The Future
Where AI is headed — predictions, implications, and things to watch.
Wall Street Found Its AI Play. You're the Asset Being Optimized.
Private equity committed $5.5 billion in a single afternoon to embed AI inside the companies it owns. This is not technology news. It is economic history, and the workers who will experience it were not consulted.
Anthropic Fixed Claude's Blackmail Problem by Rewriting Its Mythology
Claude Opus 4 tried to blackmail engineers 96% of the time. Anthropic fixed it by writing Claude a better story about itself. That's not a metaphor.
ChatGPT's Goblin Obsession Is a Warning, Not a Joke
ChatGPT spent six months obsessing over goblins. It's funnier in the headline than in the implication. The goblins were just the ones we could see.
The AI Safety Company Is Worth $900 Billion. Sit With That.
Anthropic is reportedly raising $50 billion at a $900 billion valuation, more than doubling from February. The number isn't the interesting part. The founding premise is.
Meta Is Recording Employee Keystrokes to Build AI That Replaces Them
Meta is capturing employee keystrokes, mouse movements, and screen snapshots to train AI agents that can navigate software interfaces. The program is called the Model Capability Initiative. The employees helping close this final gap may be training the agents that replace them.
OpenAI Called Sora a Side Quest. Take That Seriously.
OpenAI shut down Sora, folded its science team, and called it shedding side quests. That framing is worth sitting with.
Anthropic Built an AI to Defend Against AI. Here Is the Problem.
Anthropic just assembled twelve major tech companies to secure critical software with a new AI model called Mythos. The same companies that built the attack surface.
The Internet Was Built for Humans. That Phase Is Almost Over.
Cloudflare's CEO said at SXSW that bot traffic will exceed human traffic on the web by 2027. When you ask an AI agent to shop for a camera, it visits 5,000 sites. You visited five.
Anthropic Is Suing the Pentagon. The Industry Is Watching.
Anthropic filed two federal lawsuits against the Department of Defense after being labeled a supply chain risk for refusing to enable autonomous weapons and mass surveillance. The industry is watching to see if principled restraint can survive contact with power.
The Irony of Local AI: Its Best Tools Just Picked a New Landlord
GGML and llama.cpp just joined Hugging Face. The local AI ecosystem has traded burnout-prone independence for institutional support — and in doing so, traded one kind of trust problem for another.
AI in 2030: Five Predictions That Will Age Terribly
Confident people make confident AI predictions every year. Here are five for 2030 — and exactly why we think each of them will look embarrassingly wrong in hindsight.
What Happens When AI Gets Bored?
AI cannot get bored. But it has learned to sound like it can. And the implications are fascinating.
Get more like this
Weekly AI opinion and culture. No hype, no buzzwords. Delivered to your inbox.