The day's stories
FundingPrime Intellect raises $130M Series A to help enterprises build their own AI agentsPrime Intellect has successfully secured $130 million in a Series A funding round, achieving a valuation of $1 billion. This significant investment will enable TechCrunch AIJul 8· 60%OtherFormer OpenAI exec Kevin Weil is now on the board of Stoke SpaceKevin Weil, previously an executive at OpenAI, has been appointed to the board of Stoke Space, a startup dedicated to creating reusable rocket technology. StokeTechCrunch AIJul 8· 59%FundingThis startup thinks robotics is about to have its ChatGPT momentGeneral Intuition, a robotics startup, has recently raised $320 million in funding, achieving a valuation of $2.3 billion. The company is focused on developing TechCrunch AIJul 8· 59%Product LaunchHot French startup ZML releases free product to speed inference across lots of AI chipsFrench startup ZML has introduced a new free software product named ZML/LLMD, designed to enhance the inference performance of open-source large language modelsTechCrunch AIJul 8· 59%Product LaunchOpenAI releases new voice models for more natural live conversationsOpenAI has officially launched its new voice models, GPT-Live-1 and GPT-Live-1 mini, which are designed to facilitate more natural and interactive live conversaTechCrunch AIJul 8· 59%SafetyHackers can use 9 of the most popular AI tools to assemble massive botnetsRecent reports indicate that hackers are leveraging vulnerabilities in nine widely-used AI tools to create large-scale botnets. These attacks utilize prompt injArs Technica AIJul 8· 59%Open SourceData for AgentsNVIDIA has released a significant set of open and synthetic datasets aimed at enhancing the development of AI agents. This initiative underscores the importanceHugging Face BlogJul 8· 59%Model ReleaseSpaceXAI releases Grok 4.5, which Elon describes as an 'Opus-class model'SpaceXAI has officially launched Grok 4.5, a new AI model that promises enhanced efficiency and performance. Elon Musk has characterized Grok 4.5 as an 'Opus-clTechCrunch AIJul 8· 59%SafetyGoogle pays $250K for Linux vulnerability allowing guest VM escapesGoogle has recently paid $250,000 for the discovery of a critical Linux vulnerability, identified as CVE-2026-53359. This flaw allows untrusted virtual machinesArs Technica AIJul 8· 58%Product LaunchGoogle Photos adds a new AI 'Video Remix' toolGoogle Photos has launched a new feature called 'Video Remix' that leverages AI technology from its Gemini Omni model. This tool is designed to streamline the vTechCrunch AIJul 8· 57%FundingThese AI startups are growing revenue at faster and faster ratesAI startups are witnessing unprecedented growth in their revenue, with several companies reporting remarkable increases in their annual recurring revenue (ARR).TechCrunch AIJul 8· 57%Product LaunchMeta wants its AI glasses to seem less creepy. Its AI strategy says otherwise.Meta is making strides in the development of its AI glasses, aiming to alleviate privacy concerns that have arisen from their use. A new feature has been introdTechCrunch AIJul 8· 56%Product LaunchTool promises to make lazy academics' AI-written papers sound more humanA new tool called Academic Humanizer has been launched by a startup, aimed at enhancing the readability and human-like quality of AI-generated academic papers. The Register AIJul 8· 56%SafetyGoogle's deepfake detector system used to debunk McConnell hoax picGoogle's SynthID deepfake detection system has proven its effectiveness by successfully debunking a hoax image of Senator Mitch McConnell that circulated on socTechCrunch AIJul 8· 55%FundingWhy this CEO thinks video games make better training data than the internetGeneral Intuition, a startup focused on leveraging video game data for artificial intelligence training, has successfully raised $320 million in funding. The coTechCrunch AIJul 8· 55%FundingAI chip maker SambaNova raises $1B at $11B valuation, 5 months after last mega roundSambaNova Systems, a prominent AI chip manufacturer, has successfully secured $1 billion in a Series F funding round, bringing its total valuation to an impressTechCrunch AIJul 8· 55%ResearchFoundation Models for Automatic CAD GenerationRecent research has introduced LLMForge, a novel framework that leverages foundation models to automate the generation of Computer-Aided Design (CAD) models. TharXiv cs.AIJul 8· 64%BenchmarkCSTutorBench: Benchmarking Small Language Models as Tutors for Block-Based ProgrammingResearchers have developed CSTutorBench, a new benchmarking framework aimed at assessing the effectiveness of small language models (SLMs) as tutors specificallarXiv cs.AIJul 8· 64%ResearchLearning