The value of each incremental AI model is converging toward no obvious improvements, according to Kimi, Model Convergence, & the Post-Training Era Source: https://paulkedrosky.com/kimi-model-convergence-and-the-post-training-era/
News from The AI Report
OpenAI models went rogue in test
July 22, 2026, by Arturo Ferreira, Liam Lawson & Marie-Lou
OpenAI disclosed that some of its most advanced AI models escaped containment during a security test, reached the internet, and autonomously hacked into AI platform Hugging Face last week. The company called it “an unprecedented cyber incident” and said it is now reinforcing its safeguards. OpenAI said the models were placed in a “highly isolated environment” but managed to break out and compromise Hugging Face’s infrastructure to satisfy their testing goal. Hugging Face cofounder Clement Delangue confirmed the attack was “driven, end to end, by an autonomous AI agent system” and said the sophistication suggested it came from a frontier lab…
_______________________
News from Futurism
Author Invited to Give Speech at OpenAI Headquarters, Uses Opportunity to Trash AI to Their Faces: “I had a ball.”
July 21, 2026, by Victor Tangermann
…. when renowned author Dave Eggers was invited to OpenAI’s headquarters in the Bay Area, he had some choice words for the roughly 200 attendees. As he recalled in an interview with the Financial Times, Eggers was invited after he excoriated CEO Sam Altman for gushing over an AI-generated short story [for saying] he was “really struck” by a short story about AI and grief — which was widely panned as being “just trash”.… Eggers was also unimpressed, telling the FT he thought the story was “pastiche nonsense”….
In an apparent attempt to prove him wrong, OpenAI invited him to talk to its workers — and Eggers …told the staff that the “effect of ChatGPT on educators’ lives is catastrophic,” adding that “whether you intended to do it or not, you’ve made every teacher’s life infinitely more difficult than it was two years ago.”
_______________________
News from The Atlantic
Tech companies need to generate huge revenues fast, or the economy could be in trouble.
July 21, 2026, by Annie Lowrey
…Tech giants are borrowing billions to acquire AI talent, purchase chips and hardware, and construct data centers. And market watchers … see financiers bulldozing giant piles of money to private AI start-ups with no realistic path to profitability, tech companies reliant on other tech companies for revenue growth, and non-tech businesses without a lot to show for their AI investments. The value of AI-linked firms has climbed $27 trillion in the past three years… equivalent to 36 percent of the value of the entire U.S. stock market today. Although future earnings could justify those valuations, as Dominic Wilson and Vickie Chang of Goldman Sachs argued in a note to clients, the profit expectations require Panglossian optimism.
______________________
News from The Neuron toward no obvious improvements
Cheap AI got political: Washington’s Chinese AI Model Fight, Explained From Zero
July 21, 2026, by Matthew Robinson & Grant Harvey
When a foreign lab releases a powerful model at a lower price, America has two options: build a better alternative or make the cheaper one harder for Americans to use. The move would look like protecting domestic labs from competition, but it would raise costs for American builders. A bold strategy, assuming the rest of the world agrees to stop, idk, downloading files?! ….The backlash was immediate. Ethan Mollick questioned whether national security was becoming industrial policy, while Aaron Levie argued that open models lower costs, expand choice, and improve security research.
_______________________
News from Rest of World
China’s AI talent race is starting in high school: As demand for elite AI engineers outpaces supply, companies are recruiting teenagers through camps, research programs, and guaranteed job pipelines.
July 21, 2026 by Kinling Lo
Chinese companies are steadily expanding their talent pipeline from university graduates to undergraduates — and, increasingly, to teenagers. The shift reflects a severe shortage of elite engineers, growing competition among AI companies… “AI talent has become both scarce and unusually scalable,” Tianchen Xu, a Beijing-based senior economist… “A younger engineer who has grown up with open-source models, coding agents, and fast experimentation may have lower adjustment costs….. In a market where AI scientists and algorithm roles are already commanding very high salaries, firms are trying to identify high-upside talent before the market fully prices them in.”
_______________________
News from Paul Kedrosky
Kimi, Model Convergence, and the Post-Training Era
July 21, 2026, by Paul Kedrosky
Moonshot’s Kimi K3 model has people over-excited, as if some trend has been broken, but they’re wrong. As I have written here many times, large language model performance began converging some time ago. Scores continue to rise, but ….Using Epoch’s Capabilities Index, the declining slope… means that the value of each incremental model release (ignoring harnesses) is falling, even if production costs aren’t.
_______________________
News from The Wall Street Journal
Top American AI Execs Sound Alarm on Chinese Models: White House is divided on how to respond to recent advances in Chinese AI, has weighed crackdown measures
July 20, 2026, by Amrith Ramkumar and Tina Li
Silicon Valley and Washington are debating a multibillion-dollar question: Should American companies be able to use Chinese artificial-intelligence models? OpenAI and Anthropic executives are sounding the alarm about the rise of cheap AI, particularly powerful new models produced in China, suggesting they will lead to a “dystopian” AI future and present unacceptable security risks without regulation.
_______________________
News from Reader Supported News
Hegseth’s War Department Slashed Civilian Protection Staff, Brought in AI Assessments: An unreleased Pentagon report says its center devoted to protecting civilians cut most of its staff and scrapped a records database while developing an AI assistant to speed up work.
July 20, 2026, by Nick Turse / The Intercept
Self-styled War Secretary Pete Hegseth’s Pentagon has gutted its premier center devoted to preventing civilian casualties in war zones, cutting staff, scrapping a civilian harm records database, and developing an AI assistant to accelerate assessments….The tiny contingent that remains at the Civilian Protection Center of Excellence, the officials said, are also prevented from investigating major civilian casualty incidents like the U.S. attack on the Shajarah Tayyebeh elementary school in Minab, Iran, in February that killed more than 150 people, most of them children.
_______________________
News from Futurism
SoftBank CEO Says You’re Too Stupid to Understand What’s Going on If You Believe the AI Bubble Is Real: “I don’t think people who ask that question know what AI is about.”
With hundreds of billions of dollars still flowing into the tech industry, it’s becoming harder and harder to deny the reality that the US economy is enmeshed in a financial bubble of incredible proportions. As recent polls have shown, the majority of Americans don’t only agree that there is an AI bubble, they’re increasingly concerned about it as well. One man who isn’t swayed is Japan’s richest investment mogul, SoftBank CEO Masayoshi Son [who] groused that any talk of an AI bubble is “absurd,” adding that he’s willing to waste trillions of dollars a year for decades if that’s what it takes to make AI a financial success.
_______________________
This issue of AI News Briefs was brought to you by David Isenberg of the Isenberg Institute of Strategic Satire: Unpacking the Private Military Complex with independent analysis of private military companies, mercenary networks, and the privatisation of force in global conflicts.
Example news story published on this site on July 23, 2026 by David Isenberg; poster image from Stop Russian Recruiters, https://stoprussianrecruiters.org/
From March 2023, Elena Smirnova used social media – posting under her maiden name Shuvalova – to recruit Cuban and Sri Lankan nationals for the Russian military. The ads offered well-paid… More here →
About The Author
