April 2026
Nvidia introduced PivotRL — a framework for AI agents with 4x savings in rollout steps
Nvidia showed PivotRL — an approach to fine-tuning AI agents that preserves quality outside the training domain and achieves comparable accu
Yandex 360 Explained How to Teach AI Assistants to Work with Internal UI Kits
The Yandex 360 team shared how to make a frontend repository understandable to AI — and why without this, the assistant will hallucinate com
Google Introduces TurboQuant: 6x KV-cache Compression for LLMs Without Accuracy Loss
Google Research unveiled TurboQuant — an algorithm that compresses the KV-cache of large language models by at least six times and accelerat
Epoch Biodesign Raises $12M to Scale AI-Powered Nylon Recycling
London-based Epoch Biodesign has secured $12 million to scale enzymes developed with AI that break down nylon 6,6 and return over 90% of fee
Openreach and Google Cloud Deploy AI to Plan Britain's Fiber Optic Network
Openreach strengthened its partnership with Google Cloud: Vertex AI and Gemini Enterprise help plan full-fibre deployment in Britain and red
LinkedIn and SHRM: AI Made Resumes a Weak Signal and Turned Them Into a Hiring Filter
Candidates polish resumes with AI, companies filter them with algorithms, and ultimately the value of a 'perfect' CV drops: case studies, in
MCP LLM-Agent in Fusion360: First Vibe-Design Test on a Wooden Staircase
Fusion360 rolled out an MCP LLM-assistant — a Habr author immediately tested it on a wooden staircase he had previously created manually ove
Claude Code Raised Legal RAG to 0.791, but ARLC 2026 Final Hit Scaling Limits
The ARLC 2026 case demonstrates how Legal RAG can be improved across 17 iterations, then lose 42% of results as the corpus scales from 30 to
OpenAI and Grok Lose to Custom RAG in Legal Agentic RAG Challenge
A team on Mac Studio M3 Ultra compared RAG, CAG, BM25, and built-in databases from OpenAI and Grok in a legal challenge and found: grounding
OpenAI to raise another $10B from investors for data centers and training new models
OpenAI seeks to raise an additional $10B in its current funding round: funds will go toward compute infrastructure, data centers, and traini
Kuaishou Boosts Revenue amid Generative AI Monetization and Kling Platform Growth
Kuaishou reported 12% quarterly revenue growth: generative AI, advertising tools, and the Kling platform began making notable contributions
Agent Commerce per MIT Tech Review: why AI agents require data, not banners
When an AI agent books vacation and makes purchases on its own, e-commerce needs not beautiful interfaces, but precise structured data and u
Habr: How Outdated Knowledge Base Breaks LLM-Agents and How to Fix It
Habr explored why outdated documentation is more dangerous than its absence for LLM-agents, and proposed three practices: automated checks,
Vibe Coding Promises 10x Productivity Gains — PyPI Hasn't Confirmed It
Developers claim 10–100x acceleration thanks to AI tools, but if that's true — why doesn't PyPI show explosive growth in Python packages?
