jonharmer.medium.com/oh-no-airbnb-got-rid-of-product-managers-the-sky-is-falling-f12559ce9c9c 炎上元のツイートhttps://twitter.com/PawelHuryn/status/1677602999194009600 実際の動画 https://www.youtube.com/watch?v=Dkfijg7s76o&t=645s AWS CDKを使用したモノレポサーバーレスプロジェクトの実装https://tomoima525.hatenablog.com/entry/2024/08/20/214950 感想をぜひハッシュタグ #todayILearnedFM #tilfm でつぶやいてください! Your co-hosts: Tomoaki Imai, Noxx CTO https://twitter.com/tomoaki_imai Ryoichi Kato, Software Engineer ⁠https://twitter.com/ryo1kato

Professor Pedro Domingos, is an AI researcher and professor of computer science. He expresses skepticism about current AI regulation efforts and argues for faster AI development rather than slowing it down. He also discusses the need for new innovations to fulfil the promises of current AI techniques. MLST is sponsored by Brave: The Brave Search API covers over 20 billion webpages, built from scratch without Big Tech biases or the recent extortionate price hikes on search API access. Perfect for AI model training and retrieval augmented generation. Try it now - get 2,000 free queries monthly at http://brave.com/api. Show notes: * Domingos' views on AI regulation and why he believes it's misguided * His thoughts on the current state of AI technology and its limitations * Discussion of his novel "2040", a satirical take on AI and tech culture * Explanation of his work on "tensor logic", which aims to unify neural networks and symbolic AI * Critiques of other approaches in AI, including those of OpenAI and Gary Marcus * Thoughts on the AI "bubble" and potential future developments in the field Prof. Pedro Domingos: https://x.com/pmddomingos 2040: A Silicon Valley Satire [Pedro's new book] https://amzn.to/3T51ISd TOC: 00:00:00 Intro 00:06:31 Bio 00:08:40 Filmmaking skit 00:10:35 AI and the wisdom of crowds 00:19:49 Social Media 00:27:48 Master algorithm 00:30:48 Neurosymbolic AI / abstraction 00:39:01 Language 00:45:38 Chomsky 01:00:49 2040 Book 01:18:03 Satire as a shield for criticism? 01:29:12 AI Regulation 01:35:15 Gary Marcus 01:52:37 Copyright 01:56:11 Stochastic parrots come home to roost 02:00:03 Privacy 02:01:55 LLM ecosystem 02:05:06 Tensor logic Refs: The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World [Pedro Domingos] https://amzn.to/3MiWs9B Rebooting AI: Building Artificial Intelligence We Can Trust [Gary Marcus] https://amzn.to/3AAywvL Flash Boys [Michael Lewis] https://amzn.to/4dUGm1M

"...we wouldn't want to put somebody up on a sloped roof."—Kimberly Cheatle Once again out of the blue—or out of the red, white and blue (sorry, I couldn't resist)—I'm putting out another short podcast, this time in response to the events of last Saturday in Pennsylvania, in which Donald Trump, as we all know, was very nearly assassinated by an allegedly "lone gunman" in front of the watching world. It's amazing how the "news cycle" speeds along though, isn't it? Already it's getting harder to find mention of it in "mainstream" news outlets. But this was a momentous event—and one that, as nearly everyone is saying, may well propel Trump into the White House for a second term (assuming another assassination attempt isn't successful). Anyway, essentially what I'm doing here is to "riff" in audio form on some notes that I've been making at TMR over the last several days. It's not a deep analysis; I'm not claiming that I've got everything right, and I'm certainly open to changing my mind on any of it as I'm exposed to new information. So, please take it for what it is: thoughts, musings, trying to make sense of what happened... and basically being someone, who, like you, has the audacity (nay, effrontery) to think about matters that one is told one should leave entirely to unimpeachable organs of the state (like the FBI). [For show notes please visit https://themindrenewed.com]

