GPT-5 beats top teams at the Programmer’s World Championship — a new milestone for AI coding
GPT-5 beats top teams at ICPC 2025, solving all 12 tasks and marking a major milestone in AI coding and…
GitHub compiles a “super list” of 500+ working AI projects
GitHub releases a curated list of 500+ real AI projects, enabling developers and startups to integrate ready-made AI solutions across…
Stability AI releases Stable Audio 2.5: AI-generated music now faster and licensed for business
Stability AI releases Stable Audio 2.5 with faster generation and commercial-ready licensing, bringing AI music closer to mainstream business use.
Microsoft opens the doors: publishing apps in the Store now free for individual developers
Microsoft makes Windows Store publishing free for individual developers, removing fees and simplifying global app distribution.
The Talent War Has Reached Interns: In AI, “Adult Money” Is Now Paid Right at the Entry Level
AI companies now pay interns and early researchers up to $18K per month, showing how intense the global competition for…
Google launches AI Edge Gallery: local models now run directly on smartphones
Google launches AI Edge Gallery, enabling Gemma models to run directly on smartphones with on-device AI, offline processing, and improved…
Andrej Karpathy Summed Up the LLM Ecosystem of 2025
Andrej Karpathy’s 2025 LLM review highlights six structural shifts — RLVR, jagged intelligence, context engineering, local agents, vibe coding
“Best Value for Money” Comes to LLMs Too: Xiaomi Launches MiMo-V2-Flash — and It’s Genuinely Fast
Xiaomi introduces MiMo-V2-Flash, a 309B MoE model with 256k context and ultra-low API pricing. With strong SWE-Bench results and 50×…
Google Accelerates Its Move into Smart Glasses and Targets a Commercial Launch of Project Aura in 2026
Google plans to launch Project Aura smart glasses in 2026, combining Android XR with Gemini for real-time, context-aware augmented reality.
Grok “Hallucinates” Far Less Than ChatGPT and Gemini — but What Exactly Was Measured Matters
Claims that Grok hallucinates far less than ChatGPT and Gemini hinge on methodology. Why “hallucination rates” depend entirely on what…