The day's stories
Product LaunchApple plays catch-up at WWDCAt the recent Worldwide Developers Conference (WWDC), Apple took a significant step to address previous criticisms regarding its software offerings. Instead of TechCrunch AIJun 8· 59%ResearchThe crash that vanished: control and emergence in a five-model economyRecent research conducted by a team of scientists has explored the dynamics of a market economy through a simulation involving five different AI models. The finHugging Face BlogJun 8· 59%Product LaunchAmazon now lets you design custom merch using AIAmazon has introduced a new feature in its Shopping app that allows users to design custom merchandise using AI technology. This innovative tool enables users tTechCrunch AIJun 8· 59%Product LaunchWWDC 2026: Everything announced on Siri AI, iOS 27, Apple Intelligence and moreAt the WWDC 2026 event, Apple unveiled major updates to its Siri AI, iOS 27, and the broader Apple Intelligence ecosystem. The enhancements to Siri focus on delTechCrunch AIJun 9· 58%FundingOpenAI files confidentially for IPO, following AnthropicOpenAI has taken a significant step towards going public by filing confidentially for an Initial Public Offering (IPO). This move comes shortly after Anthropic,TechCrunch AIJun 8· 58%SecurityFor the 2nd time in weeks, Microsoft packages laced with credential stealerMicrosoft has faced another security incident involving its open source packages, with 73 packages found to contain credential-stealing code. This marks the secArs Technica AIJun 8· 58%Product LaunchApple bets cheaper AI will woo small developersAt the recent Worldwide Developers Conference, Apple unveiled a new initiative aimed at reducing AI infrastructure costs for small developers, particularly thosTechCrunch AIJun 8· 57%Product LaunchApple's long-awaited AI Siri overhaul is finally hereAt the recent WWDC 2026, Apple announced a major overhaul of its Siri assistant, marking a significant evolution in its capabilities. The revamped Siri is desigTechCrunch AIJun 8· 57%Product LaunchApple will let you build workflows using AI in its new Shortcuts appAt WWDC 2026, Apple unveiled significant updates to its Shortcuts app, integrating AI capabilities that enhance user experience by allowing the creation of workTechCrunch AIJun 8· 57%Product LaunchApple's WWDC AI demos looked more real after $250M false ad settlementAt Apple's 2026 WWDC, the company unveiled significant enhancements to its AI offerings, particularly focusing on a revamped version of Siri. This year's presenTechCrunch AIJun 8· 56%Product LaunchApple's Image Playground doesn't suck anymoreAt WWDC 2026, Apple unveiled major enhancements to its Image Playground app, focusing on its AI image generation capabilities. The updated application now allowTechCrunch AIJun 8· 55%FundingAs OpenAI files for IPO, Sam Altman's eye-scanning company is doing layoffs, report saysOpenAI has confidentially filed for an initial public offering (IPO), signaling a major step towards becoming a publicly traded company. This move could attractTechCrunch AIJun 8· 55%Product LaunchApple's Photos app is getting new AI editing featuresAt the recent WWDC 2026, Apple unveiled exciting new AI editing features for its Photos app, significantly enhancing user capabilities in image manipulation. AmTechCrunch AIJun 8· 55%Product LaunchApple just taught your iPhone to finish your sentences, your photos, and your workflowsAt the recent WWDC 2026, Apple unveiled a series of AI-powered updates designed to enhance user experience across its ecosystem. Key features include AI-driven TechCrunch AIJun 8· 54%ResearchImport AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racingRecent research has unveiled two significant advancements in the field of AI. First, a team from Kings College London has developed 'SocioHack,' a benchmark desImport AIJun 8· 69%BenchmarkMacArena: Benchmarking Computer Use Agents on an Online macOS EnvironmentMacArena has been introduced as a benchmarking tool specifically designed for computer-use agents (CUAs) operating within an online macOS environment. This toolarXiv cs.LGJun 8· 65%ResearchElmes*: Automated Construction of Fine-Grained Evaluation Rubrics for Large Language Models in Long-Tail