source: simon willison: your ai use is breaking my brain
level: technical
jason koebler writes about how ai-generated text is now unavoidable online. he calls this the zombie internet, a mess where people use ai to talk to other people, bots, and hybrid accounts. it is not just bots chatting with bots. real users deploy ai agents to interact with others, and ai-assisted writing mixes with human posts. this creates a confusing environment where it is hard to tell what is real.
the mental effort of filtering ai content is draining. koebler notes that even human writing styles are changing because of ai influence. the zombie internet includes ai summaries sold as books, fake heartfelt advice on reddit, and automated social media accounts run by marketing firms. platforms like linkedin and x are filled with this noise. the line between human and machine is blurring, making everyday reading a chore.
simon willison highlights koebler's term as a more accurate description than the dead internet theory. the dead internet imagined bots talking to bots, but the zombie internet is messier. it involves real people using ai tools, often for profit, and the result is a polluted information space. this shift affects how we trust content and how we write ourselves, as ai patterns seep into human expression.
why it matters: for ai and data science, this shows how generative models are reshaping online communication, creating challenges for content authenticity and human-computer interaction.