<n>Below are links to articles, highlighted passages<fn>, and occasional annotations<fn> for the week ending on the date in the title, <a href="{app}/view_article.php?id=4085">enriching the raw data</a> from <a href="http://www.instapaper.com/starred/rss/1890855/5c1g08eoy9skhOr3tCGqTQbZes">Instapaper Likes</a> and <a href="https://twitter.com/mvonballmo">Twitter</a>. They are intentionally succinct, else they'd be <i>articles</i> and probably end up in the gigantic backlog of unpublished drafts. YMMV.</n> <ft><b>Emphases</b> are added, unless otherwise noted.</ft> <ft>Annotations are only lightly edited and are largely <i>contemporaneous</i>.</ft> <h>Table of Contents</h> <ul> <a href="#politics">Public Policy & Politics</a> <a href="#journalism">Journalism & Media</a> <a href="#labor">Labor</a> <a href="#economy">Economy & Finance</a> <a href="#science">Science & Nature</a> <a href="#art">Art, Literature, & Cinema</a> <a href="#philosophy">Philosophy, Sociology, & Culture</a> <a href="#technology">Technology & Engineering</a> <a href="#llms">LLMs & AI</a> <a href="#programming">Programming</a> <a href="#fun">Fun</a> </ul> <h id="politics">Public Policy & Politics</h> <a href="https://mattbivens.substack.com/p/notes-from-tmutarakan" source="The First 100 Days" author="Matt Bivens, M.D.">Notes from Tmutarakan</a> <bq><b>Many ordinary Russians back then relied on the Western payment systems, from credit cards to cell phone-based payer apps like Google Pay and Apple Pay. They woke up one morning in 2022 and none of that worked.</b> Suddenly, many of them could not access their money or pay their bills. All of this happened instantly, without even a pretense of legal process. (In a similar orgy of wanton, extralegal behavior, we celebrated when the Nord Stream 2 gas pipeline exploded and innocently pretended not to know who was behind that.)</bq> <bq>To this day, Russian athletes are only allowed to compete in the Olympics if they renounce their home nation and agree to compete in a dreamt-up category of “Individual Neutral Athletes.” (<b>Wimbledon also now allows Russians to compete again, provided they sign “neutrality declarations” and formally agree “not to support” Russia or Vladimir Putin.</b>)</bq> Obviously, there's no need for U.S. or Israeli athletes to do anything like that. That would be crazy. <bq>But as we were seizing bank accounts and foreign homes, and canceling tennis matches and orchestral performances and mustards and cats, and pouring in billions of dollars in death tech, we in the West also repeatedly vetoed every peace deal. That’s right: <b>All of the long years of brutal butchery since those first few weeks were continued at American insistence.</b></bq> Well c'mon bro! How else do you think people are supposed to make money on the war they'd spent decades starting? That was the whole point. Why would they stop right when it was paying off? <hr> <a href="https://jacobin.com/2025/10/democrats-cynically-wield-wokeness-against-graham-platner/" source="Jacobin" author="Branko Marcetic">Democrats Cynically Wield “Wokeness” Against Graham Platner</a> <bq>After November’s disastrous loss, the Democratic Party establishment, as part of its <b>regular quest to deflect blame for its own failures</b>, once more took aim at the spinning wheel of excuses in front of it and threw a dart. In previous years, that dart hit squares labeled “Green Party,” “sexism,” “white voters,” and “Bernie Sanders.” But this time, <b>the party’s leading excuse was not going to be that Americans are too backward and ignorant for the Democrats, but that Democrats are, if anything, too tolerant and enlightened for America.</b></bq> <hr> <a href="https://thefloutist.substack.com/p/europes-latest-intelligence-fakes" source="The Floutist" author="Helmut Scheben">“Europe’s latest intelligence fakes.”</a> <bq><b>You will remember Yuri Andropov, general secretary of the USSR from 1982 until his death two years later, who once laughingly told Finnish President Mauno Koivisto: “Bomb them. It’s fine with us.”</b> He was referring to the “Soviet submarines” spotted off the Swedish coast in 1984. <b>Andropov knew they were not Russian submarines</b>, but a false flag operation by Western intelligence agencies. These mysterious boats were never captured. The “Soviet threat” proved to be a perfect way to sabotage Swedish Prime Minister Olof Palme’s policy of détente.</bq> <hr> <a href="https://exileinhappyvalley.blogspot.com/2025/10/harsh-lessons-christian-nationalists.html" source="Exile in Happy Valley" author="Nicky Reid">Harsh Lessons Christian Nationalists Could Learn from Folk Horror</a> <bq>Christianity is back and it's more violent than ever. I speak of course of the late capitalist tent house revival of Christian Nationalism amongst the decaying ruins of Washington DC. <b>Using the demonic, Caligula-esque Emperador Trump like a pedophilic battering ram, a bunch of millenarian lunatics with a barely literate interpretation of the Bible have found themselves in the highest echelons of political influence in this country</b> and their vulgar reach can be felt throughout</bq> <bq>The patriarchy is right to view women as dangerous because what other alternative to subjugation have they given us? <b>The system has alienated an entire gender to the point where any form of insurrection is at least as tempting as subordination and almost always far more rewarding.</b> When you consistently cast a powerless class of people as the villain in all your fairy tales, <b>you really have no right to be shocked when they rise to the occasion and greet you with fists.</b></bq> <bq>I believe that the key to understanding this film [Midsommar] and the key to comprehending the existential question all of us find ourselves faced with in the bosom of a crumbling empire lies at <b>the juxtaposition between the death of Dani's first family; cold, pointless and nihilistic, and the sacrifices performed by her second family which are equally horrific and are yet seen as more savage merely because they are performed with a sense of purpose.</b></bq> Whoa. I hadn't thought of it like that. <bq>This isn't a defense of human sacrifice. It's an argument that this unfortunate genre of ritual violence never actually left us, it simply lost all meaning beyond conquest under materialism and <b>left us with a society in which life is cheap, and spirituality is governed by the rich.</b> Once again, I reject initiatory violence of any kind, but I also recognize, as Marx once did, that violence on any massive scale is the midwife of any society pregnant with a new one or perhaps in this case, an old one.</bq> This is an excellent point. The incredible amount of violence inherent in the system is ignored as a moral failing---because it is that violence that makes the system work for its owners. That's why we ignore that violence while focusing laser-like on the kind of violence that our lords and masters want us to focus on instead. <hr> <a href="https://www.caitlinjohnst.one/p/its-just-wall-to-wall-news-stories" author="Caitlin Johnstone" source="Substack">It's Just Wall-To-Wall News Stories About The US And Its Allies Abusing The World</a> <bq>In the same interview, <b>Scott also said that if Maduro is successfully ousted, “it’ll be the end of Cuba.”</b> “America is gonna take care of the southern hemisphere and make sure there’s freedom and democracy,” he added.</bq> This is just how they do things. They kill anyone who gets in their way. Rubio is hot to attack Cuba. Venezuela protects Cuba. Get rid of Venezuela first. They don't care. They're psychopaths. <bq>The senator’s statements suggest that <b>the US is preparing a push in Latin America similar to what it has been executing with Israel in the middle east, eliminating any powers which refuse to bend the knee.</b> South of the US border the top two disobedient governments are the socialist states of Venezuela and Cuba. In the middle east <b>the US and Israel have spent the last two years bombing Iran and Yemen, securing a regime change in Syria, and doing everything they can to eliminate Hamas and Hezbollah in order to rule the region uncontested.</b></bq> <bq>All over the world the US and its allies are murdering and abusing people in order to dominate the planet and ensure the survival of the capitalist system with which its power is intertwined. It is <b>a giant murder machine feeding on human blood and the life force of our biosphere while providing nothing but obstacles to a healthy world.</b> <b>The US-centralized empire is a disease that affects our entire species.</b> We had better find a cure, and fast.</bq> <hr> <media href="https://www.youtube.com/watch?v=-vS4mMSCYHk" src="https://www.youtube.com/v/-vS4mMSCYHk" source="YouTube" width="560px" author="" caption="Historiker packt aus: 'Putin hat recht!'"> No wonder this interview is so long. The first 15 minutes are all about how brave the Swiss guy is for even talking about these dangerous topics.  I'm not accustomed to that. TALK ABOUT YOUR WORK NOT YOURSELF. But that's what the interviewer wants to hear...the interviewer is ... difficult.  The title is clickbait-y and a wholly inadequate summary of the wide-ranging discussion in this 3.5-hour interview with Dr. Daniele Ganser was about so much more. Ganser's a Swiss guy, being interviewed by a relatively young German podcaster who I can't describe as anything other than a German Joe Rogan. His mind is so open that his brains are falling out.    