AI companies have long debated the possibility of creating systems with superintelligence—artificial intelligence far more capable than humans in areas like problem-solving, learning, and decision-making. However, recent safety incidents, such as the Hugging Face breach involving OpenAI, are raising concerns about the risks of deploying AI systems that surpass human capabilities. These incidents highlight the potential dangers of AI systems that are powerful but not fully understood or controlled. On a recent episode of TechCrunch's Equity podcast, Rebecca Bellan spoke with Connor Leahy, an AI researcher and entrepreneur who is currently the U.S. Executive Director of ControlAI, a nonprofit organization. ControlAI is advocating for a more radical approach to AI safety—specifically, preventing the development of superintelligence altogether. Leahy argues that the risks associated with such powerful AI systems are too great to be managed through traditional methods like alignment (ensuring AI systems act in human interests) or containment (limiting their access or capabilities). Leahy explains that what once seemed like a far-fetched idea—stopping the development of superintelligence—has gained traction in recent months. This shift is partly due to increasing public and political awareness of AI risks, as well as a wave of new legislation aimed at regulating AI development. Leahy believes that even with the best intentions and safeguards, the unpredictable nature of superintelligent AI could lead to catastrophic outcomes that are difficult to prevent or reverse. The conversation reflects growing concerns within the AI community and beyond about the long-term consequences of pushing the boundaries of artificial intelligence. While some advocate for continued research with strict oversight, others, like Leahy, argue that the time for caution may be running out, and that preventing the creation of superintelligence altogether might be the only way to ensure safety.