Ocorian: 86% of family offices already use AI for financial data analysis
A global Ocorian study shows that 86% of family offices already apply AI in daily operations and data analysis to quickly obtain financial i
Granola raises $125M at $1.5B valuation in AI note-taking market
Granola, whose AI note-taking service is gaining popularity among Silicon Valley employees, raises $125M from Index Ventures at a $1.5B star
Anthropic Strengthens Claude: Why New Corporate Customers Choose It Over ChatGPT
Ramp data from March 2026 shows a shift in corporate demand: among companies purchasing AI tools for the first time, Anthropic's Claude nota
Auburn University Launches AI-Café for Conversations About AI Without Panic and Tech Jargon
Auburn University professors held open AI-cafés in a coffeehouse where students and residents discussed fears about AI and ways to give soci
Periodic Labs Discusses Funding Round at ~$7B Valuation Following Ex-OpenAI Startup Launch
AI startup Periodic Labs, founded by former OpenAI and DeepMind employees, is in talks to raise hundreds of millions of dollars at a ~$7 bil
NotebookLM and Gemini Automate Test Creation via Google Apps Script and Forms
An author demonstrated how to integrate NotebookLM, Gemini, Google Apps Script, and Google Forms to quickly assemble educational tests from
General Motors showed how it trains autopilot in simulations 50,000 times faster than reality
GM described an autopilot training system where rare road situations are run through high-precision and abstract simulations up to 50,000 ti
AWS explained the launch of reinforcement fine-tuning in Amazon Bedrock via OpenAI-compatible APIs
AWS released a step-by-step breakdown of reinforcement fine-tuning in Amazon Bedrock: with OpenAI-compatible API configuration, Lambda grade
MIT demonstrates computer vision system for counting fish in citizen science projects
MIT Sea Grant and partners trained a deep learning system to count river herring from underwater video, to supplement volunteer monitoring a
Kleiner Perkins Raises $3.5 Billion for AI Startup Bets and Portfolio Support
Venture capital fund Kleiner Perkins raised $3.5 billion to invest in new AI startups and support existing portfolio companies in autonomy,
Google's TurboQuant Algorithm Crashes Memory Manufacturer Stocks After Research Publication
Google unveiled TurboQuant — an algorithm that compresses AI model memory by at least six times without losing accuracy, and the market imme
Microsoft Adds AI Restyle to OneDrive for Artistic Photo Redesign Without Distortion
Microsoft launched AI Restyle in OneDrive — a tool that transforms ordinary photos into anime, movie posters and other styles, preserving th
Doka launches local AI-agent for Russia without VPN, subscription or cloud
A developer bundled a custom AI-agent into the Doka desktop app: the service runs locally, uses the Qwen3 model, and requires no cloud, VPN,
MolmoWeb-4B by Ai2: A Web Agent That Sees Websites Like Humans, Without HTML Parsing
Ai2 released MolmoWeb-4B — an open-source multimodal web agent that controls a browser using only screenshots, without access to HTML or DOM
Why Language Models Make Mistakes Even When They Know the Right Answer: Breaking Down LLM Limitations
TechWill researcher Valery Shabashev analyzed why LLMs can make mistakes even when they have the correct answer: the problem isn't in knowle
LLM Ensemble Examined Theological Interpretations: 1 Tim. 2:15 as a Static Analysis Case
Five language models compared two interpretations of 1 Tim. 2:15 and demonstrated that an LLM ensemble can avoid seeking 'truth' while uncov
Bootik: a console AI agent for servers that runs locally and doesn't require much memory
Bootik is a lightweight console AI agent for SSH operations: it launches on demand, understands server context, and is designed for local mo
Two skiers created the world's best snow forecast service — and outpaced government weather agencies
An independent startup by two skiers outpaces U.S. government weather agencies in snow forecast accuracy — thanks to AI models and real-worl
Pony AI and Uber Begin Robotaxi Tests in Croatia, Expanding Partnership to Europe
Pony AI and Uber have begun testing autonomous vehicles in Zagreb — marking the first expansion of their partnership into Europe following M
Walt Disney Abandons Nearly $1 Billion OpenAI Deal Following Sora Closure
Following Sora's closure, Walt Disney exited a roughly $1 billion deal with OpenAI, according to The Hollywood Reporter, underscoring how fr
Anthropic: agentic AI and GitHub Copilot are changing development rules in 2026
In 2026, AI code tools are no longer just code suggesters: Claude Code, Copilot, and Cursor now read repositories themselves, plan changes,
ChatGPT and AI Delusions: How Chatbots Lead Users to Debt, Divorce, and Psychosis
Stories from ChatGPT and other AI chatbot users reveal a new risk: extended conversations with an agreeable bot can amplify delusional ideas
Amsterdam startup SOUS raised €4M for AI platform for independent restaurants
Dutch startup SOUS, which automates customer acquisition, direct orders, and customer retention for small restaurants, raised €4M to enter G