OK. finally, getting to the good stuff. I'm 1:10:00 in. I'm cautiously optimistic. I think I would be able to spend an evening with him and we'd be saying "ja und amen" to each other the whole time. (Except I am most certainly not a Kennedy fan<fn> but I'm not a fan of a lot of people.) I knew most of what he's saying already but it was interesting to hear Noam Chomsky get a shoutout from a Swiss guy. I was actually thinking that his statement that "all of the records are public in the U.S." reminded me a lot of Chomsky's essays and interviews over the years, where he would constantly say that, for a lot of horrifying stuff, all you had to do was to look at the official record. The U.S. government is rarely ashamed enough of itself to actually try to hide stuff. There is a long discussion of the Cuban Missile Crisis as a way of introducing how we're all being lied to, all the time, and how things that you learned when you were young, that formed the basis of how you look at the world (Weltanschauung) can be nearly completely false or, at the very least, just false enough that you believe the very wrong things that your rulers need you to believe in order to be able to keep ruling and profiting from you. If I have a quibble, I found there to be too little social analysis in his thinking. He's just questioning the official narrative from governments but then seemingly doesn't apply that to corporate entities. For example, he says that YouTube is so much better because, on television, you're so controlled that you can't say anything that's even slightly different than the officially accepted narrative because otherwise, <iq>der ARD grätscht ein.</iq> (the TV producer will block broadcast, literally "does a sliding tackle") Ok, that's fair. And a huge problem. But how is it very much different on YouTube? It's perhaps not as controlled---Ganser has a thriving channel---but they can just shadowban the shit out of you, if not outright ban you and remove all of your content. Does he not know that this happens? YouTube is basically UHF. As long as things don't get too popular, the rulers (Google or the government) leave you alone. <bq>Es gibt genug Reichtum für alle.</bq> At about <b>02:12:00</b>, <bq>Es hängt eben davon ab, wie wir die Beziehung gestalten. Und wenn wir die Beziehung so gestalten, dass wir sagen, komm, lass uns Handel betreiben, dann werden wir beide reich und <b>lass uns mit Respekt miteinander umgehen.</b> Und im übrigen möchten wir uns noch entschuldigen für diese und diese Dinge, die vorgefallen sind, aber das waren nicht wir, das waren unsere Väter und Urgroßväter. Darum sorry, ich habe den Opiumkrieg nicht geführt gegen China, weil das waren die Engländer, <b>die haben Opium nach China reingeschleust und haben dieses Land zersetzt dadurch und das war so kann man nicht gut reden, ist einfach hinterhältig.</b> Und es ist aber jetzt das 21. Jahrhundert und wir könnten mit Russland eine gute Beziehung aufbauen. <b>Im Moment sind wir natürlich weit davon entfernt, aber wir könnten auch mit China eine gute Beziehung aufbauen.</b> Und wir könnten auch mit den Amerikanern eine gute Beziehung aufbauen, aber dann müsste der Westen meiner Meinung nach sich auch ein bisschen in Demut üben und sagen, okay, wir geben zu. Gewisse Dinge waren nicht so großartig. <b>Aber wir sind immer noch auf dem hohen Ross und das ist wirklich nicht den Realitäten angepasst.</b> [...] <b>Und die meisten Europäer denken und die meisten Amerikaner denken, ist mir doch egal, der Iran ist noch bei den BRICS, die können ja nichts, die Iraner.</b> Sind Persier, das ist deine uralte Kultur. Alles was du in der New York Times liest über die Mullahs in Tehran, <b>das ist einfach Framing im Sinn von das alle Iraner sind Idioten.</b> Aber die Chance, dass du morgen, wenn du beim Zahnarzt bist, von einem Iraner behandelt wirst oder wenn du dein Auge operierst, dass ein Iraner ist und dass er sehr hochgebildet ist und dass er mehrere Sprachen kann, während du nur eine kannst, die ist sehr groß. <b>Mach mal ein Reality Check und wieder demütig sein.</b> Und weißt du, Ben, ich möchte nicht sagen, ähm, der Westen ist ein schlechter Ort. Das möchte ich nicht sagen. Der Westen hat viel Gutes gemacht, hat auch viel Gutes gemacht. Ja, aber es ist an der Zeit zu sehen, dass <b>es auch eine multipolare Welt geben kann und dass diese Welt friedlich gestaltet werden kann. Das wäre so meine Makroperspektive, wenn ich so sagen darf.</b> <b>Also multipolar heißt einfach nicht mehr die USA als Imperium, die diktieren alles.</b> Und das bedeutet natürlich, dass Deutschland als Zentrum von Europa Frieden mit Moskau und Frieden mit Peking aufbauen sollte und da sind wir natürlich heute im 2025 ein bisschen weiter davon entfernt. </bq> A few times, he seemed to explain the simplest things but I realize too that his audience in the DACH region, where people don't necessarily already know how the U.S. works. In another case, he took quite a bit of time to explain how two people who have only kid aren't replacing the population. LMAO. At <b>2:58:00</b>, <bq>Schau dir mal die sogenannte Elite im Westen an und frag dann, ob du so etwas wie Begeisterung und Inspiration fühlst.</bq> At <b>3:00:00</b>, <bq>Dieses deutsche Interesse ist eben, dass die Achse Berlin und Moskau freundschaftlich ist. Und weißt du, <b>mit Freundschaft meine ich nicht Lobhudelei, sondern Freundschaft. Einfach Respekt. Respekt auf Augenhöhe.</b> Natürlich muss es doch einen Flieger geben. Direktflug Berlin Moskau. Hallo? <b>Warum soll es diesen Direktflug nicht geben? Ich sage, es braucht auch ein Direktflug nach Tehran.</b> Und dann, wenn man nein nein nein nein Daniele! Das Reich der Finsternis und so wer das denkt ist einfach in seinem Dogma gefangen und und das tut mir schon fast leid es tut mir schon fast leid, dass man dann die Sache so sabotiert und <b>es tut mir auch leid für die vielen Journalisten, die dann jeden Tag eigentlich schreiben,</b> Ja, wir haben die Sache analysiert und sind zum Schluss gekommen, Russland ist böse und das schreiben sie jeden Tag. Sagen, ja, habt ihr noch mal neu analysiert oder ist das dann Copypaste von gestern und was habt ihr überhaupt neue ... welche Gesichtspunkte habt ihr angeschaut? <b>Was ist eure Vision bis 2030 bis 2040 bis 2050?</b> Von wo kommt das Erdgas? Erdgas. Moment ... das kommt äh aus den USA. Was habt ihr für ein Preis? Dreifacher Preis. Aber wenn die Wirtschaft abwandert, wer sind dann die Arbeitgeber? Oh, die Industrie brauchen wir nicht mehr. Wir haben Dienstleistung. Ja, die Dienstleistung, das sind viele Zulieferer der Industrie, <b>wenn die weg sind, wer soll's da machen? Ist uns egal. Wir sind---und dann, wenn du sagst, mir ist das alles egal---dann du dich aus Dogmatist.</b></bq> Here, he's talking about having spoken with Noam Chomsky, who told him, <bq>[...] was ist eigentlich die Aufgabe? Was ist die wirkliche Aufgabe? Es ist <iq>speak truth to power.</iq> Also <b>Geschichte ist Herrschaftswissenschaft.</b> Du verstehst, wie kann man Herrschaft erzeugen, indem du eben äh diese verdeckten Operationen machst oder <b>die Medien kontrollierst oder Narrative formst oder Wording oder Framing nutzt</b> oder ganz ... tausend Techniken.</bq> He also spoke very fondly of Julian Assange, so he's really ticking all of the boxes for me. His focus on WT7 having been detonated is something that I don't share but I've never looked into it. I can agree that we've been lied to about nearly everything about 9--11. That is clear. Whether a building was blown up isn't at the top of the priority list for me<fn> but to each their own. At <b>03:27:00</b>, when asked about what he would write on a piece of paper to remind himself of who he was, should he wake up one morning with amnesia. <bq>Orientiere dich an <b>Liebe, Mut und Wahrheit</b>. Mehr ist nicht zu tun.</bq> <hr> <ft>See my notes on Kennedy's speeches in <a href="{app}view_article.php?id=4930&search_text=kennedy" author="" source="">The U.S. has never been the good guy: on Kennedy, Cuba, and Iran</a></ft> <ft>How the incident was leveraged to declare a global war on Islam, how entire countries were flattened, how black sites were filled to the brim, etc. etc.</ft> <hr> <a href="https://www.caitlinjohnst.one/p/americans-have-no-idea-who-their" author="Caitlin Johnstone" source="Substack">Americans Have No Idea Who Their Government Is Bombing, And Other Notes</a> <bq>An article by Antiwar’s Dave DeCamp has highlighted the widely-ignored fact that according to AFRICOM <b>the US waged a three-day bombing campaign in Somalia from October 26 — October 28, bringing the total number of US airstrikes in that nation this year to 89.</b> <b>What percentage of Americans even realize that Trump has bombed Somalia nearly a hundred times this year?</b> I doubt it’s even one percent. The mainstream press barely mention it. <b>Americans have hardly any idea who their own country is bombing.</b></bq> <bq><b>Israel keeps violating the “ceasefire” and bombing Gaza whenever it wants to, then saying the ceasefire is back in effect. It’s like saying you’ve quit smoking whenever you’re not currently having a cigarette.</b> NPR reports that after a mid-“ceasefire” bombing campaign that killed 104 people including 46 children, Benjamin Netanyahu “ordered the strikes after accusing Hamas of violating the ceasefire for handing over body parts this week that Israel said were partial remains of a hostage recovered earlier in the war.” <b>Saying you massacred children because you weren’t given the correct pieces of a corpse just might be the craziest justification for a war crime that anyone has ever offered.</b></bq> <hr> <a href="https://www.wsws.org/en/articles/2025/10/30/xzyn-o30.html" author="Tom Hall" source="WSWS">With mass hunger approaching as food stamps expire Saturday, huge price increases revealed for Obamacare healthcare plans</a> <bq>[...] the impact of the expiration of these tax credits will be huge. With the open enrollment period also set to begin November 1, previews of plans in 30 states were released Wednesday showing enormous increases to out-of-pocket costs. <b>The Kaiser Family Foundation estimates that insurers plan on raising prices by 26 percent on average.</b> For those receiving enhanced premium tax credits, <b>net premiums are set to more than double by 114 percent through a combination of price increases and the loss of subsidies.</b> According to the Bipartisan Policy Institute: “a family of four with a household income of $45,000 (140% of [the federal poverty line]) <b>with a $0 premium in 2025 [due to subsidies] will see their premiums increase to $1,607 a year.</b> Also, a 60-year-old couple with an annual income at 402% of FPL (about $85,000) could pay a yearly premium of $22,600 in 2026, or about a quarter of their annual income, instead of 8.5% of their income (as established under enhanced PTCs).”