For example, in 2006, when Facebook decided to open its doors to the public – not just college kids with .edu addresses – they understood that most people interested in social media already had accounts on Myspace, a service that had been sold to master enshittifier Rupert Murdoch the year before. Myspace users were champing at the bit to leave, but they were holding each other hostage.
…
Those live, ongoing connections to people – not your old posts or your identifiers – impose the highest switching costs for any social media service. Myspace users who were reluctant to leave for the superior lands of Facebook (where, Mark Zuckerberg assured them, they would never face any surveillance – no, really!) were stuck on Rupert Murdoch’s sinking ship by their love of one another, not by their old Myspace posts. Giving users who left Myspace the power to continue talking to the users who stayed was what broke the floodgates, leading to the “unraveling” that boyd observed.
Wow, I had no idea about the Facebook/MySpace message bridge bot. Definitely shows the power and importance of the various bridge/mirror projects like Bridgy. It says that the same kind of bot would now be Fedrally illegal in the US, but I haven’t seen any specifics about that, and seems like the EU just made it mandatory to enable through APIs.
I have thought a bit about this and how to breakout of silos, and it seems like now with LLM tools accessing the browser it will be nearly impossible to prevent messaging and posts from being cross-platformed, though the compute cost would be higher than by using the old API method.
https://pluralistic.net/2024/12/14/fire-exits/
Wow, I had no idea about the Facebook/MySpace message bridge bot. Definitely shows the power and importance of the various bridge/mirror projects like Bridgy. It says that the same kind of bot would now be Fedrally illegal in the US, but I haven’t seen any specifics about that, and seems like the EU just made it mandatory to enable through APIs.
I have thought a bit about this and how to breakout of silos, and it seems like now with LLM tools accessing the browser it will be nearly impossible to prevent messaging and posts from being cross-platformed, though the compute cost would be higher than by using the old API method.
You don’t need llm to import posts from another website, just an API or scraper to fetch them. Much cheaper, faster and more reliable.