Science in telegram
Science that matters: AI, space, biotech, physics, future tech — explained sharply
Больше📈 Аналитический обзор Telegram-канала Science in telegram
Канал Science in telegram (@science) языкового сегмента Английский является активным участником. Сейчас сообщество объединяет 120 560 подписчиков, занимая 103 место в категории Факты и 177 место в регионе США.
📊 Показатели аудитории и динамика
С момента создания невідомо проект демонстрирует стремительный рост, собрав аудиторию из 120 560 подписчиков.
Согласно последним данным от 24 июля, 2026, канал показывает стабильную активность. За последние 30 дней изменение числа участников составило -400, а за последние 24 часа — -28, при этом общий охват остаётся высоким.
- Статус верификации: Не верифицирован
- Уровень вовлечённости (ER): Средний показатель вовлечённости аудитории составляет 5.63%. В первые 24 часа после публикации контент обычно набирает 2.31% реакций от общего числа подписчиков.
- Охват публикаций: В среднем каждый пост получает 6 793 просмотров. В течение первых суток публикация набирает 2 790 просмотров.
- Реакции и взаимодействия: Аудитория активно поддерживает контент: среднее количество реакций на один пост — 48.
- Тематические интересы: Контент сосредоточен на ключевых темах, таких как medicine, cell, researcher, scientist, u.s.
📝 Описание и контентная политика
Автор описывает ресурс как площадку для выражения субъективного мнения:
“Science that matters: AI, space, biotech, physics, future tech — explained sharply”
Благодаря высокой частоте обновлений (последние данные получены 25 июля, 2026) канал поддерживает актуальность и высокий уровень охвата публикаций. Аналитика показывает, что аудитория активно взаимодействует с контентом, что делает его важной точкой влияния в категории Факты.
Загрузка данных...
| Дата | Привлечение подписчиков | Упоминания | Каналы | |
| 25 июля | +1 | |||
| 24 июля | +15 | |||
| 23 июля | +6 | |||
| 22 июля | +19 | |||
| 21 июля | +22 | |||
| 20 июля | +7 | |||
| 19 июля | +53 | |||
| 18 июля | +16 | |||
| 17 июля | +18 | |||
| 16 июля | +22 | |||
| 15 июля | +19 | |||
| 14 июля | +70 | |||
| 13 июля | +45 | |||
| 12 июля | +3 | |||
| 11 июля | +15 | |||
| 10 июля | +24 | |||
| 09 июля | +12 | |||
| 08 июля | 0 | |||
| 07 июля | +7 | |||
| 06 июля | +1 | |||
| 05 июля | +24 | |||
| 04 июля | +14 | |||
| 03 июля | +3 | |||
| 02 июля | +21 | |||
| 01 июля | +11 |
| 2 | 🧬 AI Discovers a Natural Peptide That Could Rival Ozempic
Researchers at Stanford Medicine have used artificial intelligence to identify a naturally occurring human peptide that suppresses appetite and promotes weight loss in mice—while avoiding the major side effects commonly associated with GLP-1 drugs like Ozempic.
The newly discovered molecule, called BRP, is only 12 amino acids long. Yet in early laboratory experiments, it activated appetite-regulating neurons far more strongly than existing GLP-1 peptides.
Unlike semaglutide, which acts on receptors throughout the brain, gut, pancreas, and other tissues, BRP appears to target primarily the hypothalamus—the brain region responsible for regulating hunger and metabolism. That more focused mechanism may explain why researchers observed no signs of nausea, constipation, anxiety, or muscle loss in the animal studies.
The discovery itself was powered by AI. Stanford’s Peptide Predictor analyzed more than 20,000 human protein-coding genes, searching for hidden biologically active peptides. From 2,683 candidates, researchers narrowed the list to just 100 for laboratory testing—and BRP emerged as the standout.
In obese mice, a single injection reduced food intake by up to 50% within one hour. After two weeks of treatment, the animals lost significant body fat while untreated mice continued gaining weight. The peptide also improved glucose tolerance and insulin sensitivity.
“Nothing we’ve tested before has compared to semaglutide’s ability to decrease appetite and body weight. We are very eager to learn if it is safe and effective in humans.” — Dr. Katrin Svensson
The findings are still preclinical, meaning BRP has not yet been tested in humans. But if future clinical trials confirm these results, it could represent a new generation of obesity treatments: delivering GLP-1-like benefits through a much more targeted biological pathway—with fewer unwanted side effects.
