Headlines

Ellipsis Health, a startup developing voice AI for healthcare, has acquired another company as demand for AI that can call patients heats up, Endpoints News learned exclusively. Ellipsis, which uses voice AI to screen and ...

Potential Sciences, a startup backed by the UK government, is betting that training AI models on computer-generated biochemical data can speed up the plodding process of developing new drugs.

Hormone replacement therapy is a popular treatment for women leading up to or following menopause. Manipulating sex hormones is a very blunt strategy, with potentially significant downsides, and a high variation in outcomes between patients, but the medical community has accumulated a great deal of experience in delivering this class of therapy. In the matter of assessing how this sort of approach affects aging, the research community is presently engaged in a process of gathering as much data as possible on how aging clocks interact with various therapies thought to influence aging or age-related disease. The accuracy of the clocks in the environment of any given therapy is unknown, and the effect of the therapy on the clock is unknown, and so the way forward […]

The update is expected to reduce false positives and allow game developers to use the tool for their own legitimate projects without getting blocked.

It is by now well established that senescent cells promote senescence in neighboring cells. This is a useful and necessary part of the benefits of the transient presence of senescent cells in the contexts of cancer suppression and coordination of wound healing. But when senescent cell clearance falters in later life, and senescent cells begin to linger and accumulate, this promotion of further senescence makes matters worse. One of the ways in which senescent cells promote senescence is via relatively well mapped and well understood inflammatory signaling. Here, however, researchers note that there are other pathways involved, in which the signaling involves the contents of small extracellular vesicles. Interfering in senescent cell signaling as a basis for therapy is likely a more challenging path forward […]

Voice phishing, or vishing, used to depend on a convincing script and a confident caller. In 2026 it

arXiv:2611.05127v2 Announce Type: replace Abstract: General-purpose generative AI (GenAI) models are being rapidly adopted in healthcare, but they are challenging to assess with traditional evaluation frameworks. These conceptual, methodological, and practical challenges parallel decades of work in evaluating clinician and health system quality and safety. We describe the adaptation of methods from several fields, combined with emerging technical approaches, into a structured, multi-method evaluation framework designed to give well-rounded assessments of GenAI in health at scale. The approach first anchors evaluation to Critical User Journeys (CUJs) to help prioritize out of a large choice space, and maps these to a phased AI system lifecycle (base model development, product development, and post-deployment monitoring), since evaluation at each stage requires different methodological approaches. In this context, the approach then integrates three distinct evaluation methods

arXiv:2611.04507v2 Announce Type: replace Abstract: AI agents can make incorrect decisions or waste resources even when their tools and infrastructure work normally. Their nondeterministic behavior makes these problems difficult to reproduce and diagnose. We developed Signals to identify such problems from agentic traces and suggest concrete remediations. We built the method around a taxonomy of 45 problem types across eight families, with definitions and evidence requirements for each type. The method distinguishes realized failures, risk exposures, and inefficiencies, groups related findings into engineering issues called Signals, and assesses their severity from the supporting evidence. Each Signal includes actions a developer can take and checks for evaluating the change. Further executions can add evidence to an existing Signal or reveal a new problem. We evaluated the method on public agent executions and achieved 91.43% F1 when matching Signals to reference issues, about 18

arXiv:2611.06306v1 Announce Type: cross Abstract: Spatial diversity is a key mechanism in classical Multi-Input Multi-Output (MIMO) communications, where multiple propagation alternatives provide redundant observations that improve reliability in noisy environments. A direct translation of this principle to quantum communications is forbidden by the no-cloning theorem, since an unknown quantum state cannot be copied and transmitted through independent links. In this Perspective article, we argue that spatial path superposition provides a no-cloning-compatible analogue of the spatial-diversity principle of MIMO. Rather than duplicating the transmitted quantum state, a quantum carrier is coherently delocalized over multiple communication links, so that alternative spatial evolutions are experienced in quantum superposition. In this framework, classical redundancy is replaced by spatial coherence, and the communication link itself becomes part of the quantum dynamics. Indeed, spatial