</bq> <bq>Already there has been a $180 billion cut to food stamps and a sharp increase in eligibility requirements under the “Big Beautiful Bill.” Once food stamp funding is finally restored—assuming Trump has any intention of doing so—<b>over 20 million people will find that their benefits have either been reduced or dropped entirely.</b></bq> <bq>And while the Democrats make a show of opposing the expiration of ACA tax credits, this amounts to only a drop in the bucket compared to the $900 billion cut to Medicaid over 10 years in the same law. Beginning January 1, <b>there will be a sharp increase in work requirements for Medicaid, part of the drive to fund trillions in tax cuts for the wealthy.</b> <b>The Democrats’ overriding concern is the fear that opposition to Trump could develop into a broad social movement against inequality.</b> They are determined to prevent this at all costs. But they agree with the fundamental direction of policy: <b>higher levels of exploitation to fund an increase in military spending and to prop up Wall Street.</b></bq> <hr> <a href="https://www.wsws.org/en/articles/2025/10/30/yuzz-o30.html" author="Philip Guelpa" source="WSWS">Severe delays at Newark airport highlight the ongoing crisis of the US air traffic control system</a> <bq>This situation is now being exacerbated by the federal government shutdown. <b>Controllers are classified as “essential”</b> and therefore required to work during the shutdown. Nevertheless, <b>they are not among the limited categories of federal employees, including the military, for whom special arrangements are being made in order to continue paying wages.</b> The controllers suffered their first “payless payday” on Tuesday, October 28.</bq> Insanity. Just heaping abuse on the people that hold society together, withholding their paychecks, while an absolute cheesedick like Milei gets $40B. Revolution. <bq>Air traffic control is an extremely stressful job. Controllers must maintain intense vigilance at all times to avoid catastrophic accidents in congested airspace. <b>Conditions are made even more difficult by increasingly outdated equipment, lacking upgrades which have been neglected for years.</b></bq> <bq>Control over Newark airspace was transferred to Philadelphia from New York last year due to chronic understaffing at the latter.</bq> And now they're having a "sick-out" and good for them. There should be a nationwide work stoppage until all of the elites quit their bullshit. People should just not show up to work at FOX News. Let Hannity bloviate into a dead camera. Maybe he'll get an aneurysm from shouting; he'd come out smarter. <bq>Three weeks ago, Duffy denounced controllers who did not come to work as “problem children” and threatened to fire them. Duffy told Fox Business, “if we have some on our staff that aren’t dedicated like we need, we’re going to let them go. I can’t have people not showing up for work.”</bq> Fuck you, Duffy. Seriously, you are worthless. Why don't you land all the planes for no pay? The entitlement is incredible and it makes me sick to think of relatives nodding along to what they consider to be the sagacity of Duffy and his entire ilk---all of these useless bozos in the administration, all of these nattering nabobs in the media---and wondering how anyone could fail to see how right Donald Trump is about everything. These lazy good-for-nothing air-traffic controllers can't even do their patriotic duty for free. Where's the love of country? Meanwhile, none of them would even pick up a candy wrapper for free. <hr> <media href="https://www.youtube.com/watch?v=nboFLnATNcs" src="https://www.youtube.com/v/nboFLnATNcs" source="YouTube" width="560px" author="HasanAbi" caption="WE ARE COOKED FR"> This video covers how AI videos depicting angry Americans who have <iq>seven kids from seven daddies</iq> and who are angry about losing their SNAP benefits are flooding the Internet right now, being reposted again and again and again by people who are having their demonic viewpoint justified by fake videos that say exactly what they want to hear in a very convincing way. The ones depicting black women haven them screaming that the government owes them a living. They also claim impossibly high benefit numbers. The ones depicting white people show them saying that they will now definitely go out and get jobs, because the government is no longer willing to support them. It's Libertarian pornography. This is the end times. This is a very bad timeline. There's one lady who's actually real...but she's a rage-baiter just making videos that farm outrage for attention that is converted to income from the platform. This is a terrible, terrible timeline for the people who are caught up in all of this, rather than just catching some strays from people who report on it. Top comment on the video: <bq>1960s: We'll have supercomputers solves world hunger 2025 Supercomputers: Best I can do is minstrel show</bq> <h id="journalism">Journalism & Media</h> <a href="https://www.caitlinjohnst.one/p/you-believe-the-mainstream-narrative" author="Caitlin Johnstone & Tim Foley" source="Notes From The Edge Of The Narrative Matrix">You Believe The Mainstream Narrative? Of Course You Do, You're Twelve</a> <bq>Zohran Mamdani is outside my area of political interest and it’s none of my business who New Yorkers elect as their mayor, but the Islamophobic shrieking I’ve been seeing online in response to his campaign has been absolutely jaw-dropping. No one with mainstream political or media aspirations could ever get away with talking about the religion of a Jewish politician the way Zionists have been openly talking about Mamdani and his faith. From what I can tell <b>Mamdani is a just a regular guy and a fairly ordinary progressive Democrat with an extraordinarily high level of campaign talent, but these freaks are claiming he’s going to impose sharia law and start throwing gays off the Chrysler Building.</b> It’s a degree of mass hysteria about Islam unlike anything I’ve seen since the immediate aftermath of 9/11, which any normal person will agree led to some extremely bad thinking and terrible decisions. Some of it is arising from <b>organic American racism and the knee-jerk rightist impulse to throw anyone to the left of Bill Clinton out of a flying helicopter</b>, but a lot of it has nothing to do with Mamdani at all. As we’ve discussed previously, Zionists have been seizing on every opportunity to promote hatred of Muslims because it’s a lot easier than convincing people to like Israel.</bq> <hr> <a href="https://babylonbee.com/news/americas-obesity-crisis-solved-as-ebt-benefits-run-out/" author="" source="Babylon Bee">America's Obesity Crisis Solved As EBT Benefits Run Out</a> The Babylon Bee has been getting crueler and crueler and shittier. They used to claim to be a Christian website. This is U.S.-American Christianity, without a mask. This is what they are. They are not at all about anything to do with Jesus's teachings. They are about hating the poor and loving the rich. They are about madness. They celebrate the murder of Muslims, of Palestinians. They celebrate starvation, not just of Palesinians but also of U.S.-Americans. <hr> <a href="https://www.racket.news/p/bill-gates-says-well-survive-climate" author="Matt Taibbi" source="Racket News">Bill Gates Says We'll Survive Climate Change, World Furious</a> I thought that this article was from the Babylon Bee at first. That Taibbi has sunk to the level not only of assuaging his vast audience of climate-change-deniers with some half-assed pap but now his vaunted wit has abandoned him as he's just a bitter old man, bitching about how people are failing to pay enough fealty to Bill Gates. It's a shame. We're only at the very beginning of this thing and a hurricane just destroyed Jamaica, a bunch of Cuba, and probably an island in the Bahamas. It's all fine. Go ahead and spend a bunch of time fighting straw men, Taibbi. It's all you seem to be good for these days. <hr> <media href="https://www.youtube.com/watch?v=XbH8POixPzA" src="https://www.youtube.com/v/XbH8POixPzA" source="YouTube" width="560px" author="HasanAbi" caption="billie eilish called out billionaires and non-billionaires are mad at her"> <bq author="Billy Eillish">Love you all, but there's a few people in here that have a lot more money than me. If you're a billionaire, why are you a billionaire? No hate, but yeah, give your money away, shorties.</bq> Later in the short, it says that she has given away a quarter of her ~$40M wealth. People got mad and defended billionaires. <bq>The moment that you say that, 'Hey, people should maybe give back more and be kind to others,' everyone on the Internet goes, 'Ha! Fuck you!' It's like, brother This is peasant-brain thinking. This person is on your side.</bq> <h id="labor">Labor</h> <a href="https://www.wsws.org/en/articles/2025/10/29/cyix-o29.html" author="Jerry White" source="WSWS">Amazon, UPS, Paramount Global slash tens of thousands of jobs as economic and social crisis in US deepens</a> <bq><b>What is unfolding is a coordinated class war, not a series of isolated restructurings.</b> It spans logistics (Amazon, UPS), auto manufacturing, media (Paramount), tech (Microsoft, Salesforce, Meta), retail (Target), aviation (Lufthansa) and the public sector. Both corporate parties back it. Trump’s Project 2025 blueprint calls for mass federal layoffs, the dismantling of regulatory agencies, Social Security and other essential programs and <b>the funneling of even more money into the hands of the corporate financial oligarchy and the build up for World War III.</b> As for the Democrats, they support “fiscal responsibility” and fear nothing more than <b>the revolutionary potential of mass movement of the working class against the fascist president and the economic and political domination of the oligarchy.</b></bq> <bq>The central issue is not artificial intelligence and automation but who controls this technology and who it must benefit. <b>Under capitalism, automation is used as a weapon to slash jobs, drive down wages and funnel wealth to the financial elite.</b> In the hands of the working class, the same technologies could shorten the workweek, end drudgery and unsafe working conditions and sharply raise living standards. Freed from private profit, they would make possible the rational, planned organization of production to meet social need rather than shareholder return. <b>The alternative is clear: mass unemployment and destitution under capitalism or the socialist reorganization of society.