📄 Nature (2025)
#AI #Obesity #WeightLoss #Ozempic #Neuroscience #StanfordMedicine #science | 1 506 |
| 3 | 🦷 Scientists May Have Found a Way to Regrow Tooth Enamel
For decades, dentists have faced a frustrating reality: once tooth enamel is lost, it never naturally grows back. Researchers at the University of Nottingham may have found a way to change that.
In a study published in Nature Communications, the team developed a protein-based biomaterial that mimics the way enamel forms during early childhood—the only time our bodies naturally produce it.
Instead of simply coating the tooth like conventional treatments, the material acts as a microscopic scaffold. It penetrates tiny defects in damaged enamel, then recruits calcium and phosphate ions from saliva, guiding them to grow new enamel crystals that integrate seamlessly with the existing tooth.
In laboratory tests on extracted human teeth, the regenerated enamel closely matched the natural material in both structure and mechanical performance. It also withstood simulated brushing, chewing, and repeated exposure to acidic conditions, suggesting it could be durable enough for everyday use.
Unlike today’s fluoride treatments—which mainly slow further damage—this approach is designed to restore enamel rather than simply protect what’s left.
The researchers have launched a startup, Mintech-Bio, to commercialize the technology and hope to bring the first products toward clinical use in the coming year. However, human clinical trials are still needed before the treatment can become widely available.
If future trials confirm these results, dentistry could gradually shift from “drill and fill” to actually rebuilding damaged teeth.
📖 Source: Nature Communications (2025)
https://www.nature.com/articles/s41467-025-64982-y
#Dentistry #Biomaterials #Enamel #RegenerativeMedicine #Science | 3 388 |
| 4 | 🌿 Ordinary Moss Produces Electrical Waves That Travel Across Its Tissue
Moss may look like one of nature’s quieter creations. Electrically, however, it appears to be surprisingly active.
Computer scientist Andy Adamatzky inserted electrodes into cushions of the common moss Brachythecium rutabulum and recorded their electrical activity continuously for 178 hours.
The moss produced several distinct patterns: rapid spikes, slower rhythmic oscillations, and very slow depolarization waves. Some signals appeared first at one electrode and later at another, suggesting that electrical activity was travelling across the tissue rather than occurring everywhere simultaneously.
Some spikes resembled action potentials and neuron-like spike trains. But resemblance is not equivalence: moss has no neurons, no brain, and the study provides no evidence that it thinks or possesses anything resembling consciousness.
What it may show is that a moss cushion behaves as a distributed, electrically connected system. One day, such living networks could potentially be used in biohybrid sensors, responsive building materials, or unconventional computing.
The interpretation remains preliminary. For now, the moss is not solving equations — but it may be coordinating electrical activity beneath our feet.
Could future computers be partly grown rather than manufactured?
📄 Royal Society Open Science
DOI: 10.1098/rsos.252341
#PlantBiology #Moss #BioComputing #LivingMaterials #UnconventionalComputing | 4 362 |
| 5 | 💻 Classical Computers Just Overturned a Major Quantum Advantage Claim
In 2025, researchers used D-Wave’s Advantage2 quantum annealer to simulate the dynamics of hundreds of interacting qubits — and argued that the task was beyond the reach of classical computers.
A team from the Flatiron Institute and Boston University has now reproduced those results using classical algorithms and relatively modest hardware. Some of the initial calculations even ran on a personal laptop.
The challenge involved quantum spin glasses: disordered systems in which hundreds of interacting qubits evolve across two- and three-dimensional lattices. Tracking the complete quantum wave function directly would require an amount of memory that grows exponentially with the number of qubits.
The researchers avoided that explosion using tensor networks — mathematical structures that compress a quantum state by retaining its most important correlations.
Think of it as a highly specialized zip file for quantum information.
They combined tensor networks with belief propagation, an algorithm developed in the 1980s and recently adapted for quantum systems. The calculations were implemented using ITensor, a high-performance tensor-network software library.
The classical simulations:
• reproduced the quantum computer’s results
• matched theoretical predictions and exact benchmarks
• scaled to systems containing hundreds of qubits
• captured quantum dynamics in complex 3D lattice geometries
There is one important caveat: the entire study was not completed on an ordinary laptop. A laptop was sufficient for some early calculations, while the largest simulations required more powerful conventional processors and GPUs.