"...we wouldn't want to put somebody up on a sloped roof."—Kimberly Cheatle Once again out of the blue—or out of the red, white and blue (sorry, I couldn't resist)—I'm putting out another short podcast, this time in response to the events of last Saturday in Pennsylvania, in which Donald Trump, as we all know, was very nearly assassinated by an allegedly "lone gunman" in front of the watching world. It's amazing how the "news cycle" speeds along though, isn't it? Already it's getting harder to find mention of it in "mainstream" news outlets. But this was a momentous event—and one that, as nearly everyone is saying, may well propel Trump into the White House for a second term (assuming another assassination attempt isn't successful). Anyway, essentially what I'm doing here is to "riff" in audio form on some notes that I've been making at TMR over the last several days. It's not a deep analysis; I'm not claiming that I've got everything right, and I'm certainly open to changing my mind on any of it as I'm exposed to new information. So, please take it for what it is: thoughts, musings, trying to make sense of what happened... and basically being someone, who, like you, has the audacity (nay, effrontery) to think about matters that one is told one should leave entirely to unimpeachable organs of the state (like the FBI). [For show notes please visit https://themindrenewed.com]

-We talk about the Trump assassination attempt, looking back at the violence and lawlessness the media and the Democrats have been pushing for right up until they got what they wanted.  -The "women are wonderful" effect is in effect all the time. Society struggles to accept the reality of violent women (abusive or murderous mothers, angel of death nurses) even as they see it with their own eyes.  -Potpourri du Moquerie: wig malfunction edition ******************************************************************************** Disaffected is sponsored by the best cured meat you'll ever have. Go to landofbiltong.com and pick up some cured beef or game. It's good for you, no-carb, and made by a family firm who's not afraid to support our spicy show! Use promo code JOSH to get free shipping!See omnystudio.com/listener for privacy information.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Towards shutdownable agents via stochastic choice, published by EJT on July 8, 2024 on LessWrong. We[1] have a new paper testing the Incomplete Preferences Proposal (IPP). The abstract and main-text is below. Appendices are in the linked PDF. Abstract Some worry that advanced artificial agents may resist being shut down. The Incomplete Preferences Proposal (IPP) is an idea for ensuring that doesn't happen. A key part of the IPP is using a novel 'Discounted REward for Same-Length Trajectories (DREST)' reward function to train agents to: 1. pursue goals effectively conditional on each trajectory-length (be 'USEFUL') 2. choose stochastically between different trajectory-lengths (be 'NEUTRAL' about trajectory-lengths). In this paper, we propose evaluation metrics for USEFULNESS and NEUTRALITY. We use a DREST reward function to train simple agents to navigate gridworlds, and we find that these agents learn to be USEFUL and NEUTRAL. Our results thus suggest that DREST reward functions could also train advanced agents to be USEFUL and NEUTRAL, and thereby make these advanced agents useful and shutdownable. 1. Introduction 1.1. The shutdown problem Let 'advanced agent' refer to an artificial agent that can autonomously pursue complex goals in the wider world. We might see the arrival of advanced agents within the next few decades. There are strong economic incentives to create such agents, and creating systems like them is the stated goal of companies like OpenAI and Google DeepMind. The rise of advanced agents would bring with it both benefits and risks. One risk is that these agents learn misaligned goals: goals that we don't want them to have [Leike et al., 2017, Hubinger et al., 2019, Russell, 2019, Carlsmith, 2021, Bengio et al., 2023, Ngo et al., 2023]. Advanced agents with misaligned goals might try to prevent us shutting them down [Omohundro, 2008, Bostrom, 2012, Soares et al., 2015, Russell, 2019, Thornley, 2024a]. After all, most goals can't be achieved after shutdown. As Stuart Russell puts it, 'you can't fetch the coffee if you're dead' [Russell, 2019, p.141]. Advanced agents with misaligned goals might resist shutdown by (for example) pretending to have aligned goals while covertly seeking to escape human control [Hubinger et al., 2019, Ngo et al., 2023]. Agents that succeed in resisting shutdown could go on to frustrate human interests in various ways. 'The shutdown problem' is the problem of training advanced agents that won't resist shutdown [Soares et al., 2015, Thornley, 2024a]. 1.2. A proposed solution The Incomplete Preferences Proposal (IPP) is a proposed solution to the shutdown problem [Thornley, 2024b]. Simplifying slightly, the idea is that we train agents to be neutral about when they get shut down. More precisely, the idea is that we train agents to satisfy: Preferences Only Between Same-Length Trajectories (POST) 1. The agent has a preference between many pairs of same-length trajectories (i.e. many pairs of trajectories in which the agent is shut down after the same length of time). 2. The agent lacks a preference between every pair of different-length trajectories (i.e. every pair of trajectories in which the agent is shut down after different lengths of time). By 'preference,' we mean a behavioral notion [Savage, 1954, p.17, Dreier, 1996, p.28, Hausman, 2011, §1.1]. On this notion, an agent prefers X to Y if and only if the agent would deterministically choose X over Y in choices between the two. An agent lacks a preference between X and Y if and only if the agent would stochastically choose between X and Y in choices between the two. So in writing of 'preferences,' we're only making claims about the agent's behavior. We're not claiming that the agent is conscious or anything of that sort. Figure 1a presents a simple example of POST-satisfying ...