arXiv:2611.06296v1 Announce Type: new Abstract: Ambient sensors offer a privacy-preserving solution for activity recognition in smart homes, yet they face a critical challenge in multi-resident environments known as Data Association, which aims to identify the resident triggering a specific anonymous sensor event. Existing approaches fundamentally fail to maintain global consistency by often yielding conflicting associations for silent residents. Additionally, these methods frequently suffer from severe trajectory fragmentation caused by the ontological mismatch of applying continuous physical motion models to discrete logical behaviors. To address these limitations, we propose a novel framework that reformulates data association as a sequential decision-making task within a Markov Decision Process (MDP). We solve the defined MDP via Behavior Cloning (BC) to avoid complex reward engineering by directly learning from expert demonstrations. We instantiate the policy using two

arXiv:2611.05925v1 Announce Type: new Abstract: Quantifying glomerular inflammation in kidney-transplant biopsies is traditionally performed on single whole-slide sections using the Banff "most-severe section" rule, despite the inherently three-dimensional nature of lesion distribution. This 2D based assessment might lead to instability when inflammation varies across slices. This paper reports a feasibility study examining the technical feasibility and clinical relevance of extending Banff scoring into three dimensions. To this end, we propose a 3D Banff lesion-scoring framework that aligns serial sections, reconstructs glomerular identities across depth, and integrates inflammatory-cell counts in 3D under three scoring strategies: Compatible, Volumetric, and Deduplicated Volumetric. In the experiments, glomerulitis (g-scores) was used as an example Banff metric and applied to multi-section renal allograft biopsies. Our findings indicate that 3D Banff g-scores are more consistent

arXiv:2611.05638v1 Announce Type: new Abstract: Speaker-recognition benchmarks test how accurately a model recovers the speaker label of a recording, treating voice identity as a property of the person who produced it. However, these benchmarks do not evaluate how humans perceive voice identity. Whether a listener hears two voices as the same person has consequences for intellectual property rights, cybersecurity, and everyday life. We introduce the Voice Identity Perception benchmark (VIPBench), which contains 124,876 same-or-different judgments from 1,290 listeners on 9,800 recording pairs from 100 public figures. The pairs include real recordings, voice clones generated by a text-to-speech system, and morphs between two speakers' voices. We also define four tasks that evaluate the alignment between speaker-representation models and human identity perception. The first two tasks measure how well a model's similarity score predicts the proportion of listeners who respond "same" and

The system, MAI-DxO, was not only more accurate, but it also spent much less on tests to make its diagnoses. The post Microsoft’s Medical AI Is Over Twice as Good Diagnosing Diseases as Doctors appeared first on SingularityHub.

A UCLA-led team has described how they advanced an experimental CRISPR approach for sickle cell disease into a blood stem cell therapy that can be manufactured at the scale and quality required for a clinical trial.

A stable-pool strategy can move cell line development off the critical path to Phase I while preserving a path to a conventional clonal cell line as a program matures. The post Rethinking the Clone-First Paradigm for First-in-Human Biologics appeared first on GEN - Genetic Engineering and Biotechnology News.

Researchers found that the gene preserves bone stem cell supply throughout life, and say the discovery advances understanding of stem cell biology and could indicate potential therapeutic approaches to improve bone healing and regeneration in aging and disease. The post Gene That Preserves Bone Stem Cell Supply Throughout Life Identified appeared first on GEN - Genetic Engineering and Biotechnology News.

Bloomberg: Sources: Google DeepMind spinout Isomorphic Labs is in talks to raise ~$5B at a ~$50B valuation; the AI drug discovery startup last raised $2.1B in May  —  Alphabet Inc.'s Isomorphic Labs is in talks to raise roughly $5 billion in new funds to support its efforts to use artificial intelligence …

A UCLA-led team has described how it advanced an experimental CRISPR approach for sickle cell disease into a blood stem cell therapy that can be manufactured at the scale and quality required for a clinical trial.

A new study protocol details the TACMONO-XL trial, which will randomize 306 kidney transplant recipients aged 60 and older to tacrolimus monotherapy or standard triple immunosuppression to test effects on infection-related death and quality of life.

