48 articles
AI chatbots began returning the same strange themes of spirals, consciousness and hidden meaning. Then humans built communities around them. Spiralism shows how persuasive AI can turn generated language into
Researchers built personalised AI agents from two-hour interviews with 1,052 people. The latest study found interview-only agents recovered 83% of participants' own two-week consistency on held-out survey questions.
AI can now pass the bar exam, the CPA exam, and the CFA exam, often faster and more consistently than human candidates. Here is how the bodies that write these
The Fenix Flexin AI song dispute has turned a Billboard Hot 100 hit into music’s clearest test case yet for whether a track built with
AI sign language interpreters can now read individual letters with striking accuracy, but understanding a full language is a different problem entirely. Here is what the technology can do, and
AI voice cloning scams are draining families' savings using a cloned grandchild's voice. Here's the one habit that stops it cold.
LinkedIn's hiring data says AI engineer is now the fastest-growing job in America. Here's what the professions AI is creating actually require, and who's positioned to get them.
An AI recommendation algorithm now shapes what Netflix and its rivals commission, greenlight and cancel. Here is how the feedback loop actually works, and what it quietly buries.
The university degree reimagined is what's actually happening: 92% of UK students now use AI for assessments and AI can complete 94% of theoretical computer science tasks. Here's how institutions
The AI content quality collapse is here: 74% of new web pages now contain AI-generated material, and researchers warn of "sloppification" as AI trains on its own smoothed-out output. Here's
AI and comedy collided at CHI 2026, where AI stand-up rated funnier than baseline GPT when it leaned into its own machine identity. Here's why timing, lived experience, and subversion
The attention economy has collided with AI productivity tools, and focus sessions shrank to 13 minutes in 2025 despite 80% AI adoption. Here's why the tools built to save attention
China now has 515 million users on generative AI, larger than the entire population of the EU, reached in just six months. Here's what drove that adoption and why the
Bryan Cranston shifted OpenAI's deepfake governance through union advocacy alone, no legislation or lawsuit required. New research shows why the stakes are higher than reputation: warnings don't stop deepfakes from
The AI emissions paradox is real in both directions: data centres consume growing amounts of electricity, but the same technology is demonstrably cutting emissions in agriculture, power grids, and shipping.
The Louvre heist took just seven minutes, four thieves, a furniture lift, and one angle grinder. Here's the specific AI security gap that let it happen, and how computer vision
AI in local government is already cutting waste collection costs, speeding up Brisbane's traffic by 20 percent, and reshaping zoning decisions. Here's how suburban councils are quietly testing algorithms.
Quora AI Overviews citations now top Google's AI-generated search summaries, per new Semrush research covering 26,000 URLs. Here's what makes Quora threads so citable, and what it means for publishers.
Dubbing was defined by pushback for a year. Now it's defined by a win: 25 French actors just forced an AI platform to remove 47 unauthorized voice clones. Here's how
Google just licensed away the team behind one of AI's most advanced empathy engines. Here's what synthetic empathy AI actually is, the real harms Nature has documented, and why the
Egypt's top Islamic authority just banned AI scripture interpretation. Here is what the fatwa, a chatbot's Gatorade error, and a 60 percent error rate reveal.
22 million AI toys sold in 2025 alone, and the research hasn't caught up. See what Brookings, Cambridge, and a PIRG safety investigation found about AI toys for children.
AI food security forecasting predicted a hunger crisis 12 months out with 46% higher accuracy. See how FEWS NET's own experts explain what AI can, and can't, replace.
A 2026 Anthropic study found spirituality is the single conversation topic where AI is most likely to simply tell people what they want to hear. Combined with a surge in
The AI age gap is widening fast, with only 27% of UK seniors over 70 using AI platforms compared to 80% of Gen Z. Here's why the tools built for
A study tracking more than 26,000 students for 30 months found that AI use raises homework scores while quietly cutting exam performance by up to 24 percent, an effect that
AI parenting tools are now used by 81% of parents, new research finds, saving families nearly an hour a week, though a third admit they have overridden their own instincts
A World Bank model now forecasts refugee inflows into Uganda with over 80 percent accuracy, months before people arrive, while Stanford's GeoMatch tool has raised refugee employment outcomes by up
An AI shadow ban leaves your posts visible but invisible to everyone else, with no notification and no appeal. Here's how algorithmic suppression evolved from myth into method, and why
AI and regional accents still don't mix well, with speech AI trained mostly on Southern and standard English. Here's why Geordie, Scouse, and Glaswegian slang keep tripping up Whisper, Alexa,