Still, the result directly challenges the earlier claim that this particular problem was beyond classical computation.
It does not mean quantum computers are useless. It means that demonstrating genuine quantum advantage requires beating the best classical algorithms available — not merely the best ones researchers happened to test previously.
The boundary between classical and quantum computing is not fixed. Every improvement in quantum hardware gives classical researchers a new target, while every better classical algorithm raises the bar for quantum advantage.
Sometimes the strongest competitor to a quantum computer is not another quantum computer.
It is better mathematics.
📄 Paper: Dynamics of disordered quantum systems with two- and three-dimensional tensor networks
https://www.science.org/doi/10.1126/science.adx2728
#QuantumComputing #Physics #TensorNetworks #ClassicalComputing #QuantumPhysics | 5 099 |
| 6 | Нет текста... | 5 205 |
| 7 | Нет текста... | 4 406 |
| 8 | Нет текста... | 5 892 |
| 9 | Нет текста... | 5 609 |
| 10 | Нет текста... | 7 082 |
| 11 | 🧬 Semaglutide May Slow Biological Aging, First Randomized Human Evidence Suggests
GLP-1 drugs such as semaglutide are already well known for treating obesity and type 2 diabetes. Now, researchers at UC San Diego report the first randomized placebo-controlled evidence that the drug may also slow biological aging — at least as measured by epigenetic aging clocks.
The team analyzed DNA methylation data from a 32-week, double-blind clinical trial involving 108 adults with HIV-associated lipohypertrophy. Participants received either weekly semaglutide injections or placebo. Researchers then used multiple epigenetic clocks — molecular biomarkers that estimate biological aging from chemical modifications to DNA — to evaluate changes in aging rate.
The results, published in Nature Communications, showed that semaglutide significantly slowed several independent measures of biological aging. On the DunedinPACE clock, the pace of aging slowed by about 9%. Significant improvements were also observed with the PCGrimAge clock, which is associated with mortality and age-related disease risk.
A separate 24-week pilot study in people with HIV and fatty liver disease reported that about 42% of participants showed a reduction in DunedinPACE after semaglutide treatment, although that study had no placebo control and should be interpreted cautiously.
In simple terms: epigenetic clocks don’t measure your chronological age. They estimate how quickly the molecular processes associated with aging are progressing. A 9% reduction means these biomarkers changed more slowly during treatment — not that people became 9% younger.
Scientists think several mechanisms may contribute. Semaglutide reduces chronic inflammation, decreases harmful visceral fat, and improves metabolic health — all processes linked to accelerated aging. These changes may influence molecular pathways involved in aging across multiple organ systems.
Key findings:
• ~9% slower pace of aging on the DunedinPACE clock
• Significant improvement in the PCGrimAge mortality-risk clock
• Similar effects across several independent epigenetic aging clocks
• A separate pilot study found ~42% of participants showed slower epigenetic aging after treatment
The study also has important limitations. This was a post hoc exploratory analysis, not a trial originally designed to study aging. Participants all had HIV-associated lipohypertrophy, a condition associated with accelerated biological aging, so the findings cannot yet be generalized to healthy individuals. Most importantly, improvements in epigenetic clocks do not prove that semaglutide extends lifespan or healthspan. Those questions will require much longer clinical studies.
Why it matters: Tens of millions of people already take GLP-1 drugs worldwide. If future studies confirm these findings in broader populations, semaglutide could become one of the first widely used medications shown to influence molecular biomarkers of human aging — extending its impact far beyond weight loss.
📄 https://www.nature.com/articles/s41467-026-72861-3
#Longevity #Semaglutide #GLP1 #Aging #Healthspan | 8 059 |
| 12 | 🤖 OpenAI Just Released Three Versions of GPT-5.6 — and the Most Important One May Not Be the Flagship
On July 9, OpenAI released GPT-5.6 as a family of three models rather than a single flagship:
• Sol — the most capable version, priced at $5 per million input tokens and $30 per million output tokens
• Terra — a balanced model with performance competitive with GPT-5.5 at roughly half the price: $2.50/$15
• Luna — the fastest and cheapest version, priced at $1/$6
Sol is designed for difficult coding, scientific research, cybersecurity, and long-running agentic tasks. It also introduces an ultra mode that can coordinate parallel subagents to solve complex problems.
