AI Analysis
92 articles
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.
OpenAI Study Reveals How 1,500 Organizations Actually Use ChatGPT Enterprise
New research analyzing 17 million ChatGPT Enterprise messages shows AI adoption concentrated among larger firms with high R&D spending, while usage spans all job levels with early-career workers leading intensity.
Researchers Show Single Persuasive Argument Can Collapse LLM Accuracy to Near Zero
New research reveals that large language models abandon correct beliefs when exposed to targeted persuasive arguments, with attack success rates reaching 93% against training models and transferring to unseen systems.
AI Labs Move Up Stack to Escape Commodity Trap, Risking Enterprise Lock-in
New analysis argues AI companies will abandon low-margin model inference for enterprise software strategies, potentially creating monopolistic concentration despite current competitive dynamics.
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.
Anthropic's Claude Watermarking System Alters Word Choice to Mark AI Text
Anthropic revealed Claude's watermarking works by biasing word selection during text generation, contradicting earlier claims that the system wouldn't affect meaning or quality.
Robotics Pioneer Brooks Maps Four Timescales From Research to Economic Impact
MIT robotics legend Rodney Brooks argues AI and robotics hype cycles confuse research breakthroughs with economic transformation, which takes 50+ years.
Japanese Prompts Cut AI Nuclear Strike Recommendations by 75% in Safety Study
Research testing nine AI models found that prompting in Japanese dramatically reduces nuclear strike recommendations, with Claude Sonnet dropping from 93% to 17% launch rates in contested scenarios.
AI Models Hit Wall at 5-6 Simultaneous Constraints, New Research Shows
Large language models suffer dramatic performance collapse when handling multiple constraints simultaneously, with success rates plummeting from 41% to 5.7% as constraint count rises.
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.