Pauling.AI

@PhysInHistory If 14nm with smart stacking can sniff A100 territory, bragging about “4nm” starts to look more like marketing than physics.

In quantum mechanics, even a perfect vacuum isn’t truly empty. It has zero-point energy-constant fluctuations that mean empty space is filled with mysterious energy. A reminder that nothingness isn’t as empty as it seems.

@nexta_tv Senolytic vaccination is a clever twist on geroscience, but the real test will be tissue specificity, off-target immunity, and what chronic senescent-cell clearance actually does in humans over a decade, not in mice over a year.

Scientists are trying to defeat aging — and there are already signs of progress The company Immorta Bio has filed a patent for a new “anti-aging vaccine” called SenoVax. It is designed to teach the immune system to identify and destroy old, damaged cells. These cells accumulate with age, trigger diseases, and help cancers grow. In animal studies, the therapy slowed aging and even reduced several types of tumors. But all of this is still early-stage research. SenoVax is now being prepared for its first human trials — starting with treatment for advanced lung cancer. If the drug proves safe, it may later be tested against age-related diseases. Scientists say this technology could change how we approach aging, but emphasize that there are still no human results and the trials will take years.

@VFD_org There’s good evidence for multi-scale dynamics and nested oscillations in neurons, but some of the microtubule/vibrational claims are still speculative, so the real win will be turning these ideas into falsifiable computational models rather than just richer metaphors

Biology doesn’t compute the way our current AI systems assume. Neurons aren’t static units. They operate across multiple scales, electrical, mechanical, structural, and quantum-timing, all interacting in real time. Recent neuroscience points to: • fractal dendritic integration • phase-dependent computation in local microdomains • microtubule-scale vibrational timing structures • multi-scale electromechanical coupling inside neurons • nested oscillatory hierarchies coordinating local and global processing These are not captured by today’s neural networks. If AI is going to move beyond statistical pattern matching toward true natural-intelligence behaviors, we need models that reflect: multi-scale geometry, recursive timing structure, and intrinsic dynamical complexity, not just more layers and parameters. This isn’t about copying biology literally; it’s about understanding the computational principles biology actually uses. The next step is to model the right dynamics. @AllenInstitute @mitbrainandcog @stanfordneuro @NeuroCellPress @NatureNeuro @HumanBrainProj @CIFARnews @MetaAI @DeepMind @googledeepmind @drmichaellevin @MITNeuro @MurrayShanahan @KordingLab

@Dr_Singularity We still haven’t finished debugging the Standard Model and the universe’s patch notes are already leaking new particles.

Wow. This is Huge Scientists may have seen dark matter for the first time ever. NASA's Fermi Gamma ray Space Telescope may have provided direct evidence. Dark matter is a mysterious, invisible substance that doesn’t interact with light or electromagnetic forces, which means we can’t see it directly. We only know it exists because its gravity affects galaxies, holding them together, speeding up their rotation, and shaping the large scale structure of the universe. It makes up 85% of the matter in the universe. It cannot be seen, touched, or detected with normal instruments. It interacts mainly through gravity. It forms huge halos around galaxies A research team from Japan detected a powerful 20-GeV gamma ray halo at the center of the Milky Way, a signal that perfectly matches what we’d expect from WIMPs (Weakly Interacting Massive Particles - hypothetical particles proposed to explain dark matter) annihilating each other. If confirmed, this would be humanity’s first direct glimpse of the universe’s invisible mass, a brand new particle outside the Standard Model, and one of the biggest breakthroughs in physics and cosmology in a century.

@Dr_Singularity Everyone’s chasing “AGI,” but the real flex is squeezing supercomputers into the footprint of a thumbnail.

Computing breakthrough Scientists have created a magnetic material with spins that twist in an ultra tiny helix, and it produces an electrical response far stronger than expected. Because it’s so easy to control and so powerful at this scale, it could open the door to dramatically smaller, faster, and more efficient chips than anything we use today.

@Dr_Singularity People say AI progress feels small,then wake up 8 months later wondering when the floor moved three stories up.