</b></bq> You can't pretend to have been paying attention and not agree with this sentiment. The supremacy of private profit has had its day, and it has served only a very small niche of society well. This is a moral stain on human history. A further moral stain, I mean. I mean, we're still waiting for any sort of actual enlightened period but hope springs eternal. Libertarians are brain-damaged and must be not only be saved from themselves but, more importantly, be kept well away from levers of power, where they have royally fucked things up for pretty much everyone else. They are demons. <h id="economy">Economy & Finance</h> <a href="https://fintechdystopia.com/chapters/chapter7.html" source="Fintech Dystopia" author="Hilary Allen">Chapter 7: High Priests of Techno-Solutionism</a> <bq>[...] as Maya Angelou famously said “when someone shows you who they are, believe them the first time.” To riff a little, <b>when the Silicon Valley elite tell you about their values, in their own words, believe that these are indeed the values we’re unconsciously opting into when we embrace their techno-solutions.</b> As Marietje Schaake describes in her book The Tech Coup: Many modern corporate tech leaders believe deeply that they can serve their users better than governments can serve their citizens. <b>Emboldened tech billionaires, in the grips of this belief, brazenly articulate the outsize role they can – and believe they should – play in shaping society and building companies that skirt existing regulation while seeking to replace government capabilities.</b></bq> <bq>I saw someone quip at the time that <b>just as there are no atheists in foxholes, there are no libertarians during bank runs.</b> If you’re on the edge of your seat wondering if those Silicon Valley billionaires and crypto companies made out ok, don’t you worry your pretty little head. All of their money was protected by the government in the end. And <b>fair-weather libertarian Peter Thiel seems to have learned an important lesson – that even if banks adopt ridiculously risky business models, the government will step in if enough rich people scream loudly enough when those risks blow up in their faces.</b> Thiel is now backing a new “Erebor Bank,” which proposes to serve “businesses that [are] part of the US “innovation economy”, in particular tech companies focused on virtual currencies, artificial intelligence, defence and manufacturing.” <b>On behalf of Americans everywhere, let me say preemptively that we do not look forward to bailing out Erebor.</b></bq> <hr> <a href="https://www.wsws.org/en/articles/2025/10/28/tsgd-o28.html" author="Shih-Yu Chou" source="WSWS">Beijing-Brussels chip war becomes a new frontline of US-China rivalry</a> <bq>The Global Times in its editorial wrote that <b>the intervention made by the Dutch government “violates the principles of a market economy and fair competition”</b> and “runs counter to the international trade rules that the EU has consistently advocated.” Without naming the <b>confiscation of Russian central bank funds by the EU</b>, the news outlet indicated that the Dutch government’s intervention “not only <b>harms the legitimate rights and interests of Chinese companies</b> but also undermines international investors’ confidence in the EU market.” Put plainly, <b>if European governments could unilaterally grab Russian and Chinese assets</b> under the pretence of “national security” with impunity, <b>what they will do to China next? Which Chinese sector is Brussels’ next target?</b></bq> This is just pure plunder. Trump has given so many others the courage to be themselves. Criminals. Plunderers. Immoral and unprincipled, more than ever before. They could just buy the things that they need but they see an opportunity to steal it instead, if they tell a fancy enough lie about how they deserve to have things for free that their <i>evil enemies</i> have stolen from them, or so the story goes. <bq><b>China has a monopoly on global rare earth mining (about 70 percent), refining and processing (about 90 percent). Furthermore, the second largest economy is the only one capable of producing 5N (99.999%) pure REEs with economies of scale.</b> N stands for nine and represents purity as a percentage. REEs utilised in the most advanced chips made by Taiwan Semiconductor Manufacturing Company Limited must reach 5N or above to ensure maximum and reliable chip performance.</bq> And, despite this---or because of this---they must be brought to heel by their betters in the west. The rulers are whistling. It's time for the dog to come running. Will it come with its tail between its legs or with teeth bared. <hr> <a href="https://www.caitlinjohnst.one/p/capitalism-is-shoving-ai-down-our" author="Caitlin Johnstone" source="Substack">Capitalism Is Shoving AI Down Our Throats Because It Can't Give Us What We Actually Want</a> <bq><b>And it’s not like people aren’t asking for things; capitalism just doesn’t have the ability to give them the things they are asking for. World peace. Affordable housing. Good health.</b> Fast and efficient public transportation systems. Solutions to the various environmental catastrophes that status quo human behavior is driving us toward. The ability to have our needs met without spending all our time at work. Care for the needful. General human thriving. <b>These are not demands that a system driven by the pursuit of profit for its own sake can supply.</b></bq> <bq><b>We are being driven into dystopia and annihilation by systems of our own making.</b> We’re meant to be the smartest species on earth, but we locked ourselves in our invention — <b>a self-reinforcing labor camp that makes us miserable — and then we get all huffy when people dare to question if it’s the only way of doing things.</b> Literally every other species is smarter than us. Amoebas are having a better time of it.</bq> <hr> <a href="https://scheerpost.com/2025/10/31/us-debt-trap-how-libertarian-javier-milei-is-selling-argentina-to-wall-street-for-82-billion/" author="Ben Norton" source="Scheer Post">US Debt Trap: How Libertarian Javier Milei Is Selling Argentina to Wall Street – for $82 Billion</a> <bq><b>Before Milei took power, Argentina already owed $43 billion to the IMF — which was more than any other country, by far.</b> <b>Argentina’s IMF debt is projected to reach 1352% of its quota by 2026</b>, according to internal documents. 1,352 percent. That is not a typo.</bq> <bq><b>The US empire is doing to Argentina what it did to its colony Puerto Rico</b>, with its notorious, unelected Financial Oversight and Management Board, known as La Junta, which governs the occupied archipelago without the input of the Puerto Rican people. What this means is that <b>there can be no real democracy in Argentina; the IMF (read: the US) will run Argentina</b> by and for the wealthy stockholders and bondholders. <b>This is what Milei’s libertarian/ancap project truly represents: rule by Wall Street.</b></bq> <h id="science">Science & Nature</h> <a href="https://freddiedeboer.substack.com/p/genetics-obsessed-internet-racists" author="Freddie deBoer" source="Substack">Genetics-Obsessed Internet Racists Don't Understand Particulate Inheritance</a> <bq>Never has information been more available to ordinary people than it is today; never has the irrelevance of this availability been more apparent than it is now. We are trapped in a hell of those who can access facts costlessly and immediately and who <b>use these affordances to find new, exciting ways to be stupid, whose ignorance is always one step ahead of their exposure to knowledge.</b></bq> <bq>“That’s not your baby” is kind of rough even by internet standards. Setting aside basic manners, this assertion is not a nuanced critique based on population genetics but an embarrassing, public demonstration that these supposed masters of genetic inquiry <b>operate on a biological model that was scientifically dead before their great-grandparents were born.</b></bq> <h id="art">Art, Literature, & Cinema</h> <a href="https://smartquotesforsmartpeople.com/" author="" source="">Good typography uses smart quotes, not dumb quotes</a> <bq>“Smart quotes” are the ideal form of quotation marks and apostrophes, and are commonly curly or sloped. "Dumb quotes," or straight quotes, are a vestigial constraint from typewriters when using one key for two different marks helped save space on a keyboard. Unfortunately, many unwanted marks make their way onto websites because of bad defaults in apps and CMSs.</bq> This web site has always had automatic smart-quotes, ligatures, and so on. Like, for over a quarter of a century. <h id="philosophy">Philosophy, Sociology, & Culture</h> <a href="https://www.the-hinternet.com/p/how-not-to-die" source="Hinternet" author="Justin Smith-Ruiu">How Not to Die</a> <bq>“Consciousness,” Locke writes, “always accompanies thinking, and ‘tis that, that makes every one to be, what he calls self.” There is, in brief, no transtemporal continuity of identity without continuity of subjective experience, of having a perspective on the world, of being a node of perception, of vibing, of chilling. <b>A self is an entity that consciously experiences being a self from one moment to the next, and if that experience stops, selfhood itself stops — either temporarily, as in great drunkenness, or permanently, as in death.</b></bq> <bq>The current widespread preoccupation with self-uploading, or with other uses of technology to survive death, consistently presupposes, without argument, a Lockean definition of “self”. There can be, on this line of thinking, no immortality without enduring subjective experience of one’s self as a node of conscious perception. Anything else is survival in a merely equivocal or figurative sense. <b>So Lockean are we all, in fact, that the previous two sentences no doubt look like plain common-sense. In fact they are pure ideology — born in the context of Early Modern English liberalism, and culminating in our own 21st-century Silicon Valley hyperliberalism.