Dragonfly Partner: AI risks may be overestimated, but “risk destruction” must not be touched

According to PANews on September 25, the managing partner of Dragonfly stated that if someone thinks that the probability of AI causing human extinction (P(doom)) is 10%, it actually means that they think that serious consequences will not occur in 90% of cases, but humans cannot judge which situation they are in. Therefore, humans should not proactively assume the risk of extinction until it can be confirmed that it has been eliminated.

Haseeb added that he was once a professional gambler, and one of the core principles of gambling is to never touch the "risk of ruin", that is, to always reserve chips for reconstruction and response to the worst-case scenario. AI risks may not be as serious as imagined, and may be easy to solve, but as long as there is still a reasonable possibility, one should not take risks without full confirmation.

Habari zinazohusiana

SpaceX President Gwynne Shotwell and her spouse donated more than 2 million SpaceX shares to a children's "Trump Account."

Codex usage details now show GPT-6.2 Astra, suspected to be a gray release test.

Sekta ya crypto ilitumia takriban dola milioni 63 kuzuia Sherrod Brown asirejee Seneti.

Nvidia, Walmart, na wengine wamejiunga na upinzani dhidi ya tokenization ya OKXICE, na kufanya idadi ya makampuni yanayopinga kufikia 50.

Binance imeanza tena uondoaji kwenye mitandao ya ZEC, TON, na SUI.