"The original field recording from Scotland was processed via Stochastic synthesis and various FX. No additional instruments were added." Glenmore stream reimagine by Scott Lawrence Whitman.

SERIES 2 EPISODE 150: COUNTDOWN WITH KEITH OLBERMANN A-Block (1:44) SPECIAL COMMENT: It was a presidential kidnapping and assassination video. It showed a pickup truck, Trump flags flapping in the wind, and on its tailgate, with an illustration of President Biden bound and gagged and thrown in the flatbed. Trump posted the video and he knew what he was doing: it is a stochastic attempt to GET Biden kidnapped or assassinated.  It all makes me again ask a chilling question I first posed months ago. Had any of Trump's various coup attempts succeeded in 2020 – if any of the coup attempts Trump is planning NOW for this election  – exactly what is Trump's plan, or the plans of his thugs and gangs cultists and psychotics and death fetishists and militias and god-and-gun lunatics – for the ACTUAL president he usurps? Now Trump has given us the answer. Have the pick-up truck mob grab Biden and tie him up and throw him in the back of the truck. And the listless, clueless, spineless media as personified as Kristen Welker of NBC, reduces it to: “It is yet another reminder that we are covering this election against the backdrop of a deeply divided nation.” Yeah - like we were divided during the Civil War. If the media had not failed us sufficiently in this case, wait'll you hear the Trump story that wasn't merely covered poorly - it wasn't covered at all. In a move I have long anticipated, Trump minions are begin to soften up potential opposition to repealing or ignoring the 22nd Amendment's presidential term limits and dare us to stop Trump running for re-election in 2028. They have produced nonsensical sophistry in "The American Conservative" that boils down to this: because Trump would not have served CONSECUTIVE terms, the 22nd Amendment wouldn't apply to him.  Disagree? Sue him. Maybe you'll get lucky and find a different Supreme Court that will agree with you and actually defend the Constitution. If not: welcome to the Trump Dictatorship. B-Block (38:30) THE WORST PERSONS IN THE WORLD: Such a pile there are EIGHT award winners. I had to expand the medal stand and hand out aluminum in addition to bronze, silver and gold. Your winners? The owner of the San Francisco Giants, the North Carolina GOP State Superintendent of Schools candidate who belongs to an organization that believes Obama is Hitler's grandson, Mayor Adams of New York, Trump's Media Company, Trump's $200 Burger Buy, Roseanne Barr, The Ohtani-deaf New York Yankees, and Representative Mike Turner. C-Block (51:42) THINGS I PROMISED NOT TO TELL: I outlasted five heads of MSNBC and three chiefs of NBC News (one of them twice) and it looks like I'm about to outlast a fourth. The Ronna McDaniel Scandal continues and the fingers and figurative knives are all pointed towards the Chairman of NBC News Cesar Condé. I have my own history with him, and with the predecessor he most reminds me for pure self-absorption: Andy Lack. Their stories in an all-new edition of Countdown.See omnystudio.com/listener for privacy information.

In this transformative episode of the On the Rocks Podcast, hosted by Emily King and sponsored by KPI Mining, experts Benny Cohen, Matheus Faria, and Giovanni Santos dive deep into the world of mining optimization through the lens of stochastic mine planning. They shed light on how traditional challenges of uncertainty in mining operations can be mitigated by embracing stochastic optimization techniques, originally honed in the oil and gas industry. The discussion explores the significant impacts of this approach, including improved Net Present Value (NPV), enhanced recovery rates, and more accurate forecasts, thereby enabling mining companies to make better-informed decisions and optimize their operations for higher efficiency and profitability.

