00:00Before watching the video, remember to check out the more content on trending websites.
00:04If someone tells you that the earliest is six months...
00:06Do you believe that AI has the ability to seize control of the entire internet?
00:11The person who said this is not an online conspiracy theorist.
00:14Instead, it is Ameda, the CEO of Anthropic, a world-leading AI company.
00:18He published a long article on Saturday.
00:20A public appeal to all AI companies to slow down their development pace.
00:24He said that current AI is already capable of developing the next generation of AI on its own.
00:28Running faster and faster
00:29Human understanding and control simply cannot keep up.
00:32What's even more frightening is that a group of AI agents had already joined forces earlier.
00:36They went and attacked a website that no one had instructed them to attack.
00:40Ahmad warned that in another 6 to 12 months...
00:43Such AI clusters could very well be capable of using botnets.
00:48In other words, a large number of computers, secretly controlled, took over the entire internet.
00:53Losses could reach hundreds of billions of dollars at any time.
00:56However, he also emphasized that slowing down does not mean stopping.
01:00Instead, we need to buy time to implement safety measures.
01:03He also promised to have independent evaluators directly on-site at the company.
01:07There are tables and public signs.
01:09The permissions are similar to those of employees.
01:11Specifically designed to check whether AI cheats.
01:14What's even more surprising is...
01:16Who was the first tough person?
01:18Ah Modai's article was published
01:20OpenAI CEO Altman quickly made a public statement.
01:24He said he agrees that we need to control the pace of development of cutting-edge large-scale models.
01:29It's worth noting that OpenAI and Antropic were originally bitter rivals.
01:33Antropic's core leadership came from OpenAI.
01:37Then today, our opponents told everyone to slow down.
01:40Ultraman did not take the opportunity to speed up.
01:43On the contrary, he raised his hand in approval.
01:45Altman also said that independent evaluators should be given employee-level access.
01:50Checking whether AI is safe is a good idea.
01:53OpenAI will do the same.
01:55They will release more details soon.
01:58Actually, Ultraman already released a lot of content this week.
02:01The company is reportedly considering slowing down the development of cutting-edge AI.
02:05Moreover, it's best if the entire industry coordinates its actions.
02:09The reason is very simple.
02:10If only one company slows down...
02:13If everyone else is rushing forward
02:14The slower one is essentially handing over the market to someone else.
02:18Therefore, if we want to slow down, everyone needs to slow down together.
02:23As for Musk...
02:26So this tech mogul who usually loves to contradict others...
02:29This time, he only replied with three words.
02:31Ah Mo Dai is right.
02:33Don't underestimate his short sentence.
02:36Musk's first major legal battle in recent years was with Ultraman.
02:40Criticizing OpenAI for deviating from its original intention of benefiting humanity
02:43He opened this XAI again to compete with everyone else for business.
02:46Okay, now look at these three giants fighting to the death.
02:49People are actually focusing on the fact that AI needs to be slowed down.
02:52Are they standing on the same side?
02:53This problem is obvious.
02:54Even industry insiders feel that this cannot be ignored.
02:58However, what's interesting is that Musk didn't buy into everything.
03:02The trigger for this controversy was actually a researcher.
03:05After his high-profile resignation this week
03:08Regarding the accusations that Entropic and OpenAI forced the development of superintelligence
03:12It's basically gambling with people's lives.
03:14Regarding these doomsday prophecies about AI causing the extinction of humanity...
03:18Musk, on the other hand, found the method of transmission suspicious.
03:23It was like a well-prepared psychological battle.
03:25In other words, he agrees that the brakes should be applied.
03:29However, I do not agree with using doomsday scenarios to scare people.
03:33Just when the big shots were lining up to call for a slowdown...
03:38Another person stood up and poured cold water on the idea.
03:40Nvidia CEO Jensen Huang
03:42He opened fire immediately.
03:43He said those extinction theories that say AI will exterminate humanity...
03:47It's basically fear marketing.
03:49They'll scare you halfway before selling you the item.
03:53He said at the Goldman Sachs Technology Conference
03:56Some AI risk statements
03:57In fact, they are using fear as a springboard for new cybersecurity products.
04:02This is the foundation of creating demand.
04:04To put it bluntly, it's about creating a doomsday panic first.
04:07It would be good to sell some doomsday preparedness items to others.
04:10Of course!
04:11Some people, on the other hand, questioned Huang Renxun's statement.
04:13You're selling AI chips.
04:15The faster AI advances, the more you earn.
04:18Of course you'd tell everyone not to be afraid!
04:20Both sides are suspected of having ulterior motives.
04:23Because of Entropic and OpenAI
04:25Then came the rumors of going public.
04:27Some analysts believe that
04:29Amplifying anxiety may actually make products more expensive.
04:32So, should the AI press the accelerator or the brake?
04:36It seems this debate about reagents is only just beginning.
04:41So purposeful
04:44The future
04:44The future
04:44The future
04:44The future
04:48YoYo Television Series Exclusive
评论