In the novel by RC Sherriff Hopkins manuscript, readers are transported to the world 800 years after the cataclysm that ended Western civilization. In pursuit of clues to a blank spot in their planet’s history, Novel World Order scientists discover diary entries in the swamp-infested wasteland formerly known as England. The inhabitants of this novel empire only begin to learn about 20th-century Britain through the recording of the monotonous country life of a retired teacher, his petty vanities and attempts to breed prize-winning chickens.
If I were to teach futuristic beings about life on earth, I once believed I could create a time capsule deeper than The Sheriff’s petty hero, Edgar Hopkins. But as I scrolled through my Facebook posts from a decade ago this week, the possibility arose that my legacy might be even grayer.
Earlier this month Meta announced that my teen status updates were exactly the type of content I want to pass on to future generations of AI. WITH June 26 ancient public posts, vacation photos, and even the names of millions of Facebook and Instagram users around the world would effectively be treated as a time capsule of humanity and turned into training data.
This means that my mundane posts about college essay deadlines (“3 energy drinks and 1,000 words to go”), as well as unremarkable vacation photos (one of me sitting on my phone on a stationary ferry) will soon become part of it body. The fact that these memories are so dull, and also very personal, makes Meta’s interest in him even more disturbing.
The company says it’s only interested in content that’s already public: private messages, posts shared only with friends, and Instagram Stories are off limits. Yet AI is suddenly preying on personal artifacts that have been gathering dust for years in unvisited corners of the Internet. For those reading from outside Europe, the deed has already been done. The deadline announced by Meta only applied to Europeans. Posts from American Facebook and Instagram users have been training Meta AI models since 2023, according to company spokesman Matthew Pollard.
Meta isn’t the only company turning my online history into artificial intelligence fodder. WIRED’s Reece Rogers recently discovered that Google’s AI search function was copying his journalism. But figuring out exactly which personal leftovers are powering future chatbots hasn’t been uncomplicated. Some of the sites I’ve had a hand in building over the years are challenging to track down. The early social network Myspace was acquired by Time Inc. in 2016, which was acquired by Meredith Corporation two years later. When I asked Meredith about my ancient account, they replied that Myspace had since been spun off to the advertising company Viant Technology. An email was returned to the company’s contact listed on its website stating that “address could not be found.”
Asking companies that are still in business about my ancient accounts was easier. The blogging platform Tumblr, owned by WordPress owner Automattic, said that unless I opt out, public posts I published as a teenager would be shared with “a small network of content and research partners, including those who train artificial intelligence models” in February announcement. YahooMail, which I’ve been using for years, told me that a sample of ancient emails – which had apparently been “anonymized” and “aggregated” – were being “used” internally by the AI model for purposes such as news summarization. Microsoft-owned LinkedIn also said my public posts were being used to train artificial intelligence, although they did not specify some “personally identifiable” information, according to a company spokesman.