Former OpenAI employee Leopold Asch warns that AI labs are neglecting security, potentially exposing key secrets for Artificial General Intelligence (AGI) to the Chinese Communist Party (CCP). Asch highlights the risks of CCP espionage, advocating for intelligence agencies to oversee AI lab security to prevent widespread proliferation of sensitive information.
Great observation. AI doomerism proved fleeting as an effective argument for AI control. Now we are shifting to national security. https://t.co/yUdLrJbglL
The justification for AI control is shifting. AI safety didn’t work. So now it’s national security. This is a general pattern. They alternate between left-wing “safety” and right-wing “security.” https://t.co/8rLdmeXj9g
AI Safety, having failed to convince government of x-risk, is pivoting to NatSec. When all else fails just yell “CHINA” in DC.
AI safety having failed to convince government of x-risk is pivoting to NatSec. When all else fails just yell “CHINA” in DC.
Calling it now, third AI control regulatory grift pivot: -first it was wokeness (need AI aligned to woke priors) -then it was AI x-risk (use something something ASI fairy tale to regulate into oligopoly) -now it will be NatSec argument (use geopolitical competition to ban OSS AI)
“At the very least, the intelligence agencies need to be running security for these AI labs. I currently do not see another way where we don't instantly proliferate this to everybody.” @leopoldasch https://t.co/0oHixGHPFz
Leopold Asch (former OpenAI) outlines the coming natsec/MIC/intel takeover of AI based on the usual fake nationalism & performative jingoism https://t.co/dywPiO1epc
How a US/China superintelligence arms race will play out: “The CCP is going to have an all-out effort to infiltrate American AI labs. Thousands of people, the full force of the Ministry of State Security. There's an enormous incentive for a first strike.” @leopoldasch https://t.co/hxAYNHLWjs
.@leopoldasch on: - the trillion dollar cluster - unhobblings + scaling = 2027 AGI - CCP espionage at AI labs - leaving OpenAI and starting an AGI investment firm - dangers of outsourcing clusters to the Middle East - The Project Full episode (including the last 32 minutes cut… https://t.co/afmkSTRoI8
Former OpenAI employee @leopoldasch says AI labs "treat security as an afterthought" and are "handing the key secrets for AGI to the CCP on a silver platter" https://t.co/SnWcCL8HV5