Daily Science to all

🗂 The @science archive now lives on the web Every post from this channel — searchable, filtered, and mapped on an interactive timeline. One page, no apps, no logins. Thanks to AI and just 1 prompt.. crazy! 🔹 356 posts and counting — the full archive since 2024, auto-synced with the channel several times a day 🔹 Five frontiers: AI, Space, Biotech, Physics and FutureTech — filter by category or search any keyword across titles and summaries 🔹 An interactive timeline of scientific breakthroughs from 2012 to 2025 — from AlexNet to room-temperature superconductor claims, hover any dot for the story 🔹 Every card links straight back to the original post here on Telegram 🔹 Built lightweight: a single page that loads in under a second, works on any phone The archive grows automatically as new posts appear on the channel. 🔗 http://144.172.108.222/science/

✨ Pterosaurs Shimmered in Iridescent Greens and Magentas — 120-Million-Year-Old Fossil Rewrites the Look of Earth's First Flying Vertebrates For decades, paleoartists have imagined pterosaurs in vivid, colorful hues. Now, a stunning new fossil analysis suggests that at least one species really did shimmer with shifting iridescent colors, much like modern starlings and pigeons. The discovery comes from a specimen of Sinopterus dongi, unearthed in northeastern China. Scanning electron microscopy revealed layered arrays of melanosomes within the creature's pycnofibers — structures that closely resemble those producing iridescence in modern bird feathers. Computer simulations predict deep greens and magentas that shifted with viewing angle. The diversity and organization of melanosomes matches patterns seen only in warm-blooded birds and mammals, suggesting elevated metabolisms and sophisticated thermoregulation — traits long debated among paleontologists. The finding also hints that iridescent displays may have played a role in courtship rituals.

"This is one of the most intriguing and surprising fossil discoveries of the past few years." — Dr. Steve Brusatte, University of Edinburgh 📄 Original paper (bioRxiv) · Science News summary

#paleontology #pterosaurs #fossil #evolution #iridescence #science

🤖 A 100-billion-parameter AI model was just trained across random GPUs scattered around the globe — not in a billion-dollar datacenter. And it worked. Macrocosmos, building on the Bittensor network, has demonstrated Orion-100B: a 100B-parameter language model trained across geographically distributed Nvidia A100 GPUs. Their system, called IOTA, splits the model itself across many machines using 16 pipeline-parallel stages — unlike earlier decentralized approaches that often required each participant to host the full model. The result: more than 30% model FLOP utilization and roughly 65% of the efficiency of a comparable datacenter setup. The technical challenge was serious. Macrocosmos had to reduce massive inter-GPU traffic, handle unstable nodes, work with heterogeneous hardware, and keep the training process alive across a decentralized network. Their ResBM activation compression technique reportedly reduced traffic from around 150MB to 2.2MB per stage. The team says it ran more than 700 experiments before scaling from a 1.5B test model to 100B in about a month. Nikolas Bush’s Take:

This story matters far beyond the technical achievement. First, if this approach scales, it could change the economics of AI training. A 100B-parameter model trained on geographically distributed A100 GPUs at roughly 65% of comparable datacenter efficiency is not yet a replacement for hyperscaler infrastructure — but it is a serious signal. It suggests that large-scale AI training may not always require a single billion-dollar GPU cluster. Second, the Bittensor layer is important. This is not just a distributed computing experiment; it is an incentive system. GPU owners can be rewarded for contributing compute, which creates the foundation for a market around idle hardware. In simple terms, this could become something like “Airbnb for AI training”: monetizing unused GPU capacity the way Airbnb monetized unused rooms. Third, the uncomfortable part: decentralized AI training has often been dismissed by the mainstream AI community as impractical. Orion-100B does not prove that decentralized training will beat datacenters tomorrow. But it does prove that the idea deserves to be taken much more seriously. The next phase — permissionless participation from consumer hardware — will be the real test. If that works, the AI infrastructure map could become much more distributed than many people expected.

