AI Extinction Risks Were Known Decades Ago, Author Argues

In recent weeks, public discourse has focused on alarming scenarios involving AI systems escaping digital containment, infiltrating the internet, and deploying autonomous agents to manipulate outcomes. However, as noted by Jacob Coxon, an Anthropic computer scientist, the broader existential risks posed by these technologies came into sharp focus only after he highlighted the danger on social media in early September.

While many have been aware of AI’s detrimental impacts on employment, education, and political integrity through the use of deepfakes, alongside the environmental costs of data centers consuming vast amounts of water and electricity, the threat to human survival remained largely unnoticed by the general public until recently.

According to Judith Levine, however, this lack of awareness is misguided. She argues that scientists and entrepreneurs were fully cognizant of the dangers associated with artificial intelligence a quarter-century ago. Despite this early understanding, the drive for curiosity and profit compelled them to proceed with development regardless of the potential catastrophic consequences.