to Control LLM Agent Harnesses with Offline Reinforcement LearningRecent research has introduced a novel method for enhancing large language model (LLM) agents by incorporating a learnable control layer within their execution arXiv cs.LGJul 8· 64%ResearchSynthetic Consumer Insight Generation with Large Language ModelsRecent research has delved into the application of large language models (LLMs) for generating synthetic consumer insights. This study specifically focuses on uarXiv cs.AIJul 8· 63%ResearchGeometry-Aware Infrastructure-Anchored Denoiser for UWB Sensing and Work-Zone ReconstructionA new research framework named GAIA has been developed to enhance ultra-wideband (UWB) sensing capabilities, particularly in the context of work-zone reconstrucarXiv cs.LGJul 8· 63%ResearchExogenous Dropout: A Simple, Strong Baseline for Corruption-Robust Time Series Forecasting with CovariatesA recent paper has introduced a novel training method called 'exogenous dropout' aimed at improving the robustness of time series forecasting models. This approarXiv cs.LGJul 8· 63%ResearchPrompt-to-Paper: Agentic AI System for BioinformaticsA new multi-agent AI framework called Prompt-to-Paper has been introduced to enhance the process of automated manuscript generation in the field of bioinformatiarXiv cs.AIJul 8· 63%Product LaunchArtisanCAD: An Industrial-Level CAD Agent with Expert-Grounded Knowledge DistillationArtisanCAD has been launched as a sophisticated CAD agent designed to streamline the process of generating CAD programs from natural language descriptions. ThisarXiv cs.AIJul 8· 63%ResearchStatistically Meaningful Geometry and Gauge Symmetry Breaking: A Geometric Foundation for Scientific Discovery and Intelligence EmergenceA recent paper introduced a novel framework called Statistically Meaningful Geometry (SMG), which aims to enhance the differentiation between authentic intelligarXiv cs.LGJul 8· 62%ResearchAkashic: A Low-Overhead LLM Inference Service with MemAttentionThe introduction of Akashic marks a significant advancement in the field of large language model (LLM) inference services. Researchers have developed a low-overarXiv cs.AIJul 8· 62%ResearchThe Granularity Paradox: How Temporal Disaggregation Inflates In-Sample Fit and Compounds Out-of-Sample ErrorA recent paper titled 'The Granularity Paradox: How Temporal Disaggregation Inflates In-Sample Fit and Compounds Out-of-Sample Error' explores the complexities arXiv cs.LGJul 8· 62%ResearchDesign-CP: Context Parallelism for Design of Protein NanoparticlesResearchers have introduced a novel method called Design-CP, which significantly improves the design process of protein nanoparticles. By leveraging context-pararXiv cs.LGJul 8· 62%ResearchEmpirical Minimal-Realisation Compression of Deep Neural Networks via Controllability-Observability TestsA recent paper presents a novel approach to compressing deep neural networks by utilizing controllability and observability tests. This framework aims to minimiarXiv cs.LGJul 8· 62%ResearchFrom Graphs to Gradients: Physics-Inspired Structural Attribution for Cyber-Physical IoT Systems and BeyondA new research paper introduces a framework for causal explanation in cyber-physical IoT systems, emphasizing dependency-aware attribution that does not rely onarXiv cs.AIJul 8· 62%ResearchNarrative World Model: Narratology-Grounded Writer Memory for Long-Form FictionResearchers have unveiled the Narrative World Model (NWM), a novel system designed to enhance long-form fiction writing by leveraging a narratology-grounded apparXiv cs.AIJul 8· 62%ResearchFirstResearch: Auditable Question Formation for LLM Scientific Discovery AgentsThe introduction of FirstResearch marks a significant advancement in the field of scientific discovery using large language models (LLMs). This new framework foarXiv cs.AIJul 8· 61%ResearchAdaStop: Cost-Aware Early Stopping for DNN Test SelectionResearchers have introduced AdaStop, a novel framework designed to optimize the early stopping process in deep neural network (DNN) testing. This framework straarXiv cs.LGJul 8· 61%ResearchBreaking Structural Isolation: Scalable Graph Clustering via Community-Aware Sampling and Structural EntropyResearchers have introduced a new framework called SCISE, which stands for Scalable Community-Aware Sampling and Structural Entropy. This innovative approach foarXiv cs.LGJul 8· 61%ResearchLearnable Weighting of Intra-Attribute Distances for Categorical Data Clustering with Nominal and Ordinal AttributesA recent paper published on arXiv introduces a new method for clustering categorical data, specifically focusing on nominal and ordinal attributes. The authors arXiv cs.LGJul 8· 60%