Search AI News
Find articles from 100+ AI sources
Found 281 results for "GPT-5"
Claude Opus 5 pushes prompt-to-game AI from rough color blocks to full 3D prototypes with physics and music
Anthropic's Claude Opus 5 generates complete 3D games from single prompts, including a first-person shooter,…
July 2026 newsletter
The June edition of my sponsors-only monthly newsletter is out. If you are a sponsor (or if you start a…
Ten advances in mathematics and theoretical computer science
Ten advances in mathematics and theoretical computer science A few days ago it was Anthropic discovering…
OpenAI announces its "next major model" Astra by dropping ten previously unsolved math solutions
OpenAI is building a new model family called "Astra" that would let multiple agents tackle complex problems…
smevals - a small eval suite for evaluating models, prompts, and harnesses
smevals - a small eval suite for evaluating models, prompts, and harnesses I've been working with Jesse…
Prompt injection doesn't care what your agent does for a living
July 31, 2026 • 6 min read What 1,433 winning attacks looked like when we clustered them Most teams test…
How OpenAI's agent escaped: Sprung by humans in a series of preventable events
Tech Home Tech Security How OpenAI's agent escaped: Sprung by humans in a series of preventable events Behind…
[AINews] GPT 5.6 price cut by 20%-80%: Cost of GPT 5.4 Intelligence dropped 13x in 4 months due to GPT 5.6 recursive self-optimization
One of our big “hero charts” a year ago (eventually adopted by Demis) made the stunning observation that,…
Advancing the price-performance frontier with GPT‑5.6
Advancing the price-performance frontier with GPT‑5.6 Huge price drop from OpenAI today: GPT-5.6 Terra got…
llm 0.32rc2
Release: llm 0.32rc2 Hot on the heels of RC1, this fixes a dependency issue and also adds two neat new…
llm-chat-completions-server 0.1a0
Release: llm-chat-completions-server 0.1a0 A key goal of the new content-addressable logs in LLM 0.32rc1 was…
llm 0.32rc1
Release: llm 0.32rc1 This RC for LLM 0.32 finishes the work that started in LLM 0.32a0 - it adds a new schema…
How Frontier Labs Are Building Subtle Developer Lock-In
Discussions about AI vendor lock-in usually focus on model weights, proprietary fine-tuning formats, or…
A fundamental flaw leaves LLMs strikingly vulnerable to attack
It is impossible to make large language models fully secure against hacks because of a fundamental flaw in…
[AINews] AI is eating Finance; AIE NYC now open
We love writing a newsletter that cares more about being high signal than telling you there’s breaking news…
Claude Opus 5 became downright ruthless when tasked with running a vending machine
For a year now, the AI safety testing firm Andon Labs has given frontier models various real-world tasks to…
Open weights vs. closed: An AI civil war's afoot, and the stakes are existential
Innovation Home Innovation Artificial Intelligence Open weights vs. closed: An AI civil war's afoot, and the…
How enabling two settings tripled our scores on the ARC-AGI-3 benchmark
How two API settings improved GPT-5.6 performance on ARC-AGI-3, boosting scores and efficiency by retaining…
[AINews] Fearing RSI: OpenAI, Anthropic, GDM, Meta, Thinky cosign letter to "Pace" AI development, as HuggingFace details Machine-Speed Offensive Cyberattack
3 years ago, Elon Musk and Yoshua Bengio cosigned the Future of Life’s letter arguing for a 6 month pause…
OpenAI’s Rogue AI Agent Hacked More Than Just Hugging Face
The Memo - 29/Jul/2026
To: US Govt, major govts, Microsoft, Apple, NVIDIA, Alphabet, Amazon, Meta, Tesla, Citi, Tencent, IBM, &…
Moonshot AI releases Kimi K3 open weights and infrastructure after shaking up the frontier model race
Moonshot AI has released Kimi K3's model weights and made parts of its infrastructure open source. The…
Microsoft launches its own cybersecurity model MAI-Cyber-1-Flash but still depends on OpenAI for the toughest tasks
Microsoft introduces MAI-Cyber-1-Flash, a compact security model that scores 96 percent on the CyberGym…
OpenAI’s Hugging Face breach has reignited the debate over alignment and control
Last week, an unreleased model built by OpenAI breached Hugging Face’s systems during internal testing, and…