AI News Archive: May 7, 2026 — Part 13
Sourced from 500+ daily AI sources, scored by relevance.
- OpenAI has new voice models that reason, translate, and transcribe as you speak
OpenAI has just released three new realtime voice models that it says will “unlock a new class of voice apps for developers.” Each new voice intelligence model has a unique speciality for different purposes. more…
- OpenAI's new voice model brings GPT-5-level reasoning to real-time conversations
OpenAI is shipping three new voice models—GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper—that can reason in real time, translate across 70+ languages, and transcribe live speech. GPT-Realtime-2 brings reasoning that OpenAI says matches GPT-5. The article OpenAI's new voice model brings GPT-5-level reasoning to real-time conversations appeared first on The Decoder .
- Anthropic’s C.E.O. Says It Could Grow by 80 Times This Year
The chief executive, Dario Amodei, said the rapid growth had exponentially increased the start-up’s need for more computing power.
- Datadog Shares Jump 30% After CEO Says AI is Boosting Sales
Datadog Shares Jump 30% After CEO Says AI is Boosting Sales The Information
- Arm sees ‘explosion’ in CPU demand on AI amid smartphone slump
Arm sees ‘explosion’ in CPU demand on AI amid smartphone slump East Bay Times
- Dexter AI
Autonomous financial research analyst
- Multimodal profiling reveals the centromedian nucleus of thalamus as a dynamic hub orchestrating staged consciousness recovery
Consciousness recovery from general anesthesia represents a fundamental brain state transition. The absence of objective, continuous metrics to monitor the shift from unconsciousness to full awareness has left its dynamic, multidimensional nature unresolved, fracturing consciousness theoretical framework and posing clinical risks due to unreliable assessment of emergence. To address this, we developed a multiscale framework integrating AI-driven behavioral analysis, continuous neuro-physiology recording, and whole-brain functional imaging. This decodes recovery into three hierarchical stages, reflex restitution, level restoration, and content re-establishment, each with unique multimodal signatures. We identified heart rate stabilization as a potential noninvasive biomarker for the restoration of consciousness level. Crucially, the centromedian nucleus of thalamus acts as a dynamic hub, actively orchestrating staged whole-brain reconfiguration via stage-dependent network routing. These
- Image-Conditioned Diffusion for Privacy-Preserving Synthetic Medical Images
Medical imaging models depend on large, shareable datasets, yet privacy constraints limit data dissemination. Current text-conditioned diffusion models fail to preserve subtle, distributed clinical signals, such as continuous physiological biomarkers, rendering synthetic data insufficient for robust downstream physiological modeling. Here, we evaluate image-to-image (I2I) diffusion as a tunable, privacy-preserving transformation that produces a synthetic counterpart of real images while preserving downstream-relevant information. We fine-tune Stable Diffusion with low-rank adapters on retinal fundus photographs and chest radiographs, assessing fidelity, clinical signal preservation, cross-site transfer, and empirical re-identification risk. I2I consistently outperforms text-to-image generation in image fidelity and in preserving biomarker information. In cross-cohort transfer to an external retinal dataset from the UK Biobank, pretraining on I2I synthetic data performs comparably to re
- AI CFD Scientist: Toward Open-Ended Computational Fluid Dynamics Discovery with Physics-Aware AI Agents
Recent LLM-based agents have closed substantial portions of the scientific discovery loop in software-only machine-learning research, in chemistry, and in biology. Extending the same loop to high-fidelity physical simulators is harder, because solver completion does not imply physical validity and m...
- UniSD: Towards a Unified Self-Distillation Framework for Large Language Models
Self-distillation (SD) offers a promising path for adapting large language models (LLMs) without relying on stronger external teachers. However, SD in autoregressive LLMs remains challenging because self-generated trajectories are free-form, correctness is task-dependent, and plausible rationales ca...
- Cross-Modal Navigation with Multi-Agent Reinforcement Learning
Robust embodied navigation relies on complementary sensory cues. However, high-quality and well-aligned multi-modal data is often difficult to obtain in practice. Training a monolithic model is also challenging as rich multi-modal inputs induce complex representations and substantially enlarge the p...
- DINORANKCLIP: DINOv3 Distillation and Injection for Vision-Language Pretraining with High-Order Ranking Consistency
Contrastive language-image pretraining (CLIP) suffers from two structural weaknesses: the symmetric InfoNCE loss discards the relative ordering among unmatched in-batch pairs, and global pooling collapses the visual representation into a semantic bottleneck that is poorly sensitive to fine-grained l...
- Towards Metric-Faithful Neural Graph Matching
Graph Edit Distance (GED) is a fundamental, albeit NP-hard, metric for structural graph similarity. Recent neural graph matching architectures approximate GED by first encoding graphs with a Graph Neural Network (GNN) and then applying either a graph-level regression head or a matching-based alignme...