The Mona Lisa is disrespected, Title IX defenders, The Stochastic Terrorism Bookfair, neo-cervix screenings, the world's largest penis, Vince McMahon is super MeToo'ed, China makes driving safer, Disney swingers, arrested for jokes in a group chat, a guy gives away money, and Madison Madnes reads the news; all that and more this week on The Dick Show!

New Discourses Bullets, Ep. 80 Entryism (https://newdiscourses.com/2023/12/communist-entryism/) is the idea of infiltrating an institution with hostile subversives. It's a name for a variety of strategies for getting them in where they're not wanted. DEI (Diversity, Equity, and Inclusion) programs achieve an entryist goal through an indirect method that involves a bit of randomness. That is, they're stochastic entryism. The way they work is to increase the probability of hiring people who are activists or who can be made activists through further training without intentionally selecting for activists in a direct way. This maintains plausible deniability about the goal of the program. Simultaneously, it pushes people who reject the Woke Marxist ideology out, but again, apparently at random. In this episode of New Discourses Bullets, host James Lindsay breaks down how DEI initiatives and programs achieve stochastic entryism for Woke Marxist activists and how this is the real goal of those programs. Get James Lindsay's book, The Marxification of Education: https://amzn.to/3RYZ0tY Support New Discourses: https://newdiscourses.com/support Follow New Discourses on other platforms: https://newdiscourses.com/subscribe Follow James Lindsay: https://linktr.ee/conceptualjames © 2024 New Discourses. All rights reserved. #NewDiscourses #JamesLindsay #DEI

New year, same Bullshit Mountain. Alex and Emily are joined by feminist technosolutionism critics Eleanor Drage and Kerry McInerney to tear down the ways AI is proposed as a solution to structural inequality, including racism, ableism, and sexism -- and why this hype can occlude the need for more meaningful changes in institutions.Dr. Eleanor Drage is a Senior Research Fellow at the Leverhulme Centre for the Future of Intelligence. Dr. Kerry McInerney is a Research Fellow at the Leverhulme Centre for the Future of Intelligence and a Research Fellow at the AI Now Institute. Together they host The Good Robot, a podcast about gender, feminism, and whether technology can be "good" in either outcomes or processes.Watch the video version of this episode on PeerTube.References:HireVue promo: How Innovative Hiring Technology Nurtures Diversity, Equity, and InclusionAlgorithm Watch: The [German Federal Asylum Agency]'s controversial dialect recognition software: new languages and an EU pilot projectWant to see how AI might be processing video of your face during a job interview? Play with React App, a tool that Eleanor helped develop to critique AI-powered video interview tools and the 'personality insights' they offer.Philosophy & Technology: Does AI Debias Recruitment? Race, Gender, and AI's “Eradication of Difference” (Drage & McInerney, 2022)Communication and Critical/Cultural Studies: Copies without an original: the performativity of biometric bordering technologies (Drage & Frabetti, 2023)Fresh AI HellInternet of Shit 2.0: a "smart" bidetFake AI “students” enrolled at Michigan UniversitySynthetic images destroy online crochet groups“AI” for teacher performance feedbackPalette cleanser: “Stochastic parrot” is the American Dialect Society's AI-related word of the year for 2023!You can check out future livestreams at https://twitch.tv/DAIR_Institute. Follow us!Emily Twitter: https://twitter.com/EmilyMBender Mastodon: https://dair-community.social/@EmilyMBender Bluesky: https://bsky.app/profile/emilymbender.bsky.social Alex Twitter: https://twitter.com/@alexhanna Mastodon: https://dair-community.social/@alex Bluesky: https://bsky.app/profile/alexhanna.bsky.social Music by Toby Menon.Artwork by Naomi Pleasure-Park. Production by Christie Taylor.

