Category: AI
News and analysis on AI models, chips, infrastructure, applications and the business shifts behind machine intelligence.
-
The Invisible Human Labor Behind Every AI Model
A Brown University student’s research maps the global workforce labeling, cleaning, and tuning AI data—revealing the human labor hidden behind machine intelligence.
-
Terran Robotics Bets AI Robots Can Master Homebuilding
Terran Robotics is using AI and robots to build houses, betting automation can ease construction’s labor shortage and speed up homebuilding.
-
Your Lawyer’s AI Chatbot May Be a Privilege Trap
AI tools are creeping into legal work, but attorney-client privilege may not survive the trip. Here’s what firms and clients need to know.
-
AI Is Learning to Hack, and Cyber Defenses Are Struggling
AI is now capable of conceiving and executing cyberattacks. Here’s why cybersecurity teams are struggling to keep pace with autonomous threats.
-
Teradyne and Sereact Bet on Robots That Think on Their Feet
Teradyne Robotics is teaming with Sereact to bring physical AI to warehouses and stores, a bet that adaptable robot brains matter as much as hardware.
-
A 10-Year-Old CNN Is Beating Frontier LLMs at Skin Cancer
A decade-old CNN reportedly outperforms frontier LLMs at skin cancer diagnosis, raising hard questions about how we evaluate and deploy AI in medicine.
-
AI Is Learning to Hack. Security Teams Are Racing to Keep Up
AI models are getting better at finding software flaws, and security teams are turning to AI to fight back. Here’s what that arms race means for defenders.
-
Lingyan’s Bionic Face Bet: Does Emotional AI Sell?
Lingyan raised over ¥10M for its bionic face tech. We examine whether emotional robotics can move past novelty into real commercial demand.
-
AI Is Coming for Your Money, Not All Money
AI is reshaping finance from trading to lending. We explore who benefits, who gets squeezed, and why the “all money” panic misses the real story.
-
AI Data Centers Have a Ghost Problem Nobody Budgeted For
The AI data center boom is leaving behind ghost infrastructure: half-finished shells, stranded capacity, and idle GPUs that nobody planned to pay for.