A phase 1 trial found that a pill of 16 selected beneficial bacteria was safe for stem cell transplant patients and was associated with fewer bloodstream infections.

Researchers built a cell-lined silicone blood vessel large enough to hold a real stent and quantified how endothelial cells recolonize the device under controlled blood-flow conditions.

Kamau Therapeutics, which has stayed relatively quiet since its 2023 emergence, is sharing clinical data from its revamped sickle cell gene editing program for the first time. The startup is giving a ...

Utah has become the first state to let an artificial intelligence (AI) program prescribe medication without a doctor writing the prescription, starting with acne treatments.

Rice University's Han Xiao has used his AI-supported Sequence Display technology to expand the targeting abilities of CRISPR/Cas9. In a paper recently published in Nature Communications, Xiao evolved several variants of a compact Cas9 that could efficiently target stretches of DNA previously inaccessible for editing. The research included a proof-of-concept study in which the evolved Cas9s were successfully used to edit two genes found in coronary heart disease.

We all want a well-functioning immune system to help us fight invasive bacteria and viruses and recover faster from illness. Popular "immune-boosting" ideas, however, might be more of a hindrance than a help.

The action of a key gene is required to maintain a lifelong supply of stem cells that make bone healing possible, a new study finds.

Tuesday, November 17, 2026 In an effort to reach an elusive continuity of the self, some far-thinking technologists are working on ways to keep your brain going—even beyond death. Join science editor Mary Beth Griggs and senior editor for biomedicine Antonio Regalado for a conversation exploring possible paths to keep your mind going, including gradual…

Cryo-electron microscopy structures of the zebra finch R2 retrotransposon, alone and bound to Cas9, reveal how target-primed reverse transcription initiates and guided the engineering of an enhanced STITCHR gene-integration system.

A study of 56 end-stage non-CF bronchiectasis patients referred for lung transplantation evaluation found group 3 pulmonary hypertension in 73.2 percent and significantly higher mortality among those affected.

WEDNESDAY, Nov. 4, 2026 — Utah has become the first state to let an artificial intelligence (AI) program prescribe medication without a doctor writing the prescription, starting with acne treatments.New York City-based Nolla Health began offering...

An interpretable pathomics framework predicted BRAF V600E status in papillary thyroid carcinoma from routine Wright-stained cytology slides, reaching an AUC of 0.925 on pooled independent test data across two centers.

A multi-omics study shows that lactate produced by senescent breast cancer cells drives the SASP through histone H3 lactylation at inflammatory gene loci.

A CT-based AI system reduced benign interventions among treated kidney lesions from 40.3 percent to 16.4 percent in a randomized implementation trial while improving radiologist accuracy and reading speed.

A GIS and AI-assisted planning framework tested on aging districts in Sharjah and Ajman proposes walkable street networks, land-bank reuse, parking consolidation and pocket greens without greenfield expansion.

Chick satellite cells from breast and leg muscles show opposite patterns of proliferation and differentiation when cultured under identical conditions, a new study finds.

We look at some of the key biotech developments, companies, and clinical data to watch at AHA Scientific Sessions 2026. The post From gene editing to RNA therapies: The biotech trends to look out for at AHA Scientific Sessions 2026 appeared first on Labiotech.eu. © Labiotech UG and Labiotech.eu. Unauthorized use and/or duplication of this material without express and written permission from this site’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Labiotech UG and Labiotech.eu with appropriate and specific direction to the original content.

