I put GPT-5.5 through a 10-round test: It scored 93/100, losing points only for exuberance (zdnet.com)

OpenAI's latest model delivers powerful results but sometimes ignores simple directions, creating a tension between intelligence and control.