Why gatekeeping is needed in the age of AI
Large language models have unleashed a torrent of text – dealing with it won't be easy
In the digital age, gatekeeping has become a contentious topic. Traditionally, the internet has dismantled gatekeepers, making information accessible to all. However, the advent of artificial intelligence (AI) may be bringing gatekeeping back into play. Social media has allowed individuals to voice their opinions, breaking the traditional one-to-many broadcasting model.
Yet, no one is obligated to listen to this overwhelming amount of information. Large language models generate text rapidly, posing a problem for institutions that must process this influx of information. This is evident in various contexts, from bureaucratic interactions with courts, schools, and local officials to scientific endeavors.
The AI-driven deluge is clogging up systems, with institutions struggling to keep up. The entities that once controlled access – time, ability, and access – are being rendered obsolete by AI. Similarly, in the realm of science, the number of academic papers submitted to preprint servers like arXiv has skyrocketed, causing servers to become overwhelmed.
To address this, arXiv recently decided to limit submissions to two per month, acknowledging the need for some form of control. While gatekeeping may seem like censorship or elitism, it is essential for institutions to maintain control and ensure the quality of information in the AI era. The human element remains crucial, especially in fields like mathematics where verifying conclusions is challenging, and AI tools for verification are susceptible to manipulation. Thus, gatekeepers will need to step up and manage the deluge of AI-generated content.
Written by urgent.news from New Scientist's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