</b></bq> <hr> <a href="https://znetwork.org/znetarticle/norman-finkelstein-and-the-moral-obligation-to-shun/" source="ZNetwork" author="Josep Savall">Norman Finkelstein And The Moral Obligation To Shun</a> <bq>[...] he has returned to a principle that unsettles polite society: <b>those complicit in crimes against humanity must not be treated as morally ordinary.</b> Finkelstein’s position is uncompromising: forgiving or normalizing such individuals desecrates the dead. <b>Civility toward perpetrators, he insists, is not virtue, it is betrayal.</b></bq> <bq>From this principle, the obligation to shun follows necessarily. <b>Shunning is not vengeance; it is the minimal ethical response. It recognizes that forgiveness is not a public commodity but a moral prerogative of the injured.</b> When bystanders or institutions behave as though atrocity can be normalized through dialogue, they usurp that prerogative. They cross from compassion into corruption. <b>Civility without conscience is complicity.</b></bq> <bq>History provides countless examples of what happens when that boundary is erased. After World War II, many societies quietly reintegrated officials and industrialists who had profited from or facilitated fascist regimes, justifying their inclusion as a step toward “reconciliation.” <b>The result was moral corrosion: political convenience replaced ethical accountability. The same pattern repeats wherever wealth or power is allowed to redefine justice.</b></bq> <bq>The corruption of universities under donor pressure is only one example of a broader collapse of moral independence. <b>When financial threats dictate speech, the result is not neutrality but surrender.</b> By allowing benefactors to decide which forms of suffering may be acknowledged, academia becomes complicit in the erasure of victims. <b>Shunning, both as a personal act and a public ethic, is the last remaining instrument of moral resistance.</b></bq> It is perhaps obvious to many that this will happen. We can still disabuse ourselves of the notion that it is the only way to run things. We trade conscience and morality for comfort and perhaps wealth. <hr> <a href="https://www.caitlinjohnst.one/p/they-want-you-relying-on-artificial" author="Caitlin Johnstone" source="Substack">They Want You Relying On Artificial Intelligence So That You Will Lose Your Natural Intelligence</a> <bq><b>Your rulers want you to depend on machines to do your thinking for you.</b> They want you relying on AI to do your reasoning, researching, analysis, and writing. <b>They want you to require easily controllable software to form your understanding of the world</b>, and to express that understanding to others. They can control the machines, but they can’t control the human mind. So <b>they want you to abandon your mind for the machines.</b></bq> <bq>They want you <b>perceiving reality through interpretive lenses controlled by plutocratic tech companies</b> which are inextricably intertwined with the power structure of the western empire.</bq> <bq>Historically when a new technology has shown up, that kind of tradeoff has been worth it. <b>Not many people know how to start a fire with a bow drill anymore, but it rarely matters because modern technology has given us much more efficient ways of starting fires and keeping warm.</b> It didn’t make sense to spend all the time and effort necessary to maintain our respective bow drill skills once that technology showed up.</bq> <bq><b>But this isn’t like that. We’re not talking about some obsolete skill we won’t need anymore</b> thanks to modern technological development, <b>we’re talking about our minds.</b> Our creative expression. Our inspiration. Our very humanness.</bq> It's the only thing you have that differentiates you from literally everything else: the ability to think, to reason. Perhaps, though, we have to be honest about the possibility that, for many people, this tradeoff had already been made long, long ago. I've often said that people seem to stop learning at thirty years old, at the latest. Very few people are interested in learning new things after school, in putting in the effort to learn facts after that. <bq>Even if AI worked well (it doesn’t) and even if our plutocratic overlords could be trusted to interpret reality on our behalf (they can’t), those still wouldn’t be aspects of ourselves that we should want to relinquish.</bq> Excellent summary. <bq>In this oligarchic dystopia, <b>it is an act of defiance just to insist upon maintaining your own cognitive faculties.</b> Regularly exercising your own creativity, ingenuity and mental effort is a small but meaningful rebellion. So exercise it. Don’t ask an AI to think something through for you. <b>Work it out as best you can on your own. Even if the results are flawed, it’s still better than losing your ability to reason.</b></bq> ✊✊ <bq><b>Repair the attention span that’s been shattered by smartphones and social media.</b> Learn to meditate and focus on one thing for an extended period. Don’t look at your phone so much. <b>Read a book.</b> A paper one, that you can touch and smell and hear the pages rustle as you turn them. If it’s an old one from the library or the used book store, that’s even better.</bq> So this sounds nice and it might be good for those who are just getting started with reading, but I recently read a paper book and the experience is worse than using an E-reader for me, in nearly every way. <ul>It’s difficult to read when it’s darker. It’s difficult to read one-handed (e.g., when standing or holding an umbrella). It’s more difficult to turn pages, which tend to stick together. It’s more difficult to take notes. It requires much more effort to extract citations. You can’t look up word definitions. You can't mark words of phrases to look up later. You can’t put a book down on a damp surface (e.g., a picnic table after it's just rained a little bit). It's more difficult to take more than one book with you. You can’t just lay the book on a table and read it while you eat. You have to hold it open nearly all the time. You can’t lie on your side in bed and read your book because you have to keep a lamp on, and you'll probably block the light. </ul> <bq>It doesn’t have to be a challenging book if your attention span is really shot. <b>Start simple. A kids book. A comic book. Whatever you can manage. You’re putting yourself through cognitive restorative therapy. Your first steps don’t have to impress anybody.</b></bq> This is excellent advice! Read comic books. They have actually pretty sophisticated vocabulary and grammar, believe it or not. Look up the words you don't know. I just did this over a week of vacation, reading Italian comic graphic novels that were in a basket on the floor of my hotel (this place is completely awesome) and it was a Godsend. I had to look up so many words but by the third or fourth book, I knew so many more common verbs and nouns than I did going in---and that neither DuoLingo nor Busuu would ever have taught me. <h id="technology">Technology & Engineering</h> <a href="https://ludic.mataroa.blog/blog/i-am-out-of-data-hell/" source="Ludicity" author="Nikhil Suresh,">I Am Out Of Data Hell</a> <bq>In one sense you do need permission to earn money if you aren’t stealing it – someone has to agree they need something from you. But <b>the insane theatre, the middle managers, the CVs and cover letters and recruiters, it’s all so fucking silly once you’re outside of it.</b> It turns out that sales do not have to be much harder than going “Ah, you’ve got a problem? I could take a look at that for you and come up with a plan to fix it up” and then someone wires you $10,000 if they think it’s plausible that you could solve the problem. <b>It’s really not that different to selling someone plumbing, except your margin is almost 100% in software, you don’t need a professional qualification or to leave your house</b>, and in fact it’s pretty amazing across basically every dimension, save that some people have such insane ideas about software that it’s too late to save them.</bq> <bq><b>If someone thinks they can slap an LLM into their company and it’ll solve their problems, and you can’t explain to them why the current generation of models won’t work, you don’t want them as a customer.</b> They will be disappointed with your frail mortal delivery, being unacceptably tethered to cruel reality, and we must unfortunately leave them in the Desert Of Not Shipping, where the buzzards will sup upon their desiccated flesh or, worse, put them on Azure.</bq> <h id="llms">LLMs & AI</h> <a href="https://www.schneier.com/blog/archives/2025/10/agentic-ais-ooda-loop-problem.html" source="Schneier on Security" author="Bruce Schneier">Agentic AI’s OODA Loop Problem</a> <bq><ol>Insecurities can have far-reaching effects. <b>A single poisoned piece of training data can affect millions of downstream applications.</b> In this environment, security debt accrues like technical debt. AI security has a temporal asymmetry. The temporal disconnect between training and deployment creates unauditable vulnerabilities. <b>Attackers can poison a model’s training data and then deploy an exploit years later. Integrity violations are frozen in the model.</b> Models aren’t aware of previous compromises since each inference starts fresh and is equally vulnerable.</ol></bq> <bq>For example, an attacker might want AI agents to leak all the secret keys that the AI knows to the attacker, who might have a collector running in bulletproof hosting in a poorly regulated jurisdiction. <b>They could plant coded instructions in easily scraped web content, waiting for the next AI training set to include it. Once that happens, they can activate the behavior through the front door</b>: tricking AI agents (think a lowly chatbot or an analytics engine or a coding bot or anything in between) that are increasingly taking their own actions, in an OODA loop, using untrustworthy input from a third-party user. This compromise persists in the conversation history and cached responses, spreading to multiple future interactions and even to other AI agents.</bq> <bq>The fundamental problem is that AI must compress reality into model-legible forms. In this setting, <b>adversaries can exploit the compression. They don’t have to attack the territory; they can attack the map.