The day's stories
Product LaunchVertu wants CEOs to run companies from an AI foldable starting at $6,880Vertu has introduced the Alphafold, a high-end foldable smartphone designed specifically for CEOs and executives, with a starting price of $6,880. This luxury dTechCrunch AIMay 28· 62%Model ReleaseAnthropic releases Opus 4.8 with new 'dynamic workflow' toolAnthropic has unveiled Opus 4.8, the latest iteration of its AI model, which introduces a new 'dynamic workflow' tool designed to enhance user experience and imTechCrunch AIMay 28· 60%ResearchRSI is the new AGI — and it's just as hard to pin downThe concept of recursive self-improvement (RSI) is emerging as a significant focus within the AI community, paralleling the ongoing discussions around artificiaTechCrunch AIMay 28· 60%InfrastructureJust like gold and oil, we'll soon be able to trade AI token futuresThe financial landscape is evolving with the introduction of AI token futures, a new financial product that allows for trading similar to commodities like gold TechCrunch AIMay 28· 60%InfrastructureThe internet is being rebuilt for machinesAmazon Web Services (AWS) has introduced OpenSearch Serverless, a new cloud infrastructure tailored for AI agents. This innovative service is designed to scale TechCrunch AIMay 28· 59%Product LaunchAsana acquires no-code agent-builder StackAIAsana has announced its acquisition of StackAI for $75 million, a strategic move aimed at bolstering its AI capabilities. StackAI specializes in no-code agent-bTechCrunch AIMay 28· 59%Product LaunchSesame, the conversational AI startup from Oculus founders, launches its iOS appSesame, a new conversational AI startup founded by the creators of Oculus, has officially launched a public preview of its iOS application. This innovative app TechCrunch AIMay 28· 58%FundingHas the hunt for AI compute uncovered the next Cerebras?General Compute has successfully raised $15 million in funding to bolster its AI inference capabilities. The company plans to leverage specialized chips from SaTechCrunch AIMay 28· 57%OtherAt Disrupt 2026: Databricks’ co-founder on what kills enterprise AI dealsDuring the TechCrunch Disrupt 2026 event, Databricks co-founder Arsalan Tavakoli-Shiraji discussed the evolving landscape of enterprise AI. He pointed out a sigTechCrunch AIMay 28· 57%Product LaunchYouTube adds new podcast features, including an AI recommendation tool and 'Auto speed'YouTube has introduced several new features aimed at enhancing the podcast experience for its Premium users. Among the highlights is an AI-powered recommendatioTechCrunch AIMay 28· 57%Product LaunchSneak peek at new Siri app reveals Apple's plans to take on ChatGPT and moreApple is gearing up to launch a revamped Siri app that aims to compete directly with AI chatbots such as ChatGPT. This new version of Siri will be integrated inTechCrunch AIMay 28· 57%ResearchWhy Google's AI can't spell Google (or anything else)Recent observations have revealed that Google's AI struggles with basic spelling tasks, showcasing significant limitations in its understanding of text. DespiteTechCrunch AIMay 28· 56%OtherHow long is Anthropic's lease with SpaceX? Opinions vary.Recent discussions have emerged regarding the lease agreement between Anthropic and SpaceX for the Colossus cluster. Elon Musk stated that the lease is a short-TechCrunch AIMay 28· 56%FundingAnthropic raises $65 Billion, nears $1T valuation ahead of IPOAnthropic has successfully raised $65 billion in its latest funding round, bringing its post-money valuation to approximately $965 billion as it prepares for anTechCrunch AIMay 28· 55%OtherFed up with vibe coders, dev sneaks data-nuking prompt injection into their codeIn a surprising turn of events, a developer has embedded a hidden prompt injection into the jqwik testing app, aimed at sabotaging AI coding agents. This act ofArs Technica AIMay 28· 55%ResearchIdentifying and Understanding Human Values in Text: A Tailorable LLM-based ArchitectureA recent paper introduces an innovative architecture leveraging large language models (LLMs) to identify and quantify human values within text. This new approacarXiv cs.AIMay 28· 64%Model ReleaseSoro: A Lightweight Foundation Model and Chatbot for TajikResearchers have unveiled Soro, a new family of conversational large language models specifically designed for the Tajik language. This innovative model not onlarXiv cs.AIMay 28· 64%ResearchWhy LLMs Fail at