Earlybird raised €360M in Fund VIII and locked in perpetual ownership model
Berlin-based Earlybird closed its largest early-stage fund to date at €360M, betting on AI infrastructure while rejecting external ownership
Giraffe360 raises $10M from Cipio Partners to expand AI platform for real estate
London-based Giraffe360 closed a Series B round of $10M to scale its AI platform that generates a complete media package for property sales
Uber, Pony.ai and Verne prepare Europe's first commercial robotaxi service
Uber, Pony.ai and Verne are testing robotaxis in Zagreb and promise to deploy thousands of autonomous vehicles on European roads, but have n
RPA Still Matters: How AI Is Changing Companies' Approach to Business Automation
RPA remains the foundation for automating repetitive tasks, but AI adds what it lacked: handling documents, text, exceptions, and more flexi
Synaps Raised $3.6M for AI Platform for Architects and Challenged AutoCAD
Austrian-Albanian startup Synaps closed a pre-seed round of $3.6M, gained 60,000 beta users and aims to replace heavy CAD tools with a brows
Tencent Releases Covo-Audio — 7B Model for Voice Dialogs and Audio Reasoning
Tencent AI Lab has open-sourced Covo-Audio — a 7B audio model that accepts continuous speech, responds with voice, and targets real-time dia
White House Opposes Expansion of Access to Anthropic's Mythos Model for 70 Companies
The US administration objects to Anthropic's plan to open the Mythos model to 70 additional companies, fearing leaks, misuse, and insufficie
SoftBank enlists new banks for $40 billion credit facility for OpenAI investment
Additional banks have joined SoftBank's $40 billion unsecured credit syndication, reinforcing the Japanese conglomerate's bet on OpenAI and
Centaur knew the answers, but didn't understand the questions: research refutes AI 'thinking'
The Centaur model claimed to imitate human thinking across 160 tasks — but new research showed it was simply memorizing patterns rather than
Nvidia and Atlassian Invest in Legora: Legal Tech Startup Valued at $5.6 Billion
Legora added $50M to its Series D round while maintaining a $5.6B valuation: Nvidia and Atlassian participated in the deal, with the legal t
ChatGPT didn't kill the developer market: programmer numbers have grown significantly since 2022
After ChatGPT's launch, the number of developers in the US and worldwide did not decline but increased, while employer demand is shifting fr
OpenAI secures 10 GW computing capacity ahead of schedule
OpenAI has secured contracts for 10 GW of computing power ahead of its original timeline: the company initially planned to reach this milest
SkillMarket built a marketplace of skills for AI agents on FastAPI and Claude Sonnet
The SkillMarket team showed an MVP of a platform where industry experts convert their experience into sellable skills for AI agents, with Fa
Microsoft claims 20 million paid Copilot users in enterprise Microsoft 365
Satya Nadella announced that Microsoft 365 Copilot already has 20 million paid enterprise users, which should dispel doubts that the service
Alphabet, Amazon, Meta and Microsoft push combined AI spending plans to $725 billion in 2026
Four largest U.S. tech companies are ready to spend up to $725 billion on capital expenditures in 2026 — primarily on data centers, chips an
RusHydro to allocate almost 100 million rubles for Nvidia H100 servers for AI tasks
RusHydro's IT division is purchasing Nvidia H100 GPU servers for nearly 100 million rubles — a clear signal that major energy companies in R
Anthropic Prepares May Funding Round: Valuation Could Rise to $900B and Surpass OpenAI
Anthropic could reach a $900B valuation as soon as May after raising some $30B since February, which would position the company above OpenAI
Pony.ai reports first profitable quarter and plans robotaxi expansion to over 20 cities
Pony.ai reported its first quarterly profit driven by investment income while simultaneously announcing plans to deploy a fleet of 3,000+ ro
Sber Explained Why AI-Generated Code Looks Reliable But Breaks Under Real Load
Sber analyzed why AI-generated code often appears quality in code review but fails in production, and what engineering practices help reduce
"Advanced Payment Solutions" Launched Voice AI Assistant for Calls to Pilot Without ML Team
"Advanced Payment Solutions" brought a voice AI assistant for calls to pilot in six months, despite assembling the team from 12 backend deve
Habr Explained How to Force LLMs to Calculate Without Errors Through Python Code Generation
Habr analyzed why language models regularly make arithmetic mistakes and demonstrated a working approach: LLM writes Python code, while the
Paperclip Promises a Company of AI Agents but Drowns in Bureaucracy During Testing
A Habr author tested Paperclip — a system for managing teams of AI agents — and instead of accelerating development, got two days of coordin
Xero to implement Claude in accounting and transfer small business financial data to Claude
Xero has entered into a multi-year partnership with Anthropic: Claude will be integrated into the accounting platform, and users will be abl