The Trident Room Podcast Senior Producer, Sara Dixon, sits down with the hosts of The Trident Room Podcast to celebrate their 50th episode. Meet hosts Sara Dixon, Karl Flynn, Steve Bakotic, Colleen Wilmington, and Dan Petersen as they share how they became members of the team, things they've learned while being podcast hosts, and a little about what it means to be part of the Trident Room Podcast. This episode was recorded on August 23, 2023. For students and alumni of the Naval Postgraduate School, the Trident Room is an integral and well-known contributor to academic success and great ideas. While a pint poured in the stout mugs that dangle from above is part of that, we'd like to think it is the unfettered conversation that really makes the difference. No matter what the topic whether it be Stochastic optimization … sub-state, social-revolutionist terrorism … applied fluid mechanics and heat transfer …, The Trident Room Podcast is the destination for illuminating, unencumbered conversation with colleagues, peers and classmates that really brews the NPS learning experience. The Trident Room Podcast … Stout, unfiltered conversation, always on tap. Sara Dixon, SENIOR PRODUCER - Sara grew up in Rushville, IN. She graduated from the United States Naval Academy in 2010 with a Bachelor's in English and commissioned in the Marine Corps. After serving as a CH-53E pilot for a few years, she laterally moved careers to become a Manpower Officer. Prior to joining NPS, she served as the G-1 Manpower-Operations Officer for 1st Marine Division on Camp Pendleton, CA. Sara is studying Manpower Systems Analysis. The Trident Room Podcast is brought to you by the Naval Postgraduate School Alumni Association and the Naval Postgraduate School Foundation • http://www.npsfoundation.org/ For comments, suggestions, and critiques, please email us at TridentRoomPodcastHost@nps.edu, and find us online at nps.edu/tridentroompodcast. Thank you!