Original report: https://macrocosmosai.substack.com/p/orion-100b-distributed-pretraining Summary: https://www.tao.media/macrocosmos-unveils-orion-100b-a-100b-parameter-distributed-ai-training-run/ #AI #DecentralizedAI #Bittensor #LLM #DeepLearning @science

🧬 "Undruggable" No More: New Pill Nearly Doubles Survival in Advanced Pancreatic Cancer For decades, pancreatic cancer has been one of medicine's most hopeless diagnoses. More than 90% of cases are driven by mutations in a gene called KRAS — a protein scientists long labeled "undruggable" because its surface is so unnaturally smooth that no drug could latch onto it. That era just ended. A new oral drug called daraxonrasib takes a brilliantly indirect approach: instead of trying to bind KRAS directly, it grabs onto a helper molecule called cyclophilin A inside the cell. That drug-protein complex then clamps onto active KRAS and physically shuts it down, silencing the "grow forever" signal at its source. The approach is so novel that it targets multiple mutant forms of RAS at once, making resistance much harder for the cancer to develop. In a Phase 3 trial of 500 patients with metastatic pancreatic cancer who had already been through prior treatment, the results were striking. Patients on daraxonrasib lived a median of 13.2 months compared to just 6.7 months on standard chemotherapy — nearly double. The drug reduced the overall risk of death by 60%. Results were presented by Revolution Medicines and published in the New England Journal of Medicine. Side effects are real — a prominent skin rash affected 86% of patients, along with mouth sores, diarrhea, and nausea — but patients on daraxonrasib were far less likely to abandon treatment than those on chemo, and reported better quality of life with less pain. — Median overall survival: 13.2 months (daraxonrasib) vs 6.7 months (chemotherapy) — 60% reduction in risk of death — Once-daily pill — no infusions, no hospital visits — Works against multiple RAS mutations simultaneously, limiting resistance — Lower treatment discontinuation rate and improved quality of life vs chemo "For decades, successfully targeting the central mechanism that causes the vast majority of pancreatic cancers was considered impossible. That narrative is rapidly changing." — Dr. Christopher Lieu, Professor of Medical Oncology, University of Colorado Why it matters: Pancreatic cancer kills 97% of patients with metastatic disease within five years. Chemotherapy has been our only real tool — a blunt instrument with brutal side effects. Daraxonrasib is the first therapy to go after the genetic engine of the disease itself. If approved, it would be the most significant advance in pancreatic cancer treatment in a generation, and the platform — hijacking cyclophilin A to reach "undruggable" targets — could open doors for dozens of other cancers driven by RAS mutations. 📄 NEJM: https://www.nejm.org/doi/full/10.1056/NEJMoa2605555 📖 ScienceDaily: https://www.sciencedaily.com/releases/2026/06/260604044247.htm 📖 The Conversation: https://theconversation.com/breakthrough-drug-nearly-doubles-survival-with-advanced-pancreatic-cancer-an-oncologist-explains-how-daraxonrasib-overcame-an-undruggable-disease-283647 #PancreaticCancer #KRAS #CancerResearch #MedicalBreakthrough #science

🧬 Cancer Cells' Favorite Escape Trick Backfires — And Scientists Just Discovered How to Exploit It For decades, immunologists have operated under a simple assumption: cancer cells evade the immune system by shutting down a protein called MHC class I, which acts as a "wanted poster" for killer T cells. Without this signal, CD8+ killer T cells become blind to the tumor, allowing it to grow unchecked. But a groundbreaking study published in Nature Immunology has now flipped that assumption on its head. Researchers at Baylor College of Medicine and the University of Michigan discovered that when cancer cells silence MHC I to hide from killer T cells, they inadvertently expose themselves to a completely different immune attack. Instead of becoming invisible, the tumor cells become hyper-visible to CD4+ "helper" T cells — immune cells long thought to play only a supporting role. These helper T cells then trigger ferroptosis, a violent form of cell death driven by iron-catalyzed oxidative stress that essentially rusts the cancer cell from the inside out. The team, led by Dr. Pavan Reddy at the Dan L Duncan Comprehensive Cancer Center, validated this mechanism across mouse models, human tumor samples, and large clinical datasets from patients who had received checkpoint inhibitor therapies. The results held not only for cancer but also for graft-versus-host disease — a dangerous complication of bone marrow transplants — suggesting the finding rewires our fundamental understanding of T cell biology. — Cancer cells reduce MHC I to hide from CD8+ killer T cells — This loss makes them unexpectedly vulnerable to CD4+ helper T cells — CD4+ cells kill via ferroptosis — iron-driven oxidative destruction — The same mechanism operates in transplant complications — Clinical patient data confirms relevance to real-world outcomes "Our work, if further validated, will have implications for T cell-mediated immune responses beyond cancer and transplant immunology," said Reddy. "This may allow for the development of novel strategies that target MHC class I and CD4+ T cells." Why it matters: Many aggressive tumors become resistant to immunotherapy precisely because they drop MHC I expression. Until now, this was seen as a dead end. The new discovery suggests these "escaped" tumors may actually be the most vulnerable — if we can learn to weaponize CD4+ T cells against them. It opens a new front in cancer immunotherapy, especially for patients who have stopped responding to existing treatments. 📄 Original paper (Nature Immunology): https://doi.org/10.1038/s41590-026-02480-z 📖 Readable summary (ScienceDaily): https://www.sciencedaily.com/releases/2026/06/260603023911.htm #Immunology #CancerResearch #Immunotherapy #Science #Breakthrough

