GPT-5.5: Autonomous Agent – A System, Not a Model
agentOpenAI's GPT-5.5: No human intervention. Scores 82.7% on Terminal-Bench 2.0. Is this a model upgrade or a systemic shift? Implications for cognitive work.
Read Report →