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.
Anthropic’s new Claude 4.1 dominates coding tests days before GPT-5 arrives
Anthropic’s Claude Opus 4.1 achieves 74.5% on coding benchmarks, leading the AI market, but faces risk as nearly half its $3.1B API revenue depends on just two customers.
AI’s last mile just got a supercomputer, courtesy of ASUS and NVIDIA
ASUS and NVIDIA redefine local AI performance with Grace Blackwell-powered desktops that fit on your desk—or in your hand.
How AI amplifies these other tech trends that matter most to business in 2025
The top frontier technologies for companies in 2025 include agentic AI and the fast-moving evolution of AI solutions in the marketplace.
Are Marketplaces Defensible In The Age Of AI Purchasing Agents?
This article is the second in a two-part series examining the evolution of marketplaces. In the first, we journeyed through the marketplace innovations that lowered friction in the purchasing process....
The Clock Is Ticking On AI Security In A Quantum World
AI systems are scaling fast but quantum threats are catching up. Security leaders need to act before the clock runs out.
Microsoft’s AI Prototype Can Reverse Engineer Malware, No Human Needed
Although Project Ire is a prototype, Microsoft says the ‘AI agent’ can (in some cases) reverse engineer any type of software on its own to determine if it’s malicious.
OpenAI open weight models available today on AWS
Customers will have access to advanced open weight models in Amazon Bedrock and Amazon SageMaker AI, bringing together OpenAI and the world’s most comprehensive and broadly adopted cloud.
The Problem
This is a new introduction to AI as an extinction threat, previously posted to the MIRI website in February alongside a summary. It was written indep…
Nvidia Pushes Back on AI Chip Tracking, Kill Switches, But US Might Do It Anyway
The US is looking at requiring ‘software or physical changes’ to enable chip-based location tracking to stop Nvidia’s most advanced AI chips from being smuggled into China.
Google embeds AI agents deep into its data stack – here’s what they can do for you
Google’s new agents aren’t chatbots. They’re autonomous problem-solvers for the enterprise. Here’s how they integrate with BigQuery, Spanner, and Gemini to transform analytics, coding, and infrastructure.
Why the AI era is forcing a redesign of the entire compute backbone
Guest Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now The past few...
Persona Vectors – Anthropic Paper
> In a new paper, we identify patterns of activity within an AI model’s neural network that control its character traits. We call these persona vecto…
7 major AI updates this week — including one that could change search forever
AI is moving quickly, here’s what’s new
LLMs break into networks with no help, and it’s not science fiction anymore
AI model replicated the Equifax breach without a single human command
Big tech has spent $155bn on AI this year. It’s about to spend hundreds of billions more
Tech giants have spent more on AI than the US government has on education, jobs and social services in 2025 so far
Anthropic beats OpenAI as the top LLM provider for business – and it’s not even close
A new report explores the state of LLMs in the enterprise.
“We want AWS to be the place where everyone runs enterprise AI agents” – the agentic era is here for your business, so be prepared for the new age
AWS wants to be the central place for AI agents to run
AI is already replacing thousands of jobs per month, report finds
The tech industry ‘is being reshaped by the advancement of artificial intelligence,’ firm says
Tim Cook says Apple ‘must’ figure out AI and ‘will make the investment to do it’
“This is sort of ours to grab.”
New vision model from Cohere runs on two GPUs, beats top-tier VLMs on visual tasks
Cohere’s Command A Vision can read graphs and PDFs to make enterprise research richer and analyze the documents businesses actually rely on.
At $250 million, top AI salaries dwarf those of the Manhattan Project and the Space Race
A 24 year-old AI researcher will earn 327x what Oppenheimer made while developing the atomic bomb.
Why open-source AI became an American national priority
To reflect democratic principles, AI must be built in the open. If the U.S. wants to lead the AI race, it must lead the open-source AI race.
AI in a vat: Fundamental limits of efficient world modelling for safe agent sandboxing
Drawing inspiration from the ‘brain in a vat’ thought experiment, this blogpost investigates methods to simplify world models that remain agnostic to…
EU AI Act: one year on, new measures enter effect
With the implementation schedule still half-developed, the first anniversary of the European Artificial Intelligence law serves to review a legislation whose biggest challenge is to remain relevant in the face...
Most Read: NiCE Acquires Agentic AI Company in $955M Deal; Department of Energy Unveils First Four AI Data Center Sites
Also inside, Zuckerberg unveils “personal superintelligence” vision as Meta spends big; Intel slashes 15% of workforce as new CEO battles financial crisis, and more.
The AI-Driven Cloud Market Share Shift by @ttunguz
What force could dethrone AWS after more than a decade of unchallenged dominance? For years, Amazon Web Services ruled the cloud infrastructure market. It was the default choice without a...
Wall Street delighted with Microsoft as it spends $100bn on AI
Company’s second-quarter financial results showed a booming cloud business and enormous capital expenditures
First NHS AI-run physio clinic in England halves back-pain waiting list
AI platform Flok Health to be rolled out to other NHS trusts but experts warn tech-based care could increase health inequalities
OpenAI taps Norway for major AIDC project
OpenAI also said that the AI data center will aim to deploy 100,000 Nvidia GPUs by end-2026.
The Week in AI Governance
There was enough governance related news this week to spin it out. • THE EU AI CODE OF PRACTICE …
Gemini 2.5 Deep Think rolling out now for Google AI Ultra
After previewing at I/O 2025 in May, Google is now rolling out Gemini 2.5 Deep Think to AI Ultra subscribers.
Research Areas in Economic Theory and Game Theory (The Alignment Project by UK AISI)
By considering AI agents in a strategic setting, we hope to ensure alignment approaches are robust to capabilities improvements and that these framings can inspire new ideas for alignment techniques.
Research Areas in AI Control (The Alignment Project by UK AISI)
The field of AI Control aims to reduce this risk through techniques such as monitoring of untrusted AI systems by other AI models and restricting the affordances of untrusted AI...
EU rules on general-purpose AI models start to apply tomorrow, bringing more transparency, safety and accountability
Tomorrow, the AI Act obligations for providers of general-purpose AI (GPAI) models enter into application across the EU.
Mark Zuckerberg Says Meta’s AI Seems to Be Getting Smarter on Its Own
The Meta CEO just shared some big thoughts about ‘personal superintelligence’ and how it could help people.
We Asked 100+ AI Models to Write Code. Here’s How Many Failed Security Tests.
Application Security for the AI Era | Veracode
From Zero to MCP: Three Lessons I Learned Building Tools for LLMs
A Blog post by r3nz0 on Hugging Face
Replit’s New Release Address Most of The Challenges We Hit Vibe Coding. But Is ‘Prosumer’ Vibe Coding Really Ready for Commercial Apps Yet?
We recently had a database deletion incident involving Replit’s AI Agent and an app we were working to build for the SaaStr community. After 100+ straight hours of vibe coding,...
Microsoft Backs Up A.I. Spending With $27.2 Billion Quarterly Profit
The company’s results are likely to ease investors’ worries that big investments in artificial intelligence aren’t paying off.
Meta Forecasts Continued Spending as It Races to Build ‘Superintelligence’
The Silicon Valley giant, which has been on a spending spree to hire A.I. researchers, also posted a 36 percent quarterly increase in profit and a 22 percent jump in...