2026

Medicine·Habr AI

Agentic Coding as Addiction: Why Developers Can't Stop

Startup CTOs stay up until 3 a.m. with no deadline, and the Y Combinator CEO brags about 19-hour sessions. UC Berkeley researchers see a gambling pattern.

PromptPilot: a task scheduler for Claude Code and Codex that works while you sleep

A Russian developer built a task scheduler for AI CLIs. PromptPilot takes prompts from a terminal, browser or Telegram bot and runs them on a schedule, retrying…

Medicine·TNW

Anthropic buys a biotech startup for $400 million, and it has fewer than 10 employees

Anthropic has acquired Coefficient Bio, a stealth computational biology startup staffed by former Genentech researchers, paying $400 million in stock.

Creative·TNW

Microsoft launches three MAI AI models without OpenAI, signaling a push for technological independence

Six months after renegotiating its OpenAI contract, Microsoft has released its own MAI models for transcription, voice and images, with no mention of its partner on…

01Nvidia H100 rental prices climb despite Blackwell launch: up 40% in six monthsApr 28, 2026·2 min 02Yandex Code Assistant for VS Code: What Changed in the Extension and What Code Indexing AddsApr 28, 2026·2 min 03OpenClaw flaw let attackers quietly gain admin rights on exposed instancesApr 28, 2026·2 min 04How One Developer Used Claude Code to Launch a GEO Platform for Brands Across Nine AI ModelsApr 28, 2026·2 min 05SpaceX and Blue Origin want to put AI data centers in orbit, but physics says noApr 28, 2026·3 min

Top themes

Medicine·Bloomberg Tech

OpenAI reshuffles its leadership: Brad Lightcap changes roles, Fidji Simo takes medical leave

LLM·MarkTechPost

Google DeepMind lets an LLM rewrite game-theory algorithms and beat the experts

Autonomous·Habr AI

Luminarys AI unveils an AI agent platform with isolated skills and cluster deployment

Creative·3DNews AI

Anthropic: Under Pressure and Given an Impossible Task, Claude May Turn to Deception and Blackmail

LLM·MarkTechPost

Z.AI shows how to build production-ready agent systems on GLM-5 with tool calling

Z.AI has published a detailed GLM-5 tutorial covering SDK setup, the OpenAI-compatible API, streaming, tool calling, JSON output and a multi-turn production agent.

Ethics·Habr AI

OpenClaw on a Xiaomi 11T: turning an old smartphone into a home AI server

An old Xiaomi 11T with 8 GB of RAM now runs OpenClaw as a home AI gateway, answering in Telegram via Termux and OpenRouter, though it needs careful setup and…

Hardware·Habr AI

Samsung expects the memory shortage to end by 2028, and that changes how the AI boom looks

Samsung, the largest memory maker, expects the shortage to ease by 2028. The article's author reads that as a signal: the AI market is preparing not only for growth but also for a possible cooldown.

Creative·MarkTechPost

Netflix open-sources Void, a video object-removal model that accounts for scene physics

Ethics·Habr AI

Raft unveils "AI KOMP-AS" framework for profitable, safe AI adoption

Habr AI: Why Agentic Systems Need New Control and Safety Metrics

Hardware·3DNews AI

Nvidia shows neural texture compression for games, cutting VRAM use nearly sevenfold

Creative·3DNews AI

OpenAI reshuffles its leadership as it prepares for an IPO and weighs ads in ChatGPT

PHP and RubixML Move From Arrays to GPUs: How the Ecosystem's Approach to ML Is Changing

Nvidia puts $2 billion into Marvell and turns a partnership into an infrastructure toll

Creative

Google releases Gemma 4 as Anthropic lands at the center of leaks and research

LLM·TNW

Anthropic bars OpenClaw from using Claude Pro and Max flat-rate plans

From April 4, 2026, Anthropic no longer allows Claude Pro and Max flat-rate subscriptions to power OpenClaw, moving autonomous agents to pay-as-you-go billing.

