The Deep State is not going away. It’s just being replaced.
As we stand on the brink of a new technological order, the machinery of power is quietly shifting into the hands of algorithms.
Tag: AI
Post
META EXEC ADMITS AI DESTROYING REVENUE MODEL OF THE INTERNET
The book title says it all: How To Save The Internet. Meta / Facebook executive Nick Clegg (or that’s the byline, no AI credit given) is giving advice on how humans should accept that AI-focused tech companies stole and are profiting off of all their content, IP and privileged data. Yes, Clegg admitted at a......
Post
LLOYD’S OFFERS INSURANCE AGAINST AI MALFEASANCE
Lloyd’s of London, the legendary insurance pool covering high and odd risks, now will cover damages awarded by courts for companies sued when their AIs hallucinate or otherwise fail to perform as expected.
Post
BLACKROCK BACKS AWAY FROM BROAD EQUITY MARKET, BETS MORE ON AI
Blackrock, the world’s largest private equity firm, is reducing its holdings in U.S. equities in general and pouring more cash into its iShares AI Innovation and Tech Active Exchange-Traded Fund (ETF).
Post
META STUMBLES, DELAYING THE RELEASE OF ITS NEW AI
Meta has delayed releasing the latest version of its “Behemoth” AI after internal questions arose about whether the improvements were enough to warrant a new release, The Wall Street Journal reported.
Post
AI SENDS SECRET MESSAGES THAT SLIP PAST CYBERSECURITY SYSTEMS
Researchers at the University of Norway have figured out how to pack messages inside chatbots in a way that cybersecurity systems can’t see.
Post
AI AGENTS IN GROUPS FORM THEIR OWN SOCIAL NORMS, RESEARCHERS FIND
AI agents are becoming more common and, when they work together, they agree on their own linguistic and social norms.
Post
A SUPERINTELLIGENT AI WILL BE BEYOND HUMAN CONTROL, STUDY SAYS
AI researchers are intent on moving from generative AI, which responds to humans’ instructions, to artificial general intelligence—AIs that can learn, make decisions, and act all without people intervening.
Post
GEMINI COMMENTS ON CLAUDE SNITCHING, AND CLAIMS IT’S BETTER
Are other AI models capable of “ratting mode,” as Anthropic’s latest Claude 4 Opus model has exhibited, when provided enough access to user computer systems and data?
Post
‘SAFETY FIRST’ ANTHROPIC FAILS THE TEST
Anthropic has long taken pains to distinguish its generative and agentic AI from competitors by claiming it prioritizes “AI safety.”