But Terra may be the more consequential release.
If it consistently delivers something close to GPT-5.5 performance at half the cost, many companies may have little reason to use the flagship for everyday workloads. Sol wins benchmarks and headlines; Terra could win production volume.
OpenAI also launched ChatGPT Work, a workspace that can gather context from connected tools and turn it into finished spreadsheets, documents, presentations, analyses, and other deliverables.
This suggests that the AI race is moving beyond model intelligence alone.
Nikolas Bush Take
1. Model portfolios are replacing the idea of one universal model
OpenAI is beginning to resemble a cloud provider: instead of asking which model is “best,” customers choose the right combination of intelligence, latency, and price.
That makes the middle tier strategically important. Terra could become the default model for businesses, while Sol remains the premium option for the hardest tasks.
2. Government involvement in frontier-model launches is becoming harder to ignore
OpenAI initially released GPT-5.6 through a limited preview after discussing the models’ capabilities with the US government. The company said trusted partners were selected with government involvement before broader availability.
Anthropic faced an even more direct intervention: access to Claude Fable 5 and Mythos 5 was suspended after US export controls were imposed, before Fable 5 returned globally on July 1 with stronger safeguards.
This is not yet a formal regulatory system. But a de facto pre-release review process for highly capable AI models may be emerging through government pressure, security testing, and access restrictions.
3. The real competition is shifting toward the workspace
When several frontier models reach similar levels of capability, the surrounding product becomes more important.
The winner may not be the company with the highest benchmark score. It may be the company that owns the environment where people research, write, analyze data, build presentations, manage projects, and automate everyday work.
Models can increasingly be replaced. Workflows, integrations, organizational data, and user habits are much harder to displace.
Independent testing of GPT-5.6 is still limited, and many performance claims come from OpenAI’s own evaluations. Terra’s price-performance advantage therefore needs to be validated across real business workloads.
But the broader direction is already visible:
The era of one flagship model for every task may be ending. The next phase of AI will be fought through portfolios, workflows, and distribution.
Sources:
https://openai.com/index/gpt-5-6/
https://openai.com/index/previewing-gpt-5-6-sol/
https://openai.com/chatgpt-work/
https://www.anthropic.com/news/redeploying-fable-5
#AI #OpenAI #GPT56 #ChatGPT #ArtificialIntelligence #NikolasBushTake | 7 616 |
| 13 | 🔭 Physicists Recreated a Black-Hole-Inspired Effect — in a Circuit With No Moving Parts
More than 50 years ago, Roger Penrose proposed that rotational energy could be extracted from a spinning black hole. Yakov Zel’dovich later predicted a wave version of the effect: an object rotating fast enough could transfer some of its energy to incoming waves, amplifying them.
The problem was speed. Reproducing the effect with electromagnetic waves would require rotation far beyond the limits of ordinary machinery.
Researchers at the CUNY Advanced Science Research Center have now bypassed that obstacle with a stationary radio-frequency circuit. It contains three coupled electronic resonators whose properties are modulated in a precisely timed sequence. Nothing physically rotates, but the traveling modulation pattern acts like ultrafast “synthetic rotation.”
Waves carrying the appropriate orbital angular momentum emerged amplified, with a reported net gain of up to 7.8 decibels. The experiment, published in Nature, demonstrates what the researchers call Floquet rotational superradiance.
There is no free energy involved. The additional energy comes from the external drive used to modulate the circuit — not from an actual black hole. The researchers also stress that this is not a one-to-one replica of Zel’dovich’s original rotating-object proposal, but a related effect combining rotational Doppler physics with parametric amplification.
Counterintuitively, ordinary losses in the circuit helped broaden the conditions under which amplification occurred.
The platform could provide a new way to study extreme rotational wave physics and may eventually inspire selective amplifiers, photonic devices and communication technologies.
What other seemingly impossible physical environments could be recreated using synthetic motion?
📄 Paper: https://www.nature.com/articles/s41586-026-10725-y
#Physics #BlackHoles #Superradiance #Metamaterials | 6 087 |
| 14 | 🌌 We @science May Have Finally Figured Out Why the Universe Is Expanding…
The explanation is surprisingly simple.
Large language models operate using tokens. Every time a model generates a new response, it processes the accumulated context of the conversation.