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Learning-theoretic agenda reading list, published by Vanessa Kosoy on November 9, 2023 on The AI Alignment Forum. Recently, I'm receiving more and more requests for a self-study reading list for people interested in the learning-theoretic agenda. I created a standard list for that, but before now I limited myself to sending it to individual people in private, out of some sense of perfectionism: many of the entries on the list might not be the best sources for the topics and I haven't read all of them cover to cover myself. But, at this point it seems like it's better to publish a flawed list than wait for perfection that will never come. Also, commenters are encouraged to recommend alternative sources that they consider better, if they know any. General math background "Introductory Functional Analysis with Applications" by Kreyszig (especially chapters 1, 2, 3, 4) "Computational Complexity: A Conceptual Perspective" by Goldreich (especially chapters 1, 2, 5, 10) "Probability: Theory and Examples" by Durret (especially chapters 4, 5, 6) "Elements of Information Theory" by Cover and Thomas (especially chapter 2) "Lambda-Calculus and Combinators: An Introduction" by Hindley "Game Theory: An Introduction" by Tadelis AI theory "Machine Learning: From Theory to Algorithms" by Shalev-Shwarz and Ben-David (especially part I and chapter 21) "Bandit Algorithms" by Lattimore and Szepesvari (especially parts II, III, V, VIII) Alternative/complementary: "Regret Analysis of Stochastic and Nonstochastic Multi-armed Bandit Problems" by Bubeck and Cesa-Bianchi (especially sections 1, 2, 5) "Prediction Learning and Games" by Cesa-Bianchi and Lugosi (mostly chapter 7) "Universal Artificial Intelligence" by Hutter Alternative: "A Theory of Universal Artificial Intelligence based on Algorithmic Complexity" (Hutter 2000) Bonus: "Nonparametric General Reinforcement Learning" by Jan Leike Reinforcement learning theory "Near-optimal Regret Bounds for Reinforcement Learning" (Jaksch, Ortner and Auer, 2010) "Efficient Bias-Span-Constrained Exploration-Exploitation in Reinforcement Learning" (Fruit et al, 2018) "Regret Bounds for Learning State Representations in Reinforcement Learning" (Ortner et al, 2019) "Efficient PAC Reinforcement Learning in Regular Decision Processes" (Ronca and De Giacomo, 2022) "Tight Guarantees for Interactive Decision Making with the Decision-Estimation Coefficient" (Foster, Golowich and Han, 2023) Agent foundations "Functional Decision Theory" (Yudkowsky and Soares 2017) "Embedded Agency" (Demski and Garrabrant 2019) Learning-theoretic AI alignment research agenda Overview Infra-Bayesianism sequence Bonus: podcast "Online Learning in Unknown Markov Games" (Tian et al, 2020) Infra-Bayesian physicalism Bonus: podcast Reinforcement learning with imperceptible rewards Bonus materials "Logical Induction" (Garrabrant et al, 2016) "Forecasting Using Incomplete Models" (Kosoy 2017) "Cartesian Frames" (Garrabrant, Herrman and Lopez-Wild, 2021) "Optimal Polynomial-Time Estimators" (Kosoy and Appel, 2016) "Algebraic Geometry and Statistical Learning Theory" by Watanabe Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The Stochastic Parrot Hypothesis is debatable for the last generation of LLMs, published by Quentin FEUILLADE--MONTIXI on November 8, 2023 on LessWrong. This post is part of a sequence on Model Psychology.@Pierre Peigné wrote the details section in argument 3 and the other weird phenomenon. The rest is written in the voice of @Quentin FEUILLADE--MONTIXI Intro Before diving into what model psychology is, it is crucial to clarify the nature of the subject we are studying. In this post, I'll challenge the commonly debated stochastic parrot hypothesis for state-of-the-art large language models (GPT-4), and in the next post, I'll shed light on the foundations from which I am building model psychology. The stochastic parrot hypothesis suggests that LLMs, despite their remarkable capabilities, don't truly comprehend language. They are like mere parrots, replicating human speech patterns without truly grasping the essence of the words they utter. While I previously thought this argument had faded into oblivion, I often find myself in prolonged debates about why current SOTA LLMs surpass this simplistic view. Most of the time, people argue using examples of GPT3.5 and aren't aware of GPT-4's prowess. Through this post, I am presenting my current stance, using model psychology tools, against that hypothesis. Let's delve into the argument. Central to our debate is the concept of a "world model". A world model represents an entity's internal understanding and representation of the external environment they live in. For humans, it's our understanding of the world around us, how it works, how concepts interact with each other, and our place within it. The stochastic parrot hypothesis challenges the notion that LLMs possess a robust world model. It suggests that while they might reproduce language with impressive accuracy, they lack a deep, authentic understanding of the world and its nuances. Even if they have a good representation of the shadows on the wall (text), they don't truly understand the processes that lead to those shadows, and the objects from which they are cast (real world). Yet, is this truly the case? While it is hard to give a definitive proof, it is possible to find pieces of evidence hinting at a robust representation of the real world. Let's go through four of them.[1]Argument 1: Drawing and "Seeing" GPT-4 is able to draw AND see in SVG (despite having never seen as far as I know) with an impressive proficiency. SVG (Scalable Vector Graphics) defines vector-based graphics in XML format. To put it simply, it's a way to describe images using a programming language. For instance, a blue circle would be represented as: in a .svg file.Drawing GPT-4 can produce and edit SVG representations through abstract instructions (like "Draw me a dog", "add black spots on the dog", … ). GPT-4 drawing a cute shoggoth with a mask: "Seeing" More surprising, GPT-4 can also recognize complex objects by looking only at the code of the SVG, without having ever been trained on any images[2] (AFAIK) I first generated an articulated lamp and a rendition of the three wise apes with GPT-4 using the same method as above. Then, I sent the code of the SVG, and asked GPT-4 to guess what the code was drawing. GPT-4 guessed the articulated lamp (although it thought it was a street light.[3]): And the rendition of the three wise apes (It can also recognize a car, a fountain pen, and a bunch of other simple objects[4]) The ability of seeing is interesting because it means that it has some kind of internal representation of objects and concepts that it is able to link to abstract visuals despite having never seen them before.Pinch of salt It's worth noting that these tests were done on a limited set of objects. Further exploration would be beneficial, maybe with an objective scale for SVG diffi...

