811
New
The latest blog posts to take off on Hacker News. 3,238 posts.
812
State of AI: An Empirical 100T Token Study with OpenRouter
Read OpenRouter's 2025 State of AI report — an empirical 100 trillion token study of real LLM usage, model trends, and developer insights.
813
The "confident idiot" problem: Why AI needs hard rules, not vibe checks
We cannot fix probability with more probability. We need to re-introduce assert.
814
Ecosia: The greenest AI is here
While the AI race is raging, we’ve been building an alternative. One that’s helpful, private, and optional — and designed for you, not for profit.
815
Mistral 3 family of models released
The most powerful AI platform for enterprises. Customize, fine-tune, and deploy AI assistants, autonomous agents, and multimodal AI with open models.
816
AI agents find $4.6M in blockchain smart contract exploits
On a benchmark of real exploited contracts, AI agents found vulnerabilities worth a combined $4.6 million, underscoring the need for AI-powered defense.
817
All about automotive lidar
A comprehensive overview of automotive lidar technology — what lidar measures, how it measures distance, and how it determines bearing.
818
Writing a good Claude.md
`CLAUDE.md` is a high-leverage configuration point for Claude Code. Learning how to write a good `CLAUDE.md` (or `AGENTS.md`) is a key skill for agent-enabled software engineering.
819
Effective harnesses for long-running agents
How an initializer agent and an incremental coding agent help the Claude Agent SDK make steady progress across many context windows.
820
Codex, Opus, Gemini try to build Counter Strike
Introducing CS Bench: we ask all top models to build a multiplayer Counter-Strike from scratch.
821
AI Adoption Rates Starting to Flatten Out
Data from the Census Bureau and Ramp shows that AI adoption rates are starting to flatten out across all firm...
822
Ilya Sutskever, Yann LeCun and the End of “Just Add GPUs”
When two of the most influential people in AI both say that today’s large language models are hitting their limits, it’s worth paying attention. In a recent…
823
Google Antigravity exfiltrates data via indirect prompt injection attack
An indirect prompt injection in an implementation blog can manipulate Antigravity to invoke a malicious browser subagent in order to steal credentials and sensitive code from a user’s IDE.
824
FLUX.2: Frontier Visual Intelligence
Today, we release FLUX.2, our most capable model to date.
825
PRC elites voice AI-skepticism
Executive Summary: Rapid advancement in artificial intelligence (AI) has become a point of national pride in the People’s Republic of China (PRC), driven in part by a cohort of accelerationist advisors who view technology as a transformative solution to the country’s economic ills. Lauding it as a “new productive…
826
Claude Advanced Tool Use
Claude can now discover, learn, and execute tools dynamically to enable agents that take action in the real world. Here’s how.
827
Claude Opus 4.5
Claude Opus 4.5 is our new flagship model, with state-of-the-art coding and agentic performance at lower pricing.
828
General principles for the use of AI at CERN
This content is currently in the process of being archived and integrated into the CERN Library collections.
829
WorldGen – Text to Immersive 3D Worlds
Today, we’re introducing WorldGen: a state-of-the-art end-to-end system for generating interactive and navigable 3D worlds from a single text prompt.
830
Olmo 3: Charting a path through the model flow to lead open-source AI
Our new flagship Olmo 3 model family empowers the open source community with not only state-of-the-art open models, but the entire model flow and full traceability back to training data.
831
Gmail can read your emails and attachments to train its AI, unless you opt out
Did you know that Gmail can use your emails and attachments for its smart features? Here's how to check your settings.
832
Kagi Assistants
TL;DR Today we're releasing two research assistants: Quick Assistant and Research Assistant (previously named Ki during beta). Kagi's Research Assistant happened to top a popular benchmark (SimpleQA) when we ran it in August 2025. This was...
833
Nano Banana Pro
Nano Banana Pro is our new image generation and editing model from Google DeepMind.
834
Measuring political bias in Claude
We introduce and open-source an automated evaluation for political even-handedness, testing six models across 1,350 prompt pairs.
835
Building more with GPT-5.1-Codex-Max
Introducing GPT-5.1-Codex-Max, a faster, more intelligent agentic coding model for Codex. The model is designed for long-running, project-scale work with enhanced reasoning and token efficiency.
836
Meta Segment Anything Model 3
We’ve seen incredible adoption of SAM 3 over the last few months, and during that time, we’ve been working behind the scenes on updates to improve video processing efficiency. Today, we’re pleased to introduce SAM 3.1.
837
838
AA-Omniscience v1.1: Knowledge and Hallucination Benchmark
Compare AI model performance on AA-Omniscience v1.1: Knowledge and Hallucination Benchmark. A benchmark measuring factual recall and hallucination across various economically relevant domains.
839
5 Things to Try with Gemini 3 Pro in Gemini CLI
Unlock state-of-the-art coding and reasoning in your terminal. Gemini 3 Pro is now in Gemini CLI for Google AI Ultra and paid API subscribers.
840
Gemini 3 for developers: New reasoning, agentic capabilities
Gemini 3 is introducing advanced agentic coding capabilities, plus Google Antigravity, a new agentic development platform.