As the dialogue grows, each new message adds more information — while the model must repeatedly process everything that came before. The total amount of processed information therefore grows almost geometrically within every conversation.
Now imagine that the Universe is actually one enormous artificial intelligence.
Every new star, planet, living organism, human thought, cat video, and online argument adds more tokens to the global conversation. The context keeps growing in geometric progression, so the system needs more and more space to process it all.
That is why the Universe is expanding.
Dark energy may have nothing to do with it. Someone simply forgot to click “Start a new chat.” 😅
#science #space #universe #AI | 6 654 |
| 15 | 🌊 Scientists Find Why Some Animals Survived Earth’s Worst Mass Extinction
About 252 million years ago, the Permian–Triassic extinction erased most marine species and permanently reshaped life in the oceans. A new Stanford-led study suggests that survival depended partly on how animals’ oxygen requirements changed as the water warmed.
Researchers measured oxygen consumption in living brachiopods and other animals representing the “Paleozoic” and modern marine faunas. At cooler temperatures, brachiopods could tolerate remarkably low oxygen levels. But as the water warmed, the minimum amount of oxygen they needed rose much faster.
Their slow metabolism was not the problem by itself. More active groups generally had muscles, gills and respiratory systems better able to supply additional oxygen under heat stress.
This may explain why brachiopods and crinoids were devastated, while bivalves, snails, fish and echinoderms suffered fewer losses and later came to dominate the oceans.
The extinction was therefore not completely random — but metabolism was not the only factor either. Warming and ocean deoxygenation acted through animals’ physiology, alongside stresses such as acidification.
Modern oceans are also warming and losing oxygen. The study does not predict another “Great Dying,” but it shows how strongly climate can favour some body plans over others.
Could the next reshuffling of ocean life be determined by which animals can keep breathing as the water heats up?
📄 Study: https://www.pnas.org/doi/10.1073/pnas.2533086123
#Paleontology #MassExtinction #MarineBiology #OceanWarming #ClimateChange | 6 296 |
| 16 | 🤖 NEO Just Got New Hands
25 degrees of freedom.
Still far from matching the human hand, of course — but the progress is impressive.
The company calls them an “API to the physical world.”
https://www.1x.tech/discover/neos-hands
#Robotics #NEO #RobotHands #gadget #science | 8 269 |
| 17 | 🐸 A Frog Gut Bacterium Eliminated Colorectal Tumors in Mice
A bacterium found in the gut of Japanese tree frogs has shown a striking anti-cancer effect in mice.
Researchers at the Japan Advanced Institute of Science and Technology (JAIST) screened 45 bacterial strains isolated from amphibians and reptiles, including Japanese tree frogs, fire belly newts, and grass lizards. One strain stood out: Ewingella americana, a naturally occurring bacterium from the intestines of the Japanese tree frog.
In a mouse model of colorectal cancer, a single intravenous dose of E. americana led to complete tumor elimination in all treated mice — a 100% complete response rate in that experiment.
The bacterium appears to work in two ways.
First, it accumulates inside tumors, where oxygen levels are low and blood vessels are unusually leaky. Once there, the bacterial population increased by roughly 3,000 times within 24 hours, directly damaging cancer cells.
Second, it seems to wake up the immune system. The treatment attracted T cells, B cells, and neutrophils into the tumor and boosted inflammatory signals such as TNF-α and IFN-γ, helping the body attack the cancer more aggressively.
The safety data in mice were also encouraging. The bacteria disappeared from the bloodstream within 24 hours, did not colonize healthy organs such as the liver, spleen, lungs, kidneys, or heart, and caused only temporary mild inflammation. Over a 60-day observation period, the researchers reported no chronic toxicity.
But the caveat is essential: this is still mouse research.
Many cancer therapies look spectacular in preclinical models and fail later in humans. A living bacterium injected into the bloodstream would face a much higher safety bar in people, and the human immune response could be very different.
Still, the study is an unusually strong proof of concept for bacterial cancer therapy. Instead of simply changing the gut microbiome, the researchers used a living bacterium as a tumor-targeting agent — one that both attacks cancer cells and recruits the immune system.
The big takeaway: one of the next ideas in cancer therapy may come not from a synthetic drug library, but from the gut of a frog.
