Feedback on ChatGPT’s Read Aloud feature’s recent tonal changes

The recent changes to the ChatGPT Read Aloud feature have significantly impacted its quality and usability. The voice now sounds robotic, with unnatural fluctuations in tone and rhythm, compared to previous versions that were much more natural and comfortable to listen to. This degradation in listening quality affects usability and makes the feature less effective for users who rely on it.

Additionally, the functionality itself has persistent technical issues. The feature often resets after 2.5 minutes, requiring users to manually skip ahead to resume where it left off. It also does not read beyond 6–7 minutes, leaving longer responses unfinished. These problems disrupt the experience, making the feature unreliable for extended use.

Instead of improving over time, the quality of the Read Aloud feature has declined, which is disappointing for a product from a company known for innovation. Combined with the addition of irrelevant features, such as emoticons in responses, these changes give the impression that maintaining and improving core functionalities is no longer a priority. Users expect reliability and polished performance, and the current state of Read Aloud does not meet these expectations.