Lords: * JohnB * CisHetKayFaber Topics: * The human body is more similar than I'd like to shitty software (humans have chitinase in our DNA, immune system works like an antivirus) * The origin of hiccups * https://derinthescarletpescatarian.tumblr.com/post/723161770948231168/okay-so-the-thing-about-hiccups-is-that-you-have * Twenty years ago there were like three real books about videogames that weren't strategy guides and now there are hundreds. * "untitled" by Prostetnic Vogon Jeltz * https://allpoetry.com/poem/13310612-Prostetnic-Vogon-Jeltz-by-Bollox * TV shows that have inspired original mathematics - yes, it has happened more than once Microtopics: * Whether it's okay to catch up. * Air conditioning on hot days. * A steel mill catching fire and the whole east bay smells like the back of a cathode ray tube. * Shaking yellow sand out of your laundry. * Importing brushfire smoke from Canada. * Using bad metaphors to explain RAM. * Stochastic immunology. * Alphabetizing all your gut bacteria. * Whitman's Anal Sampling Mechanism. * The Bonzi Buddy of the human body. * Your inkjet printer waking up at 5am to cycle some ink. * The optimal humidity to reflow your ink cartridges. * Norton Commander. * Selecting files with the arrow keys and viewing them with F3. * How Thunking ruined Jim's Norton Commander clone. * Your fish nervous system trying to be a fish again. * Reciting the anti-hiccup poem to cure your hiccups. * Fish reminding each other to not take ten small sips of water or they'll stop breathing. * Scaly swimmers. * Working in a shark lab. * Whether a fish sandwich is a fish or a sandwich. * Offering people a "tuna roll" and when they say yes guessing whether they think you mean the sushi or tuna salad on a bready roll. * Meditating on the concept of hiccups. * Congratulations, you won the game! * How to Beat the Atari Home Video Games. * Where a library would shelve those books of type-in BASIC programs. * Dave (2018) and Dave Demo (2018) * Which Boss Fight Books are good. * Starting your kid on Mario Maker. * Game Engine Black Book: Doom and Game Engine Black Book: Wolfenstein 3D. * Mario Galaxy's camera-oriented level design. * Your body telling you when it's time to drink water. * Using Final Fantasy 4 as a bridge to talk about localization. * Developing a light gun game to install in all the bowling alleys now that the bowling fad has faded. * The Countdown to Irrelevance. * Books that used to be blogs. * Inventing criteria for how you beat Galaga. * Learning what micturations are twenty years after reading that Vogon poem and wondering which seeming nonsense words you'll learn in the next twenty years. * The BBC TV Hitchhiker's Guide series. * The second-worst poetry in the galaxy. * Red Dwarf airing on Dave. * A better way to think about it. (If you are a math person.) * Drive-by Mathematics. * Nesting your watch orders. * Watching every episode in every possible order to get the maximum possible amount of context for every scene. * Marylin vos Savant. * The Monty Hall Problem. * A game design that presumes the player does not want a goat. * Whether you want to switch to the other goat. * Opening 98 doors, revealing goats behind each one. * Pushing limits and talking about calculus. * Learning Japanese because you want to read the smutty manga. * X: The Everything App.

Dr. Jim interviews Dr. Josh WilsonDr. Josh Wilson is the Co-Founder of EBMC, Psych Inc, WCMI, and Organizational Solutions. Prior to becoming a licensed psychologist and President of the aforementioned businesses (and all around cool guy) Dr. Josh worked with kids on the spectrum and their parents.Dr. Josh worked at Quest Therapeutic Camp. While where he discovered his passion for working with children on the (Autism) spectrum, including gifted kids. It is also where he met his current wife. Two discoveries he is quite impassioned about.In this episode Dr. Wilson discusses his passion for 2E kids (or twice exceptional children) who present with a high IQ and an accompanying cognitive challenge and/or learning disability, i.e. ADHD, Autism, Dyslexia, Dyscalculia. Dr. Josh helps us become more humane by understanding and esteeming this extraordinary population.WCMI networking group A networking group for mindfulness-focused clinicians dedicated to learning together & collaborating for more information click here

