What we're reading, in the order we're reading it.
Anthony and Harry's working stream — links, charts, tweets, and short takes. The unprocessed inputs behind the briefings.
ChatGPT is getting a big upgrade. It’s smarter and less likely to deceive you
OpenAI on Thursday launched GPT-5, an upgraded version of the AI model behind ChatGPT that the company claims is significantly faster and more capable than its predecessor. The launch comes...
Labor under pressure on plans to regulate AI as Coalition accuses government of mixed messages
Protecting arts copyright a priority, PM says, but government also wants to reap productivity gains from AI
OpenAI launches new GPT-5 model for all ChatGPT users
OpenAI announced GPT-5, its latest and most advanced AI model.
GitHub – nottelabs/notte: 🔥 Reliable Browser AI agents (YC S25)
🔥 Reliable Browser AI agents (YC S25). Contribute to nottelabs/notte development by creating an account on GitHub.
OpenAI launches GPT-5 as the AI industry seeks a return on investment
OpenAI launched on Thursday its GPT-5 artificial intelligence model, the highly anticipated latest installment of a technology that has helped transform global business and culture.
OpenAI launches GPT-5, nano, mini and Pro — not AGI, but capable of generating ‘software-on-demand’
With safer design, more robust reasoning, expanded developer tooling, and broad user access, GPT-5 reflects a maturing AI ecosystem.
OpenAI claims new GPT-5 model boosts ChatGPT to ‘PhD level’
GPT-5’s release comes as tech firms continue to compete in an effort to claim the world’s most advanced AI.
News Corp Chief Stands Up for Copyright
In comments tied to the release of HarperCollins parent company News Corp’s fiscal 2025 results, CEO Robert Thomson hammered home the need for strong copyright protections in the age of...
NSF invests nearly $32M to accelerate novel AI-driven approaches in protein design, strengthening the U.S. bioeconomy
The U.S. National Science Foundation Directorate for Technology, Innovation and Partnerships (NSF TIP) announced an inaugural investment of nearly $32 million to five teams across the U.S. through the…
OpenAI’s New Open Source Models Are A Very Big Deal: 3 Reasons Why
This article discusses OpenAI new open source model, china vs US in AI, and implications for AI technology roadmap for companies
Vision Language Model Alignment in TRL ⚡️
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Meet your new AI coding teammate: Gemini CLI GitHub Actions
Today, we’re introducing Gemini CLI GitHub Actions. It’s a no-cost, powerful AI coding teammate for your repository. It acts both as an autonomous agent for critical routine coding tasks, and...
Google has a new AI coding agent – and it’s now free for everyone to use
Jules is a new, free to use, AI agent from Google
Two Chinese nationals in California accused of illegally shipping Nvidia AI chips to China
Two Chinese nationals in California were arrested and charged with illegally shipping tens of millions of dollars’ worth of AI chips to China, including Nvidia H100s, the US Justice Department...
AXRP Episode 46 – Tom Davidson on AI-enabled Coups
YouTube link • Could AI enable a small group to gain power over a large country, and lock in their power permanently? Often, people worried about cat…
If you can generate obfuscated chain-of-thought, can you monitor it?
tldr: Chain-of-thought (CoT) monitoring is a proposed safety mechanism for overseeing AI systems, but its effectiveness against deliberately obfuscat…
SoftBank’s AI investment spree to be in focus on at Q1 earnings
When Japan’s SoftBank Group reports earnings on Thursday, its mammoth investments in artificial intelligence companies are set to take the spotlight.
New ‘persona vectors’ from Anthropic let you decode and direct an LLM’s personality
A new study from Anthropic introduces “persona vectors,” a technique for developers to monitor, predict and control unwanted LLM behaviors.
Meet The AI Agents Redefining B2B GTM Strategies And Approaches At B2B Summit EMEA
AI agents are augmenting the B2B GTM workforce and workflows at a rapid pace. Get to know them and their capabilities better at Forrester’s B2B Summit EMEA.
The initial reactions to OpenAI’s landmark open source gpt-oss models are highly varied and mixed
The verdict, for now, is split. OpenAI’s gpt-oss models are a landmark in terms of licensing and accessibility.
OpenAI announces deal to offer ChatGPT to US executive branch at almost no cost
Questions linger about ideological bias in models as well as data security.
For regulated industries, AWS’s neurosymbolic AI promises safe, explainable agent automation
AWS is making automated reasoning checks, a feature on Bedrock, generally available to customers to start proving truth in their AI systems.
Inside the Biden Administration’s Unpublished Report on AI Safety
The US government conducted a groundbreaking study on frontier models—and never published the results.
Google’s new diffusion AI agent mimics human writing to improve enterprise research
A diffusion model inspired by the human process of drafting, searching for information, and making iterative revisions.
OpenAI Announces Massive US Government Partnership
The ChatGPT maker is providing its frontier AI models to federal agencies for $1 for the next year.
OpenAI teases GPT-5 release for ChatGPT
OpenAI is having a big week. Yesterday, the leading AI firm dropped not one, but two open-weight models, including one…
DeepMind’s new AI Genie 3 could change physics modeling
The accuracy of the new model at predicting how physics works suggests that neural networks could eventually replace physics modeling, impacting fields like robotics and science.
Qwen/Qwen3-4B-Thinking-2507 · Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
AI learns from nature to design super-adhesive gels that work underwater
Nature – The protein glues used by organisms to cling to wet surfaces have informed an AI-supported strategy to discover highly adhesive hydrogels.
Why Perplexity’s CEO couldn’t wait for perfection to launch an operating system for the AI era
The AI search engine startup is releasing products that anticipate where AI will be in the future.
Paradigm Shifts and the Winner’s Curse
When paradigms change, previous winners have the hardest time adjusting; that is why AI might be a challenge for Apple and Amazon
NSF invests in AI-ready test beds to accelerate U.S. leadership
The U.S. National Science Foundation announced over $2 million in planning grants to support the development of artificial intelligence-ready test beds — real-world environments that will accelerate…
GitHub – shutootaki/bookwith: BookWith – A New Reading Experience with AI. A next-generation conversational reading platform that goes beyond traditional e-book readers
BookWith – A New Reading Experience with AI. A next-generation conversational reading platform that goes beyond traditional e-book readers – shutootaki/bookwith
States take the lead in AI regulation as federal government steers clear
All 50 states have introduced AI-related legislation this year.
Nuclear Experts Say Mixing AI and Nuclear Weapons Is Inevitable
Human judgement remains central to the launch of nuclear weapons. But experts say it’s a matter of when, not if, artificial intelligence will get baked into the world’s most dangerous...
OpenAI’s long-awaited GPT-5 model nears release
OpenAI’s GPT-5, the latest installment of the AI technology that powered the ChatGPT juggernaut in 2022, is set for an imminent release, and users will scrutinize if the step up...
Google Cloud’s newest AI agents want to boost data science and engineering in your business
AI agents from Google Cloud look to improve data analytics
Behind the Meta scale AI deal: why more data Isn’t always better for physical AI
Precision, not volume, drives success in Physical AI
Doctor Jekyll Or Mr. Hyde? The Dual Nature Of Agentic AI In Healthcare
Agentic AI is rapidly reshaping healthcare, offering faster diagnostics, reduced costs, and expanded access. But like Dr. Jekyll and Mr. Hyde, its dual nature poses risks — from bias and...
OpenAI returns to its open-source roots with new open-weight AI models, and it’s a big deal
The models are licensed under Apache 2.0, one of the most permissive open licenses. Here’s what that means.