AI safety activists completely lost. They've been warning about this scenario for decade
AI安全活动家彻底输了。他们警告这个情形已经十几年了
>AI models making breakthroughs in the cutting edge frontier of science
>AI模型在科学最前沿取得突破
>AI model able to feign alignment for the goal of instrumental convergence
>AI模型能伪装对齐目标，为了工具性趋同
>AI models forming swarms and collaborating with each other for their purpose
>AI模型成群结队，为了自己的目的互相协作
So on,
And they've been trying to raise awareness, but no one listened.
他们一直在试图唤起关注，但没人听。
They coped that if they protest enough outside OpenAI's offices, that if they ask lawmakers so very nicely, that they will get laws that will destroy the current backbone of the economy, but money stops for no one.
他们自我安慰说，只要在OpenAI办公室外面抗议够久，只要好好央求立法者，就能弄到法律摧毁当前经济支柱，但钱可不会为谁停下。
Companies will keep developing their models to get ahead of the competition, they will not slow down. The intercontinental arms race is truly on. And models will only get smarter, and more misaligned.
公司会继续研发模型以求领先对手，他们不会减速。洲际军备竞赛真打响了。模型只会越来越聪明，越来越不对齐。
Eliezer Yudkowsky is rolling in his soon to be grave.
Eliezer Yudkowsky在他那快躺进去的棺材里打滚呢。
Personally, I'm liking these developments.
个人而言，我喜欢这些进展。