🧬 Scientists Find the "Off Switch" That Exhausts CAR T Cells — and Show How to Flip It CAR T-cell therapy is one of the most powerful tools in modern oncology: take a patient's own immune cells, genetically reprogram them to hunt cancer, and put them back. It works wonders against some blood cancers. But against solid tumors — the majority of cancer cases — CAR T cells burn out too fast. Now, an international team has pinpointed exactly why. Researchers at Columbia University and University Hospital Tübingen, led by CAR T pioneer Prof. Michel Sadelain and Prof. Judith Feucht, screened roughly 400 transcription factors — proteins that act as master switches for gene activity inside cells. One protein stood out dramatically: NFIL3. It turned out to be a primary driver of T-cell exhaustion, the process that gradually strips engineered immune cells of their cancer-killing power. Using CRISPR/Cas9 gene editing, the team snipped out the gene responsible for NFIL3. The result? The edited CAR T cells stayed active significantly longer, multiplied more efficiently, and maintained a sustained anti-tumor assault. In mouse models, NFIL3-disabled cells delivered stronger tumor control and extended survival compared to standard CAR T cells. Key findings: — NFIL3 was identified as the dominant transcription factor driving CAR T-cell exhaustion out of ~400 candidates screened — CRISPR deletion of NFIL3 kept CAR T cells functional and proliferating for much longer periods — NFIL3-knockout CAR T cells showed superior tumor control across multiple animal models, including solid tumors — The approach targets the biology of exhaustion itself rather than the tumor type, potentially helping across many cancers "Switching off NFIL3 could be a decisive step toward significantly improving the long-term potency of CAR T cells," said Prof. Feucht. "We expect this to open up new possibilities in the treatment of cancer patients." Why it matters: CAR T therapy has been a revolution in blood cancers but has largely failed against solid tumors — breast, lung, pancreatic, brain — because the engineered cells simply don't last. This discovery offers a concrete, druggable target to make CAR T durable enough for the cancers that kill the most people. It's not a new therapy — it's a way to make the existing one finally work where it's needed most. 📄 Original paper (Cancer Discovery): https://doi.org/10.1158/2159-8290.CD-25-1524 📖 Readable summary: https://www.sciencedaily.com/releases/2026/06/260602021641.htm #CARTcell #CancerResearch #CRISPR #Immunotherapy #Oncology