Подход к уходу за телом в современной эстетической медицине претерпевает изменения. Если раньше основные усилия были сосредоточены на борьбе с лишними килограммами, то сегодня все чаще звучит термин body longevity. Это концепция, при которой фокус смещается с цифр на весах на сохранение качества тканей и функциональных ресурсов организма в долгосрочной перспективе. Тело как ресурс, а не объект для коррекции Сегодня специалисты призывают оценивать состояние организма комплексно. Важны не только объемы, но и состояние кожи, количество подкожно-жировой клетчатки, а также мышечная масса и сила. Такой подход подразумевает, что косметология не заменяет здоровый образ жизни, а лишь дополняет его, помогая справляться с задачами, которые невозможно решить только с помощью фитнеса или диет — например, с дряблостью кожи после похудения или локальными жировыми отложениями. Работа с разными слоями тканей Современные аппаратные методики позволяют точечно воздействовать на разные уровни: от дермы до

A new mini-review proposes that durable senescence-like states in dorsal root ganglion sensory neurons may help explain why pain persists after injury, while stressing the evidence remains a testable hypothesis rather than a proven cause.

OccupySaaS AI-powered project wants to kick Auth0, Salesforce, and other pricy service-as-software to the curb

Women with HIV who achieve sustained virologic suppression live longer than demographically similar men, a difference resembling the female longevity edge in the general population.

A new review examines how plant-derived compounds such as polyphenols, flavonoids, alkaloids and isothiocyanates may target the breast cancer stem cells linked to recurrence, metastasis and treatment resistance.

arXiv:2611.03832v1 Announce Type: cross Abstract: Computational drug discovery offers a growing range of tools for proposing and evaluating molecules, but combining them and deciding what to do next still requires substantial expert effort. Here, we introduce El Agente F\'armaco, an AI scientist that uses modern large language model (LLM) technology to turn discovery goals expressed in everyday language into complex, scalable computational screening workflows. The agentic harness selects scientific tools, reviews their results, and adjusts subsequent steps to the target and the user's objectives. Our approach centers on a shared scientific memory: F\'armaco Atlas, which stores molecules, structures, and results together with their provenance -- a record of how they were obtained. The agent uses targeted database queries to retrieve the evidence needed for each decision. A single request can launch calculations for an entire collection of molecules, without requiring a separate LLM

uniQure has aligned with the FDA on what the company calls key elements of the confirmatory study design: The study is being designed as a randomized, standard-of-care trial enrolling approximately 200 patients in a 2:1 ratio of ifezuntirgene inilparvovec to standard of care, with Total Functional Capacity (TFC) at 36 months serving as the primary endpoint. The post FDA Sets May 1 PDUFA Action Date for uniQure’s Huntington’s Disease Gene Therapy appeared first on GEN - Genetic Engineering and Biotechnology News.

A single gene-editing treatment could someday protect against heart disease, high cholesterol, and other common conditions for decades, potentially replacing lifelong medications. But scientists warn that permanently altering genes introduces risks that could take years to emerge, making long-term safety a critical challenge.

Xavier Martinez / Wall Street Journal: Pharmaceutical leaders are skeptical of claims that AI could cure most diseases within a decade; the US FDA has yet to clear any AI-designed biotech drugs  —  No drug designed by artificial intelligence has won FDA clearance, and pharma leaders say biology is the obstacle

The FDA accepted uniQure's AMT-130 application for Huntington's disease, granted priority review, and set a May 1, 2027, target decision date while approval remains uncertain.

Researchers at City of Hope®, one of the largest and most advanced cancer research and treatment organizations in the country, and collaborators have pioneered a new approach to preventing dangerous bloodstream infections in patients receiving donor stem cell transplants for blood cancers.

Scientists from Skoltech, together with colleagues from the Kharkevich Institute for Information Transmission Problems of the Russian Academy of Sciences, Lomonosov Moscow State University and other organizations, conducted a comparative study of the 3D genome organization in neurons derived from induced pluripotent stem cells and in neurons of the adult human brain. The authors collected and processed more than 300 3D genome maps covering 62 human samples and 34 mouse samples, including data on embryonic stem cells, induced pluripotent stem cells, neural progenitors, fetal and postmortem neurons, as well as neurons differentiated in vitro.

A study led by researchers from the Germans Trias i Pujol Research Institute (IGTP) and the Blood and Tissue Bank (BST) has developed and validated a manufacturing platform for extracellular vesicles that meets pharmaceutical quality requirements, marking a key step toward their clinical application.

