Tech

• AI is giving bad advice to flatter its users, says new study on dangers of overly agreeable chatbots
• Can AI understand literature? Researchers put it to the test
• AI model excels in single image reflection removal
• New detector chip compresses X-ray data 100- to 200-fold in real time
• Highly performing AI agents can still fail to spot deception, study finds
• A better method for identifying overconfident large language models
• Sheepdogs reveal a better way to guide robot swarms
• New model uses team interactions to predict next steps in collaborative work
• What flocking birds can teach AI about reducing noise
• Optimization method may expand impedance relays into medium-voltage distribution networks
• AI gets a D: ChatGPT struggles with scientific true-or-false, study shows
• AI's game-playing still has flaws: AlphaZero-style self-play tested on Nim
• Shortest paths research narrows a 25-year gap in graph algorithms
• Human brain and AI speech recognition decode speech in similar step-by-step stages, study finds
• AI is homogenizing human expression and thought, computer scientists and psychologists say