Dragonfly Partner: As Long as AI Still Poses an "Extinction Risk," We Should Not Rush to Accelerate

Odaily reports: Dragonfly Managing Partner Haseeb stated on the X platform that he was once a professional "gambler," and his core principle was to never touch the "risk of ruin"—that is, always keep chips in reserve for rebuilding and coping with the worst-case scenario. Similarly, AI risk may not be as severe as imagined, and may be easy to solve, but as long as a reasonable possibility exists, we should not take the gamble before fully confirming otherwise.

Haseeb said that if someone believes the probability of AI causing human extinction is 10%, it actually means they believe that in 90% of cases no serious consequences will occur—but humanity cannot determine which scenario it is actually in. Therefore, until it can be confirmed that the extinction risk has been eliminated, humanity should not voluntarily take on this risk.