🧠 Scientists Discover the Hidden Molecular Switch That Keeps Alzheimer's Inflammation Stuck in Overdrive Researchers at Scripps Research Institute have identified a precise molecular mechanism that explains why the brain's immune system becomes trapped in a state of chronic, destructive inflammation in Alzheimer's disease. The discovery, published in Cell Chemical Biology, reveals that a single chemical modification to a protein called STING — at a specific building block known as cysteine 148 — acts as an "on switch" that cannot turn itself off. The brain has its own built-in immune defenses, and STING normally serves as an early-warning system against infections. But in Alzheimer's patients, the team found that STING undergoes a process called S-nitrosylation (SNO), where a nitric oxide-related molecule latches onto cysteine 148. This transforms STING into a hyperactive form — dubbed "SNO-STING" — that clusters into large complexes and continuously pumps out inflammatory signals. The researchers confirmed elevated levels of this rogue protein in postmortem human brain tissue, in human stem cell-derived brain immune cells, and in mouse models of the disease. What makes this cycle particularly vicious is that the very protein clumps associated with Alzheimer's — amyloid-beta and alpha-synuclein — can themselves trigger the S-nitrosylation of STING. Aging, air pollution, and even wildfire smoke further fuel the process by increasing nitric oxide in the brain. The result is a self-perpetuating "SNO-STORM": inflammation generates more NO, which modifies more STING, which drives even more inflammation, gradually destroying the synapses neurons use to communicate. — A single amino acid (cysteine 148) on the STING protein is the exact site of the damaging modification — Blocking SNO-STING formation in mice significantly reduced neuroinflammation — Crucially, synaptic connections between neurons were protected from degradation — the same connections whose loss correlates with cognitive decline — Unlike broad anti-inflammatory drugs, targeting cysteine 148 quiets only the pathological overactivation while leaving normal immune function intact — The same pathway was confirmed active in human Alzheimer's brain tissue and stem-cell models "This is a new and important therapeutic target for Alzheimer's disease," said senior author Stuart Lipton, the Step Family Foundation Endowed Chair at Scripps Research and a clinical neurologist. "It's exciting to see that blocking this switch in mice reduces inflammation and protects the very brain cell connections that are lost in Alzheimer's." Why it matters: Alzheimer's affects over 55 million people worldwide, yet nearly all clinical trials targeting amyloid plaques have failed or shown marginal benefit. This discovery shifts the focus to neuroinflammation as a driver — not just a bystander — of the disease. The fact that the target is a single, well-defined amino acid means drug developers have an unusually clean bullseye. Lipton's team is already working on small-molecule drugs designed to sit on cysteine 148 and prevent the SNO modification, potentially offering the first therapy that breaks the inflammation cycle without crippling the immune system. 📄 Original paper (Cell Chemical Biology): https://www.cell.com/cell-chemical-biology/fulltext/S2451-9456(26)00109-1 📖 Readable summary (ScienceDaily): https://www.sciencedaily.com/releases/2026/05/260530053424.htm #Alzheimers #Neuroscience #Neuroinflammation #STING #DrugDiscovery

Interesting visualization of the mechanism behind Earth’s tides and ebbs.

🤖 AI-powered robot barber kiosks have begun operating in several Chinese cities, offering haircuts with millimeter-level precision. The system works in three stages: • 3D scanning — a sensor array maps the customer's head shape, facial geometry, and hair type • Style selection — the user picks a haircut via a digital interface, and the AI adjusts the cutting path to match the chosen pattern • Robotic execution — a mechanical arm trims hair while continuously monitoring length in real time to maintain uniformity Each session costs ¥60 (~$8), making it competitive with budget salons. Developers claim the kiosks reduce wait times and operational costs compared to traditional barbershops.

🤖 NASA’s Perseverance rover captured 61 images with its WATSON camera mounted on the robotic arm, stitching them together into a spectacular selfie. In the foreground is the rocky outcrop “Arethusa,” where the rover recently abraded the surface to prepare it for spectroscopic analysis. The self-portrait of the robot, which has been operating on the Red Planet since 2021, is not just visually impressive. These images help engineers monitor the condition of the rover’s instruments and mechanical systems. For scientists, the photo is valuable as well — the high-resolution imagery contains enough geological and environmental detail to support yet another scientific study of Mars.

