Survey Shows Americans Struggle to Spot AI-Generated Deepfakes
A recent quiz and survey reveal that Americans can barely distinguish AI-created deepfake images from real ones, with accuracy rates close to random guessing
Archive 
Google Updates Gemini AI Usage Rates and Subscription Tiers
Google has revamped how usage quotas are calculated for its Gemini AI, affecting free and paid users with new limits based on computing power…
MIT’s Bailey Flanigan Advances Democratic Participation Through Computational Research
Bailey Flanigan, a multi-disciplinary researcher at MIT, develops computational tools to enhance democratic processes and decision-making through her joint computer science and political science…
The AI-Powered World Cup and Its Hidden Data Labor Chain
The 2026 World Cup has embraced AI-driven technologies like VAR and real-time analytics, transforming soccer while exposing a complex labor and economic chain behind…
ORCA Computing’s Quantum AI Advances Peptide Discovery for Drug Development
British startup ORCA Computing’s quantum computer has accelerated AI-driven peptide prediction, enhancing precision in drug research, including vaccines and immunotherapies
Air Force Cadet Develops Military AI Tool Using Leading Chatbots
U.S. Air Force cadet Joshua Lynch created a prototype military application with AI chatbots from Anthropic, OpenAI, and Google, demonstrating novice-driven coding in defense…
MIT Exhibit Explores Algorithmic Aesthetics and Computational Design
MIT researcher Alexandros Haridis presents “Beyond Data-Driven Aesthetics,” an exhibition translating computational theories and AI models into interactive design installations through June 30
Understanding Agentic AI: Current Developments and Implications
Agentic AI systems can take actions digitally or physically and are rapidly deployed by businesses, raising important questions on capabilities, risks, and future directions
Research Warns AI Risks Making Human Choices Predictable and Dull
New academic study reveals large language models may push users toward common, risk-averse decisions, potentially reducing diversity of preferences and creativity
MIT’s New AI Method Helps Robots Follow Vague User Instructions
MIT researchers developed Masked IRL, a system using large language models to help robots understand ambiguous commands and prioritize key task details with less…