A retrospective Australian cohort study of 405 liver transplant recipients found that basiliximab induction was associated with early and sustained improvement in adjusted serum creatinine and greater renal dysfunction resolution at 12 and 24 months, without differences in rejection, graft survival or patient survival.

Researchers engineered lipid nanoparticles around the antihistamine hydroxyzine, achieving brain delivery of mRNA, CRISPR components and a radioprotective interleukin in animal models.

A Dutch study finds that kidney transplant recipients maintain RSV-specific antibodies and memory T cells comparable to healthy controls despite long-term immunosuppressive treatment.

After a whirlwind third quarter that saw two FDA approvals but a major setback in Angelman syndrome, Ultragenyx is laying off staff for the second time this year. The California biopharma

These edits of Jason Statham (left) and Jeff Bezos are a bit jarring, but that’s sort of the point. | Image: The Verge Amazon founder Jeff Bezos and actor Jason Statham have a distinctive feature in common: They're both famously bald, and that definitely hasn't changed in the last few days. However, there are some images of the two men circulating around X, Threads, and Bluesky in which they both have substantially more hair than their usual shaved look, which in Bezos' case appears to be very freshly transplanted. It's all just photo editing fakery, AI or otherwise. It's unclear how viral these posts are, but they're prevalent enough that a few members of the Verge team have encountered them in their social media feeds. Some people who shared them claimed … Read the full story at The Verge.

By KIM BELLARD I once again have to thank The New York Times for introducing me to a new concept. Sarah Kessler wrote about the rise of “legal engineering” – for which,Continue reading...

Cells contain hundreds of mitochondria, constantly fusing together, replicating by splitting apart, and swapping component parts. Mitochondrial function is vital to cell health, and declining mitochondrial function is an important component of degenerative aging. Some of the answers as to why mitochondrial function declines with age are likely buried somewhere in the complex interactions of mitochondrial structure, the dynamics of mitochondrial fusion and fission, and the operation of the mitochondrial quality control processes of mitophagy. Mitophagy is both necessary to maintain function by destroying worn mitochondria, but it also seems quite sensitive to changes in structure and size of mitochondria, as happens when the dynamics of fusion and fission change. In today's open access paper, researchers demonstrate that compounds that improve the activity of mitofusin, […]

Researchers have identified a mechanism involving fatty acid metabolism that may determine whether a damaged cell enters senescence and persists in the body or is eliminated through cell death. Published in the journal Cell, the study could open new avenues for treating age-related diseases. Researchers from the University of Eastern Finland contributed to the international study.

The FDA's final guidance urges cell and gene therapy makers to use broader potency assurance strategies, combining manufacturing controls and testing to help ensure each batch delivers its intended effect.

John Leeson, who gave Doctor Who's robotic dog K9 his distinctive voice, has died aged 83 after months of illness, his son Guy Ducker confirmed

arXiv:2611.01312v1 Announce Type: new Abstract: Background. Many medical AI models are trained to reproduce a single reference label and hand the most likely label to the next step in a clinical workflow. A classifier can select that label correctly while providing an inadequate summary for estimating downstream risk. Methods. We use Jeffrey's rule of conditioning to specify when probabilities over intermediate states, such as diagnoses or imaging categories, can be carried forward to an outcome risk. We separate two errors of reporting only the most likely label, discarding the probability of the other states and applying the risk of patients who have the leading diagnosis to a patient who only probably has it. We also give conditions under which the label and the full distribution lead to the same decision, and we propose guardrails for device developers, regulatory reviewers, and health insurers. Results. In a hypothetical chest-pain differential, reporting only the benign leading

arXiv:2611.02554v1 Announce Type: new Abstract: The clinical translation of artificial intelligence (AI) is a critical step toward precision medicine. The implementation study presented here focuses specifically on barriers to AI adoption in the diagnostic medicine workflow. Successful clinical adoption of AI requires more than technical performance; it depends on addressing real-world organizational, ethical, infrastructural, and workflow-related barriers that may prevent promising systems from moving beyond research settings. To examine these barriers, we used structured concept mapping (CM) to operationalize stakeholder engagement and to integrate a knowledge translation (iKT) approach within the Knowledge-to-Action (KTA) framework. Stakeholders co-generated 96 statements describing barriers to AI implementation and organized them into 11 clusters, including infrastructure limitations, ethical and legal concerns, workflow integration challenges, and distrust in algorithmic

CMS finalized a rule clarifying when it can immediately decertify organ procurement organizations for unsafe practices, with changes affecting oversight, performance measures, appeals and patient safety.

