Bug Report: Issues with Video Upload and Playback in ChatGPT
Issue Description
- Inability to Watch YouTube Videos Directly
- Previously, ChatGPT could summarize or analyze YouTube videos based on links. Now, it cannot access them at all, requiring the user to manually describe the content.
- Google Photos Video Links Are Not Viewable
- When sharing a Google Photos video link, ChatGPT cannot fetch or preview the video, making it difficult to discuss its content.
- Uploaded Videos Cannot Be Viewed or Processed
- After uploading a video file directly to ChatGPT, the AI cannot play or preview the content.
- There is no auto-transcription, keyframe extraction, or video summary, making the uploaded file almost useless within the chat.
- Users must describe the video manually, which negates the purpose of file uploads.
Suggested Improvements
- Allow ChatGPT to fetch basic metadata or summaries for public YouTube videos (e.g., title, description, transcript).
- Enable Google Photos video preview or basic metadata extraction to facilitate discussions.
- Improve video upload functionality by adding:
- Auto-transcription (speech-to-text conversion).
- Keyframe extraction (generating representative snapshots).
- Basic video summarization (content analysis).
- Provide an embedded video player to watch uploaded videos within the chat.
This issue significantly impacts usability and user experience, especially when discussing video content. Hope OpenAI can address these limitations in future updates!
more idea?
contact me: doffwu0409@gmail.com
doffwu (Taiwan) 250311