The day's stories
Product LaunchChatGPT's market share slips below 50% for first timeChatGPT has experienced a notable decline in market share, dropping below the 50% threshold for the first time, now sitting at 46.4%. This shift is attributed tTechCrunch AIJun 16· 60%FundingMalaysia's AI agent-powered messaging app Respond.io raises $62.5M, eyes acquisitionsRespond.io, a Malaysian startup specializing in AI agent-powered messaging solutions, has successfully raised $62.5 million in a Series B funding round. This inTechCrunch AIJun 16· 59%Product LaunchAndroid 17 launches with new multitasking tools as Google expands Gemini featuresGoogle has officially launched Android 17 along with Wear OS 7, marking a significant update that introduces enhanced multitasking tools and expanded AI featureTechCrunch AIJun 16· 59%FundingProbably raises $9M to build a more reliable kind of AIProbably, a startup focused on enhancing the reliability of AI systems, has successfully raised $9 million in seed funding. The company aims to address the commTechCrunch AIJun 16· 58%Product LaunchYear of free HPE software a “step in the correct direction” in VMware rivalryHewlett Packard Enterprise (HPE) has announced a significant promotional initiative by offering its virtualization software, 'HPE Morpheus Software—VM EssentialArs Technica AIJun 16· 58%Product LaunchThe new Siri makes one of Apple's most convenient OS features a cumbersome messApple's recent update to Siri, unveiled at WWDC, aimed to enhance user interaction by introducing AI capabilities that allow for more conversational exchanges aThe Register AIJun 16· 58%SafetyCritical Copilot vulnerability allowed hackers to seal 2FA code from usersA critical vulnerability in Microsoft's M365 Copilot AI platform has been identified, allowing hackers to extract two-factor authentication (2FA) codes and sensArs Technica AIJun 16· 58%FundingSpaceX passes Amazon as valuation balloons to $2.7TSpaceX has recently achieved a remarkable milestone by surpassing Amazon in valuation, now standing at $2.7 trillion, making it the fifth-most valuable company TechCrunch AIJun 16· 58%OtherSixty percent of US consumers say 'AI' in brand messaging is a turnoff, survey findsA recent survey has revealed that a significant 60% of U.S. consumers are turned off by brands that incorporate 'AI' into their messaging. This finding underscoTechCrunch AIJun 16· 57%Product LaunchPlaud says its software business topped $100M in ARR after shipping over 2M AI notetakersPlaud has announced a significant milestone in its business, achieving over $100 million in annual recurring revenue (ARR) after successfully shipping more thanTechCrunch AIJun 16· 57%OtherAnthropic's latest feud with the Trump admin may actually help it, sales data suggestsAnthropic has recently made headlines by surpassing OpenAI in business spending market share for the first time, achieving a remarkable $65 billion in funding aTechCrunch AIJun 16· 57%OtherRobinhood's note on 10% layoffs shows blaming AI isn't cutting itRobinhood has announced a significant workforce reduction, laying off 10% of its employees, which translates to around 290 positions. Unlike many tech companiesTechCrunch AIJun 16· 56%Product LaunchSpaceX to acquire Cursor for $60B in stock, days after blockbuster IPOSpaceX has made headlines with its announcement of acquiring the AI coding startup Cursor for a staggering $60 billion in stock. This acquisition comes on the hTechCrunch AIJun 16· 54%Open SourcePrologMCP: A Standardized Prolog Tool Interface for LLM AgentsPrologMCP is a newly introduced open-source server that provides a standardized interface for integrating Prolog with large language models (LLMs). This tool isarXiv cs.AIJun 16· 65%ResearchRemember, Don't Re-read: Stateful ReAct Agents for Token-Efficient Autonomous ExperimentationA recent paper introduces a novel stateful ReAct agent designed to optimize autonomous experimentation in machine learning. This agent is capable of retaining earXiv cs.LGJun 16· 64%ResearchQPILOTS: Efficient Test-Time Q-Steering for Flow PoliciesResearchers have introduced QPILOTS, an innovative method designed to enhance the efficiency of test-time Q-steering for flow policies in reinforcement learningarXiv cs.LGJun 16· 64%ResearchPolicy Regret