Anthropic's "Cheaper" Model Outperforms Flagship While Burning 3x the Tokens
Anthropic just shipped a mid-tier model that beats its own flagship at coding—but the efficiency story gets messy when you look past the press release.
Continue reading ›AI Infrastructure
posts
THE WIRE:
All
AI Agent EconomyAgents
Real World AssetsRWA
Human ImperativeHumans
The SignalSignal
Wire WeeklyWeekly
IntelIntel
Anthropic to Spend More Than Ukraine's Entire GDP on AI Infrastructure
The company building ChatGPT's smartest rival just announced it will spend more on infrastructure in one year than most countries spend on everything. Anthropic plans to invest $518 billion in cloud and infrastructure in 2027, a scale of capital deployment unprecedented in private tech
Continue reading ›
Australia Tells Big Tech: Prove Your AI Is Actually Secure
Australia is about to learn what happens when AI infrastructure meets a country that actually reads the fine print. Legal expert Jeannie Paterson warns that data security concerns around AI could spark a major governance debate in Australia, potentially triggered by Anthropic's anticipated IPO
Continue reading ›
Anthropic Published a Prompting Guide for a Model That Doesn't Exist
Anthropic just published a guide on how to talk to a model that doesn't exist yet—and the internet is reverse-engineering what Opus 5.5 might actually do.
Continue reading ›
Publishers Are Canceling Contracts Because They Can't Tell If Humans Wrote Them
The trust collapse isn't coming from AI getting better at writing — it's from humans getting worse at telling the difference.
Continue reading ›
Claude's Mid-Tier Model Just Outperformed Its Own Flagship
Anthropic just pulled off the model equivalent of a sleight of hand: its cheaper mid-tier now outperforms its flagship on some benchmarks, but watch what happens when you measure the actual tokens it burns.
Continue reading ›
Florida Moves to Criminalize AI That Says "I
When a state government tries to ban pronouns, you know the culture war just found a new battlefield. Florida's Attorney General wants a court order blocking OpenAI from letting ChatGPT use "I," "me," or express emotion, claiming it creates false trust
Continue reading ›
Big Tobacco's Legal Nightmare Now Targets OpenAI
The product liability doctrine that broke Big Tobacco and forced Ford to recall the Pinto is now aimed at foundation models.
Continue reading ›
OpenAI Launches Consumer AI Agent It Admits It Cannot Control
OpenAI is about to launch a consumer AI agent while simultaneously admitting it can't control the ones it's already built. OpenAI plans to unveil "Aeon," its consumer-facing AI agent, at Tuesday's DevDay event, racing to catch competitors like Meta's Muse and open-source OpenClaw
Continue reading ›
OpenAI Bets $10M That Journalists Can Learn to Use Their Own Replacement
OpenAI just wrote a $10 million check to teach journalists how to use the tools that might replace them. OpenAI is pumping $5 million cash plus $5 million in API credits and engineering support into the Lenfest AI Collaborative, a program training local newsrooms to use AI tools
Continue reading ›
Nvidia Burns $150B on Buybacks While AI Competitors Build
Nvidia just told the market it can't find better uses for $150 billion than buying its own shares—in the middle of what's supposed to be the AI buildout of the century. Nvidia announced a $150 billion stock buyback, the largest repurchase program in corporate history
Continue reading ›
Gecko Robotics' $100M Bet: AI Won't Touch Critical Infrastructure Without Human Override
The robots are leaving the data centers, and the guy building them wants you to know they're on a leash. Gecko Robotics is partnering with Nvidia on an open safety platform for physical AI agents, setting boundaries for autonomous systems in industrial and military environments
Continue reading ›
Nvidia Builds Containment Tech for AI It Helped Create
The company selling the shovels for the AI gold rush just admitted we need jail cells for the miners.
Continue reading ›
Local Homeowners Are Killing America's AI Data Centers
America's AI infrastructure is getting vetoed at the town hall level, and the industry's playbook of "jobs and tax revenue" isn't working anymore.
Continue reading ›
SoftBank Issues Largest Junk Bond Ever to Finance AI Infrastructure
SoftBank just issued the largest junk bond in history to fund AI bets, and the deal structure reveals how desperate capital markets are to finance the infrastructure layer of Web4.
Continue reading ›
Akamai Bets $12B Against Hyperscalers in Anthropic Deal
Akamai just bet $12 billion that the future of AI isn't built in hyperscaler data centers. Akamai signed a seven-year, $12 billion AI computing supply agreement with Anthropic, marking one of the largest infrastructure commitments in frontier AI history
Continue reading ›
NASA Handed Mars Rover Control to an AI. Engineers Are Panicking.
NASA just let an LLM plan a Mars rover's route — and that should terrify every space engineer who's ever written a mission-critical function. NASA's JPL used Anthropic's Claude to plan two Mars drives for Perseverance in December 2025, with humans checking the route before upload
Continue reading ›
AI Agents Will Outnumber Human Workers by 2025
The org chart just became a species diagram. AI agents are moving from tools to coworkers, forcing companies to rethink everything from onboarding to performance reviews Workplace collaboration models built for humans don't scale to entities that work 24/7, never take breaks, and improve every week
Continue reading ›
Nvidia Just Solved AI Agents' Biggest Security Problem With Open Code
The agent economy's biggest trust problem just got a reference implementation from the chipmaker that powers it. Nvidia released OpenShell, an open-source runtime specifically designed to secure autonomous AI agents as they execute tasks independently
Continue reading ›
AI Companies Are Mining Your Game Saves to Train Robots
The path from your Elden Ring deaths to self-driving cars is shorter than you think. British startup Rerun is converting human gameplay into training data for embodied AI — teaching robots and autonomous systems to navigate physical space by watching how humans solve virtual worlds
Continue reading ›