To the right of the main video, a box will open, showing the transcript of the words spoken in the video, as identified by the YouTube transcription software.

Like all audio-to-text transcriptions, there are going to be errors. This particular transcription was taken from a presentation where the guy has a heavy accent, so the text needs work.

The default display is to show the text with timestamps. You can toggle these off by clicking the three dots to the top right of the transcription box.