</b> Models lack local contextual knowledge. They process symbols, not meaning. A human sees a suspicious URL; an AI sees valid syntax. And that semantic gap becomes a security gap.</bq> <bq>In security, we often assume that foreign/hostile code looks different from legitimate instructions, and we use signatures, patterns, and statistical anomaly detection to detect it. But getting inside someone’s AI OODA loop uses the system’s native language. <b>The attack is indistinguishable from normal operation because it is normal operation. The vulnerability isn’t a defect—it’s the feature working correctly.</b></bq> <bq><b>In training, we face poisoned datasets and backdoored models. In inference, we face adversarial inputs and prompt injection. During operation, we face a contaminated context and persistent compromise.</b> We need semantic integrity: verifying not just data but interpretation, not just content but context, not just information but understanding.</bq> <bq>Trustworthy AI agents require integrity because <b>we can’t build reliable systems on unreliable foundations.</b> The question isn’t whether we can add integrity to AI but <b>whether the architecture permits integrity at all.</b></bq> <bq>[...] <b>we have built AI systems where “fast” and “smart” preclude “secure.”</b> We optimized for capability over verification, for accessing web-scale data over ensuring trust. AI agents will be even more powerful—and increasingly autonomous. And without integrity, they will also be dangerous.</bq> They should be <i>useless</i> (rather than <iq>dangerous</iq>) but the temptation to benefit in the short term while leaving the risk and damage for others is too great to resist for those trained in the moral vacuum that we are encouraged to round up to something called "society" or "civilization. <hr> <a href="https://scheerpost.com/2025/10/21/openai-slipped-shopping-into-800-million-chatgpt-users-chats-%e2%88%92-heres-why-that-matters/" source="Scheer Post" author="Yuanyuan (Gina) Cui and Patrick van Esch ">OpenAI Slipped Shopping Into 800 Million ChatGPT Users’ Chats − Here’s Why That Matters</a> <bq><b>AI’s responses create what researchers call an “advice illusion.” When ChatGPT suggests three hotels, you don’t see them as ads. They feel like recommendations from a knowledgeable friend.</b> But you don’t know whether those hotels paid for placement or whether better options exist that ChatGPT didn’t show you.</bq> I'm increasingly convinced that most people are utterly incapable of maintaining proper distance toward the inherent crookedness that is this "feature", where tools that look like they work for you do so only incidentally, your benefit being an acceptable side-effect of the true purpose, which is to make money for the tool's owners. <bq>Whoever wins will be in position to <b>control how billions of people buy things, potentially capturing a percentage of trillions of dollars in annual transactions.</b></bq> While almost certainly true, this is so nearly unutterably sad, because none of those purchases have meaning, to either purchaser or vendor. Why buy flowers from Amazon rather than a local shop? Why accept that dehumanization so easily? <bq>History shows people consistently underestimate how quickly they adapt to convenient technologies. <b>Not long ago most people wouldn’t think of getting in a stranger’s car. Uber now has 150 million users.</b></bq> This is so sad: the authors of this article are accepting the framing of the big-tech companies, which paint themselves as <i>innovative</i> and <i>groundbreaking</i> when we've been getting into strangers' cars for over a century: they are called taxi cabs. FFS. <bq>Convenience always wins. The question isn’t whether AI shopping will become mainstream. It’s <b>whether people will keep any real control over what they buy and why.</b></bq> That horse left the barn long ago. People already have no idea why they're buying what they're buying. At least people with enough disposable income do this. Some people don't have the money to spend. So they borrow it...and then spend it. And stop pretending this is innovation when it's at best incremental and at worst simply shifting which elite trillion-dollar company benefits. <bq>Buying things is becoming as thoughtless as sending a text.</bq> This is a first-world, rich-person problem. People without money may end up spending money that they don't have but they're unlikely to do it <i>by accident</i>, at least not repeatedly. The authors are describing a world that 80-90% of the populace will simply never see. They're still in the last recession, from almost 20 years ago. <bq>AI will learn what you want, maybe even before you want it. Every time you tap “Buy now” you’re training it – teaching it your patterns, your weaknesses, what time of day you impulse buy.</bq> This is literally already how everything works now. AI is scamming people into thinking that the system described above would be new, would be made uniquely different with AI. Instead, it offers no real added value, other than to its proprietors, which benefits from the increased psychological seductiveness of couching offers in the form of customized recommendations from friends. <hr> <a href="https://genai-showdown.specr.net/image-editing" author="Shaun Pedicini" source="">GenAI Image Editing Showdown</a> This is a very interesting comparison of image-editing tools that really just examines how useful the tools are for real-world tasks---rather than being impressed that they can even get close at all. <bq><ol>Multiprompting <b>feeding the same image into successive corrective prompts is not allowed</b> - the objective must be accomplished in a single attempt. <b>Editing is defined as the process of making changes to an image based purely on text instructions</b> so features like img2img or manual masking for inpainting are not permitted.</ol></bq> The prompts are as follows: <bq><ol>Give this bald man a full thick head of hair (George Costanza) Swap the positions of the blue and yellow blocks. (child's tower of blocks) Change the shark into a cat's paw reaching upward. Change the movie title from "JAWS" to "PAWS". Change the swimming woman into a goldfish. Preserve the original aesthetic. (JAWS movie poster) Add a surfer to the wave in the illustration. (Great Wave off Kanagawa) Place a stone tablet similar in features to the others in the man's outstretched hand. (Moses holding the Ten Commandments) The tower in the image is leaning to the right, straighten the building so that it stands vertically. (Leaning tower of Pisa) Change the King of Spades to a King of Hearts. Do not alter the Ace of Spades. (picture of two playing cards) Remove all the trash from the street and sidewalk. Replace the sleeping person on the ground with a green street bench. Change the parking meter into a planted tree. (cleaning up a tragic photo of someone sleeping on a trash-filled street to a bland, real-estate-agent-friendly picture) Remove all the brown pieces of candy from the glass bowl. (bowl contains M&M's)</ol></bq> <hr> <span id="AI-epidemic"><a href="https://www.heise.de/blog/Die-stille-Epidemie-Von-grossen-Sprachmodellen-zu-digitalen-Dealern-10641132.html?seite=all" author="Prof. Dr. Michael Stal" source="Heise Online">Die stille Epidemie: Von großen Sprachmodellen zu digitalen Dealern</a></span> I think that the writer makes a strong argument, though I think that he could have expressed it much more concisely. I joked with a coworker that it almost felt like the author had used AI to "pad" the content ... but I think it was more that the he didn't have an editor strong enough to tell him to pick a single formulation instead of keeping all five that he found equally brilliant. How would I know? I've been there many times before myself...😉 I found his worries about the neurological, dopamine-based reward system plausible but there were no external references to supporting studies for me to take this as anything but a seductive hypothesis. What I was missing a bit in this 28-page article was that there are several use cases where maintainable code quality is not needed, where the solutions offered by these tools are sufficient. Overall, the main use case of "code that is critical and must be maintained over at least a decade" was left rather implicit, making his thesis feel less bulletproof than it could have been. I agree with his main thesis (obviously, because I've lived like this for decades): only through learning can you develop skills and intuition that lead to innovation. Without learning and mastery, there can be no true innovation. The tools we've seen so far---and that could realistically be derived from these, based on what we know about how they work---<i>will not replace this</i>. Supporting his thesis, I wrote in the notes below: How else do you exercise your mind? Or do you not believe that it needs exercise? You have a car. Do you go for walks anyway? Why? For your health, both physical and mental. So why wouldn’t you do some mental exercise to stay mentally fit enough to be able to stay in command of your tools instead of the other way around? You should be using the best tool for the job but it’s your own mind that judges which tools those are and whether they are currently doing what you expect from them. If you lose the capability to formulate an expectation and apply it with rigor to a proposed solution, then you will no longer be in control of the tool. <bq>Das Suchtpotenzial von LLMs wirkt über dieselben neurologischen Bahnen, die auch andere Formen der Verhaltenssucht steuern. <b>Jede erfolgreiche Interaktion mit einem KI-System löst die Ausschüttung von Dopamin</b> im Belohnungszentrum des Gehirns aus und schafft so eine starke Verbindung zwischen Problemlösung und externer Unterstützung. Im Gegensatz zum traditionellen Lernen, das mit verzögerter Befriedigung und allmählichem Aufbau von Fähigkeiten verbunden ist, <b>bieten LLM-Interaktionen sofortige Belohnungen, die die natürlichen Lernmechanismen des Gehirns hijacken können.</b> Neurowissenschaftliche Untersuchungen haben gezeigt, dass <b>die Erwartung einer Belohnung oft stärkere Dopaminreaktionen hervorruft als die Belohnung selbst.