Martin Ott @ottpops

Lithuanian composer and conductor Mindaugas Piečaitis, directs his orchestra on the notes of Nora the cat playing the piano. She earns a standing ovation.

Prince was such a massive fan of THE MUPPETS, he asked if he could host the Muppets Tonight show in 1997, leading to one of the rare instances we saw a more comedic version of himself instead of the quiet and reserved persona fans were used to 😂👏

Scientists have identified a reversal of the long-standing Flynn effect—the roughly 200-year trend of rising average intelligence (measured via IQ and cognitive tests) across generations. For the first time in modern recorded history, Generation Z (born roughly 1997–2012) shows lower performance than previous generations in key cognitive domains, including attention, memory, literacy, numeracy, executive function, problem-solving, and general IQ—despite spending more years in formal education than ever before. Neuroscientist and educator Dr. Jared Cooney Horvath, PhD, MEd, testified before the U.S. Senate Committee on Commerce, Science, and Transportation on January 15, 2026, highlighting this shift. In his written testimony, he stated that cognitive development in children across much of the developed world has stalled or reversed over the past two decades, with declines evident in international assessments (e.g., PISA, TIMSS) and other large-scale data starting around the mid-2000s and accelerating post-2010. Horvath attributes the primary driver not to reduced schooling, but to the widespread integration of digital screens and educational technology (EdTech) in classrooms. He argues that human brains evolved for deep, focused learning through face-to-face interaction and sustained attention, not fragmented skimming or constant task-switching encouraged by devices. Key points from his testimony include: - Teens now spend over half their waking hours on screens, with significant portions in school involving computers or tablets—often leading to off-task behavior and shallower processing. - Evidence from meta-analyses and national/international studies shows a consistent pattern: higher classroom screen exposure correlates with weaker outcomes in reading, math, science, and higher-order reasoning. - Digital tools may aid narrow, repetitive skill practice in controlled settings, but in core academic contexts, they tend to reduce depth of understanding, retention, and critical thinking. Horvath describes this as a "structural mismatch" between human cognition and how digital platforms are designed (to capture and fragment attention), warning that unchecked EdTech adoption risks long-term harm to workforce skills, innovation, and societal reasoning. [Horvath, J. C. (2026). Written testimony before the U.S. Senate Committee on Commerce, Science, and Transportation. U.S. Senate]

This is why buildings don’t sink! Soil science is wild. Watch this👇

NASA 的火星车传回了令人惊艳的画面。 远在2.25亿公里外的火星夜空，呈现出了完全不同的景象。

🇫🇷🎻🏰 INSOLITE | Ce groupe normand de musique médiévale reprend « Beat It » de Michael Jackson avec des instruments d’époque. Le résultat est bluffant : on a l’impression d’entendre un tube joué dans une taverne du XIVe siècle. 😂 (Insta : @coursevalbardcore)

Yesterday scientists published proof that AI has a fundamental cognitive weakness. Not on a hard problem. Not on a complex benchmark. On a test designed for undergraduate psychology students. The paper was published June 10, 2026 in PNAS Nexus, one of the most rigorous scientific journals in the world. Three researchers from Texas Tech University and the City University of New York. One finding that reframes every confident claim about AI capability you have heard in the last two years. Here is the test they used. The Stroop task. Invented in 1935. Used in psychology labs for 90 years. You see a word, the word says RED but the ink is printed in blue. Your job is to name the ink color. Not read the word. Just name the color. Your brain fights itself. The obvious answer, RED is wrong. The correct answer, blue requires suppressing the thing your mind wants to do automatically. That suppression is called executive control. It is one of the most fundamental measures of cognitive function humans possess. It underlies everything important following complex instructions, maintaining a rule across a long task, catching when a later piece of information contradicts an earlier one, noticing when you are about to give the wrong answer because the obvious answer is wrong. Researchers gave top AI models the classic attention test and found a major flaw. While the models could correctly name colors in short lists, their performance deteriorated sharply as the task became longer and more complex. Short list. The AI is fine. Gets it right. Looks capable. Impressive even. Longer list. Performance collapses. The paper describes the finding as deficient executive control in transformer attention. Not slower. Not less accurate. Deficient. Here is what makes this alarming beyond the benchmark score. Every enterprise AI deployment in the world right now is built on an assumption. The assumption is that if AI performs well on the demonstration the controlled test, the curated benchmark, the polished proof of concept, it will perform comparably on the real task. The Stroop finding breaks that assumption at its foundation. Short task. Looks fine. Long task. Deficient. The demo is always short. The real work is always long. A legal AI reviewing a 300-page contract needs to maintain a rule flag this clause type,across hundreds of pages of text. A medical AI analyzing a complex patient history needs to hold context across dozens of symptoms, test results, and medications without letting the obvious pattern override the correct one. A financial AI auditing a large dataset needs to catch the exception buried in page 47 of a 60-page report. These are Stroop tasks. They are tasks where the obvious answer is wrong and executive control is the only thing that catches it. And the paper published yesterday says that executive control in transformer-based AI models is deficient under exactly the conditions where it matters most. The AI does not know the task is getting harder. It does not experience cognitive load increasing. It does not know it is failing. It generates the wrong answer in the same confident tone it used for all the right ones. There is no uncertainty signal. No "this is getting complicated." No slowdown that would prompt a human to pause and check. Just a confident answer that is increasingly wrong as the task grows. AI passed the bar exam. Scored 90% on elite mathematical competitions. Achieved human-level performance on medical licensing exams. Every benchmark the AI industry uses to demonstrate capability is a short Stroop task. The real work is the long one. Yesterday researchers published proof that the long one is where AI breaks. Source: Patel, Wang, Fan · Texas Tech University + City University of New York · "Deficient Executive Control in Transformer Attention" · PNAS Nexus ( Link in the comments)

