On September 26th, according to a foreign media report tonightMicrosoft AI CEO MUSTAFA SULAIMAN WARNED THAT AI MIGHT BE STRONG ENOUGH TO REQUIRE MILITARY-LEVEL INTERVENTION TO STOP IT。
Sleiman stressed the importance of being able to audit and regulate these complex capabilities to prevent them from going out of control. He said, “We should not panic. Regulation is necessary, but we mustAt the right time and in the right wayI DON'T KNOW. IF AI HAS THE ABILITY TO ADAPT ITS OWN CODE, TO SET ITS OWN TARGETS, TO ACT INDEPENDENTLY AND TO ACCUMULATE RESOURCES, IT WILL BECOME AN EXTREMELY POWERFUL SYSTEM, AND IT IS POSSIBLE THAT IT WILL BE ABLE TO DO SO 5 to 10 years will require military-level intervention to control. "
SULAIMAN HAD PREVIOUSLY WARNED OF THE POTENTIAL DANGER TO HUMANS BY AI. HE SAID, "WE HAVE TO..BUILD AI FOR HUMANSNotTURN AI INTO A MANI DON'T KNOW. HE STRESSED THAT THE DESIGN OF AI FOR HUMANS IS MORE IMPORTANT THAN THE USE OF TECHNOLOGY AS AN ADULT DIGITAL PERSON. WHILE PROMOTING THE AI PARTNERSHIP PROGRAMME, HE REMAINED CONVINCED THAT WHILE ENSURING THE VALUE OF TECHNOLOGY, PROTECTION MECHANISMS SHOULD BE PUT IN PLACE TO GUARANTEE HUMAN SECURITY。
1AI learned from the reports that Google DeepMind CEO Demes Hashabis also believes that humans are "near" to AGI, butSociety is not ready to respondTHIS REALITY KEEPS YOU UP ALL NIGHT. THE DIRECTOR OF THE CYBERSECURITY LABORATORY AT LOUISVILLE UNIVERSITY, AND AN AI SECURITY EXPERT, ROMAN YEUNG POLSKI, CLAIMED, "THERE'S AN A99.99999% PROBABILITYTHE ONLY WAY TO AVOID THIS RISK IS TO STOP DEVELOPING AI FROM THE START。