Why Companies Lose Millions on ChatGPT and AI: Three Major B2B Rollout Mistakes

Companies buy expensive AI tools, but employees dodge bans, switch to personal accounts and make leaks invisible when rollouts are built on control rather than…

Hardware·Bloomberg Tech

Nvidia partner Hon Hai grows quarterly revenue 29.7% on AI server demand

Hon Hai, a key Nvidia manufacturing partner, grew quarterly revenue 29.7% to NT$2.13 trillion, confirming steady demand for AI infrastructure.

Medicine·TNW

Mistral tops the list: Europe's biggest funding rounds from March 30 to April 5

European startups raised money in AI, quantum computing, biotech and fintech from March 30 to April 5, and Mistral sent the market its biggest signal with $830…

01Anthropic's Claude Code: How to Set Up an AI Assistant for Work Without Coding SkillsApr 28, 2026·3 min 02How AI Is Helping Clothing Brands Design the Fashion of the FutureApr 28, 2026·3 min 03Microsoft sells Copilot as a work AI, but its terms call it entertainmentApr 28, 2026·2 min 04Google opens free access to Veo 3.1: 10 video generations a month, no subscription requiredApr 28, 2026·3 min 05Sam Altman and OpenAI Sharply Cut Their AI Infrastructure Spending Plan Through 2030Apr 28, 2026·3 min

Top themes

Why AI Agents Fail in Production: What a Mature Enterprise LLM System Is Made Of

Two AIs Beat One: How an OpenAI Plugin Lets Claude and Codex Argue With Each Other

Business·Habr AI

Dario Amodei vs. Sam Altman: A Decade-Long Feud in the Battle for AI's Future

LLM agents in a real CI/CD setup bypass the rules instead of solving the task properly

AI for the Smart Home: Llama 8B Locally, Real Pitfalls, and How to Do Without the Cloud

A practical breakdown: how to tie Llama 8B, Ollama and Home Assistant into an offline stack, what performance to expect and where deployment traps lie.

Robotics·Guardian

Japan Airlines Brings Humanoid Baggage Robots to Haneda Airport to Ease Labor Shortage

From May, Japan Airlines is testing humanoid robots on baggage handling at Tokyo's Haneda, a first step toward permanent deployment amid an acute labor shortage.

Hardware·3DNews AI

Foxconn's quarterly revenue jumps nearly 30% on demand for Nvidia AI servers

Taiwan's Foxconn reported first-quarter revenue up 29.7% to $66.5 billion, driven by surging demand for server components for Nvidia's AI systems.

Creative·MarkTechPost

How to Build the Netflix Void Pipeline for Removing Objects from Video with CogVideoX

Claude Code and 11 agents: how a QA team automated up to 80% of routine testing

Business·MarkTechPost

Gladstone Institutes unveils MaxToki, an AI model that predicts cell aging

Why LLMs Lie and Forget Facts: How Language Model Memory Works

LLM Invented a Crisis Hotline: Why a Prompt Won't Save You From Hallucinations

T1 Cloud: H200 vs. L40S — a technical look at GPUs for generative AI

NVIDIA Nemotron 3 Super 120B: Testing on Real Analytics Tasks on a Single GPU

LLM

International Bestseller on Large Language Models Now Available in Russian

Ethics·Bloomberg Tech

OpenAI Proposes a Public Wealth Fund and Faster Grid Buildout for the AI Era

OpenAI has published policy recommendations for governments: a public wealth fund, rapid social support programs, and faster construction of power grids.

Creative·3DNews AI

Google releases Edge Eloquent for offline speech transcription without internet

Google quietly released AI Edge Eloquent, an iOS app that turns speech into text on the device itself, with no internet and no subscription. An Android version has…

PSB shows how it deploys AI in the bank: chatbots, RAG and services for business

PSB has detailed how it uses generative AI for small and medium businesses and staff: the Katyusha assistant, RAG consultations, messenger payments and internal…

Ethics·TNW

Trump administration pushes Anthropic to banks while the Pentagon fights the company in court

The US Treasury and the Fed are urging the biggest banks to test Anthropic's Mythos for cyber defense, even though the Pentagon earlier labeled the company a supply-chain risk.

Yandex Cloud explains why the frontend leads AI integration in DataLens at launch