#Code Generation
The Fast Waterfall Trap: Why Detailed Specifications for AI Agents Slow Down Development
Automated code generation using neural networks shifts the bottleneck from writing code to verification and approvals. How to use SDD without the risk of building the wrong system.
Опыт M&T Bank: как масштабная модернизация IT открыла дорогу генеративному ИИ для 15 000 сотрудников
Американский региональный банк M&T Bank развернул корпоративные ИИ-инструменты для большей части своего персонала. Эффект уже заметен на практике: обработка обращений в колл-центре ускорилась на 6 минут на каждый звонок.
Neural Networks Taught to Port Mobile Apps to Aurora OS
The Aurora OS team conducted an experiment using AI agents to port the major enterprise messenger Mattermost to Flutter. The project was successfully built and published in RuStore.
Experiment: Can AI Models Catch All Bugs and Merge External Code into a Flawless Solution?
A practical test demonstrated why developers shouldn't blindly trust a single language model: four LLMs searched for seven bugs in a script, while seven others attempted to assemble a flawless version from external fixes.
Why Accelerating Development with AI Does Not Guarantee Business Growth: Key Takeaways from the MWS Meetup
Experts at the ninth MWS Meetup analyzed the AI integration paradox: localized developer productivity gains often hit systemic constraints and do not automatically accelerate the business as a whole.
OpenAI Ends Model Access for Cursor Following SpaceX Takeover
OpenAI is severing ties with AI coding editor Cursor by mid-November after the startup's acquisition by Elon Musk's SpaceX, citing past contractual disputes.
From Assistants to Virtual Employees: How AI Agents Are Integrated into Corporate Processes
Experts from MWS and Orion Soft discussed the transformation of generative AI into full-fledged digital employees: why context is more important than the ability to write code and how engineering tasks are evolving.
Unified Environment for Six Computers: An Alternative to Claude Code Team Plans
A practical setup for creating a synchronized workspace across three developers and six devices using Claude Code, Syncthing, and Obsidian without paying for team tiers.
Unreal Engine Multiplayer Optimization with AI: Successes, Pitfalls, and the Developer's Real Role
A game developer shared hands-on experience using neural networks for multiplayer network optimization in Unreal Engine: how Claude dramatically sped up development, where AI hit a dead end, and why humans still need to keep architecture under control.
GLM-5 Released: Zhipu AI's Open Model Reaches Top-Tier Performance in Agentic Coding
Zhipu AI and Tsinghua University have introduced GLM-5, a powerful open MoE model with 744 billion parameters capable of autonomously writing code, conducting deep web search, and tackling complex multi-step tasks on par with proprietary flagships.
Anthropic Releases Claude Sonnet 5: Focus on Autonomous Agents, Coding, and Hidden Costs
Anthropic has introduced Claude Sonnet 5, a new language model optimized for multi-step autonomous workflows and coding. Despite attractive base pricing, reasoning patterns and a redesigned tokenizer may make real-world usage more expensive than expected.