- NeuroAgent: LLM Agents for Multimodal Neuroimaging Analysis and Research
Multimodal neuroimaging analysis often involves complex, modality-specific preprocessing workflows that require careful configuration, quality control, and coordination across heterogeneous toolchains. Beyond preprocessing, downstream statistical analysis and disease classification commonly require ...
- Improved techniques for fine-tuning flow models via adjoint matching: a deterministic control pipeline
We propose a deterministic adjoint matching framework that formulates human preference alignment for flow-based generative models as an optimal control problem over velocity fields. One can directly regress the control toward a value-gradient-induced target under the current policy, leading to a sim...
- Directional Consistency as a Complementary Optimization Signal: The GONO Framework
We identify and formalize an underexplored phenomenon in deep learning optimization: directional alignment and loss convergence can be decoupled. An optimizer can exhibit near-perfect directional consistency (cc_t -> 1, measured via consecutive gradient cosine similarity) while the loss remains high...
- Coordination Matters: Evaluation of Cooperative Multi-Agent Reinforcement Learning
Cooperative multi-agent reinforcement learning (MARL) benchmarks commonly emphasize aggregate outcomes such as return, success rate, or completion time. While essential, these metrics often fail to reveal how agents coordinate, particularly in settings where agents, tasks, and joint assignment choic...
- Continuous Latent Diffusion Language Model
Large language models have achieved remarkable success under the autoregressive paradigm, yet high-quality text generation need not be tied to a fixed left-to-right order. Existing alternatives still struggle to jointly achieve generation efficiency, scalable representation learning, and effective g...
- On the Implicit Reward Overfitting and the Low-rank Dynamics in RLVR
Recent extensive research has demonstrated that the enhanced reasoning capabilities acquired by models through Reinforcement Learning with Verifiable Rewards (RLVR) are primarily concentrated within the rank-1 components. Predicated on this observation, we employed Periodic Rank-1 Substitution and i...
- Learning to Cut: Reinforcement Learning for Benders Decomposition
Benders decomposition (BD) is a widely used solution approach for solving two-stage stochastic programs arising in real-world decision-making under uncertainty. However, it often suffers from slow convergence as the master problem grows with an increasing number of cuts. In this paper, we propose Re...
- Is One Layer Enough? Understanding Inference Dynamics in Tabular Foundation Models
Transformer-based tabular foundation models (TFMs) dominate small to medium tabular predictive benchmark tasks, yet their inference mechanisms remain largely unexplored. We present the first large-scale mechanistic study of layerwise dynamics in 6 state-of-the-art tabular in-context learning models....
- PACZero: PAC-Private Fine-Tuning of Language Models via Sign Quantization
We introduce PACZero, a family of PAC-private zeroth-order mechanisms for fine-tuning large language models that delivers usable utility at $I(S^*; Y_{1:T})=0$. This privacy regime bounds the membership-inference attack (MIA) posterior success rate at the prior, an MIA-resistance level the DP framew...
- Robotic perturbation proteomics and AI agents enable scalable drug mechanism discovery
Large-scale mass spectrometry-based proteomic screening could reveal cellular mechanisms of drug action at systems resolution but remains limited by experimental complexity and the difficulty of extracting insight from high-dimensional datasets. Here, we describe an end-to-end platform that combines semi-automated sample preparation, rapid LC-MS/MS, and AI agent-based data analysis to enable scalable proteomic screening. In a screen of 172 compounds in HepG2 cells, we generated 1,232 proteomes with more than 8,700 quantified proteins in approximately three weeks. Agentic AI reduced data analysis and interpretation time to less than one day while translating proteomic measurements into structured mechanism-oriented summaries and experimentally testable hypotheses. Guided by this framework, we validated: (1) a cholesterol-lowering effect of methylene blue in vitro and (2) an association between loratadine exposure and increased circulating iron in matched electronic health record analyse
- Virtual brain twins guide personalized treatment decision in schizophrenia
Schizophrenia is a complex psychiatric disorder whose pathophysiology spans multiple spatial and temporal scales. Structural and functional neuroimaging studies have identified a broad range of disease-associated markers encompassing cortical atrophy, white matter disruptions, and aberrant functional connectivity patterns. Their application to personalized diagnosis and treatment selection has remained elusive. Here, we introduce the first Virtual Brain Twin (VBT) pipeline that integrates individual connectome-based network models with multimodal neuroimaging data, incorporating patient-specific structural connectivity, cortical thickness, and resting-state fMRI features to construct personalised whole-brain dynamical models. Dopaminergic and serotonergic signaling pathways are embedded within a mean-field framework, and simulation-based inference (SBI) is used to recover key pathophysiological parameters from individual patient data. The validity of this inference is first established
- ChatIBD: design, safeguards, and early international use of a guideline-grounded generative AI tool for inflammatory bowel disease (IBD) professionals
Objectives To describe the design, operational safeguards, and early use of ChatIBD, a specialty-specific generative AI platform for inflammatory bowel disease (IBD), during its first 6 months of live deployment. Methods ChatIBD is an online question-answering platform that uses retrieval-augmented generation over a curated corpus of IBD guidelines. Queries undergo hybrid semantic and keyword retrieval with query expansion and reranking, and the model is instructed to answer only from retrieved material and return linked citations. Safeguards include fixed medication dosing information from European Medicines Agency (EMA), user feedback capture, and clinician review of flagged outputs. We performed a descriptive service evaluation of aggregated, de-identified platform metrics collected between 1 October 2025 and 1 April 2026. Results During the study period, ChatIBD registered 913 users and processed 7,222 messages across 3,855 conversations. Activity was recorded across 69 countries a