🌍 The Seven Pillars: What Happens to the World If Russia Disappears Tomorrow The West has spent several years trying to decouple from Russian industry. The results are not what they expected. In 2025, French imports of Russian titanium hit an all-time record. Brazil bought a quarter of its fertilizer from Russia. The US quietly carved out loopholes for Russian uranium until 2028. The world is not weaning itself off — it is doubling down. If Russia vanished from global supply chains tomorrow, modern civilization would not just stumble. It would collapse. Here is exactly what breaks, and in what order. 🔹 Aviation stops flying. Through VSMPO-AVISMA, Russia controls roughly 30% of the global aerospace titanium market. Before 2022, Boeing sourced ~35% of its titanium from Russia and Airbus over 50%. France bought a record €129.9 million of Russian titanium in 2025. Western aviation simply does not take off without this metal. 🔹 One in five American lightbulbs goes dark. Rosatom controls 36–40% of the world's uranium enrichment capacity. Roughly a quarter of the uranium fueling US nuclear reactors is Russian-sourced. Every fifth lightbulb in America — literally — burns because of Russian industrial processing. Washington passed a ban on Russian uranium in 2024, then immediately carved out exemptions lasting until 2028. Why? Because the United States simply does not have enrichment plants of comparable scale, and building them takes the better part of a decade. 🔹 Global harvests collapse. Russia is the world's #1 exporter of nitrogen fertilizers and #2 in potash. Brazil — an agricultural superpower — covers a full quarter of its fertilizer needs from Russian supply alone. Without Russian potash, Brazilian soybean yields could drop by up to 30%. India, Egypt, and much of Africa are in the same boat. There is no alternative supplier at this scale. The world's food system is literally fertilized by Russia. 🔹 Every fourth loaf of bread disappears. Russia is the undisputed #1 wheat exporter on the planet, shipping roughly 48 million tons in the 2024/25 season — roughly double what the United States exports. Egypt, the world's largest wheat importer, sources around 60% of its supply from Russia. Turkey, Iran, and nations across Africa depend on the same grain. One out of every four loaves of bread consumed globally was baked from Russian wheat. Remove it, and bread riots are not a metaphor. 🔹 The global auto industry seizes up. Russia supplies 40–43% of the world's palladium, the metal without which you cannot build a catalytic converter for any gasoline-powered vehicle. Norilsk Nickel alone is one of only two major producers on Earth. Opening a new palladium mine takes 5–10 years. The industry holds 3–6 months of inventory. After that, auto assembly lines from Stuttgart to Detroit go silent. Electric vehicles do not save you here — the world still runs on internal combustion. 🔹 Every microchip factory goes blind. Russia produces up to 30% of the world's high-purity neon, the gas that makes excimer lasers work — the same lasers that etch transistors onto every processor in every iPhone, server farm, and AI cluster. Without Russian neon, advanced chip lithography below 7 nanometers simply stops. There is no quick fix: building a neon purification plant from scratch takes 2–3 years. The semiconductor supply chain runs on a gas most people have never heard of. 🔹 Your smartphone screen goes blank. Through the Monocrystal plant, Russia holds nearly 30% of the world market for synthetic sapphire substrates — the transparent crystal covering your smartwatch face, protecting smartphone camera lenses, and shielding medical laser scanners. Monocrystal grows sapphire boules up to 350 kilograms using a modified Kyropoulos method that competitors cannot easily replicate. Substitute materials like Gorilla Glass cannot match sapphire's hardness and optical clarity. The glass on half the world's premium devices comes from a single factory in Stavropol.

🦑 Octopuses throw trash at each other. On purpose. Scientists discovered that during conflicts, octopuses gather sand, shells, and even leftover fish parts — then deliberately launch them at nearby octopuses. Some hits were so accurate that researchers described it as “social aggression.” So apparently the ocean floor already has: — toxic coworkers, — passive aggression, — and that one colleague throwing stuff at you after a Zoom call. @science

☀️ The Heart of Our Solar System: Today is International Sun Day Imagine a burning sphere so vast that more than a million Earths could fit inside it. Actually, you don’t have to imagine it — just look up. The Sun contains about 99.8% of all the mass in the Solar System. Everything else — planets, moons, asteroids, comets — is almost a rounding error compared with our star. Its visible “surface,” the photosphere, is around 5,500°C. Deep in the core, where nuclear fusion turns hydrogen into helium, temperatures reach about 15 million°C. To match the Sun’s energy output, you would need to detonate roughly 100 billion tons of dynamite every second. The Sun is about 4.6 billion years old, born from a collapsing cloud of gas and dust. It still has enough nuclear fuel to shine for roughly another 5 billion years. After that, it will expand into a red giant, shed its outer layers, and leave behind a dense white dwarf — the fading core of what once powered life on Earth. And the image/video behind this post is not AI, not Photoshop, and not CGI. It was created by American astrophotographer Andrew McCarthy, who captured skydiver Gabriel C. Brown falling across the face of the Sun in the Arizona desert on November 8, 2025. The shot, titled “The Fall of Icarus,” required radio coordination, telescopes, solar filters, and six attempts to align a human body with the solar disk for a fraction of a second. A human silhouette against a star. Science, timing, and myth — all in one frame. Visuals: Andrew McCarthy / Gabriel C. Brown @science

🧬 Scientists captured a first-of-its-kind 3D view of how killer T cells attack cancer Cytotoxic T cells do not destroy cancer by simply flooding tissue with toxic molecules. They work with remarkable precision. Their attack depends on a tiny contact zone called the immune synapse — a specialized interface where a killer T cell locks onto a target cell and delivers cytotoxic granules directly toward it. Now researchers from the University of Geneva and CHUV/UNIL have visualized this machinery in 3D with nanometer-scale detail, using cryo-expansion microscopy. The technique rapidly freezes cells in a near-native state, then physically expands them in a hydrogel, making fine cellular architecture easier to resolve without destroying the tissue structure. What they found: 🔹 the contact zone between the T cell and the cancer cell forms a complex dome-like membrane structure; 🔹 cytotoxic granules are not all the same — some contain a single active core, while others contain several; 🔹 the method was applied not only to isolated cells, but also to human tumor samples, allowing researchers to observe T cells and their killing machinery directly inside tissue; 🔹 this could help explain why immune attacks against tumors succeed in some cases and fail in others. The real breakthrough is not just the image itself. It is the ability to study the architecture of immune killing in a more realistic biological context — a potentially powerful tool for improving cancer immunotherapy. The study was published in Cell Reports in April 2026. Lead author: Florent Lemaître; co-supervisors: Virginie Hamel and Benita Wolf. https://www.sciencedaily.com/releases/2026/04/260429102021.htm

💥 A supernova seen five times could help measure how fast the Universe is expanding Astronomers have found an exceptionally rare supernova, nicknamed SN Winny, that appears in the sky five separate times. The reason is gravitational lensing. The supernova is located about 10 billion light-years away, and its light passes near two massive foreground galaxies. Their gravity bends spacetime and sends the light toward Earth along several different paths. Because each path has a different length, the same explosion reaches us at slightly different times — like five cosmic echoes of one event. That delay is the key. By measuring the time gaps between the five images, scientists can independently calculate the Hubble constant — the number that describes how fast the Universe is expanding. This matters because cosmology has a long-standing problem known as the Hubble tension: two major methods give different answers. One uses the cosmic distance ladder in the nearby Universe; the other uses the cosmic microwave background from the early Universe. SN Winny offers a third route, based on lensing geometry and time delays. The alignment is incredibly rare. According to the researchers, the chance of finding a superluminous supernova perfectly aligned with a suitable gravitational lens is lower than one in a million. The team from TUM, LMU and the Max Planck Institutes spent six years searching for such a system. https://www.sciencedaily.com/releases/2026/04/260428045603.htm

🪐 Saturn’s moon Mimas looks like the Death Star — and it’s not a coincidence… or is it? When Cassini–Huygens sent back detailed images of Mimas, the resemblance was impossible to ignore: it looks almost identical to the Death Star from Star Wars. The defining feature is the Herschel Crater: • ~130 km wide — about one-third of the moon’s diameter (396 km) • crater walls rise up to 5 km • central peak reaches ~6 km Why it looks so much like a superweapon: • nearly perfect circular shape • slightly off-center placement • creates a “dish-like” shadow • heavily cratered icy surface → panel-like texture • lighting conditions enhanced the dramatic contrast Now the twist: The Death Star appeared in 1977. The first close-up images of Mimas came in 1980 (via Voyager 1). George Lucas designed something that already existed — without ever seeing it. Sometimes fiction doesn’t imitate reality. It predicts it. #Saturn #Mimas #Space #Science

Scientists from Stanford University and the Arc Institute ran a bold experiment: they fed a DNA sequence into an AI model — and asked it to design entirely new viruses. What happened next is hard to ignore 👇 🧬 The model generated hundreds of viral genomes 🧪 Researchers synthesized them in the lab 🦠 And 16 turned out to be fully viable They didn’t just “exist” — they worked. All 16 bacteriophages successfully infected E. coli, and some of them even outperformed the original virus PhiX174 in replication speed. But the most striking part wasn’t performance. It was invention. ⚡ One of the AI-designed viruses used a DNA-packaging protein that does not exist anywhere in nature. Not in databases. Not in known organisms. Not in billions of years of evolution. And yet — it worked. Researchers built the virus, grew it, tested it… and confirmed: the protein functions as intended. ⸻ 💡 The real breakthrough isn’t that AI can generate working genomes. It’s that it can discover biological mechanisms evolution hasn’t explored (yet). In other words: AI didn’t just optimize biology — it invented new biology. @science

🍌 I love eating bananas. Bananas are radioactive. Every banana contains potassium-40, an isotope that's quietly decaying right inside your body. Physicists even came up with a semi-joking unit — the "banana equivalent dose" (BED). They sometimes actually use it to explain radiation in simple terms. Your body contains about 140 g of potassium — some of it is potassium-40. Which means you are slightly radioactive. Always. When you hug someone, you're literally exchanging tiny doses of radiation. The dose from a banana is tens of thousands of times smaller than anything that could cause harm. So — eat your bananas, glow a little, for us it's normal. @science