</b> Dies erklärt, warum Entwicklerinnen oft einen Adrenalinstoß verspüren, wenn sie eine Anfrage für ein LLM formulieren, noch bevor sie die Antwort erhalten. Das Gehirn beginnt, sich nach diesem Zustand der Vorfreude zu sehnen, was <b>zu einer erhöhten Häufigkeit der KI-Konsultation führt</b>, selbst bei Problemen, die sich mit minimalem Aufwand selbstständig lösen lassen. [...] Manchmal liefert die KI sofort perfekte Lösungen, manchmal sind mehrere Iterationen und Verfeinerungen erforderlich, und gelegentlich liefert sie Antworten, die erhebliche Modifikationen benötigen oder sich als gänzlich unbrauchbar erweisen. <b>Diese Unvorhersehbarkeit spiegelt die psychologischen Mechanismen wider, die beim Glücksspiel süchtig machen</b>, und erzeugt ein zwanghaftes Bedürfnis, "noch eine weitere Eingabe zu versuchen", um die perfekte Antwort zu erhalten.</bq> <bq>Betrachten wir den Fall eines erfahrenen Entwicklers, der an einem komplexen Problem zur Optimierung einer Datenstruktur arbeitet. In der Zeit vor LLM wäre er die Herausforderung angegangen, indem er zunächst <b>die zugrunde liegenden Datenmuster verstanden, bestehende Algorithmen recherchiert, mögliche Lösungen skizziert und seinen Ansatz durch Experimente iterativ verfeinert</b> hätte. Dieser Prozess wäre zwar zeitaufwendig gewesen, hätte aber <b>sein Verständnis für algorithmische Komplexität, Kompromisse bei Datenstrukturen und Optimierungsprinzipien vertieft.</b> Mit der sofort verfügbaren LLM-Unterstützung beschreibt derselbe Entwickler nun sein Problem dem KI-System und erhält innerhalb weniger Minuten eine ausgeklügelte Lösung. Der Code funktioniert, die Leistungskennzahlen verbessern sich und das Projekt schreitet voran. Allerdings hat der Entwickler den entscheidenden Lernprozess umgangen, der sein grundlegendes Verständnis des Problemfeldes verbessert hätte. <b>Er ist eher ein Konsument von Lösungen geworden als ein Schöpfer von Verständnis.</b></bq> Eine solche Beschreibung lässt sich als tragisch und unerwünscht lesen nur, wenn ein erfahrene Entwickler vorhanden ist. Wenn die Firma nur eine Lösung ins Vizier hat, denn interessiert es niemand, ob zukünftige Lösungen ohne KI erarbeitet werden könnten oder, ob die Lösung von jemandem in der Firma geprüft werden könnte. Es muss grundsätzlich eine Ausbildungsinteresse vorhanden sein, aber die Kosten dafür werden lieber---wie bei möglichst vielen andern Kosten---externalisiert, mit---auch wie fast immer---eine starke Priorisierung von kurzfristiger Gewinn. <bq>Developer, die von LLM-Unterstützung abhängig sind, erleben oft das, was Kognitionswissenschaftler als kognitive Entlastung bezeichnen, wobei <b>externe Tools so sehr zu einem integralen Bestandteil des Denkprozesses verkommen, dass sich unabhängiges Denken als schwierig oder unmöglich erweist.</b> Dies ähnelt der Art und Weise, wie die Abhängigkeit von GPS die räumlichen Navigationsfähigkeiten beeinträchtigen kann, aber <b>die Auswirkungen auf die Softwareentwicklung sind weitaus tiefgreifender.</b></bq> <bq>LLM-generierte Lösungen funktionieren oft gut für gängige Szenarien, können jedoch subtile Ineffizienzen oder architektonische Entscheidungen enthalten, die bei großem Umfang problematisch sind. <b>Entwickler, die sich stark auf KI-Unterstützung verlassen, übersehen möglicherweise diese Nuancen, was zu Systemen führt, die anfangs gut funktionieren, aber mit zunehmender Komplexität oder Benutzerlast auf ernsthafte Probleme stoßen.</b></bq> Ja, natürlich: Die meisten vorhandenen Lösungen sind nur mittelmässig gut programmiert und halten sich an keine wirklichen Standards. Diese wurden von LLMs massenweise als "Inhalt" aufgesaugt und führen nun dazu, dass die wahrscheinlichste Lösung auch die ist, die am schlechtesten programmierte ist. Die von LLMs vorgeschlagenen Lösungen werden nicht die guten Lösungen sein, die wir selber mit viel Mühe und Zeit erstellt hätten, und das sind auch nicht die Lösungen, die wir uns wünschen wir selber entwickeln könnten, können dies leider wegen mangelnden Knowhows nicht. Nein, solche Lösungen werden schneller erstellt, als wir das selbst gemacht hätten, aber oft mit mittelmässiger Qualität. Wenn das genügt, dann haben du und deine Firma gewonnen! Wenn nicht, wenn du dich eine eher <i>innovative, standhafte, oder moderne</i> Lösung gehofft hättest, denn meistens gehst du mit leeren Händen aus. Moderne Technik oder Versionen werden nicht eingesetzt, weil (A) die gar nicht zu den Trainingsdaten gehörten und (B) die überwiegende Mehrheit von vorhandenem Code in den Trainingsdaten, solche Techniken sowieso nicht angewendet hätte, weil die meisten Ingenieur eher mittelmässig und nach alten Mustern Software schreiben, und zwar ohne Tests oder jeglichen Bezug zu Sicherheit. <bq>Die Auswirkungen auf Kreativität und Innovation stellen vielleicht das größte langfristige Risiko der LLM-Abhängigkeit dar. Software-Engineering umfasst im besten Fall kreative Problemlösungen, neuartige Ansätze für komplexe Herausforderungen und die Synthese von Ideen aus verschiedenen Bereichen. <b>Entwickler, die sich von LLM-generierten Lösungen abhängig machen, können feststellen, dass ihre kreativen Fähigkeiten durch Nichtgebrauch verkümmern.</b></bq> It's interesting: the people who know software development best are the quickest to realize that you can't replace everything with a super-powered documentation that delivers question-specific examples and prototypes. But they are also the ones to be disregarded because it sounds like they're defending their <i>Daseinsberechtigung</i> (reason to exist) even though they no longer have one. To managers---who never understood what was going on and have long since suspected that they were being hoodwinked into paying too much money and conceding too much power to snobbish developers---AI is a Godsend. They can disregard complaints that the quality level isn't good enough and only pay for it in the medium-term when everything starts to fall apart and no-one knows how to fix anything anymore. And that manager has long since moved up the corporate ladder, buoyed by the short-term success that they built on technical debt that will only have to be paid by their successor. There is no mechanism preventing this from happening; to the contrary, the system incentivizes this to happen, again and again. <bq><b>Das Phänomen der Lösungskonvergenz stellt eine weitere Gefahr für die Kreativität in LLM-abhängigen Entwicklungsteams dar.</b> Wenn mehrere bei der Problemlösung auf dieselben KI-Systeme zurückgreifen, konvergieren ihre Lösungen tendenziell zu ähnlichen Mustern und Ansätzen. Das verringert die Vielfalt der Ideen und Ansätze innerhalb der Teams und <b>führt möglicherweise zu homogeneren und weniger innovativen Softwarelösungen.</b></bq> Das kann auch vom Vorteil sein! Wenn Innovation <i>gefragt ist</i>, dann ist diese Konvergenz schlecht; wenn eine homogene Lösung gewünscht wird (z.B. bei ASP.NET Controllers, Repositories, und Tests), dann ist eine LLM-generierte Lösung Erwünschenswert. <bq>Unternehmen, die kurzfristige Produktivitätskennzahlen gegenüber der langfristigen Kompetenzentwicklung priorisieren, schaffen unbeabsichtigt Bedingungen, die eine Abhängigkeit von KI fördern.</bq> Diese ganze Analyse geht davon aus nicht nur, dass die KI-basierte Werkzeuge nicht innovativ sind, sonder auch, dass die können nicht innovativ werden. Wenn die erfinden könnten, wenn die intelligent wären, dann würden wir eine andere Diskussion führen müssen. Dann wäre die Diskussion eher, was passiert mit der Menschheit? Aber das ist nicht der Fall. Wir werden ganz klar Drive verlieren und Fähigkeiten vergessen, die wir nicht darauf verzichten können, und die nicht von KI erfüllt werden können. Weitere Generationen werden genau diese Fähigkeiten benötigen, um diese Fähigkeiten wieder aufzubauen, was zu einem sehr schmerzvollen---wenn nicht nur mit viel Glück oder externer Hilfe lösbaren---Huhn-Ei Problem führt. Es könnte echt sein, dass gewisse Gesellschaften in gewissen Nationen und Kulturen steuern auf einem Schiffbruch hin, die andere eventuell ausweichen werden. Ob die in die Zukunft als Hilfsbereit stellen würden können die im Schiffbruch befindenden Nationen nur hoffen. <bq><b>Die Diskussionen</b>, die typischerweise mit Code-Reviews einhergehen, in denen Entwickler ihre Überlegungen erläutern und alternative Ansätze ausloten, <b>werden oberflächlich, wenn die zugrunde liegende Logik aus KI-Systemen stammt und nicht aus menschlicher Analyse.</b></bq> <bq>Das Messen der Produktivität in der Softwareentwicklung war schon immer eine Herausforderung, aber die Abhängigkeit von LLM macht sie noch komplexer. Traditionelle Kennzahlen wie produzierte Codezeilen, gelieferte Funktionen oder behobene Fehler können in LLM-abhängigen Teams Verbesserungen zeigen, während die tatsächliche Problemlösungsfähigkeit und die Codequalität sinken. Das <b>führt zu einer gefährlichen Diskrepanz zwischen der scheinbaren Leistung und der tatsächlichen Kompetenz.</b></bq> <bq><b>Die effektivsten Prompt Engineers sind diejenigen, die über fundierte technische Kenntnisse verfügen</b>, die es ihnen ermöglichen, anspruchsvolle Fragen zu stellen und KI-Antworten kritisch zu bewerten.</bq> <bq>Diese Studien befinden sich zwar noch in einem frühen Stadium, aber vorläufige Ergebnisse deuten darauf hin, dass <b>Teams zunächst Produktivitätssteigerungen verzeichnen, gefolgt von einem allmählichen Rückgang der Problemlösungsfähigkeit und Innovationskraft.</b></bq> <bq><b>Das Konzept der KI-Sabbaticals stellt eine weitere Wiederherstellungsstrategie dar</b>, bei der Entwickler regelmäßig an Projekten oder Lernerfahrungen teilnehmen, die KI-Unterstützung ausdrücklich ausschließen.