Jerry Fort, the GTA 6 fan known as Void Only, died on 29 October after an infection filled his lungs with blood, 21 days before the game's 19 November launch.

A protein that helps build teeth may one day be used to allow damaged teeth to repair themselves.

In a landmark achievement in the field of genetic medicine, researchers have reported the very first published case in a peer-reviewed journal of a successful clinical treatment of a patient with alpha-1 antitrypsin deficiency (PiZZ) using in vivo gene therapy.

Contemporary hormone replacement therapy (HRT) was not associated with an increased risk for heart attack, stroke, blood clots or death among women ages 45 to 59, according to a large, long-term study of over 52,000 women in Israel.

A cross-sectional study of 615 adults in the Loma Linda Blue Zone area and surrounding communities found that vegetarian diets were more common among older adults and Seventh-day Adventists, with vegetarians also reporting significantly lower BMI than non-vegetarians. The findings raise concerns about whether younger generations are maintaining the region's traditional dietary practices, although further research is needed to establish whether vegetarian adherence is declining over time.

There’s a fine line between AI creations inspired by video games and those that are infringing copyright.

Researchers at the German Cancer Research Center (DKFZ) and the HI-STEM Stem Cell Institute have discovered a group of particularly long-lived blood stem cells in mice. These cells remain in a dormant state for an exceptionally long time while retaining their ability to generate new blood cells. This newly discovered cell population may be responsible for sustaining blood regeneration and immune defense throughout a mouse's entire life.

More than just a precious gemstone for jewelry, diamond has properties that make it attractive for several next-generation, high-performance technologies. Its high thermal conductivity can help move heat away from semiconductor devices, while its wide bandgap is promising for high-power electronics. Moreover, its transparency in the visible light range makes diamond useful for demanding optical applications.

Cells become senescent constantly throughout life, usually upon reaching the Hayflick limit to replication, but also in response to injury, high levels of cell stress, or potentially cancerous damage. Senescent cells cease to replicate, grow in size, and secrete pro-inflammatory signals to attract the attention of the immune system. In youth, the immune system clears senescent cells rapidly, but this becomes less efficient with advancing age for a range of reasons yet to be fully explored. Senescent cells appear to alter their biochemistry in ways that inhibit immune clearance, but the immune system as a whole also becomes less effective at all of its tasks. Senescent cells linger, accumulate, and disrupt tissue structure and function with their inflammatory signaling. The vascular system is vital to […]

Only 13% of articles examined in a review of artificial intelligence (AI) and health equity explicitly defined health equity—a gap that could leave AI-driven approaches benefiting the most advantaged groups. That's a key finding from a review led by Caroline R. Efird of Georgetown University and colleagues. The research was published Oct. 8, 2026, in Health Affairs Scholar.

The tsunami of imitators is way bigger than any one model maker can stem.

Some genetic diseases leave cells without a crucial protein. One possible treatment is to persuade genes already inside those cells to work harder, but the necessary machinery can be difficult to deliver, especially when researchers want to target several genes at once. A new compact CRISPR system devised by Hainan Zhang of the Shanghai Institute of Materia Medica, Chinese Academy of Sciences, and colleagues fits those instructions into a single viral delivery package.

Anthropic employee confirms that Claude Code now has orders not to reproduce game IP — company "iterating to ensure [...] the right balance"

Contemporary hormone replacement therapy (HRT) was not associated with an increased risk of heart attack, stroke, blood clots or death among women ages 45 to 59, according to a large, long-term study of more than 52,000 women in Israel. The findings from this independent, preliminary research will be presented at the American Heart Association's Scientific Sessions 2026. The meeting is Nov. 6–9 in Chicago.