for Embedding Model Routing: Contextual Bandits with Low-Rank ExpertsRecent research has introduced the Hypentropy Policy Gradient (HPG) algorithm, which formalizes the concept of embedding model routing within recommendation sysarXiv cs.LGJun 16· 63%ResearchGRASP: Gradient-Aligned Sequential Parameter Transfer for Memory-Efficient Multi-Source LearningThe recent paper on GRASP presents a novel approach to multi-source transfer learning that addresses the challenges of memory limitations in machine learning moarXiv cs.LGJun 16· 63%ResearchFusion is not one-size-fits-all: Cross-Modal Representation Alignment for Time-to-Event ModelingA recent study published on arXiv introduces a novel framework for cross-modal representation alignment aimed at improving time-to-event modeling in clinical searXiv cs.AIJun 16· 63%BenchmarkOSGuard: A Benchmark for Safety in Computer-Use AgentsResearchers have unveiled OSGuard, a new benchmark suite designed to assess the safety of computer-use agents. This innovative tool evaluates how these agents marXiv cs.AIJun 16· 63%ResearchDr-DCI: Scaling Direct Corpus Interaction via Dynamic Workspace ExpansionResearchers have unveiled a novel framework known as DR-DCI, which aims to enhance Direct Corpus Interaction by implementing dynamic workspace expansion. This iarXiv cs.AIJun 16· 63%ResearchTrust Between AI Agents: Measuring Formation, Breakage, and Recovery, with Implications for Governing Multi-Agent SystemsRecent research has focused on the dynamics of trust between AI agents, particularly in cooperative settings. A new behavioral measure has been introduced to asarXiv cs.AIJun 16· 62%ResearchMetric Match: A Subset Selection Approach to Evaluating LLM Judge ReliabilityA recent study has introduced 'Metric Match', a novel approach aimed at improving the evaluation of large language model (LLM) judges. This method focuses on searXiv cs.AIJun 16· 62%ResearchAI Engram: In Search of Memory Traces in Artificial IntelligenceRecent research has introduced a novel geometric framework aimed at identifying 'AI engrams', which are essentially memory traces within deep neural networks. TarXiv cs.AIJun 16· 62%ResearchSeparable Neural Architectures as Physical World Models: from Mathematical Theory to ApplicationsRecent research has unveiled the Separable Neural Architecture (SNA), a novel approach that integrates neural approximation techniques with tensor decompositionarXiv cs.LGJun 16· 61%BenchmarkBenchmarking Instance-Dependent Label Noise with Controlled CorruptionsA new framework called CILN has been introduced by researchers to address the challenges posed by instance-dependent label noise (IDN) in machine learning. ThisarXiv cs.LGJun 16· 61%ResearchA Comparative Study of Graph Neural Network Layer Selection for Interaction Modelling in Driving Trajectory PredictionA recent study published on arXiv delves into the effectiveness of various graph neural network (GNN) layers for modeling interactions in driving trajectory prearXiv cs.LGJun 16· 61%ResearchGRAPE: Guided Parameter-Space Evolution for Compact Adversarial RobustnessThe recent introduction of GRAPE (Guided Parameter-Space Evolution) presents a novel training framework aimed at enhancing the adversarial robustness of neural arXiv cs.LGJun 16· 61%ResearchLeveraging Physiological Signals to Predict Exam Outcomes with Machine LearningA recent study published on arXiv investigates the potential of using physiological signals to predict academic exam outcomes through machine learning techniquearXiv cs.LGJun 16· 61%ResearchSemantics-Enhanced Retrieval-Augmented Time Series ForecastingA new framework called SERAF has been introduced, which stands for Semantics-Enhanced Retrieval-Augmented Time Series Forecasting. This innovative approach combarXiv cs.AIJun 16· 60%ResearchRelational Structural Causal ModelsRecent research has introduced relational structural causal models, which improve upon traditional causal models by incorporating varying objects and their intearXiv cs.AIJun 16· 60%ResearchA Definition of Good Explanations and the Challenges Explaining LLM OutputsA recent paper published on arXiv delves into the complexities of providing good explanations for outputs generated by large language models (LLMs). It proposesarXiv cs.AIJun 16· 58%