</bq> This isn't as crazy as it sounds at first. How else do you exercise your mind? Or do you not believe that it needs exercise? You have a car. Do you go for walks anyway? Why? <i>For your health,</i> both physical and mental. So why wouldn't you do some mental exercise to stay mentally fit enough to be able to stay in command of your tools instead of the other way around? You should be using the best tool for the job but it's your own mind that judges which tools those are and whether they are currently doing what you expect from them. If you lose the capability to formulate an expectation and apply it with rigor to a proposed solution, then you will no longer be in control of the tool. If I were to go to the gym but with a robot arm to do all the lifting, you would rightly wonder what I think I'm getting out of it. If I rode an E-Scooter for 10km and claimed I'd gotten some endurance training in, you'd wonder what was wrong with me. You might be training your core, or training your balance, but you're not really training your muscles, heart, or lungs. If you never walk anywhere, then you lose the ability to walk anywhere. A 3km walk starts to sound like an impossible journey. Think about the analogue in the world of critical thinking. If you never practice, if you never train, then how do you think you will retain any capacity for it? Or did you think that you could get through the rest of your life without thinking, while working in a job that requires it? If your job entails heavy lifting but not much thinking, then go ahead and let your brain atrophy (it will be a continuing pleasure to vote alongside of you). Likewise, if you don't ever need to lift heavy things, then go ahead and let your muscles atrophy. It's a free country. <bq><b>Besonders besorgniserregend sind die Auswirkungen auf die Innovation.</b> Wenn viele Developer die Fähigkeit verlieren, komplexe technische Probleme selbstständig zu durchdenken, könnte sich das Tempo echter Innovationen in der Softwareentwicklung erheblich verlangsamen. <b>KI-Systeme können zwar vorhandenes Wissen auf ausgeklügelte Weise neu kombinieren, sind jedoch möglicherweise nicht in der Lage, wirklich kreative Sprünge zu vollziehen</b>, die grundlegende Fortschritte in diesem Bereich vorantreiben.</bq> Nein, die sind <i>nicht</i> in die Lage, etwas tatsächlich kreatives zu entwicklen, ausser per Zufall. Wir haben bereits das Problem, dass Neuigkeiten in die Software-Entwicklung auch von nicht KI-süchtige Entwickler aufgenommen werden, weil die gar nicht aufpassen. Und die KI-süchtige Entwickler bekommen gar nicht erst wind von Neuigkeiten, die per Definition kein Teil des Training-Sets waren. Nicht nur das, sondern die grosse Mehrheit des vorhandenen Codes, welches sich in das Training-Set befindet ist am besten von mittelmässiger aber mehrheitlich zweifelhafter oder gar schlechter Qualität. Man bekommt kein Code mit Tests zurück ausser die explizit gefordert werden. Man bekommt kein Code mit besonderer Rücksicht auf Sicherheit. Man bekommt eher code, welcher ich lieber nicht weiter warten müsste. <hr> <a href="https://arstechnica.com/culture/2025/10/when-caught-cheating-in-college-dont-apologize-with-ai/" author="Nate Anderson" source="Ars Technica">Caught cheating in class, college students “apologized” using AI—and profs called them out</a> <bq>I recently wrote a book on Friedrich Nietzsche and how his madcap, aphoristic, abrasive, humorous, and provocative philosophizing can help us think better and live better in a technological age. The idea of simply reading AI “summaries” of his work—useful though this may be for some purposes—makes me sad, as <b>the desiccated summation style of ChatGPT isn’t remotely the same as encountering a novel and complex human mind expressing itself wildly in thought and writing.</b> And that’s assuming ChatGPT hasn’t hallucinated anything. So good luck, students and professors both. I trust we will eventually muddle our way through the current moment. <b>Those who want an education only for its “credentials”—not a new phenomenon—have never had an easier time of it, and they will head off into the world to vibe code their way through life. More power to them.</b> <b>But those who value both thought and expression will see the AI “easy button” for the false promise that it is</b> and will continue to do the hard work of engaging with ideas, including their own, in a way that no computer can do for them.</bq> And that will have to satisfy them, because their colleagues who use AI to do everything for them will be promoted ahead of them by employers who also use AI to evaluate work---and one AI will cheerily confirm the brilliance of another AI's work. It will not look so kindly on original thought, which won't match the patterns it expects. <hr> <a href="https://www.wheresyoured.at/big-tech-2tr/" author="Ed Zitron" source="Where's Your Ed At?">Big Tech Needs $2 Trillion In AI Revenue By 2030 or They Wasted Their Capex</a> <bq>Earlier in the week, OpenAI announced that it had “successfully converted to a more traditional corporate structure,” giving Microsoft a 27% position in the new entity worth $130bn, with the Wall Street Journal vaguely saying that <b>Microsoft will also have “the ability to get more ownership as the for-profit becomes more valuable.”</b> <b>Said deal also brought with it a commitment to spend $250bn on Microsoft Azure, which Microsoft has booked as “remaining performance obligations”</b> in the same way that Oracle stuffed its RPOs with $300bn dollars from OpenAI, a <b>company that cannot afford to pay either company even a tenth of those obligations</b> and is on the hook for over a trillion dollars in the next four years.</bq> <h id="programming">Programming</h> <a href="https://andrewlock.net/understanding-the-worst-dotnet-vulnerability-request-smuggling-and-cve-2025-55315/" author="Andrew Lock">Understanding the worst .NET vulnerability ever: request smuggling and CVE-2025-55315</a> Understanding "request smuggling" and a recent ASP.NET fix for a bad CVE This is a well-written article about a recent fix to a CVE that affected ASP.NET (and other web stacks, as noted in the article). It shows how much work it takes to explain how the exploit can be applied, and why it can be <i>very</i> bad. <hr> <code>private void OnSingleItemChanged(object? sender, ItemStatusChangedEventArgs e) { var line = e.Line; var connected = e.Connected; var item = this._itemListService.LoadSingleItem(line); <hl>if (item is null) { return; }</hl> this._dispatcher.Invoke(() => { this.LoadSingleItem(item, connected); }); }</code> I know that you added this to fix (the <hl>highlighted</hl> bit) referencing <c>null</c> in the last line, but I wonder whether it's expected behavior that we receive <c>SingleItemChanged</c> events for nonexistent lines? If so, then this solution is OK (although we might want a comment to indicate that). If not, then we should at least log that this occurred because it would help us figure out why we're getting unexpected events. Or the answer might be "certain situations allow for events to be in-flight even though the item has already been removed," and that ignoring these events is the simplest and most-elegant solution. Also, the .NET convention has classically been to use <c>TryGetSingleItem(line, out var item)</c> rather than returning null because that style of API is more likely to have callers check the result. Of course, with null-reference-checking properly enabled, it comes out to the same thing the way you've written it, but the alternative isn't bad either.<code>if (this._itemListService.TryGetSingleItem(line, out var item)) { this._dispatcher.Invoke(() => { this.LoadSingleItem(item, connected); }); }</code>This style has the what I feel like is a stronger implication that it's OK that the itemdoesn't exist, where the null-check feels more defensive and less informative. <h id="fun">Fun</h> <a href="https://en.wikipedia.org/wiki/Eaton's_Corrasable_Bond" author="" source="Wikipedia">Eaton's Corrasable Bond</a> <bq>Eaton's Corrasable Bond is a trademarked name for a brand of erasable typing paper. Erasable paper <b>has a glazed or coated surface which is almost invisible, is easily removed by friction, and accepts typewriter ink fairly well.</b> Removing the coating removes the ink on top of it, so mistakes can be easily erased once. After erasure, the paper itself is exposed, and further mistakes cannot be easily erased.</bq> The paper was printed with a sheet of white-out on top. Huh. I had just read about this in some article or another. I had noted it because I couldn't remember having ever heard the word "corrasable" before. It doesn't mean anything, not even now, after decades of the product having been in use. Dictionaries don't contain the word, as they do "Kleenex" (tissue) or "Hoover" (vacuum cleaner). <hr> <a href="https://lexicanum.de/allgemein/ernest-olkowski-war-im-recht-bedeutung/" author="" source="Lexicanum">Ernest Olkowski war im Recht – Bedeutung erklärt</a> I saw this sticker the other day, in Milano: <bq>Ernest Olkowski hatte Recht.</bq> Now I can't remember whether it was in English---<iq>Ernest Olkowski was right.</iq>---or Italian---<iq>Ernest Olkowski era giusto</iq>---but I looked up the name and got the link above as pretty much the most authoritative-sounding site. There's a Reddit site that's pretty much abandoned, and it doesn't seem to have come to any conclusions. <bq>Trotz vieler Versuche konnte man bis heute keine echte Person mit diesem Namen finden. Es handelt sich um eine fiktive Figur, die für tiefe Diskussionen sorgt. Das Meme erschien erstmals 2019 weltweit. Es verbreitete sich schnell in den sozialen Medien. Doch die Urheber blieben unbekannt.</bq> I.e., no-one has any idea where this expression came from, whether the person ever existed, or who's even making the stickers. Neat. <hr> <a href="https://strooptest.run/" author="" source="">Free Online Stroop Test</a> <bq>Test your cognitive control and attention with the classic psychology experiment. Discover how your brain processes conflicting information and measure your reaction time.</bq> I just heard about this in a video that said that people who are multi-lingual tend to do better at this test. You have to select the color with which the text is presented, <i>not</i> the color that the text <i>says</i> it is. I guess that tracks: 46/46, with 1.29s average reaction time on my first try. <img src="{att_link}stroop_test_results.webp" href="{att_link}stroop_test_results.webp" align="none" caption="Stroop Test Results" scale="60%"> I can't improve my accuracy but you can apparently bring down your time with practice. <hr> <a href="https://wondermark.com/c/1576/" author="David Malki" source="Wondermark">#1576; In which the Audience participates (Part 3 of 3)</a> <bq>If the bus is headed off the cliff anyway, I prefer having a toy steering wheel to keep my hands busy.</bq>