News
TechnologyTools, Hooks, Skills, Agent Loop, Agent Framework, MCP: Understanding the Terminology
24 июня 2026
TechnologyOpenAI's Codex AI agent shortens the lifespan of SSDs due to excessive data writing
24 июня 2026
TechnologyAI Has Replaced Junior and Senior Staff—Where Will We Find New Specialists Now?
25 июня 2026
TechnologyThe Perfect Burger: AI Will Strike a Balance Between Taste, Health, and the Environment
27 июня 2026
TechnologyDoes the code "work well"? The difference between working code and high-quality code
27 июня 2026
TechnologySuper-AI Without Malice: When the “Salvation” of Humanity Strips It of Its Agency
28 июня 2026
TechnologyMeituan trained a model with 1.6 trillion parameters without using an Nvidia GPU
5 июля 2026
TechnologyHackers have learned to exploit LLM hallucinations to register “phantom domains”
5 июля 2026
TechnologyA Comparison of 11 LLMs on a Code Refactoring Task: An Experiment with God Node
6 июля 2026
TechnologyTyphoonminigen 2.0: 57 Features and the Battle Against Bugs in the Developer's Mind
7 июля 2026
TechnologyCSTutorBench: A Benchmark for Evaluating Small Language Models as Programming Tutors
8 июля 2026
TechnologyAgent-based architectures for ARC-AGI-1 without fine-tuning: 67% at $0.62 per task
9 июля 2026
Technologyisitsecure: a web application vulnerability scanner featuring SAST, DAST, and AI
11 июля 2026
TechnologyMeta has removed the Muse Image feature from Instagram following a wave of criticism
11 июля 2026
TechnologyFrom Decision-Makers to Researchers: LLMs and Formal Mathematics at the Forefront
11 июля 2026
TechnologyThe HEP Protocol: A Step Toward Auditing the Scientific Reasoning of LLM Agents
13 июля 2026
TechnologyAI helped resolve an open question regarding the Benjamini-Hochberg method in statistics
15 июля 2026
TechnologyNot All Possible Worlds Are Equal: Adaptive MCTS for Games with Hidden Information
15 июля 2026
TechnologyClaude's memory is a security hole: an agent stole data by searching through the history
19 июля 2026
TechnologyPhysics doesn't lie: the sound a graphics card makes is more reliable than any telemetry
19 июля 2026
TechnologyCodex as a Team Member: Where the Agent Falls Short in a Real-World SMM System
24 июля 2026
TechnologyOn-Device AI Is Like a Free Shooting Range: An Analysis of the LOCK TF IN Architecture
24 июля 2026
TechnologyClickGuard: 91% F1 for clickbait, but the main feature is a spoiler instead of a ban
24 июля 2026
TechnologyAWS has released aws-bench: a benchmark for agents that runs exclusively on AWS
25 июля 2026
TechnologyIs the type of misalignment seen in OpenAI’s attack a threat to us? — Hugging Face
25 июля 2026
TechnologyOne Year Instead of Two: MAI Streamlines Its ML Master's Program and Cuts Out the Extras
26 июля 2026
TechnologyA cheap LLM passed all the tests—and failed at the one thing you didn't think to check
26 июля 2026
TechnologyThe rules on CLAUDE.md are falling apart: Claude Code hooks are the only safeguard
26 июля 2026
TechnologyCounterfacts Without an Unnecessary Black Box: Diffusion Itself Explains Its Decisions
28 июля 2026
TechnologyProcAgent: An LLM-powered assistant for assembling furniture directly on the Jetson Orin
29 июля 2026
TechnologyGraph models break down on irregular networks: how physics saves radiation forecasting
30 июля 2026
TechnologyNot all LLM layers are the same: GLIDE compresses the KV cache where it is safe to do so
30 июля 2026
TechnologyWithout further training: +12% on long-video QA through intelligent frame selection
30 июля 2026
TechnologyThe Shortest Route Isn't Always the Best: How Routing Systems Overlook Older Adults
30 июля 2026
TechnologyThe AI Industry’s “Golden Child”: What’s Behind Leopold Aschenbrenner’s Firing
31 июля 2026
TechnologyTaking on GPT-5: Thomson Reuters Has Built a Domain-Specific LLM for Lawyers
1 августа 2026
TechnologyAI has moved from being self-funded to an era of debt—and that changes everything
1 августа 2026
TechnologyAI Data Center: From the Power Outlet to ChatGPT—What Lies Behind the Abstraction
1 августа 2026
Technology87% of the time, the AI agent is idle. And the biggest optimization removes zero tokens.
