X @Polyhedra

AI Model Complexity & Verifiability - Model mode complexity highlights the importance of verifiable AI [1] - Cryptography can prove which model was used and that it executed correctly, eliminating the need to trust labels blindly [1] OpenAI Model Confusion - OpenAI's attempts to simplify model selection may have inadvertently increased confusion regarding model usage [1] - It's unclear how the different models available in ChatGPT map to those available in the API [1] - The relationship between "ChatGPT-5 Thinking" and GPT-5 is ambiguous [1]