#AI News

1841 articles

Peter Smith Peter Smith Industry
2 weeks ago

Robot AI Breakthrough: 4,000-Hour Co-Training Model Cuts Data Costs

Robots trained on 4,000+ hours of co-training data generalize better across tasks without costly data collection.

Saad Ullah Saad Ullah Industry
2 weeks ago

LLM Alignment Reduces Real-World Human Behavior Accuracy by Nearly 10:1

Aligned LLMs diverge from real human decisions. Base models outperform aligned versions 10:1 in strategic scenarios.

Victoria Bazir Victoria Bazir Industry
2 weeks ago

Brain-Controlled Robotics Advances With EEG Breakthrough - Users Now Move Robot Arms Using Thought Alone

EEG and eye tracking let users control robots with thought alone, marking a leap in AI-driven human-machine interaction.

Eseandre Mordi Eseandre Mordi Benchmarks
2 weeks ago

Baidu's VEGA-3D Beats 4 AI Models in 3D Spatial Reasoning Without Geometric Supervision

Baidu's VEGA-3D uses video diffusion models to boost 3D spatial reasoning in multimodal AI - no explicit geometric supervision needed.

Saad Ullah Saad Ullah Trends
2 weeks ago

Ethereum Leads $1T Trust Economy as AI Erases Digital Authenticity

Ethereum powers Web3's trust economy as AI-driven synthetic identities threaten digital authenticity across global networks.

Peter Smith Peter Smith AI models
2 weeks ago

Why Multi-Agent AI Systems With Specialized Roles Are Outperforming Single Models

Specialized multi-agent AI systems are replacing single models, boosting efficiency through coordinated roles and shared memory layers.

Marina Lyubimova Marina Lyubimova Industry
2 weeks ago

12+ Monthly Articles Drive 5.1% Organic Traffic Growth, NP Digital Data Shows

NP Digital survey of 300 marketers reveals publishing 12+ articles monthly delivers 5.1% organic traffic growth.

Saad Ullah Saad Ullah Industry
2 weeks ago

V-Co Cuts Image Generation FID from 15.15 to 2.52 with Dual-Stream Co-Denoising

V-Co diffusion model uses dual-stream co-denoising to cut FID from 15.15 to 2.52 with fewer training epochs.

Peter Smith Peter Smith Industry
2 weeks ago

LEAD Cuts Hallucinations in Multimodal AI With Zero Extra Training

LEAD is a training-free decoding method that reduces hallucinations in multimodal AI by adapting reasoning to uncertainty levels.

Victoria Bazir Victoria Bazir Benchmarks
2 weeks ago

Anthropic Captures 73% of New Enterprise AI Clients, Overtaking OpenAI in 60 Days

Anthropic surpassed OpenAI in enterprise AI adoption, capturing 73% of new clients in just two months.

Eseandre Mordi Eseandre Mordi Industry
2 weeks ago

Kinema4D Unveils 4D Generative Robotics Simulation With Robo4D-200k Dataset

Kinema4D introduces action-conditioned 4D robotic simulator with Robo4D-200k dataset for realistic robot-world interactions.

Peter Smith Peter Smith AI models
2 weeks ago

MiniMax M2.7 Hits 66.6% Medal Rate With Self-Evolving AI Across 22 Competitions

MiniMax M2.7 self-improves across 22 ML competitions, hitting 66.6% medal rate with no human input.

Saad Ullah Saad Ullah AI models
2 weeks ago

Google Open-Sources Colab MCP Server, Giving AI Agents Cloud Code Execution

Google open-sources Colab MCP Server, letting AI agents run cloud code via Claude, Cursor, Codex, and Gemini.

Eseandre Mordi Eseandre Mordi Industry
2 weeks ago

InCoder-32B Enters Industrial AI Code Market with 32B-Parameter Unified Model

InCoder-32B is a 32B-parameter industrial code model unifying chip design, GPU kernels, and embedded systems - outperforming rivals on domain benchmarks.

Marina Lyubimova Marina Lyubimova AI models
2 weeks ago

Google and OpenAI Lead AI Image Race With 60+ Models Ranked

Google and OpenAI dominate Design Arena AI image benchmarks. Gemini 3.1 Flash and ChatGPT models top rankings.

Usman Salis Usman Salis AI models
2 weeks ago

MiniMax M2.7 Launch: 200K Context and 56% SWE-Pro Score Enter the Agent Race

MiniMax M2.7 launches with 200K context and 56.22% SWE-Pro score, targeting AI agent workflows.

Saad Ullah Saad Ullah AI models
2 weeks ago

Cursor AI Cuts Token Use 5x With Self-Summarization RL

Cursor's RL self-summarization beats compaction, scoring 47.9 on CursorBench Hard 80k with 5x fewer tokens.

Marina Lyubimova Marina Lyubimova Benchmarks
2 weeks ago

GPT-5.4 Mini Scores 72.1% on OSWorld as Gemini 3 Flash Hits 90.4% on GPQA Diamond

OpenAI's GPT-5.4 Mini tops execution benchmarks with 72.1% on OSWorld, while Gemini 3 Flash leads reasoning at 90.4% GPQA Diamond.

Peter Smith Peter Smith Benchmarks
2 weeks ago

OpenAI GPT-5.4 Mini Hits 72.1% on OSWorld, Outpaces Rivals in 2025 Benchmark Race

GPT-5.4 mini scores 72.1% on OSWorld-Verified, beating Claude Haiku 4.5's 50.7% in efficiency benchmarks.

Victoria Bazir Victoria Bazir Benchmarks
2 weeks ago

OpenSeeker AI Beats Alibaba on 4 Key Open-Source Search Benchmarks

OpenSeeker open-source search agent tops Alibaba's Tongyi DeepResearch on BrowseComp-ZH using just 11.7K training samples.

Posts feed
OpenAI's $122 Billion Superapp Strategy Signals a Major AI Platform Shift
2 days ago
BitVLA Shrinks Robot AI by 11x While Boosting Speed 4.4x
2 days ago
LFM2.5-350M Debuts With Strong Benchmark Results Across 6 Evaluation Suites
2 days ago
Grok 4.20 Multi-Agent Beta Ranks #7 in Search Arena Leaderboard
2 days ago
Claude Code Leak Reveals 4 Key Features Powering Autonomous AI Systems
2 days ago
RISE Robot Framework Boosts Task Success Rates by 35% Using AI-Driven Imagination
2 days ago
Multi-Agent AI Systems Face 15 Emerging Social Risks, New Study Finds
2 days ago
Mistral AI Raises $830M to Build Paris AI Data Center
2 days ago
OpenAI Codex Adds Claude Code CLI Integration in Cross-Platform Move
2 days ago
AI Coding Boom: Claude Code NPM Downloads Surge 70x, Codex Installs Hit 5M+
2 days ago
Follow us on Twitter

Disclaimer: Aigazine provides content for informational and educational purposes only. We do not offer financial, investment, or legal advice. Any decisions made based on the information published on this website are solely at the reader’s own discretion and risk. We encourage all readers to conduct their own research and seek professional guidance when necessary.

Aigazine is a news platform focused on artificial intelligence, covering global AI trends, technology, and innovation. Aigazine is based in Tbilisi (0179, Georgia, Tbilisi City, Vake District, 49 Besarion Zhghenti Street, VAT 305786600).

© 2025 aigazine.com

We use cookies to improve your experience on our site and to show you relevant advertising. To find our more, read our privacy policy and cookie policy