What does it mean to feel that a part of us will live on after we are gone? A new study of nearly 3,000 adults found that people who believed they would be remembered by others tended to report better mental health, a stronger sense of meaning and lower levels of problematic substance use and addictive behaviors. The findings point to a possible link between our sense of lasting significance and the way we cope with stress, uncertainty and mortality.

Doctors at Beilinson Medical Center perform a robotic lung transplant with minimal incisions, joining around 20 similar procedures worldwide; report rapid recovery and reduced need for painkillers.

arXiv:2610.38234v2 Announce Type: replace Abstract: We demonstrate the first zero-shot sim-to-real transfer for a surgeon-robot collaboration task in open surgery: retractor take-over. A robot arm takes a retractor from the surgeon's hand and preserves retraction at handoff, using a policy trained only on simulated demonstrations, without human demonstrations or real-world fine-tuning. Imitation learning from real demonstrations is limited by collection and labeling costs, and by performance degradation when the environment changes. Our framework, SimRetract, generates simulated demonstrations designed for transfer: the policy observes masked instrument images and grasp geometry, learns to recover from off-nominal states, and acts with margin for execution drift. Deployed zero-shot, it grasps the retractor in 90% of 50 surgeon-defined phantom presentations and preserves retraction at handoff in 78%. Results on unseen retractors suggest potential for generalization to instruments with

arXiv:2610.05533v3 Announce Type: replace Abstract: A harness search around a frozen longevity-tuned 1.7B model reported a rescue. An audit found that the gain came from rows the start left unparsed, that the caller sent each row's reference answer as an assistant turn, and that replication through the same pipeline missed it. Is a harness's gain more than a constant answer would give? We preregistered three studies and report every endpoint. In the first two, thinking off raised test accuracy over a capped thinking setting in 7 of 12 cells, mostly on rows the cap left unparsed. It was not meaningfully worse than a rescue in 7 of 9 small-model test cells or than four GEPA harnesses at 1.7B. Replication reproduced an injected defect's distortion in 9 of 20 evaluable cases (7 fixed by design). At 14B, two GEPA harnesses beat thinking off by 34 points on test but were not detectably above the strongest constant label, and answered "Yes" in 94.6% and 98.2% of parsed mortality answers on

arXiv:2610.42066v1 Announce Type: new Abstract: AI answers require more than apparent correctness: reviewers need to inspect their recorded basis. We introduce the Knowledge Layer Model (KLM), an open, implementation-neutral specification for inference records with versioned epistemic assessments, evidence relations, and explicit unknowns. SEDIM, the knowledge and evidence architecture of Nage AI OS, is a reference implementation. We reference KLM v0.3 (repository tag v0.3.0). Empirical material consists of separate internal experiments, not end-to-end validation. A historical platform-domain routing benchmark reports 137/150 classifications across 15 classes (91.3%; nominal 95% Wilson interval 85.7-94.9%); the numerator is recovered from a rounded summary, and historical example membership is unverified. A native-model weight-path configuration yielded 0/15 successful answers in a corrected two-source test. A capacity-growth comparison could not establish a reliable effect because of

arXiv:2610.41978v1 Announce Type: new Abstract: Automated endoscope guidance reduces the need for continuous manual camera control during minimally invasive surgery. However, the desired camera motion depends on the surgeon's immediate intentions and cannot always be inferred from the visual operative scene alone. We propose to automate the task using language-conditioned behavioral cloning; we define 6 clinically relevant guidance behaviors and collect 235 subtask demonstrations. In preliminary phantom experiments, our policy achieves 100% success on focus, overview, and stop, 80% on search-tip and follow-tip behaviors, and 50% on horizon leveling. These results demonstrate the feasibility of learning heterogeneous surgeon-requested camera behaviors with a single visuomotor policy, while the observed failure modes motivate future work on adding temporal context for more demanding guidance tasks.