- Artificial Intelligence Driven Support and Self Care Competence as Determinants of Medication Adherence in Diabetes Care, A Cross-sectional Nigerian Study
Medication adherence among patients with diabetes remains suboptimal in low and middle income countries, including Nigeria. Emerging digital health interventions such as AI powered virtual support may be associated with improved adherence behaviours. This study examined self care competence and perceived AI powered virtual support as predictors of medication adherence among patients with diabetes. A cross sectional survey was conducted among 450 patients recruited through multistage sampling across hospitals in Benue State, Nigeria. Standardised measures of self care competence scale, perceived AI support scale, and medication adherence scale were analysed using correlation and regression analyses. Results showed that, self-care competence significantly predicted medication adherence, although some components (glucose management, physical activity, healthcare use) showed negative associations. Perceived AI-powered support demonstrated stronger predictive power, with social presence and
- Conserved neuroectodermal aging encodes primate health and longevity
Neuroectoderm-derived tissues are highly metabolically active and exhibit minimal regenerative turnover, rendering them uniquely vulnerable to age-related stress while preserving undiluted degenerative signals. Yet aging dynamics in these tissues remain elusive in living primates. Here, we introduce an in vivo neuroectodermal aging clock and trace its trajectory in 66,602 human adults and six rhesus macaques across nine health and disease cohorts using an in situ optical biopsy. Through a digital histology atlas integrated with artificial intelligence, we resolve tissue representations of neuroectodermal aging within the human retina, predominantly localized to the metabolically active ganglion and bipolar cell populations and the photoreceptor complex, while demonstrating their evolutionary conservation across primate species. Neuroectodermal aging predicts health and longevity, scales across space and time, and captures preclinical aging signals within and beyond the neuroectodermal
- Toward Early Diagnosis and Therapeutic Discovery in CLN3 Disease: A Computational Biomarker Discovery Framework
Background CLN3 disease, also known as juvenile neuronal ceroid lipofuscinosis, is a rare and neurodegenerative disorder characterized by the accumulation of lipopigments in the cells, progressive cognitive decline, seizures, and vision loss. Biomarker discovery in CLN3 disease is essential for enabling early and accurate diagnosis, which is critical given its neurodegenerative course. Biomarkers provide objective measures to track disease progression, stratify patients, and serve as surrogate endpoints in clinical trials, thereby accelerating therapeutic development. They also offer valuable insights into underlying disease mechanisms and treatment response, ultimately advancing individualized medicine and improving clinical outcomes. Methods We developed various machine learning models to predict potential protein biomarkers in CLN3 disease using proteomics data and laboratory tests collected from participants in a prospective, observational cohort. To prioritize and evaluate these c
- Generating synthetic tau-PET scans in Alzheimer's disease from MRI, blood biomarkers and demographics with deep learning
Tau protein aggregation in the brain is a hallmark of Alzheimer's disease (AD). Positron emission tomography (PET) is the only in vivo method to visualize tau pathology and estimate both its burden and regional distribution, but the use of tau-PET is constrained by high cost and limited accessibility. Here, we develop a deep learning model to synthesize tau-PET scans from more accessible data: structural magnetic resonance imaging (MRI), demographics, and when available, blood biomarkers. We included 5,191 participants across the AD continuum or with another neurological disorder from 13 cohorts (mean age 70 years, 51% female) and optimized a 3D U-Net neural network with residual and attention units for this task. In held-out test data, synthetic tau-PET reliably modeled tau burden, with correlations of R=0.77-0.86 with true tau-PET across individuals in common AD regions of interest. Spatial similarity between synthetic and true tau-PET was likewise high, with mean regional correlatio
- Natural Language Autoencoders: Turning Claude’s thoughts into text
Natural Language Autoencoders: Turning Claude’s thoughts into text
- Financial stability risks are rising as AI fuels cyber-attacks, IMF warns; oil below $100 on Iran peace hopes – as it happened
Rolling coverage of the latest economic and financial news Climate campaigners attack Shell over ‘windfall’ profits from Iran war The Danish shipping giant Maersk has maintained its profit guidance for the year, even as it reported a spike in fuel costs and warned that traffic through the strait of Hormuz “remains at a near standstill”. The company, which transports goods around the world via sea, road, rail and air, said demand for shipping containers remained strong, but that war in the Middle East was ramping up costs. “The reopening of the strait of Hormuz, whether it happens in the days to come or the months to come, will have limited impact on cargo flows. What really are the most important factors to consider: first is our ability to mitigate the cost increases we have been suddenly faced with. And I would say so far we have been successful with both our cost measures and the revenue, the commercial measures that we have put in place to mitigate the impact of these increases to
- 'You would be able to say to it: 'Make a better version of yourself.' And it just goes off and does that completely autonomously': Anthropic Co-Founder on our wild 'recursive' AI future
Anthropic warns of AI capable of 'recursive self-improvement,' and we'd better prepare for this scary, near future.
