00:00China has a very different view on AI Armageddon than anthropic or open AI.
00:05And it's a clear sign that even a pause on development won't be enough.
00:09In a recent essay about AI's biggest dangers,
00:12a Chinese official didn't mention human extinction once.
00:15Instead, the top concern is that deep fakes and influence operations
00:19could ruin social stability.
00:21After that, it's cyber security threats.
00:24In China's eyes, humanity, and not AI itself, is the real threat.
00:28Consider a recent bug AI researchers created that could take over an account on WeChat,
00:32China's super app, in seconds.
00:35All a user sees is an incoming call from one of their contacts.
00:38Even if they don't answer, a major breach occurs.
00:41That would be catastrophic in China,
00:43where WeChat is used by more than a billion people for texting,
00:47calling, payments, and even getting around.
00:49Thankfully, the Palo Alto-based company Responsible only simulated the threat,
00:54and WeChat's parent company was able to mitigate it.
00:56But the whole episode proves that cross-border risk sharing isn't just possible,
01:01it's necessary.
01:02Slowing down or pausing the AI race isn't a bad idea.
01:06But without global buy-in, it will be very hard to enforce.
01:10President Xi Jinping, meanwhile, emphasized Beijing's pitch
01:13for an open ecosystem for AI during a recent BRICS summit in New Delhi.
01:18Under that framework, it wouldn't just be the US deciding the rules.
01:21China and the US might not be able to agree on much when it comes to AI governance,
01:25and China almost certainly won't abide by a US-mandated pause on AI.
01:31But both countries need to make transparency a priority.
01:35That might not halt the AI race, but it's a baby step towards making it far less reckless.
Comments