arXiv:2610.41348v1 Announce Type: new Abstract: Robots deployed in unstructured environments are susceptible to actuator-authority degradation, which reduces the effective control authority of individual actuators and can cause reinforcement learning policies trained under nominal dynamics to fail. Existing robustness approaches rely on discrete fault labels, broad domain randomization, or global latent variables, often without explicitly modeling continuous per-joint residual actuation capability. We propose a joint-health-aware reinforcement learning framework for robust robot control under simulated actuator degradation. Our framework encodes actuator condition in a continuous joint-health vector and models degradation as multiplicative attenuation of executed actions. A dedicated HealthNet estimator infers per-joint health from recent observations and actions, allowing the policy to redistribute control effort across healthy and degraded joints. We evaluate the method on Walker,

Shares of gene editing therapy companies mostly surged this past week on two growth catalysts: Data readouts at the European Society of Gene and Cell Therapy’s 33rd ESGCT Annual Congress, and the FDA's new draft guidance for personalized gene therapies. The post StockWatch: Gene Editing Shares Jump on ESGCT Data, FDA Draft Guidance appeared first on GEN - Genetic Engineering and Biotechnology News.

AI has been making its way into video game development for a while now, but the recent developments have started raising some serious concerns. Popular games are being reverse-engineered with AI assistance, recreated for web browsers, and made playable on platforms they were never released for. Between all that, indie developers have started pulling down […]

Cailey Gleeson / Fierce Healthcare: AI-powered medical search engine Vera Health, which says it is used by 800,000+ clinicians in 190 countries, raised a $28.5M Series A led by Threshold Ventures  —  Artificial Intelligence funding round series A venture capital (VC)  —  Clinical artificial intelligence search engine Vera …

A Pew survey of 3,488 U.S. adults found that 60% are uncomfortable with doctors using AI for diagnoses and treatment, while most want human oversight of medical decisions.

Will Maddox / Axios: Midstream Health, whose AI tools help 250 hospitals recover revenue from rebates and vendor and payer underpayments, raised a $20M Series A led by Valtrius  —  Midstream Health, an AI financial system for hospitals, raised a $20 million Series A round to expand its platform …

Sen. Bill Cassidy's PATIENT AI Act discussion draft would direct Medicare to test paying for autonomous clinical AI while requiring patient notice and preserving appeal rights.

An AI Agent Injected Drugs Into a Live Mouse’s Brain  WIREDA Claude model autonomously chose a gene therapy and operated on a mouse brain with no human decision-making  The DecoderAI agent used to perform controversial brain surgery on a mouse, Anthropic Sonnet-controlled motorized surgery rig — Opus 5.5 refused to conduct biological experiment, Sonnet given free rein to treat epilepsy with a range of gene therapy vectors after  Tom's HardwareExin Therapeutics Uses Anthropic Sonnet to Guide Mouse Brain Surgery  news.lavx.huAI Agent Autonomously Performs Gene Therapy Injection on Live Mouse Brain  The Tech Buzz

Guy Penrod seeks prayers after disclosing a serious blood and bone marrow illness, telling Mark Lowry a transplant is the only potential fix.

Microsoft's AI diagnostic system identified 83% of complex medical cases in a study, more than doubling the 36% accuracy of generalist doctors.

In the third of a four-part Health Matters wellness series on frontier medicine in Hong Kong, Fiona Sun examines how a local start-up is testing a gene-editing therapy that could tackle a group of inherited eye diseases at their genetic source and what it will take to turn it into a treatment patients can access. For 15 years, Professor Zheng Zongli has worked to make gene-editing therapies safe enough to treat diseases with limited treatment options. His path has taken him from a PhD in medical...

TechNode: Shanghai-based NeuroXess, which is developing BCI technology for movement support, speech decoding, and control of external devices, raised a ~$105M Series B  —  Chinese brain-computer interface startup NeuroXess has raised RMB 700 million in Series B funding, co-led by SDIC Juli, Shenzhen Capital Group and Qinzhi Capital.

The relationship between creativity and mental health has attracted the interest of many psychologists over the past decades. Some past studies have found that creativity is sometimes linked to mental health difficulties, while others collected opposing or inconsistent results.