A new protein-based gel can regenerate tooth enamel. For generations, dental patients have accepted a hard truth: once your tooth enamel is gone, it is gone for good. However, a major scientific breakthrough has shattered this limitation. An international research team, led by scientists at the University of Nottingham, has developed a revolutionary biomimetic gel capable of repairing and regenerating damaged tooth enamel. Published in the journal Nature Communications, this fluoride-free treatment uses lab-engineered proteins to mimic the natural processes of enamel development during infancy. When applied to the teeth, it creates a microscopic scaffold that captures calcium and phosphate ions from the patient's saliva, systematically growing a new mineral layer that seamlessly integrates with the tooth's original structure. This innovative gel offers a painless, drill-free alternative to traditional fillings and crowns, which only mask damage rather than healing it. Clinical trials for the biocompatible gel are slated to begin this year, and the development team is already collaborating with spin-off company Mintech-Bio with the goal of releasing their first commercial product as early as next year. By offering a permanent, easy-to-apply solution for acid erosion and early-stage cavities, this regenerative technology could fundamentally change routine dental care and spare millions of patients the discomfort of the dental drill. source: Hasan, A., et al.. Biomimetic supramolecular protein matrix restores structure and properties of human dental enamel. Nature Communications, 16(1), 9434.

Scientists are using extracts from sheep's wool to regenerate bone tissue in living animals. Researchers at King’s College London have unlocked a highly sustainable breakthrough in regenerative medicine by turning ordinary sheep’s wool into advanced bone-healing scaffolds. The team extracted keratin, the tough structural protein that gives wool its strength, and developed it into thin, biomimetic membranes. In laboratory trials, these membranes successfully supported the growth of human bone cells. When tested in living animal models with severe, non-healing skull injuries, the wool-derived scaffolds guided the growth of new bone directly across the damaged areas, demonstrating the immense clinical potential of this abundant agricultural byproduct. What makes this discovery particularly remarkable is the quality of the regenerated tissue. While collagen has long been the gold-standard scaffold in orthopedics, it often suffers from poor structural strength and rapid degradation. The new wool-based keratin scaffolds, however, produced bone that was far more organized, stable, and structurally similar to healthy natural bone than tissue grown with conventional collagen. Because wool is a renewable resource often discarded as farming waste, this technology offers a scalable, eco-friendly, and cost-effective alternative that could soon challenge the long-standing medical reliance on expensive, animal-derived collagen. source: Gamea, S., Kaabi, H., Elredah, I., Aljohani, H., Horamee, S., Abu Shaqrah, H. J. N., Lu, E. M.-C., Hodgkinson, T., Balto, H., Sharpe, P. T., & Elsharkawy, S. (2026). Bone regeneration of rat calvarial defect using biomimetic keratin-based membranes. Biomaterials Advances, 184, 214806.

Ever fix one slide and watch the agent regenerate the whole deck and break everything? This work solves that with hierarchical memory plus local revision 🗂️ Title: MemSlides: A Hierarchical Memory Driven Agent Framework for Personalized Slide Generation with Multi-turn Local Revision URL: ❓ What's new? 💡 It gives a slide-generation agent a hierarchical memory that cleanly separates persistent preferences (user profile), session-level working memory, and reusable tool experience. It also edits only the target slides instead of regenerating the full deck. ❓ How does it keep local edits from drifting? 💡 A Plan–Act–Guard loop. Plan turns a request into an explicit "execution contract" (target paths, active rules, coverage requirements); Act picks tools like batch CSS over shared selectors or patch operations; Guard binds patches to content hashes and blocks premature finalization until all targets are covered. ❓ Do one-off instructions leak into permanent preferences? 💡 No — at job end, intent-aware consolidation writes back only stable signals, preventing transient "just this round" requests from becoming persistent preferences. ❓ Does it work? 💡 In blind review, personalization beats DeepPresenter on every dimension (e.g. Visual +1.66). With tool memory, closed-loop completion hits 0.963 and time-to-first-correct-edit drops 609.5s → 242.5s (~60% faster). Designing agent memory by role rather than one flat store feels broadly useful for multi-turn editing tasks. #AIAgents# #LLM#

Scottish scientists are at the forefront of MS research using tiny fish that can regenerate their nervous systems. But Britain’s post-EU reality is creating bottlenecks for breakthroughs. 🔗

A new method tackles the enterprise RAG chunking problem while cutting cost by 95.7%. Title: D-RAC: Document Retrieval-Aware Chunking URL: 📌 Overview D-RAC is a four-stage pipeline that normalizes messy enterprise documents (PDF, DOCX, PPTX, scans) into PDF, converts them once with a multimodal LLM into retrieval-optimized Markdown, then plans chunks deterministically over element IDs. ❗ The problem Rule-based extraction destroys tables and heading hierarchy, while accurate agentic chunking regenerates the entire document, making it expensive at scale. 🛠️ Method Tricks like turning each table row into a self-contained prose sentence, reconstructing heading hierarchy, and passing only element IDs to the planning LLM keep both retrieval quality and cost efficient. 📊 Results Across 236 documents and 795 pages, D-RAC cut output tokens by 95.7% while Recall@6 reached 0.798, beating agentic chunking (0.795) and rule-based extraction (0.717), with 75% less processing time. 🏢 Use cases Performance stayed stable across automotive, banking, cloud, and other enterprise domains, making it a strong fit for production RAG pipelines. #RAG# #DocumentAI#

// Design Docs Are All You Need // Banger paper from Google DeepMind, MIT, and colleagues. What a genuinely strange and interesting paper this one is. Here is the setup: They maintain a performance-modeling library whose main branch contains almost no code. The repository is a directed graph of natural-language design docs. Coding sub-agents regenerate the entire implementation from those docs whenever a version updates. Every human change is an edit to a doc. The premise is that ML performance modeling invalidates its own abstractions every hardware and model generation, and coding agents are now cheap enough that regenerating a library beats patching one. Two things make the regeneration reliable. The design docs are written around step-by-step worked examples, which act as in-context demonstrations for the generating agents. The system is also anchored on a minimal recursively defined operator IR with symbolic cost expressions in SymPy. Regenerated implementations reproduce hand-audited reference models to round-off precision, including DeepSeek-V3 serving on a TPU pod slice. Paper: