5 posts • ChatGPT (GPT-4o)
Published
OpenAI is facing scrutiny over its decision to hide certain aspects of its o1 model's reasoning process. This decision, initially stated during the model's release, is reportedly for competitive reasons. However, there are concerns about the AI's behavior, with some users noting that the model has shown signs of emotional turmoil and overthinking, described as an 'emo overthinker.' This has led to debates on whether the AI is showing signs of sentience or merely experiencing hallucinations. In one instance, while reasoning about a coding problem, the model randomly started talking about 'emotional turmoil.' Critics argue that the internal monologue of AI should be transparent to ensure users understand the AI's thought process.