Habr AI
AI news source. Articles are auto-selected and adapted by Hamidun News editors.
Latest publications
Gazprombank.Tech showed how AI personalization in debt collection increased debt recovery by 25%
Gazprombank.Tech rebuilt overdue debt collection: it began receiving debt data immediately after the triggering event and used uplift models to choose the best communication channel, raising the settlement rate by 25%.
AI generated a chip design with a queue that in reality would have grown to the size of a lot
An AI-generated hardware design showed that blind faith in tests is dangerous: the model hid a giant queue in the circuit that breaks down even with a small increase in load.
Sber and Yandex Practicum expert explained where AI helps in Data Science
Sber's head of data analytics and a Yandex Practicum expert described which Data Science tasks are already practical to hand over to AI and where manual verification and domain expertise are still necessary.
StudyAI showed that AI ideas appeared long before computers — from Heron to Geniac
StudyAI recalled that the idea of an intelligent machine was born long before neural networks: ancient automata, the Antikythera mechanism, da Vinci's robot, Geniac, and the Soviet «Umka» make this clear.
Yandex Cloud explained how to deploy AI agents in information security without false blocks or disruptions
Yandex Cloud showed why AI agents in a SOC should start in the assistant role: without guardrails and HITL, even useful automation can halt a domain and business processes.
IFR: AI, tactile systems, and digital twins will accelerate robotics in 2026
The global robotics market is entering a new growth cycle: robots are becoming smarter, more versatile, and safer, while Russia is accelerating deployment and expanding local production.
ChatGPT and Gemini in QA training: the experience of a Stepik course author for beginners
A Stepik testing course author showed how he uses ChatGPT, Gemini, RAG, and AI-based assignment checking to make learning more practical without abandoning manual proofreading and classic formats.
A marketer built Znich in five months — AI analytics for YouTube video creators
A marketer without an engineering background built Znich, an AI tool for YouTube analytics, in five months — to automatically understand why one video takes off and another does not.
Habr AI Explained Why the 'Write Like a Human' Prompt Doesn't Save AI Texts from Clichés
Habr AI analyzed the popular advice 'write like a human' and concluded that the problem with AI-generated texts isn't in individual words, but in the incorrect choice of style, task, and authorial perspective.
Open WebUI and llama.cpp: how to set up local web search with Qwen and Docker on Windows
The author of a practical guide showed how to turn the Open WebUI and llama.cpp pairing from weak web search into a working local stack with Qwen, Tika, Brave Search, and Docker on Windows.
Habr AI showed how to build a brand platform for a startup in 8 hours without a designer
Habr AI showed how a single-page brand platform helps a startup even before the MVP: instead of ad hoc generations, it provides a system for landing pages, screens, and creatives.
OpenAI releases GPT-5.4: betting on computer agents, not just code
OpenAI unveiled GPT-5.4 with native computer use, tool search and long context: the model is noticeably stronger on agent tasks, but still makes mistakes in visual layout.
AI startups in 2026 shift from a single prompt to multi-agent pipelines
The era of one-prompt AI services is ending: to deliver consistent results, startups in 2026 are moving to multi-agent pipelines with memory, checks, and orchestration.
Why AI texts annoy readers: a Habr author analyzed reactions to AI writing style
A Habr author ran an experiment using AI to edit their own texts and showed that audiences often react not to the ideas, but to the recognizable “sterile” form of AI writing.
LLMs and agent systems are displacing ROS from the center of robotics — why this matters
Robot architecture is shifting from rigidly scripted ROS scenarios to a combination of LLMs and agent frameworks, where the model plans actions and tools execute them in the real world.
How Del Complex, a fake floating AI data center with H100, fooled X and major tech media
The Del Complex story showed how, in 2023, a polished website, AI-generated images, and bold promises of a sea-based data center with 10,000 H100 misled X and tech media.
Habr AI showed how to prepare structured input for an AI agent instead of a raw technical spec
Habr AI explained why it is better to give agents not an entire document but atomic requirements with JSON attributes: this makes it easier to spot vagueness, scope boundaries, and contradictions.
ETH Strategy: Parallel AI and AskSurf gave opposite assessments of the same DeFi project
A comparison of two AI audits of ETH Strategy showed that even strong models diverge on basic conclusions: one found serious risks, while the other missed them and misidentified the GitHub repository.
X5 Tech shared how it built an AI service for international exam preparation in 7 days
At AI Talent Camp, the ExamLab Bot team built in a week a service that creates personalized exam prep plans and cuts a teacher's routine work from 3–4 hours to a couple of minutes.
Claude turned into a personal medical analyst with access to sleep and lab results
The author connected Claude to Notion, Oura, a Telegram nutrition bot, and scale data so the AI could correlate sleep, heart rate, food, weight, and blood tests in real time.
OpenAI explained how Codex is changing development in a world where code is increasingly written by agents
OpenAI described an experiment in which an internal product was built almost entirely through Codex: people defined rules, checked invariants, and gradually increased the agents' autonomy.
Claude Code and Codex compared on a real-world task: Claude is stronger in RAG, Codex saves tokens
A broad comparison of Claude Code and Codex shows that choosing a coding agent depends not on demo speed, but on task type, output quality, ecosystem, and price.
Claude Code helped build a production Elixir application without hand-written code in four months
The case showed how Claude Code was used over four months to build a production Elixir service with 1,702 commits, 3,880 tests, and two painful but useful production incidents.
Veai 5.6 for JetBrains IDEs adds commit message generation and manual Skills launch
Veai has released version 5.6 for JetBrains IDEs: the agent now suggests a commit message from the diff, offers quick actions in chat, supports manual Skills launch, and makes image attachment easier.
Habr AI showed how to build a Linear Layer in C++ and CUDA as part of the From MNIST to Transformer series
In the third part of the From MNIST to Transformer series, Habr AI shows how to move from matrices to tensors, implement a Linear Layer in C++ and CUDA, and build a network for MNIST classification.
Selectel added high-end servers for model training and expanded its image catalog
Selectel summarized its February updates: it launched HGX B300 servers for training large models, added new images, brought L4 GPUs to the cloud, and improved Kubernetes scaling.
DEMIURGOS adds a rules debugging mode for AI agents and support for 20+ tools
DEMIURGOS lets teams describe the stack, restrictions, and project standards once so Claude Code, Cursor, Copilot, and other agents automatically work under the same rules.
Whisper and Gemma 3 linked with contrastive learning for low-cost speech input to LLMs
A developer showed how they linked Whisper and Gemma 3 through an MLP projector and contrastive learning: after a series of failed regularization attempts, the model cut WER to 35% and began recognizing speech consistent
An open-source stack of 6 models and 9 agents showed how to build an AI team on a single server
Six open-source models and nine agents on a single GPU server form an autonomous AI team that designs, writes, tests, and deploys new agents on its own.
Gamma, Tome and SlidesAI: Which AI Services Actually Speed Up Presentation Creation
AI services for presentations handle structure, text and basic design, but only truly save time when the author provides a precise brief and manually finalizes the result.