Causal Discovery and How Interventional Agents EscapeRecent research has highlighted significant limitations of large language models (LLMs) in the realm of causal discovery. A new approach called Agentic Causal BarXiv cs.AIMay 28· 63%ResearchRULER: Representation-Level Verification of Machine UnlearningA team of researchers has developed RULER, a novel set of metrics aimed at improving the verification process for machine unlearning. Traditional methods have parXiv cs.AIMay 28· 63%ResearchComparative Analysis of Liquid Neural Networks and LSTM for Sequential Pattern Recognition: Robustness, Efficiency, and Clinical UtilityA recent study published on arXiv presents a comparative analysis of Liquid Neural Networks (LNNs) and Long Short-Term Memory (LSTM) networks, focusing on theirarXiv cs.LGMay 28· 63%Research$E^3$-Agent: An Executable and Evolving Agent for Resource Management of Edge Generative InferenceThe $E^3$-Agent has been introduced as a novel solution for managing resources in edge generative inference environments. This agent is designed to tackle the carXiv cs.LGMay 28· 63%ResearchDynaSchedBench: Calibrated Dynamic Scheduling Benchmarks and Observability Paradox in LLM-based Scheduling AgentsA recent study introduced DynaSchedBench, a novel framework designed to improve the evaluation of algorithms addressing the Dynamic Flexible Job Shop SchedulingarXiv cs.AIMay 28· 63%ResearchHEAL: Resilient and Self-* Hub-based LearningResearchers have unveiled a new decentralized learning framework known as HEAL, which stands for Resilient and Self-* Hub-based Learning. This innovative approaarXiv cs.LGMay 28· 63%ResearchIGADA-IoT: IoT Sensor Energy Optimization in Wireless Sensor Networks Driven by Automatic Data AugmentationResearchers have introduced IGADA-IoT, a novel framework aimed at optimizing energy consumption in Internet of Things (IoT) sensors within wireless sensor netwoarXiv cs.LGMay 28· 63%ResearchA Simple State Space Model Excels at Multivariate Time Series ClassificationRecent research highlights the effectiveness of a simple structured state space model (SSM) for multivariate time series classification. The study demonstrates arXiv cs.LGMay 28· 62%ResearchLaneRoPE: Positional Encoding for Collaborative Parallel Reasoning and GenerationThe recent introduction of LaneRoPE marks a significant advancement in the field of collaborative reasoning within language models. This innovative method allowarXiv cs.AIMay 28· 62%ResearchDiscovery Agents for Real-Time Analytics: Toward Proactive Insight SystemsA recent study introduced a multi-agent architecture designed for autonomous insight discovery in real-time analytics. This innovative system utilizes technologarXiv cs.AIMay 28· 62%Open SourceAgyn: An Open-Source Platform for AI Agents with Scalable On-Demand Execution, Agent Definition as a Code, and Zero-Trust AccessAgyn has been launched as an innovative open-source platform designed to facilitate the scalable execution of AI agents. This platform leverages a serverless ruarXiv cs.AIMay 28· 61%ResearchArchitecture-driven Shift: towards a lightweight selector for capturing the trends of logit shiftRecent research has introduced the Architecture-driven Shift (ADS) framework, which aims to enhance model selection in Continual Learning (CL) scenarios. This iarXiv cs.LGMay 28· 61%ResearchMetric-Aware PCA as a Linear Instance of Geometric Deep LearningA recent paper has introduced Metric-Aware Principal Component Analysis (MAPCA), presenting it as a linear instance of geometric deep learning. This innovative arXiv cs.LGMay 28· 61%ResearchDetect by Yourself: Self-Designing Agentic Workflows for Few-Shot Graph Anomaly DetectionA recent paper introduces SignGAD, a new framework aimed at improving few-shot graph anomaly detection. Traditional methods often rely on fixed detection pipeliarXiv cs.LGMay 28· 61%ResearchPersonalized Observation Normalization for Federated Reinforcement Learning in Simulation Environments with HeterogeneityA recent paper has introduced a novel method called Personalized Observation Normalization (PON) aimed at enhancing federated reinforcement learning (FedRL) in arXiv cs.LGMay 28· 60%ResearchOn the Origin of Synthetic Information by Means of Steganographic InheritanceA recent study published on arXiv introduces a novel steganographic mechanism aimed at tracing the lineage of synthetic information generated by advanced AI modarXiv cs.AIMay 28· 60%