OpenAI

"OpenAI's defensive patent pledge faces skepticism as they block competitor access to training data while defending against copyright infringement suits."

ChatGPT showed minimal bias across gender, race, ethnicity. LMRA analysed millions of interactions. <0.1% cases had harmful stereotypes. Newer models improved.

OpenAI's MLE-bench tests AI on ML tasks. Best AI setup achieved bronze-level in 16.9% of competitions. Study explores resource-scaling and contamination effects.

OpenAI partners with Hearst to integrate content from 20+ magazine brands and 40+ newspapers into ChatGPT, enhancing information base.

OpenAI secures $6.6B in funding and $4B credit facility, totaling $10B+ in liquidity to fuel growth, research, infrastructure expansion, and talent acquisition.

OpenAI's Canvas offers inline AI feedback, shortcuts for tasks like adjusting text length and debugging code, and allows users to highlight sections for focused AI assistance.