Weeks into filming The Breakfast Club, John Hughes made a decision: Judd Nelson had to go. It wasn't a passing frustration. Molly Ringwald later described Hughes as angrier than she had ever seen him — not the managed tension of a director having a rough day, but something final. Something absolute. The set went still. Everyone understood what it meant. Nelson had arrived with a commitment that made people uncomfortable. Trained in the method tradition at the Stella Adler Conservatory, he had decided before shooting began that he would never step out of character. Not between takes. Not at lunch. Not for anyone. The character was John Bender — hostile, wounded, probing for weakness in everyone around him, turning every room into a test of who would break first. Staying in that character continuously meant staying sharp and needling and unpredictable. The set, much of the time, felt like the library in the film. His primary target was Ringwald. She later said she understood. She could see the method behind the provocation — that the discomfort Nelson created between takes was the same discomfort that made their scenes together feel raw instead of rehearsed. She was a professional. She recognized a professional working. Hughes did not see it that way. When he said he wanted Nelson removed, the cast faced a quiet crisis. Bender wasn't a supporting role. He was the engine — the force pushing every other character toward the truth the whole film was built around. Recasting weeks in, with the chemistry already forming, wasn't a minor adjustment. It was a potential collapse. So they walked in together. Ringwald went first — the person Nelson had been targeting, the one whose discomfort had set everything off. Ally Sheedy went. Anthony Michael Hall went. Emilio Estevez went. They told Hughes that what looked like provocation was commitment. That the discomfort was the point. That Bender was working because Nelson was working. Paul Gleason — the veteran playing Principal Vernon, a man who had spent a career learning to tell the difference between someone causing problems and someone solving them unusually — added his voice. Nelson deserved to finish. Hughes listened. He changed his mind. Nelson stayed. He fought for every corner of the character — including the scene where Bender describes his father burning him with a cigar. Hughes wanted the cruelty to read as careless, accidental. Nelson pushed back hard. He had built this character from the inside, and the inside required deliberate damage — a father who chose to hurt his son, because only that specific intentional cruelty could explain the specific intentional damage Bender aimed at everyone around him. He won that argument too. The scene plays exactly as Nelson wanted it. Bender describes his father with the flat affect of someone who processed horror by draining all the feeling out of it. The emptiness is more disturbing than anger would have been. Roger Ebert, reviewing the film on its release in February 1985, called Bender the strong center of the movie. The Breakfast Club cost around a million dollars to make. It earned fifty-one million. It became one of the defining films of a generation — not as a time capsule, but as something that kept finding new audiences for decades, because what it said about identity, and the gap between who we appear to be and who we actually are, was true enough to stay true. Nelson said afterward that Hughes was one of the only directors he ever worked with who genuinely wanted collaboration — who would hear a better argument and change his position. Working with others later made him understand how rare that was. He also said, simply, that he never wanted to be difficult. He just didn't want to fail the character. He didn't fail him. But the Bender that ended up on screen — the one generations of teenagers recognized as the most honest portrait of a certain kind of hurt they had never seen named before — exists because four young actors walked into a hard conversation with their director and asked him to give a fifth one more time. A film about strangers choosing to protect each other was saved by strangers choosing to protect each other. *That's either a coincidence. Or it was always the only way it could have worked.

