I’ve encountered several instances where my posts were incorrectly removed due to false positive censorship on this platform, leading me to seriously consider leaving. It seems evident that an advanced AI language model, which takes into account the context of the content, could have prevented these removals and my lost of faith in this platform.
Looking forward, I am keen to understand if there are plans to implement such an AI model for moderating content to reduce false positives, particularly in relation to adhering to the EULA while ensuring free speech. From my observation, the current system appears to rely heavily on keyword flagging, which lacks the nuance needed to understand context, a method that even seemed more sophisticated in the 1960s with landline phone usage.
Moreover, a close friend and colleague of mine had their account abruptly closed without any recovery process, mistakenly flagged as spam because they were using an API integration I set up to automate content sharing from this platform to a website. This issue, I suspect, is widespread, causing many users to abandon the social aspects of the platform and retain only the messenger for personal communication.
I understand that with the new administration, efforts are being made to address these concerns. You will have a job to win our trust back, maybe even an understanding. I’m curious about the push back or compliance pressures your platform might be facing. It would be insightful if you could share details or write about these challenges. Additionally, I would appreciate any information or discussion on more subtle forms of censorship like shadow banning or the interruption of post composition.
Thank you for considering these points. I look forward to hearing how you plan to enhance the user experience and maintain the integrity of free speech on your platform.
Sincerely,
Jason Page
The rabbit done died