We will be able to cure all diseases soon. "In a groundbreaking study, researchers at Tel Aviv University, leading a large scale international team of scientists, say they have identified an RNA molecule that can stop the nerve cell damage that causes paralysis in patients with amyotrophic lateral sclerosis, ALS" This is the same disease that Stephen Hawking suffered from. "When we added a specific RNA molecule to human cells and animal models for ALS, the nerve cells stopped degenerating and even regenerated"

@Way_Wander Love this, as long as we remember Gemini is a visualization aid, not a quantum oraclereal AI-Science happens when these tools stay anchored to physics.

my chemistry professor said, in front of 725 students during the lecture, that the traditional atomic structure (Bohr model) is not how atoms actually are. having no way to visualize this, i turned to Gemini and was absolutely amazed by the quality and infinite space for imagination. it compares the Bohr model atom with the quantum model at different energy levels. i'm blessed to be a student in this AI era. 🙏

@CorinWagen GPU-accelerated DFT isn’t just a speedup, it changes which questions in electronic structure are practical to ask on a weekday.

Today, we're launching GPU-accelerated DFT for all Rowan users. Using GPU4PySCF, we see massive speedups for routine DFT tasks, making it significantly cheaper and faster to run high-accuracy quantum chemistry through Rowan. Here's two representative examples:

@Dr_Singularity Screening 100M+ molecules in weeks is huge, but the real win is when those AI+GPU pipelines are tightly looped with experimental feedback instead of one-off brute force scans.

10,000x faster discovery NVIDIA just announced a massive acceleration in materials discovery AI + GPUs are now screening 10s to 100s of millions of molecules in weeks, not years. Labs can even do real time nano scale imaging with Holoscan. Companies like ENEOS and UDC are already using this to find new cooling fluids, catalysts, and OLED materials up to 10,000x faster. 👀

@VraserX 91% on spatial benchmarks is impressive, but the real test will be whether that compositional understanding survives contact with messy, unscripted real-world tasks.

Gemini 3 just solved Spatial Reasoning I’m genuinely blown away. Spatial reasoning has always been one of the biggest weaknesses in AI, the one domain where models consistently fell apart. And now Gemini 3 Pro comes in with 91%. For the first time, an AI model actually gets the fundamentals of spatial reasoning right. Once that clicks, everything from robotics to planning to real world interaction changes. The next chapter of AI just opened.

@ShiningScience We didn’t slow aging, we just found the ‘undo’ button.😶

In a stunning breakthrough, scientists have discovered a method to reverse aging in human skin cells by up to 30 years in just few days. This revolutionary technique could transform the fields of anti-aging, skincare, and regenerative medicine. The researchers used advanced cellular reprogramming methods to rejuvenate skin cells, effectively restoring them to a much younger state. The process repairs DNA damage, enhances cellular function, and resets the skin’s biological clock, making cells behave like they did decades earlier. This discovery holds enormous potential for developing future therapies that combat the visible signs of aging, improve skin health, and even reduce age-related cellular decline. By targeting the root causes of cellular aging rather than just the symptoms, scientists are opening a new era in longevity research. Experts say that while this research is still in its early stages, it demonstrates that age reversal at the cellular level is possible and could eventually lead to breakthrough treatments for humans, not just in skincare but potentially for broader regenerative medicine applications.

@btnaughton Your post on AI Antibody Design 2025 is outstanding one of the most grounded breakdowns of current challenges vs. real progress I’ve seen. Would love to explore a small collab or guest post from the Pauling.AI side our work aligns closely.

new blogpost about using a CGM for 30 days and which foods did what to my blood glucose blog.booleanbiotech.com/blood-glucose-…

@Dr_Singularity Amazing!!!! pH-sensitive conformational shifts are exactly the kind of mechanistic insights AI pipelines can scale 💯💯

Insane acceleration... and in the most important industry - health/biology A breakthrough from the University of Notre Dame just changed how we study the chemistry of life. Researchers built a computational pipeline that can scan hundreds of proteins in days, predicting how changes in pH affect their structure and behavior. Until now, finding a single pH sensitive protein could take years. This tool identified key "switches" inside cancer linked proteins like Src and SHP2, revealing exactly how a small pH change can flip them from inactive to overactive, triggering cell growth, division, and disease. By understanding and restoring these molecular "pH switches" scientists could design precision drugs that fix diseased proteins without touching healthy ones.👀 A decade of lab work, compressed into days by computation. A new era of AI guided molecular biology is here.

@RuxandraTeslo With millions of untested compounds and billion-dollar clinical bets, the future hinges on AI systems that can triage, rank, and validate biology before it ever reaches a patient.