A mathematician coined the term "artificial intelligence" in 1955, built the language that dominated AI research for 30 years, and predicted cloud computing 40 years before AWS existed and almost nobody outside the field knows his name. His name was John McCarthy. He was born in Boston in 1927, earned his PhD in mathematics from Princeton in 1951, and spent the next 55 years working on a single question that most of his peers considered either impossible or insane. Can a machine think? In the summer of 1955, McCarthy sat down and wrote a two-page proposal for a workshop at Dartmouth College. The proposal opened with one sentence that changed everything: "every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it." He needed a name for the field he was proposing. He chose "artificial intelligence." Before that document, no such field existed. After it, every researcher working on thinking machines had a name for what they were doing, a home discipline to publish in, and a founding document to point to. McCarthy did not just contribute to AI. He created the container it lives in. The Dartmouth Conference ran for eight weeks in the summer of 1956. It was the moment AI became a real scientific discipline. McCarthy kept building. In 1958 he invented LISP, the second oldest high-level programming language still in use today, older only than FORTRAN by one year. LISP was designed for a specific purpose: symbolic reasoning. It could manipulate ideas, not just numbers. It became the language every serious AI researcher wrote in for the next three decades. From 1958 through the late 1980s, if you were working on AI, you were almost certainly working in LISP. Inside LISP he invented garbage collection in 1959, the technique that automatically frees up memory a program no longer needs. Java uses it. Python uses it. JavaScript uses it. Every modern language that manages memory automatically is using the idea McCarthy worked out while building LISP. In 1961 he stood at a centennial celebration at MIT and said something that everyone in the room thought was science fiction. He proposed that computing would one day be delivered as a public utility, the same way electricity or water is delivered to a home. You would not own the computer. You would pay for access to it over a network. AWS launched in 2006. Azure launched in 2010. Google Cloud launched in 2011. What McCarthy described in 1961 is now a trillion-dollar industry. He was 45 years early. In 1962 he founded the Stanford Artificial Intelligence Laboratory, SAIL, which became one of the most important research centers in the history of the field. The researchers who trained there shaped the next 40 years of AI. He won the Turing Award in 1971. The National Medal of Science in 1990. The Benjamin Franklin Medal in 2003. He retired from Stanford in 2000. He died on October 24, 2011, at his home in Stanford, California. He was 84. The researchers at OpenAI, Google DeepMind, and Anthropic building the models you use today are working in a field McCarthy named in 1955, using memory management he invented in 1959, inside an industry structure he predicted in 1961, toward a goal he spent his entire career insisting was not only possible but inevitable. He was right about all of it. He just did not live to see the part where the rest of the world finally believed him.

William Shatner’s unforgettable performance of Elton John’s ‘Rocket Man’ at the 1978 Saturn Awards.

Who in their right mind puts AI in a giant dinosaur animatronic?!

🚨🇺🇸BREAKING: The CIA officer caught with $40 million in gold bars allegedly invented an entire fake top-secret spy program to steal the money. As if this story couldn't get and wilder: -David Rush allegedly built a sham "special access program," the blackest box in U.S. intelligence, so secret even top-clearance officials couldn't look inside without authorization -The fake program posed as "continuity of government" work, the doomsday planning that keeps Washington running after a nuclear war -He allegedly read in two colleagues as unwitting accomplices and used a made-up government contract to funnel millions, persuading a defense contractor to buy huge amounts of gold -The FBI raid on his home seized 303 gold bars worth roughly $40 million, $2 million in cash, and 35 luxury watches -Investigators say he lied about his college degrees, faked being a Navy pilot, and still sailed through the CIA's notoriously brutal vetting -A judge ordered him held as a flight risk, and several CIA officials are now on leave as the probe widens The scheme worked because of the system, not in spite of it. The secrecy walls built to hide operations from China and Russia hid the fraud from the CIA itself. A man with a fake résumé ran a fake doomsday program inside the most paranoid institution in America, and for years nobody noticed... Source: Washington Post

keyboard_arrow_left Previous keyboard_arrow_right Next

🚨 PHYSICISTS JUST CONFIRMED “NEGATIVE TIME” IS REAL IN A MIND-BENDING QUANTUM EXPERIMENT. Light can exit a cloud of atoms before it even enters. In a new experiment, researchers fired photons through a dense cloud of ultra-cold atoms and measured something that shouldn’t be possible in classical physics. Some photons appeared to spend a negative amount of time inside the cloud effectively leaving before they had fully arrived. Why this matters: • This isn’t time travel it’s a quantum effect involving how light interacts with matter at the deepest level • It comes from “weak measurements” that let scientists observe the system without fully disturbing it • The atoms themselves “report” spending negative time in an excited state • It challenges our everyday intuition about cause and effect in quantum systems The deeper implication is enormous: We are seeing the strange, non-intuitive nature of quantum mechanics play out in real experiments. Time at the quantum scale doesn’t always behave like the arrow we experience in daily life. Effects can appear to precede causes in measurable ways without breaking relativity or causality. This is one of the clearest experimental windows yet into how reality works at its most fundamental level. What do you think does “negative time” change how you see reality, or is it just another quantum quirk we’ll eventually get used to? Follow for more frontier physics and reality-bending discoveries.

