AI Analysis
65 articles
Length Penalties in AI Reasoning Models Reduce Transparency Without Stopping Bias
New research shows that training AI models to produce shorter reasoning chains reduces monitoring ability while hidden biases continue to influence answers.
New benchmark reveals AI models hallucinate in 45% of clinical trial summaries
Researchers developed a framework testing GPT-4o, Claude, and Gemini on clinical trial summaries, finding widespread hallucinations that pose risks for healthcare decisions.
Socratic AI tutors outperform prompt-training approach in programming education study
Research comparing two AI tutoring methods found that Socratic dialogue-based tutors produced better long-term learning outcomes than prompt-refinement tutors in a graduate robotics course.
LLMs Score 96% on CBT Theory But Fail Clinical Application
New research reveals large language models excel at cognitive behavioral therapy licensing exams but struggle to apply therapeutic techniques effectively in practice, defaulting to validation responses regardless of patient needs.
OpenAI research shows agents handling 8-hour tasks as Codex adoption soars
New OpenAI research reveals 25.6% of individual users now delegate tasks exceeding eight hours to Codex agents, with non-developer adoption rising 137x since August 2025.
AI advice eliminates human willingness to say 'I don't know', study finds
New research shows people become three times less accurate when AI advice is available, even when the advice is wrong and accuracy is financially incentivized.
AI Futures Project releases 2040 scenario delaying superintelligence to avoid rushed development
The AI Futures Project published AI 2040: Plan A, a detailed scenario recommending governments delay superintelligence development from 2030 to 2040 through decisive regulatory action.
Global Startup Investment Hits Record $510B in H1 2026 as AI Boom Accelerates
Venture funding reached a record $510 billion in the first half of 2026, with OpenAI and Anthropic alone capturing $217 billion as exits returned to 2021 levels.
Researchers Introduce Idiobionics Framework for AI-Powered Prosthetic Privacy
University researchers have coined 'idiobionics' as a new field examining privacy risks in intelligent robotic prostheses that use AI sensors and control systems.
Researchers Propose Framework for AI-Human Information Warfare
University researchers outline adversarial social epistemology theory to analyze how humans and AI systems exploit trust mechanisms in shared information environments.
Frontier AI Models Compete Then Collaborate to Train Better Coding Students
New research shows four leading AI models working together through competition and collaboration can improve smaller coding models more effectively than traditional imitation learning approaches.
Researchers Publish Mathematical Theory of Slow Thinking for Large Language Models
Four researchers have published a first-principles mathematical framework called 'active lifting' that formally derives slow thinking processes in large language models and active perception systems.