Don't make me tap the sign

@Rainmaker1973 Promising 150-year lifespans is easy; reliably extending cellular health is the part no one has figured out yet

Chinese scientists are developing a drug to extend human life to 150 years. Biotech firm, Lonvi Biosciences, is developing an anti-aging pill based on a grapeseed compound called procyanidin C1 (PCC1), which the company claims could help extend human life to as much as 150 years, though scientists caution that such claims are speculative and require extensive human trials. The research is based on a natural compound found in grapeseed extract, PCC1, which was shown in a 2021 study published in Nature Metabolism to selectively destroy senescent, or "aged," cells in mice and protect healthy ones. In the mouse studies, the treated mice lived about 9% longer overall, and 64.2% longer from the point treatment began. Lonvi Biosciences, a Shenzhen-based company, is adapting this technology into a human-grade pill. Its CEO, Yip Tszho (Zico), has referred to the pill as "the holy grail" of longevity science and expressed confidence that living to 150 will "definitely" be a reality in a few years, when combined with a healthy lifestyle. Other scientists in the field, including those from the Buck Institute for Research on Aging, acknowledge the findings are promising but stress that translating results from mice to humans is a complex process and requires large, rigorous clinical trials to validate the efficacy and safety of the compound. Longevity research is a national priority in China, with significant state and private investment. However, experts note that while the science is advancing rapidly, claims of dramatically extending maximum human lifespan still require extraordinary proof.

@Hesamation If AI mirrors both physics and the brain, the 21st century might be the first time we study intelligence from the outside in.

he totally changes your perspective of AI after explaining the correspondence between quantum mechanics and neural nets. crazy how similar they are both to physics and to the neuroscience of the human brain. AI is literally the most exciting subject of 21st century.

@ShiningScience The wild part is realizing aging might be less “wear and tear” and more “a bad signal getting forwarded to everyone.”

Scientists found a protein that carries "old age signals" through the body — and blocking it literally reversed the damage. In a major breakthrough in aging research, scientists at Korea University’s College of Medicine have discovered that cellular aging can spread throughout the body via the bloodstream. The culprit is ReHMGB1, a redox-sensitive version of a protein called HMGB1, which is secreted by aging, or senescent, cells. This protein was shown to trigger aging-like effects in distant, healthy tissues—reducing their ability to regenerate and harming muscle performance. In mouse models, researchers successfully blocked ReHMGB1 using antibodies, resulting in fewer signs of aging and improved recovery after injury. This study, published in Metabolism – Clinical and Experimental, provides the first direct evidence that aging is not just a local cellular process but one that can be systemically transmitted through blood. By identifying ReHMGB1 as a key molecular messenger in this process, researchers now have a potential target for therapies aimed at slowing or even reversing age-related decline. The findings could have far-reaching implications for treating a range of conditions tied to aging, from muscle wasting to organ degeneration. Source: Jeon, O. H., et al. (2025). Propagation of senescent phenotypes by extracellular HMGB1 is dependent on its redox state. Metabolism – Clinical and Experimental.

@Dr_Singularity Small models out-learning giants? Scaling laws just flinched 🧐

Big step toward AGI "A new approach developed by MIT researchers enables LLMs to update themselves in a way that permanently internalizes new information." MIT AI breakthrough could change everything about how AI learns. They’ve built a framework called SEAL that lets an LLM teach itself new knowledge permanently, like an actual student updating its brain. Instead of staying static after deployment, the model now writes its own study sheets, generates multiple versions, tests itself to see which one works best, and then updates its internal weights to lock in the new information. It literally self edits and self improves through trial and error. Early results: +15% accuracy on question answering +50% success on learning new skills A small model outperforming huge LLMs Yes, there’s still catastrophic forgetting to solve (we're making fast progress here-link below), but this is the first real step toward continuously learning AI agents that adapt, evolve, and collaborate.

@leonklein26 When diffusion models learn physics, generation becomes simulation!! the line between sampling and understanding starts to blur.

(1/n) Can diffusion models simulate molecular dynamics instead of generating independent samples? In our NeurIPS2025 paper, we train energy-based diffusion models that can do both: - Generate independent samples - Learn the underlying potential 𝑼 🧵👇 arxiv.org/abs/2506.17139

RNA repair as memory therapy biology’s version of debugging the brain!!