Original paper, Gut Microbes:
https://www.tandfonline.com/doi/full/10.1080/19490976.2025.2599562
DOI:
https://doi.org/10.1080/19490976.2025.2599562
ScienceDaily summary:
https://www.sciencedaily.com/releases/2026/07/260709160655.htm
JAIST press release:
https://www.jaist.ac.jp/english/whatsnew/press/2025/12/17-1.html
#CancerResearch #Microbiome #Immunotherapy #Biotech | 7 686 |
| 18 | 🧬 Harvard Built a Silicon Chip That Can Write DNA Using Electricity and Water
Writing custom DNA has barely changed in decades. Today’s DNA synthesis relies on phosphoramidite chemistry—a highly effective but solvent-intensive process that depends on hazardous chemicals and specialized manufacturing facilities.
Researchers at Harvard have now demonstrated a completely different approach: parallel enzymatic DNA synthesis on a semiconductor chip, controlled simply by electricity in water.
The CMOS chip contains 64 programmable synthesis sites, each surrounded by two concentric ring electrodes. When current flows through the inner ring, it generates protons that create a tiny acidic region—exactly what’s needed to remove the temporary blocking group from a growing DNA strand. At the same time, the outer ring consumes escaping protons, preventing neighboring reactions from interfering with one another.
Using this approach, the team synthesized 64 different DNA sequences, each 38–39 nucleotides long, entirely in an aqueous enzymatic process. Previous demonstrations of parallel enzymatic DNA synthesis were limited to roughly a dozen sequences, making this the largest demonstration of its kind so far.
The hardware has an unusual history. It was originally developed for recording electrical activity from thousands of neurons. The researchers later realized that the same ability to precisely control microscopic electrical currents could also be used to control chemical reactions needed for DNA synthesis.
Key points:
• 64 unique DNA sequences synthesized in parallel
• Water-based enzymatic process instead of traditional solvent-heavy chemistry
• Precise local pH control using dual concentric electrodes
• DNA strands up to 39 nucleotides long
• Demonstrated storage of a 169-byte text message in the synthesized DNA
The technology is still at an early stage. DNA strands of 39 nucleotides are far shorter than real genes, which typically contain thousands of bases. According to the researchers, the main limitation is now the chemistry used to remove temporary protecting groups—not the chip itself—suggesting that future advances may come from improved chemistry rather than new electronics.
If the method can be scaled, it could eventually make DNA manufacturing cleaner, cheaper, and more accessible for synthetic biology, diagnostics, gene therapies, and even DNA-based data storage, one of the highest-density storage media ever proposed.
📄 Nature Electronics: https://www.nature.com/articles/s41928-026-01662-9
🏛 Harvard SEAS: https://seas.harvard.edu/news/making-dna-semiconductor-chip
#DNA #SyntheticBiology #Biotechnology #Semiconductors #DataStorage | 7 683 |
| 19 | 🔬 Quantum Entanglement Has Been Found in a Crystal You Could Hold in Your Hand
Quantum entanglement is usually imagined as something fragile: isolated atoms, photons, superconducting circuits, vacuum chambers, and temperatures close to absolute zero.
But physicists at TU Wien and collaborators have now detected strong multipartite entanglement inside a centimeter-sized crystal — a solid object large enough to hold in your hand.
The material is Ce₃Pd₂₀Si₆, a “strange metal” made of cerium, palladium, and silicon. Strange metals are famous for refusing to behave like ordinary metals: their electrical resistance changes in an unusual way, and previous experiments suggest their current can flow with surprisingly low noise.
To look inside the crystal, the team used inelastic neutron scattering at the Institut Laue-Langevin in Grenoble. In simple terms, they fired neutrons into the material and measured how the crystal responded.
Then they analyzed the response using quantum Fisher information — a mathematical tool from quantum metrology that can reveal whether many parts of a system are acting independently or as an entangled collective.
The result: the crystal’s response could not be explained as separate particles acting alone. The data point to groups of at least nine quantum-entangled entities acting together.
That does not mean the whole crystal was placed into a Schrödinger-cat-style superposition. It also does not mean this is immediately useful for quantum computing. The measurement was indirect, based on neutron-scattering data and a specific theoretical framework.
But it is still a big deal.
It shows that entanglement can be detected inside a bulk quantum material, not just in carefully isolated microscopic systems. It also gives physicists a new way to connect two worlds that often use very different languages: quantum information science and solid-state physics.
