A follow-up query in regards to the ultimate rating was answered accurately, however Gemini bought the identify of the scorer of the primary landing mistaken: The AI prompt it was Johan Dotson. Dotson was proven getting a landing within the highlights with the scores at 0-0, nevertheless it was dominated out—an instance of the nuances that AI does not essentially choose up on.
Gemini did efficiently determine when the Kansas Metropolis Chiefs bought their first factors, and even included a timestamp linking straight to the landing within the YouTube clip. It additionally bought the identify of the scorer proper. It appears Gemini is closely reliant on the commentary for sports activities clips, which is not stunning.
Summarize Video Contents
Subsequent, we tried placing Gemini up in opposition to a behind-the-scenes featurette for The Grand Budapest Lodge, directed by Wes Anderson. The clip runs to four-and-a-half minutes, and Gemini fired again some replies virtually immediately: It recognized the identify of the movie being talked about, and the principle beats of the clip’s narrative.
Nonetheless, it is all reliant on the audio (or the transcript) once more—there does not appear to be any evaluation of the particular video contents. The AI could not say who the speaking heads had been within the video, regardless that their names had been proven on display screen, and wasn’t in a position to say who the director was (regardless that this was additionally talked about within the video description).
On the plus facet, Gemini did do a formidable job of summing up the audio of the video. It accurately recognized a few of the filmmaking challenges that had been talked about all through, and supplied timestamps to them — from in search of a set to symbolize the Grand Budapest, to filling it with extras.
Summarize Interviews
Lastly, we tried Google Gemini with an interview: Channel 4 within the UK chatting with Charlie Brooker and Siena Kelly in regards to the newest sequence of Black Mirror (maybe acceptable for an article on AI). Gemini proved itself very succesful at selecting out the speaking factors, and including timestamps, although after all the entire video is usually speaking.
Once more although, there is not any context about something exterior of the audio or the transcript. Gemini AI could not say the place the interview befell, or how the individuals had been appearing, or the rest in regards to the visuals of the video—which is price allowing for in the event you use it your self.
For movies the place the solutions you need are within the audio of a YouTube video, and its related transcript, Gemini works rather well at summarizing and offering correct solutions (supplied the commentators point out when a landing is dominated out, in addition to when one is scored). For any sort of visible data, you are still going to have to look at the video your self.