All news
11218
How AI Companies Are Changing the Internet: The Traffic Drain
New data shows Google's AI summaries cutting clicks to websites, while a flood of machine-made content thins out the web and degrades future training data for…
Face recognition AI: accuracy can't be reduced to a single number
Spetslab explains that face identification systems have no single "accuracy" figure: results depend on angle, use case, match threshold and database quality.
AI for Work: How to Run Neural Networks Locally on Your PC
In 2026, Russian users are buying powerful GPUs and NVMe drives in bulk to run and train neural network models at home, without cloud services or foreign providers.
AI Code Generators: Why AI-Written Code Becomes Debt
ChatGPT for Studying: 15 AI Tools for Students in 2026
How to Avoid Data Leakage in Machine Learning: Target Encoding
AI costs: investors demand reports from Amazon, Microsoft and Google
MWS AI and SberAI propose DRAGOn, a dynamic benchmark for evaluating RAG systems
Researchers from MWS AI, SberAI and universities describe DRAGOn, a dynamic RAG evaluation approach where the corpus is refreshed regularly and questions and the…
YouTube Neural Network for Video: How to Create an AI Avatar With Your Voice
YouTube has opened personal AI avatars for Shorts to all creators: a digital double with a cloned voice can now appear in videos in your place.
How AI Coding Agents Work: Memory, Tools and Context
Copilot for work and business: what Microsoft said
AI in software development: how Snowflake built a 24/7 cycle with AI agents
AI Research Agent: How to Use Google Antigravity
Amazon may sell its AI chips to outside companies, and the business is already heading toward $20 billion
Amazon is considering selling its own AI chips to other companies. According to Andy Jassy, its in-house chip unit is already on track for more than $20 billion in…
OpenClaw gets a Go reimplementation: one 35 MB binary instead of 800 MB of dependencies
The popular OpenClaw now has an independent Go version. It runs as a single 35 MB file, needs 3–5 times less memory and simplifies deployment on weak VPS and…