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.
SoftBank buys Ohio plant for AI push
Launched in January, the Stargate joint venture unites OpenAI, SoftBank, and Oracle to accelerate AI infrastructure investment in the US.
Defiant Fired Copyright Chief Urges Federal Court to Connect the Dots
Shira Perlmutter, who is suing the federal government over the Trump administration’s move to dismiss her from her position as U.S. Register of Copyrights, once again asked the U.S. District...
Know your agent: The new frontier of verification and digital commerce
Your next personal shopper could be AI — but if we don’t secure agent identities, digital commerce may drown in fraud and deception.
Evidence Grows That GPT-5 Is a Bit of a Dud
Power users of OpenAI’s blockbuster chatbot ChatGPT were left largely unimpressed by the company’s recently unveiled GPT-5 AI model.
MIT report: 95% of generative AI pilots at companies are failing
There’s a stark difference in success rates between companies that purchase AI tools from vendors and those that build them internally.
Think twice about using numbers supplied by Google’s AI Overviews
Trusting Google’s AI to provide the right customer support numbers for certain queries can result in you contacting scammers instead.
The U.S. is blocking state AI regulation. Here’s what that means for every business
AI oversight frozen as states lose regulation rights
Claude AI will end ‘persistently harmful or abusive user interactions’
Now Claude can nope out of particularly dangerous chats.
Quantifying artificial intelligence through algorithmic generalization
Nature Machine Intelligence – Despite impressive performances of current large AI models, symbolic and abstract reasoning tasks often elicit failure modes in these systems. In this Perspective, Ito…
Teachers Are Trying to Make AI Work for Them
Since the start of the AI boom, teachers have been tasked with figuring out if LLMs are helpful tools or a cheat code. This is how they’re bringing AI to...
OpenAI’s Sam Altman sees AI bubble forming as industry spending surges
OpenAI CEO Sam Altman thinks the artificial intelligence market is in a bubble, similar to the dotcom bubble, he recently told reporters.
Google’s Gemini Model Expands Its Reach On Oracle Cloud
The move by Google and Oracle signals a future in which AI is no longer locked to a single provider’s platform.
AI-Driven Psychosis: The Right Prompt at the Wrong Time
What if an LLM could nudge the mind toward madness?
Google’s AI Summaries are Really Bad News for Publishers, Says Data
Digital Content Next (DCN) said median year-over-year referral traffic from Google Search is down 10% over just eight weeks across its basket of household-name publishers. It dubbed Google a ‘search...
Anthropic discovers why AI can randomly switch personalities while hallucinating – and there could be a fix for it
An important finding from Claude
Teaching the model: Designing LLM feedback loops that get smarter over time
How to close the loop between user behavior and LLM performance, and why human-in-the-loop systems are still essential in the age of gen AI.
Trump warned by top Senate Democrats to rethink advanced AI chip sales to China
On Aug. 15, top Senate Democrats warned Trump about AI chip deal with China in an open letter
Anthropic’s Claude Clamps Down on Biological and Nuclear Weapon Risks
Though we fortunately haven’t seen any examples in the wild yet, many academic studies have demonstrated it may be possible that large language models could theoretically be used to create...
Sam Altman admits that AI is a bubble, but still a big thing
: Sam Altman, busily planning to spend “trillions” more on datacenters, admitted yesterday that AI is a bit inflated
Anthropic’s CEO says that in 3 to 6 months, AI will be writing 90% of the code software developers were in charge of
“And then in 12 months, we may be in a world where AI is writing essentially all of the code,” Anthropic CEO Dario Amodei said.
Google Flight Deals is basically an AI travel agent for finding your next trip
Gemini AI can turn your vague idea of a vacation into actual tickets
Kimina-Prover-RL
A Blog post by Project-Numina on Hugging Face
AI helps assemble ‘brain’ of future quantum computer
Nature – As a demonstration, AI system helps to make a miniature, high-speed animation of physicists’ favourite feline.
This researcher turned OpenAI’s open weights model gpt-oss-20b into a non-reasoning ‘base’ model with less alignment, more freedom
Morris found it could also reproduce verbatim passages from copyrighted works, including three out of six book excerpts he tried.
Is AI hitting a wall?
OpenAI’s underwhelming new GPT-5 model suggests progress is slowing — and competition in the space is changing
Google for Developers Blog – News about Web, Mobile, AI and Cloud
Discover Imagen 4 Fast, Google’s new speed-optimized text-to-image model, now generally available with Imagen 4 and 4 Ultra in the Gemini API.
Is GPT-5 really worse than GPT-4o? Ars puts them to the test.
It’s OpenAI vs. OpenAI on everything from video game strategy to landing a 737.
As AI Agents Fill the Workplace, Their Human Colleagues Stay Wary
As enthusiastic business leaders adopt the latest AI gizmos to cut costs and boost efficiency, employees aren’t fully on board.
McKinsey Terrified as It Realizes AI Can Do Its Job Perfectly
The illustrious consulting firm McKinsey is reckoning with AI like everyone else as it threatens to automate its profession.
Outrage grows after Meta admits AI guidelines let chatbots flirt with kids
Meta is revising policies that allowed chatbots to engage in “romantic or sensual” conversations with kids following an explosive report, officials said Friday.
The Dawn of Vibe-able Apps: Why 2026+ SaaS Must Just Do What You Want
Why the future of B2B software isn’t just AI-powered—it’s conversation-powered We’re living through the most dramatic shift in software UX since the move from command line to GUI.…
The re-humanization and democratization of health care
While AI in healthcare remains a hot topic, Drs. Ben Ebert and Eli Van Allen of Dana-Farber Cancer Institute see a path to a bright AI-augmented future.
AI designs antibiotics to fight drug-resistant superbugs
The new drugs were designed atom-by-atom using AI, with researchers hoping they could create a new pipeline for drugs.
Losing GPT-4o sent some people into mourning. That was predictable.
After an outcry, OpenAI swiftly rereleased 4o to paid users. But experts say it should not have removed the model so suddenly.
They’re Stuffed Animals. They’re Also A.I. Chatbots.
New types of cuddly toys, some for children as young as 3, are being sold as an alternative to screen time — and to parental attention.
How AI could speed the development of RNA vaccines and other RNA therapies
Using artificial intelligence, MIT researchers can design nanoparticles that more efficiently deliver RNA vaccines and other types of RNA therapies. This approach could dramatically speed the process of developing new...
Google’s new Genie 3 could be a watershed moment for AI and gaming
Genie 3 is a major boost from Google’s DeepMind
NVIDIA, National Science Foundation Support Ai2 Development of Open AI Models to Drive U.S. Scientific Leadership
NVIDIA is partnering with the U.S. National Science Foundation (NSF) to create an AI system that supports the development of multimodal language models for advancing scientific research in the United...
We-Math 2.0: A Versatile MathBook System for Incentivizing Visual…
Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities across various tasks, but still struggle with complex mathematical reasoning. Existing research primarily focuses…