As we have become more connected than ever, we find ourselves becoming less social than ever before. The internet has heralded the corrosion of working time and free time, self and other, and human versus machine. As users confront these different contradictions they are forced to reassess themselves, and try to generate an explanation with limited information and resources. If anything is clear, it seems that the internet is no longer the progressive, democratizing force that it was once promised as. The mutual co-creation of new social worlds has been replaced by duelling paranoia and narcissism, all while our data are bought, sold and analyzed behind our backs. Some users now claim that the internet is dead, populated only by bots, algorithmically generated content, and AI lurking behind a facade of endless feeds and targeted ads. This thesis seeks to understand this perspective, sympathetic to those who find themselves disenfranchised by the internet today, and critical of the infrastructure that enables it. In order to shed light on and describe this phenomenon, it investigates three case studies, each of which represents different instances that are cited as reasons for the supposed death of the internet, and involve artificial intelligence. These are, respectively, the presence of bots on social media platforms, the proliferation of AI generated media, and finally the ubiquity of AI powered algorithms and decision making systems in our lives. In his 1984 cyberpunk classic Neuromancer, William Gibson wrote “For thousands of years men dreamed of pacts with demons. Only now are such things possible” (p.165). This thesis seeks to understand the terms of that pact, what the demons are, and what we are giving up in exchange.