EPISODE 240: COUNTDOWN WITH KEITH OLBERMANN A-Block (1:44) SPECIAL COMMENT: One of Trump's stochastic terrorists went to Barack Obama's house to assassinate him. Anybody else notice this, or was it just me? And that a Washington judge was having a hard time last night finding an excuse to hold him without bail? Hold the WOULD-BE ASSASSIN without bail? And that Trump had doxxed Obama and finally found a sucker willing to do his bidding for him? AND park his van full of ammo and weapons a few blocks from Obama's home. AND livestreamed himself threatening Jamie Raskin and Kevin McCarthy and oh he was going to take his van to Maryland to a federal facility containing a nuclear research reactor and blow up the van and the nuclear reactor and… I mean am I the only one seeing this anywhere? Trump – crazier and more dangerous by the minute – is finally clearly and inextricably linked to a guy ready to answer Trump's constant drumbeat to try to get one of his cultists to kill, without TELLING them to kill and thus leaving himself open to prosecution FOR killing or incitement-to-kill, and the Department of Justice remains asleep at the switch. The US Magistrate Judge in DC, Zia M. Faruqui managed to avoid having to release Trump's figurative flying monkey – his name is Taylor Taranto – by keeping him in stir overnight and putting the onus on the government to answer the judge's question: if the original charges against this would-be political murderer Taranto were simply left over trespassing accusations from January 6th, and the government wanted him held without bail because he's a flight risk, what if the Judge finds he's NOT a flight risk? Is it enough that threatening key Democratic and Republican congressmen and talking about going to a government nuclear reactor and blowing up your van next to – oh and responding to one former president's publishing of another former president's address by going there with a van full of weapons of death – is THAT quote “clear convincing evidence” of a danger to the public and even if it is, does that mean the judge can really hold him in jail? Yeah, you guys let Taranto loose while the lot of you argue over how many angels can dance on the head of this proverbial pin, and, you know, if he goes back to Obama's neighborhood or, I dunno, tries to ram the White House with his van while screaming “TRUMP IS GOD” as he does so, maybe then prosecutors can indict him for various vehicular and parking offenses. B-Block (20:00) POSTSCRIPTS TO THE NEWS: Miles Taylor says Trump negotiated with Erik Prince to form his own private mercenary army to dispatch to Afghanistan and Venezuela; RFK needs money more than power; Steve Bass's new horizon; farewell to Dr. Frank Field and Marvin Kitman. (25:50) THE WORST PERSONS IN THE WORLD: Laura Ingraham wants to defund Jack Smith. How about demoting him to 7 PM? The warmest day since July 4, 122,977 BC. And GQ Magazine first neuters a piece that offended David Zaslav of Warner BROS, then kills the piece completely, then it turns out GQ's Editor is producing a film FOR WARNER BROS. C-Block (34:10) EVERY DOG HAS ITS DAY: Angelina, L'il Bit and Prince in Georgia (35:20) THINGS I PROMISED NOT TO TELL: Nothing makes me smile more than a week with July 4th in the middle. It's because of that kind of calendar that I got my first network job just 44 years ago. The holiday ate up nearly all the business days they would've had to look for anybody besides me! (Plus that first broadcast from July 1979).See omnystudio.com/listener for privacy information.

Rep. Jayapal says if the GOP doesn't raise the debt ceiling, there will be backlash in the streets. Killer orcas attacking Clarence Thomas. The difference between how Trump and DeSantis treat the media. Getting out of the military. See omnystudio.com/listener for privacy information.

So-called “artificial intelligence” is one of the most divisive topics of the year, with even those who understand it in total disagreement about its potential impacts. This week, A.I. reseachers and authors of the famous paper “On the Dangers of Stochastic Parrots”, Emily Bender and Timnit Gebru, join Adam to discuss what everyone gets wrong about A.I. Learn more about your ad choices. Visit megaphone.fm/adchoices See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

So-called “artificial intelligence” is one of the most divisive topics of the year, with even those who understand it in total disagreement about its potential impacts. This week, A.I. reseachers and authors of the famous paper “On the Dangers of Stochastic Parrots”, Emily Bender and Timnit Gebru, join Adam to discuss what everyone gets wrong about A.I. Learn more about your ad choices. Visit megaphone.fm/adchoices

Episode 2361: The Art Of The Stochastic Terrorism At Twitter; Judas Pence Helped Promote Lockdowns

We're seeing protests in China over Covid lockdowns not seen since the days of Tiananmen Square. And a new phrase has been introduced into civil debate: Stochastic Terrorism. What is it and where did it come from?

In a culture so captive to a “critical theory mood,” the only right and loving response to the hijacking of language is to oppose falsehood and advance truth. This will mean, among other things, refusing to accept or use words and phrases like “stochastic terrorism” which assume all kinds of false and dangerous ideas about who we are.