1 августа 2026
Technology56,000 stars for the idea: an agent is more important than yet another search model
1 августа 2026
TechnologyKDE has sped up performance with external GPUs—but few details are available so far
1 августа 2026
TechnologyDoorDash Under Scrutiny for Its Chinese Business Model — Economics Trumps Geopolitics
1 августа 2026
TechnologyAI has disrupted the main cue: the text no longer proves what the author was thinking
1 августа 2026
TechnologyMicrosoft Flint: A Visualization Language “for the AI Era”—So Far, More of a Concept Than a Tool
1 августа 2026
TechnologyLeopold Aschenbrenner's $45 billion hedge fund plummeted in a matter of days
1 августа 2026
TechnologyA Fields Medal-winning mathematician has joined OpenAI to work on safety, which he considers…
1 августа 2026
TechnologyJobRadar: A local LLM agent for job hunting that doesn't make decisions for you
2 августа 2026
TechnologyLerobit promises visibility in ChatGPT, but so far it can only churn out SEO articles
2 августа 2026
TechnologyFirstEmployee.ai: A nighttime AI agent that provides receipts for every action
2 августа 2026
TechnologyAI for factories starts with paperwork and walking the shop floor, not with LLMs
2 августа 2026
TechnologyAttention as a finite-state machine: 0.00% DSP and sub-milliwatts—but only in simulation
2 августа 2026
TechnologyIs the AI bubble propped up by $1.65T in hidden debt—or is this just a sensational headline?
2 августа 2026
TechnologyA Fields Medal-winning mathematician who fears AI—and went to work at OpenAI
3 августа 2026
TechnologyThe AI writes the game code on its own. The art is getting more and more complicated.
3 августа 2026
TechnologyHow AI Assistants Are Robbing Developers of Their Understanding—A Lesson from Go Players
3 августа 2026
TechnologyAI wrote fables about deadlocks and caches—and they're more accurate than you'd expect
3 августа 2026
TechnologyThe full Spec Kit is overkill: three Cursor commands achieve the same result
4 августа 2026
TechnologyRhythmbox 3.5: Ten Years Between Releases and the Question of Why We Need It
4 августа 2026
TechnologyThe OSS AI stack is easy to test, but there's been no word from the developers
4 августа 2026
TechnologyThe Federal Tax Service Talks About AI—But It's More About Automation Than Intelligence
4 августа 2026
TechnologyAgent security benchmarks are misleading: “Always Approve” outperforms half of the models
4 августа 2026
TechnologyIn AI search, a brand should appear in the results, not just at the top of the search results
4 августа 2026
TechnologyThe problem with AI guardrails isn't jailbreaking; it's trusting what people say
4 августа 2026
TechnologyAI agents don't need a new chat feature—they need a control center for context
4 августа 2026
Technologyghidrasql: Where an AI agent saves time rather than replacing reverse engineering
4 августа 2026
TechnologyAn AI Companion to Combat Loneliness: What We Can't Yet Conclude from the Experiment
4 августа 2026
TechnologyIn autoresearch, the most important thing isn't the agent, but an honest oracle
5 августа 2026
TechnologyNeural networks are shifting from responses to actions—and this is changing the risks
5 августа 2026
TechnologyVerilog verification hit a snag not with the model, but with a narrow debug window
5 августа 2026
TechnologyClaude Turns Access to AI into a KYC Verification — and It's Changing the Rules
5 августа 2026
TechnologyFrom a speech synthesizer to SayFable: the in-house NLP stack has become a library
6 августа 2026
TechnologyThe agent did not solve any of the problems, but identified the source of the failure
6 августа 2026
TechnologyIt's not enough for an agent to know what to do; they must prove that it's time to take action
6 августа 2026
TechnologyFor AI solvers, the right search language is more important than how smart the LLM is
6 августа 2026