- Memory Efficient Full-gradient Attacks (MEFA) Framework for Adversarial Defense Evaluations
This work studies the robust evaluation of iterative stochastic purification defenses under white-box adversarial attacks. Our key technical insight is that gradient checkpointing makes exact end-to-end gradient computation through long purification trajectories practical by trading additional recom...
- SwiftI2V: Efficient High-Resolution Image-to-Video Generation via Conditional Segment-wise Generation
High-resolution image-to-video (I2V) generation aims to synthesize realistic temporal dynamics while preserving fine-grained appearance details of the input image. At 2K resolution, it becomes extremely challenging, and existing solutions suffer from various weaknesses: 1) end-to-end models are ofte...
- NavOne: One-Step Global Planning for Vision-Language Navigation on Top-Down Maps
Existing Vision-Language Navigation (VLN) methods typically adopt an egocentric, step-by-step paradigm, which struggles with error accumulation and limits efficiency. While recent approaches attempt to leverage pre-built environment maps, they often rely on incrementally updating memory graphs or sc...
- Render, Don't Decode: Weight-Space World Models with Latent Structural Disentanglement
Training world models on vast quantities of unlabelled videos is a critical step toward fully autonomous intelligence. However, the prevailing paradigm of encoding raw pixels into opaque latent spaces and relying on heavy decoders for reconstruction leaves these models computationally expensive and ...
- Eulerian Motion Guidance: Robust Image Animation via Bidirectional Geometric Consistency
Recent advancements in image animation have utilized diffusion models to breathe life into static images. However, existing controllable frameworks typically rely on Lagrangian motion guidance, where optical flow is estimated relative to the initial frame. This paper revisits the same optical-flow p...
- Zyka.ai
35+ AI media models - API, SDK + studio under one platform
- Sol
The AI Mental Health Therapist
- EU Nations Approve Deal to Roll Back AI Restrictions
The provisional agreement is awaiting formal endorsement from the European Parliament.
- Verifier-Backed Hard Problem Generation for Mathematical Reasoning
Large Language Models (LLMs) demonstrate strong capabilities for solving scientific and mathematical problems, yet they struggle to produce valid, challenging, and novel problems - an essential component for advancing LLM training and enabling autonomous scientific research. Existing problem generat...
- Optimizer-Model Consistency: Full Finetuning with the Same Optimizer as Pretraining Forgets Less
Optimizers play an important role in both pretraining and finetuning stages when training large language models (LLMs). In this paper, we present an observation that full finetuning with the same optimizer as in pretraining achieves a better learning-forgetting tradeoff, i.e., forgetting less while ...
- When No Benchmark Exists: Validating Comparative LLM Safety Scoring Without Ground-Truth Labels
Many deployments must compare candidate language models for safety before a labeled benchmark exists for the relevant language, sector, or regulatory regime. We formalize this setting as benchmarkless comparative safety scoring and specify the contract under which a scenario-based audit can be inter...
- Are We Making Progress in Multimodal Domain Generalization? A Comprehensive Benchmark Study
Despite the growing popularity of Multimodal Domain Generalization (MMDG) for enhancing model robustness, it remains unclear whether reported performance gains reflect genuine algorithmic progress or are artifacts of inconsistent evaluation protocols. Current research is fragmented, with studies var...
- PawModel
AI portraits and reels that actually look like your pet
- Marketing Stack
Your automatic AI marketing team, running every day.
- Runflow
Ship AI image features. Without the complexity.
- AgentLayers
The trust layer for the agentic economy
- Oplif
AI platform to open and run any hospitality business
- CodeWhispr
Paste any code → get a plain-English explanation instantly