#LLM
Curated coverage on «LLM»: releases, research, benchmarks, open-source.

ChatGPT Plus at $20: A Habit Tax or a Real Necessity?
Is it still worth paying Sam Altman when the free version of ChatGPT has become suspiciously capable? A look at the tiers and some uncomfortable truths about the…

Vidu Q2: AI Now Copies Not Just the Image, but the Acting Too
While Western giants hold back their releases, China's Vidu rolls out Reference Pro. Show the model an example and you can control everything from facial…

TTT-Discover: Stanford and NVIDIA Make AI "Think" on the Fly and Outpace Scientists
Stanford and NVIDIA unveil TTT-Discover, an algorithm that solves scientific problems twice as fast as human experts by fine-tuning itself while it works.

Ads in ChatGPT: OpenAI Puts a Premium Price Tag on Access

Distilling Reasoning: Why Teachers That Are Too Smart Hurt Small Neural Networks

Anthropic vs. Clawdbot: Why Claude No Longer Tolerates Namesakes

LingBot-VLA: Ant Group Teaches Robots to Understand the World Without Extra Instructions

The Multimodal Shift: How AI Stopped Being Blind, and Why It Matters
The era of text-only chatbots is officially over. Modern neural networks now see, hear, and understand video directly. Here's why that changes the rules.

Amazon Bedrock: How Pushpay Learned to Keep Its Agents' Hallucinations in Check
A look at Pushpay's case: how to build a reliable evaluation system for AI agents on AWS and stop relying on luck with every prompt update.

Tax Agency Texts: How Neural Networks Made Phishing Indistinguishable From the Real Thing

Search Without Borders: AI Is Finally Eating the Blue Links

Google AI Plus Goes All In: Global Rollout Complete

Google's Gemini 3 takes over search: it's no longer a results page but an endless conversation

ChatGPT thinks the grass is greener in the West (and the people are better-looking)
Oxford researchers found that OpenAI's chatbot has a Western-centric bias: it assumes rich countries are both safer and home to more attractive people.

Beeline and red_mad_robot: Why a Telecom Wants Its Own AI Agents
A telecom giant and a tech company are building a platform for agentic AI, an attempt to turn neural networks from toys into autonomous employees for business.