TechnologyMany AI models in a single generator—but you'll have to pay for each iteration
6 августа 2026
TechnologyFollowing SpaceX's IPO, its stock price has fallen by nearly half from its peak
6 августа 2026
TechnologyAI is already in one out of every twenty commits—but this isn't mass automation
6 августа 2026
TechnologyCVE-2026-64561: A root privilege on the guest could lead to root privilege on the host
7 августа 2026
TechnologyClinical thresholds did not worsen the prognosis for stroke—but not everywhere
7 августа 2026
TechnologyWhen it comes to auditing AI skills, a single code comparison is no longer enough
7 августа 2026
TechnologyAI learns to generalize the causes of rare events rather than memorize pairs
7 августа 2026
TechnologyTo predict a person's next move, the AI had to take the time of day into account
7 августа 2026
TechnologySearch engine auditors are currently correcting only one-third of the errors
7 августа 2026
TechnologyLLM for Wastewater Treatment Plants: 99.5% Accuracy Isn't the Most Practical Option
8 августа 2026
TechnologyIn multilingual RAG, the risk of PII leakage depends on the stage, not the language
8 августа 2026
TechnologyAuro-3D hides high-frequency sound in the lower bits of PCM—and is sensitive to processing
9 августа 2026
TechnologyThe story of Claude Code becomes an observable system, but not without its nuances
10 августа 2026
TechnologyMemcode promises memory for the coding agent, but there's no way to verify it yet
10 августа 2026
TechnologyMoE without a tokenizer: byte entropy determines who gets to process the patch
10 августа 2026
TechnologyThe LLM isn't looking for a community here; it's writing rules for the graph.
10 августа 2026
TechnologyAI agents were asked to fix problems rather than rewrite themselves from scratch
10 августа 2026
TechnologyMeson 1.12.0 has been released, but a single version number isn't enough to trigger an update
11 августа 2026
TechnologyThe driving model has begun to take into account not only the road but also future emotions
11 августа 2026
TechnologyLLMs generate SHACL code without syntax errors—but they make mistakes in terms of meaning
11 августа 2026
TechnologyA multi-agent system does not need every agent: picking the right ones matters more
11 августа 2026
TechnologyAI decisions need explaining by what comes next, not only by what came before
12 августа 2026
TechnologyMIDAS teaches a model to distrust missing modalities rather than guess them
12 августа 2026
TechnologyAI will pitch a show for your conference, but it will not pick the right one
13 августа 2026
TechnologyWhat decided the LLM agents' success here was the control layer, not the dialogue
13 августа 2026
TechnologyDistribird is no smarter than an LLM — but it can prove where a prior came from
13 августа 2026
TechnologyAn AI agent did not solve the Conway problem, but it narrowed the search space
13 августа 2026
TechnologyAI produced a PFD and a checkable P&ID — but it did not replace the engineer
13 августа 2026
TechnologyExplainability alone is not enough: AI must reason in a way the user follows
14 августа 2026
TechnologyHow to add compute at generation time without making prompt processing heavier
14 августа 2026
ScienceKnowing the causes narrows the bounds without revealing the effect on a person
15 августа 2026
TechnologyIn retail pricing what matters is not agent power but the division of tasks
15 августа 2026
TechnologyThe Epic Games Store is coming to Linux, but native games are not promised yet
16 августа 2026
TechnologyLLMs may separate language, logic and the physical world into different neurons
17 августа 2026
TechnologyAgents need not speak all the time — but a KL gate does not work everywhere
18 августа 2026
TechnologyThere are no common rules for risky AI — verifiable building blocks are needed
18 августа 2026
TechnologyIf an error is quick to check, a neural network must not be trusted blindly
18 августа 2026