All Updates

Page 380 of 888

March 27, 2026

📊
Bloomberg Technology34d ago

AI Reshaping the Battlefield

AI is fueling a global arms race, with algorithms, sensors, and autonomous systems redefining warfare. Bloomberg Tech: Asia episode examines AI's impact on geopolitics and raises urgent ethical questions about military applications.

#military-ai#geopolitics#autonomous-systems
🐯
虎嗅34d ago

Apple Laptops Excel in Hidden Value

Apple notebooks seem expensive but deliver superior value via right comparisons: MacBook Pro undercuts pro workstations plus displays, while MacBook Air beats ultrabooks like LG Gram in battery and portability. High margins persist because they target specific needs effectively. Users without matching demands shouldn't buy.

#apple-strategy#laptop-value#price-anchoring
🔥
36氪34d ago

Zhiyun Launches AI-Satellite Tech Firm

Shenzhen Huida Zhiyun Technology Co., Ltd. was recently established with 15M RMB capital and Feng Bin as legal rep. Scope covers AI application/theory software dev, IoT manufacturing, satellite integration, semiconductors. Fully owned by Dalian Zhiyun Automation Equipment.

#ai-company#satellite#iot
🏠
IT之家34d ago

China Launches Xiangshan RISC-V Chip Stack & Ruyi OS

Chinese Academy of Sciences released Xiangshan, the world's only fully open-source 5th-gen RISC-V high-performance processor system, and Ruyi native OS. Xiangshan covers core to tools for custom high-end chips. Ruyi ensures software compatibility and RVA23 performance support.

#open-processor#chip-ecosystem#china-rd
⚛️
量子位34d ago

QoderWork Integrates DingTalk, WeChat, Feishu

Alibaba's QoderWork now connects with DingTalk, WeChat, and Feishu for seamless AI access. Mobile users can summon desktop AI assistants directly to handle remote tasks. Capabilities include file organization, data processing, and document generation.

#ai-assistant#cross-platform#remote-productivity
🔥
36氪34d ago

Meituan Open-Sources Multimodal LongCat-Next

Meituan released and fully open-sourced its native multimodal large model LongCat-Next and core component dNaViT visual tokenizer on March 27. It unifies image, speech, and text into discrete tokens, breaking language-centric architectures. The model uses pure Next Token Prediction for native vision and speech handling.

#multimodal#open-source#tokenizer
🗾
ITmedia AI+ (日本)34d ago

GitHub Copilot to Train on User Data

GitHub will use input/output data from GitHub Copilot for AI model training in individual plans. An opt-out option is available for users. This policy change aims to improve the tool's capabilities.

#data-usage#privacy-policy#opt-out
⚛️
量子位34d ago

Qiu Xipeng Startup Secures Major Angel Funding

Qiu Xipeng's new company announces hundreds of millions RMB in angel funding. Focuses on evolving from tokens to 'ciyuan' as base units in multimodal AI era. Positions as foundational models and interaction gateways for all-modal AI.

#funding#multimodal#tokenization
🔥
36氪34d ago

Musk Announces Major Grok Imagine Release

Elon Musk stated on X on March 26 that Grok Imagine will have a major release next week. This teases significant updates to xAI's image generation feature within Grok.

#xai#image-generation#launch
⚛️
量子位34d ago

AI Coding Deployment Proves Major Pain Point

AI coding tools lower entry barriers but struggle with deployment complexities. 'Capacities' highlights frustrations in deployment phase. Industry shifts to automating deployments in AI programming's next phase.

#deployment#automation#ai-devops
📰
New York Times Technology34d ago

Helium Shortage Threatens AI Chips

A third of global helium supply is offline due to the war in Iran, creating a bottleneck for the chip industry. This shortage directly impacts critical AI chip makers. Gas companies are racing to ensure no production disruptions.

#helium-shortage#supply-chain#semiconductors
📱
Ifanr (爱范儿)34d ago

vivo: Make AI Understand World

vivo President Hu Baishan states in an interview that AI is already smart but needs to truly understand the world. vivo plans to drive this advancement. The piece explores the company's direction over the next decade.

#ai-strategy#perception-ai#interview
📄
ArXiv AI34d ago

Standardized Benchmarks Bridge MOS Evaluation Gap

Researchers launch the first comprehensive benchmark suite for multi-objective search (MOS) to fix fragmented evaluations. It covers road networks, synthetic graphs, game grids, and robotic roadmaps with fixed instances and reference Pareto sets. This enables robust, reproducible comparisons across correlated to independent objectives.

#benchmarks#optimization#robotics
📄
ArXiv AI34d ago

RL Environments: Pixels to Semantic Agents

This empirical study analyzes over 2,000 RL publications to map the evolution from physical simulations to LLM-driven agents. It proposes a multi-dimensional taxonomy revealing a bifurcation into 'Semantic Prior' (LLM-dominated) and 'Domain-Specific Generalization' ecosystems. The work uncovers cognitive fingerprints and offers a roadmap for next-gen embodied simulators.

#environments#taxonomy#embodied-ai
🗾
ITmedia AI+ (日本)34d ago

NIST Launches AI Agent Standards Initiative

NIST has launched the AI Agent Standards Initiative to promote standardization. It focuses on ensuring reliability and interoperability for autonomous AI agents. This addresses barriers beyond mere convenience for widespread adoption.

#standardization#ai-agents#interoperability
📄
ArXiv AI34d ago

Multi-Agent Boosts Medical AI Calibration

A multi-agent framework uses four specialist agents (respiratory, cardiology, neurology, gastroenterology) based on Qwen2.5-7B-Instruct for medical MCQA. It employs two-phase verification for consistency and S-score weighted fusion to improve calibration and discrimination. Evaluations on MedQA-USMLE and MedMCQA show 49-74% ECE reduction.

#multi-agent#medical-qa
🗾
ITmedia AI+ (日本)34d ago

Microsoft Agent Framework RC for Type-Safe AI

Microsoft released the release candidate of Agent Framework for AI agent development. It enables type-safe code invocation and supports both .NET and Python. Formal release preparation is complete.

#ai-framework#type-safety#agent-development
📄
ArXiv AI34d ago

MAPUS: LLM Agents for Fair Urban Sensing

MAPUS is an LLM-based multi-agent framework for personalized and fair participatory urban sensing. It models participants as autonomous agents with profiles and schedules, coordinated by a fairness-aware agent via language negotiation. Experiments on real datasets demonstrate competitive coverage with improved satisfaction and fairness.

#multi-agent#urban-sensing#fairness
📄
ArXiv AI34d ago

INF: Normal Form for Self-Reference

Introduces incongruent normal form (INF) to represent self-referential sentences as finite families of locally satisfiable but jointly unsatisfiable non-self-referential sentences. Demonstrates that incongruence preserves semantic informativeness despite completeness precluding it, and links incompleteness to incompatible theory extensions. Develops a quantitative Fourier-analytic framework modeling semantic energy and deriving uncertainty bounds.

#self-reference#liar-paradox#model-theory
📄
ArXiv AI34d ago

EMoT: Bio-Inspired LLM Reasoning with Dormancy

EMoT introduces a bio-inspired hierarchical reasoning framework for LLMs with four levels (Micro, Meso, Macro, Meta), strategic dormancy, and mnemonic Memory Palace encoding. It matches CoT performance on complex tasks with better stability and cross-domain synthesis but underperforms on simple problems due to overthinking and 33x compute cost. Ablations confirm dormancy's critical role.

#prompting#mnemonic-encoding#dormancy
Page 380 of 888