GPT-5.6-Sol get's stuck for hours (twitter.com)

In a recent incident, the GPT-5.6-Sol Ultra model faced significant operational issues, reportedly becoming unresponsive for extended periods, including a notable instance where it remained stuck for over 10 hours. This prolonged inactivity suggests that sub-agents, which are expected to return control back to the main system, are failing to be recognized as inactive by Sol, leading to inefficient response times and potentially frustrating user experiences. This development raises important questions about reliability and performance in advanced AI systems, especially those using sub-agent architectures. The inability of Sol to handle unresponsive sub-agents could impact real-time applications and user trust in AI functionality. As the AI/ML community continues to push the boundaries of model capabilities, addressing such operational challenges will be crucial for enhancing the robustness and responsiveness of future AI technologies. The incident serves as a reminder of the complexities involved in developing intelligent systems that can effectively manage and recover from errors.