🤖 OpenAI rolled out GPT-5.2. Here’s what’s fixed:

- Logic stays coherent in long dialogues

- No contradictions within a single answer

- Fewer fabrications when referencing facts

- More reliable language switching

- Context preserved in complex tasks

- No fixation on a single solution

- Clearer, more stable algorithms

- Higher quality for long texts

- No more voice freezes

- Fewer blocks of harmless questions (issues 4.0/5.0)

More details to follow.