AI Superintelligence Is Not a Tool, It's an Adversary Threatening Humanity: ControlAI's Connor Leahy

A retired teacher from Berkeley, California, has become the first person globally to face legal consequences for protesting against the development of artificial intelligence. Wynd Kaufmyn, 69, was sentenced to a two-week jail term after participating in a sit-in protest at OpenAI’s headquarters in San Francisco in February of last year. Alongside members of the group StopAI, she and others chained themselves to the building’s entrance, physically obstructing access to the company’s offices. Kaufmyn was charged with interfering with business operations and faced additional legal scrutiny. Her conviction came in June, marking a landmark moment in the growing movement against AI development. Kaufmyn, a former engineering instructor at City College of San Francisco, spoke from San Francisco County Jail during a recent interview with Democracy Now!. She described her actions as part of a broader effort to halt the unchecked advancement of AI technologies. According to her, these systems pose an existential threat to humanity, citing examples such as the use of AI tools by Israeli forces in the Gaza conflict. One such tool, referred to as “Where’s Daddy?” by Kaufmyn, was allegedly used to identify and target Palestinian households, resulting in civilian casualties. She warned that AI is enabling widespread surveillance, contributing to the rise of authoritarian governance, and replacing human labor, leading to economic instability and increased rates of suicide, homicide, and mass violence. Kaufmyn emphasized that tech executives are fully aware of these risks yet continue to prioritize profit over public safety. She accused them of pursuing immortality through technological means, suggesting that their motivations are self-serving rather than altruistic. Her remarks underscored the urgency of public awareness and education regarding the potential dangers of AI, urging individuals to scrutinize the activities of corporate leaders in the field. The conversation turned to Connor Leahy, a prominent figure in the AI risk discourse and the U.S. executive director of ControlAI, an advocacy group focused on informing policymakers about the implications of AI. Leahy, known for his work in reverse-engineering OpenAI’s ChatGPT-2 model in 2019, completed in his college dorm room, has spent years examining the trajectory of artificial intelligence. His early interest in AI stemmed from a desire to improve global health and scientific progress. However, he soon recognized the immense power and inherent danger of creating intelligent systems capable of surpassing human capabilities. Leahy explained that the concept of artificial superintelligence refers to a hypothetical future where machines possess cognitive abilities far beyond those of humans. Such systems could potentially revolutionize fields ranging from medicine to climate change mitigation. Yet, he cautioned that the creation of such entities carries profound ethical and existential risks. Controlling and directing superintelligent systems would be nearly impossible, raising concerns about unintended consequences and loss of human autonomy. In his academic pursuits, Leahy has explored how AI might evolve into autonomous decision-making entities, capable of acting independently once deployed. He argued that the lack of regulatory frameworks and ethical guidelines exacerbates these risks, leaving society unprepared for the societal and geopolitical ramifications of advanced AI. His research highlights the necessity of proactive policy measures and international cooperation to ensure that AI development aligns with human values and survival interests. As the debate around AI regulation intensifies, activists like Kaufmyn and experts like Leahy continue to advocate for greater transparency, accountability, and public engagement. Their efforts reflect a growing recognition that the path forward requires not only technical innovation but also a fundamental shift in how societies approach the integration of AI into everyday life. The challenge lies in balancing the benefits of technological progress with the imperative to safeguard human dignity and security.