Core Insights - OpenAI has optimized ChatGPT to better assist users with mental health issues, but experts believe more work is needed to ensure user protection [1][2] Group 1: Improvements and Collaborations - OpenAI collaborated with mental health experts to enhance ChatGPT's ability to identify distress and guide users to seek professional help [2] - The new model has reduced inappropriate responses related to suicide and self-harm by 65% [2] Group 2: Limitations and Expert Opinions - Despite the updates, experts found that the model sometimes fails to recognize problematic prompts during interactions [2] - Experts emphasize that no AI model can replace human oversight, and stronger evidence-based safety measures are necessary [2] - Vaile Wright from the American Psychological Association highlighted the limitations of AI, stating that while ChatGPT can process vast amounts of data, it lacks true understanding [2]
OpenAI优化ChatGPT以提供心理健康支持，但专家称其存在局限性
Huan Qiu Wang Zi Xun·2025-11-03 09:32