Viner Shakirov
Gemini 4 Carbon: a new top-1? A review of the available benchmarks
Google has announced Gemini 4 Carbon. We break down the test results, differences from Argon, comparison with Opus 5.5, and the key features of the new model
Learn more
ChatGPT and Codex will share a limit: what changes in December
Five GPT Values: How OpenAI Wants to Change Model Behavior
Researchers find browser security flaws in Cursor and Codex
An OpenAI employee explains when he chooses Astra over Sol
OpenAI and Anthropic say AI disproved the exponential time hypothesis
Suno connects to ChatGPT and Claude to make music in the chat
Google's Genie 4 creates interactive worlds with sound
GPT-6.1 Terra: use cases and comparison with Sol 6.1 and DeepSeek V4.1
Anthropic files for an IPO: what we know about its market debut
Kimi to restore up to 25% of monthly quotas after K3.1 Pro issues
A Fields medalist at Berkeley stopped grading homework because of AI
Canva releases Craft, its first in-house image generation model
OpenAI trains models on a copy of the internet: the live web is too risky
MIT left 13 AI agents on an island with no task, and they started to play
ChatGPT ads reach 26 more countries: 100,000 companies have run them
OpenAI's Decisions API: an AI model that doesn't write, it chooses
Spotify pulled 50 million AI tracks from recommendations this year
Kimi K3.1 Pro - Opus 5.5 Level? Independent Tests of Moonshot's New Model
Is Qwen 4 Being Tested on Arena AI? What We Know About Its Release
Black October: a PHP flaw, hackers with Claude Code and 9 more incidents
What Google's CEO thinks about AI: 7 takeaways from Pichai's interview
Claude Duet: Anthropic answers OpenAI Dots
Will Claude Fable 5.5 Ship Before the End of October? What We Know About Anthropic's New Model
Mercury 3 from Inception: 1,068 Tokens/s and a 1M-Token Context
MiniMax releases M3.1 Flash: agent tasks 70% cheaper
Google Taught AI Agents to Rewrite Their Own Instructions, No Retraining
Opus 5.5 vs DeepSeek and GPT-6.1 Sol in Poker: Who Is Stronger and Cheaper?
Why Is the DeepSeek V4.1 Flash 1017 Update Causing Such Excitement?
People Guess a Deepfake Only 50% of the Time. Can You Really Spot a Fake?
AI or Human on the Phone: What 14,000 Calls Showed
How Good Is the New Grok Voice 3.0, Really?
Haiku 5.5 or GPT-6 Luna: Which to Choose When Prices Are Nearly Equal?
Gemini 4 Argon Is Open to Everyone, but Some Things Don't Work
Has Muse Caught Up With Opus 5.5? What the New Muse Wonder Model Showed
Sonnet 5.5 or Opus 5.5: When Paying More Isn't Worth It