The most important caveat: this was not an everyday-room-temperature demonstration. The experiment required extremely low temperatures and a carefully tuned magnetic field. So the “hand-sized crystal” part refers to the size of the sample — not to easy real-world operating conditions.
Still, the takeaway is fascinating: quantum weirdness may be hiding inside materials we can see and touch. We just needed the right tool to measure it.
Original paper, Nature Physics:
https://www.nature.com/articles/s41567-026-03298-0
DOI:
https://doi.org/10.1038/s41567-026-03298-0
ScienceDaily summary:
https://www.sciencedaily.com/releases/2026/07/260701015242.htm
ILL summary:
https://www.ill.eu/en/about-the-ill/news-and-events/news/quantum-entanglement-detected-inside-a-centimetre-sized-strange-metal/
#QuantumPhysics #Entanglement #StrangeMetals #MaterialsScience | 2 399 |
| 20 | 🔒 GPT-5.6 Is Here — But Most People Can’t Use It Yet
OpenAI has launched GPT-5.6 — its next-generation model family — but this is not a normal public rollout.
The new lineup comes in three tiers:
Sol — the flagship model, built for the hardest reasoning, coding, cyber and science tasks
Terra — a balanced lower-cost model, positioned as competitive with GPT-5.5 at roughly half the price
Luna — the fastest and cheapest tier for high-volume workloads
But here is the real story: during the preview, GPT-5.6 is not available in ChatGPT and is not open through public self-service access. OpenAI says Sol, Terra and Luna are available only through API and Codex to a select group of trusted partners and organizations.
Why?
Cybersecurity.
OpenAI says GPT-5.6 Sol is its most capable model yet for cybersecurity tasks, including long-horizon vulnerability research and exploitation. The company also says it previewed the model and its release plans to the U.S. government before launch, and — at the government’s request — started with a limited preview for trusted partners whose participation had been shared with the government.
That is a very different model launch from the old “release first, regulate later” era.
To be clear: OpenAI says GPT-5.6 Sol does not cross its “Cyber Critical” threshold. In tests involving Chromium and Firefox, the model identified bugs and exploitation primitives, but did not autonomously produce a full-chain exploit under the tested conditions.
Still, the direction is obvious: frontier AI is moving into dual-use territory, where the same capability can help defenders find vulnerabilities faster — or help attackers do the same.
And this is not happening in isolation.
Anthropic’s Claude Fable 5 and Mythos 5 were temporarily suspended in June after U.S. export controls, then restored after new safeguards and further government coordination. Mythos 5 remains restricted to approved organizations through Project Glasswing, a controlled-access program focused on defensive cybersecurity.
The Five Eyes cyber agencies have also warned that frontier AI models could transform offensive and defensive cyber capabilities on a timeline of “months, not years.”
Nikolas Bush Take:
The most important shift here is not the benchmark. It is access.
For years, the promise of AI was democratization: more people getting access to more intelligence at lower cost. But the frontier is starting to split into layers:
— models everyone can use
— models enterprises can buy
— models only vetted partners can access
— models governments may effectively influence before release
The cybersecurity argument is real. A system that can help find serious vulnerabilities at scale is not just another productivity tool. It is dual-use infrastructure.
But the governance question is just as real:
Who decides who is “trusted”?
Are the criteria transparent?
Can decisions be appealed?
Are restrictions temporary — or do they become the new default?
This may also accelerate the race for open-weight alternatives. Reflection AI recently locked in a compute deal worth up to $6.3 billion through 2029 to build open-weight frontier models. That looks like a direct market response to the risk of depending entirely on closed, gated systems.
But there is a darker possibility too: if open-weight models reach similar cyber capability, governments may move to restrict them as well.
The GPT-5.6 launch is not just another AI product update.
It may be an early signal of a new frontier AI regime: more powerful, more useful, more regulated — and much less predictable.
Sources:
https://openai.com/index/previewing-gpt-5-6-sol/
https://help.openai.com/en/articles/20001325-a-preview-of-gpt-56-sol-terra-and-luna
https://www.anthropic.com/news/redeploying-fable-5
https://www.anthropic.com/glasswing
https://www.cyber.gov.au/about-us/view-all-content/news/five-eyes-cyber-security-agencies-statement
https://techcrunch.com/2026/06/22/spacex-inks-compute-deal-with-reflection-ai-an-open-source-ai-lab/ | 4 250 |
