Dragonfly Partner: AI Won't Become Safe on Its Own, Humans Must Solve Safety and Control Issues

Odaily reports that Dragonfly Managing Partner Haseeb posted on X platform stating that human history has faced numerous risks such as tiger attacks, the Black Death, smallpox, famine, earthquakes, and wars, but these problems did not disappear on their own—rather, humans gradually solved them through vaccines, sewer construction, alliances, and Earth science research.

Haseeb stated that just because humanity has solved many risks in the past, we cannot assume that AI risks will also disappear on their own; AI will not become safe by itself, and humans must genuinely solve AI safety issues and maintain control over AI. He noted that humanity may soon hand over almost all critical control in society to AI, so it is essential to ensure that humans still hold control at that time.