Educational ScenariosA recent study has introduced Elmes*, a novel framework designed to automate the construction of fine-grained evaluation rubrics for large language models (LLMsarXiv cs.LGJun 8· 64%ResearchLean4Agent: Formal Modeling and Verification for Agent Workflow and TrajectoryThe introduction of Lean4Agent marks a significant advancement in the field of formal modeling and verification for agent workflows, particularly in the contextarXiv cs.AIJun 8· 63%ResearchSafeGene: Reusable Adapters for Transferable Safety AlignmentThe introduction of SafeGene marks a significant advancement in the field of AI safety, particularly for open-weight large language models (LLMs). This innovatiarXiv cs.AIJun 8· 63%ResearchDiBS: Diffusion-Informed Branch SelectionResearchers have developed a new method called DiBS (Diffusion-Informed Branch Selection) aimed at enhancing the efficiency of solving Sudoku puzzles. This innoarXiv cs.AIJun 8· 63%ResearchSkip a Layer or Loop It? Learning Program-of-Layers in LLMsRecent research has introduced a groundbreaking method known as program-of-layers (PoLar) for large language models (LLMs). This innovative approach allows for arXiv cs.LGJun 8· 63%ResearchCrowdMath: A Dataset of Crowdsourced Mathematical Research DiscussionsThe introduction of the CrowdMath dataset marks a significant advancement in the evaluation of collaborative mathematical reasoning. This dataset, developed froarXiv cs.AIJun 8· 62%ResearchGaussian Process Latent Factor Regression for Low-Data, High-Dimensional Output ProblemsRecent research has introduced a novel approach known as Gaussian Process Latent Factor Regression (GPLFR), which addresses the challenge of predicting high-dimarXiv cs.LGJun 8· 62%ResearchFAIR-Calib: Frontier-Aware Instability-Reweighted Calibration for Post-Training Quantization of Diffusion Large Language ModelsResearchers have developed a new calibration framework called FAIR-Calib, aimed at improving Post-Training Quantization (PTQ) for Diffusion Large Language ModelarXiv cs.LGJun 8· 62%ResearchMulti-Scale Feature Attention Network for Polymer Classification using THz Dual-Comb SpectroscopyA recent study has introduced a new deep learning architecture known as the Multi-Scale Feature Attention Network (MSFAN), specifically designed for the classifarXiv cs.LGJun 8· 62%ResearchPrinciples and Practice of Deep Representation Learning: or a Mathematical Theory of MemoryA new book titled 'Principles and Practice of Deep Representation Learning' has been released, focusing on the critical area of representation learning within darXiv cs.LGJun 8· 61%ResearchAttack Selection in Agentic AI Control Evaluations Meaningfully Decreases SafetyRecent research published on arXiv highlights the critical impact of attack selection on the safety of AI control evaluations. The study demonstrates that when arXiv cs.AIJun 8· 61%ResearchGenerative Models Erode Human Temporal Learning Through Market SelectionRecent research has revealed that the rise of generative models is significantly impacting Human Temporal Learning (HTL). The study indicates that as these modearXiv cs.LGJun 8· 61%ResearchPosition: Don't Just "Fix it in Post": A Science of AI Must Study Training DynamicsA recent paper published on arXiv advocates for a deeper scientific exploration of AI training dynamics, arguing that current practices often treat AI models asarXiv cs.AIJun 8· 61%ResearchWAV: Multi-Resolution Block Residual Routing for Deep Decoder-Only TransformersA recent paper has introduced WAV v1, a novel multi-resolution block residual routing method designed specifically for decoder-only Transformers. This innovativarXiv cs.LGJun 8· 60%ResearchDetecting and Mitigating Bias by Treating Fairness as a Symmetry OperationRecent research has introduced a novel framework for addressing bias in machine learning systems by conceptualizing fairness as a symmetry operation. This approarXiv cs.AIJun 8· 60%ResearchAre you sure? A Comprehensive and Comprehensible Survey of Uncertainty Quantification in Symbolic RegressionA recent survey published on arXiv provides an in-depth examination of uncertainty quantification (UQ) in the field of symbolic regression (SR). The researchersarXiv cs.LGJun 8· 59%