🦔GitHub Copilot switched to token-based billing this morning and users are already out of credits. Pro+ subscribers paying $39 a month are reporting 60% of their credits gone in two hours of normal use. One user lost 20% of their allowance from a single file review with no code changes. Another hit their monthly cap before the calendar even flipped to June. Orgs with shared token pools have no way to see individual usage, so entire teams get cut off when one person runs a heavy prompt. Users are canceling and moving to Claude Code and Codex. GitHub community forums are on fire. My Take Flat-rate AI subscriptions were always subsidized. Everyone in the industry knew it. Today the subsidy ran out for a few million developers at once. The problem is a lot of companies already restructured around these tools. They cut headcount and told remaining engineers to lean on Copilot instead of building skills internally. Those companies now depend on a tool whose cost just became unpredictable and whose usefulness completely changes when you have to ration prompts to stay under budget. The developers moving to Claude Code and Codex will hit the same wall eventually. Every AI provider faces the same unit economics. Anthropic filed its S-1 this morning, and the durability of its revenue depends on whether customers stick around once real pricing kicks in everywhere. If a $39 subscriber cancels after one day because the tool became unusable, multiply that across millions of seats and the churn risk becomes very real. Today showed what happens when AI pricing meets reality. The companies that built their workflows around cheap tokens just discovered the tokens aren't cheap anymore and the people who knew how to do the work without them are already gone. Hedgie🤗

في عام 1983 ظهر مزارع أسترالي نحيف عمره 61 عام اسمه Cliff Young على خط انطلاق سباق ألترا ماراثون بطول 875 كيلومتر بين سيدني وملبورن. كان يرتدي ملابس عمله وحذاء مزرعة، لدرجة أن البعض ضحكوا عليه، بعضهم قال له بشكل مباشر: (ستنهار وقد تموت قبل أن تنهي السباق ) فرد عليهم: كنت أطارد أغنامي وأبقاري أيام العواصف أحيانا ليومين أو ثلاثة أيام متواصلة، أعتقد انني أستطيع فعل هذا أيضا ثم حدث شيء غريب، بينما كان الجميع يركضون ثم ينامون لساعات، واصل كليف الركض، ليلة بعد ليلة، يوما بعد يوم كأنه آلة لا تعرف التوقف ويأكل ويشرب وهو يركض، لدرجة أنه لم يتوقف الا لقضاء حاجته فقط وفي النهاية أنهى ال875 كيلومتر خلال 5 أيام و15 ساعة فقط، محطما بذلك الرقم القياسي ومتقدما على أقرب منافسيه بحوالي 10 ساعات كاملة. كتبت الصحف أن رجل عجوز من مزرعة نائية أعاد تعريف حدود التحمل البشري، وأثبت أن بعض البشر تدربهم الحياة بطريقة لا تستطيع أي صالة رياضية في العالم مجاراتها

@Gonzohall We really like GoFit in Campo Grande and walk 20 minutes because of facilities, enormous pool etc. - cheaper than Holmes by a lot.

Mark Hamill says he wanted Luke, Leia, and Han to reunite in the Star Wars sequels but he got rejected. “I said, ‘It’ll only take 30 seconds.’ JJ said, ‘Well, Mark, it’s not Luke’s story anymore”

🚨 MATHEMATICIANS ARE NOW CHALLENGING DARK ENERGY ITSELF. For decades we’ve explained the universe’s accelerating expansion by adding a mysterious force called dark energy. But a new mathematical paper argues it might not be needed at all. Researchers at UC Davis say the standard cosmological model may be unstable under the Einstein–Euler equations — the mathematical framework that combines general relativity with the dynamics of cosmic fluids. Their bold claim: the acceleration we observe could emerge naturally from instabilities in Einstein’s own equations… without any invisible dark energy. Why this matters: Dark energy is supposed to make up ~70% of the universe’s energy budget. Yet we have never detected a single particle of it, nor do we know what it actually is. It’s simply something we inserted into the equations to make the numbers work. The deeper implication is massive: Maybe the universe isn’t being pushed apart by an exotic substance. Maybe the standard model of cosmology is sitting on an unstable solution — like a pencil balanced on its tip and the slightest perturbation naturally drives cosmic acceleration. This does not mean dark energy is dead. But it does mean one of the deepest assumptions in modern physics is being seriously challenged again. And that is exactly how revolutions in cosmology begin. What if the biggest “mystery force” in the universe was never a force at all? Follow for more frontier physics and cosmic discoveries.

Hahahahahha!

Sir Escanor (𝘏𝘰𝘱𝘪𝘶𝘮 𝘚𝘭𝘢𝘺𝘦𝘳) @EscanorReloaded

4 weeks ago

CEOs are quietly realizing the AI replacement plan has a problem. Two problems, actually. One: the token costs for running AI agents are now exceeding what they were paying the employees they fired. Two: when the tokens run out, the AI stops. Just stops. No continuity. No

1K 14K 53K 3.3M 9K