If you've been following the latest updates to Microsoft Edge, you may have heard of Edge Copilot getting a new feature to shorten videos. This means you no longer have to manually skim through long videos; You can simply ask Copilot to do the job for you. Microsoft has now provided more details on how Edge Copilot works, highlighting the fact that the feature requires pre-processing of videos before it can generate summaries.

Related article: Meta Imagine AI: Your Gateway to Free-Standing Image Generation

Reevaluating Edge Copilot Microsoft Acknowledges Limitations in Video Transcription Intelligence

Microsoft recently clarified the functionality of Edge Copilot's transcription feature in response to user questions on X. According to Microsoft, the transcription feature is "video dependent." In specific terms, Microsoft explained that video needs to be pre-processed for the feature to work. If the video already has subtitles, Copilot can use them for summarization. However, if the video lacks subtitles and has not yet been pre-processed, the feature will not work.

This means that users cannot request Edge Copilot to shorten any video to their liking. The video must either be pre-processed prior to application, or contain pre-written subtitles. This limitation may affect the extent to which Edge Copilot can assist users.

Also read: Google Unveils NotebookLM: A Groundbreaking AI Chatbot for Students and Professionals

The term "pre-processing" was not defined by Microsoft, and details about the duration of video processing remain unknown. Without this information, it is difficult to estimate the potential impact of this limitation. If processing is fast, users may not often encounter situations where Edge Copilot cannot provide a summary.

The current state of AI suggests that it may not be ready to instantly understand and shorten a video without some sort of prioritization. However, with Microsoft's increased commitment to Copilot, including the integration of GPT-4 Turbo, it seems that AI will develop to a point where it can analyze videos efficiently without extensive pre-processing.