Workflow
马斯克深夜挥刀,Grok幕后员工1/3失业,谷歌AI靠人肉堆起,血汗工厂曝光
AlphabetAlphabet(US:GOOG) 3 6 Ke·2025-09-14 23:42

Core Insights - The rapid advancement of AI technology is heavily reliant on the labor of data annotators, who often work under high pressure and low pay, facing significant psychological stress [1][2][10] - xAI has recently laid off approximately 500 data annotators, which constitutes one-third of its annotation team, while planning to recruit ten times more specialized mentors [3][6] - Google’s data annotators are also experiencing harsh working conditions, being forced to handle violent and inappropriate content, leading to concerns about the quality of AI models [8][9] Group 1 - xAI's recent layoffs of 500 data annotators highlight a shift in strategy towards hiring specialized mentors instead of general annotators [3][6] - The layoffs resulted in a drastic reduction in the Slack group from 1500 to 1000 members, indicating a significant workforce contraction [3] - Data annotators are facing increasing workloads and psychological pressure, with reports of being forced to review disturbing content [2][9] Group 2 - Google’s data annotators are often misled about their roles, initially believing they would be engaged in writing or analysis, only to find themselves reviewing harmful content [9] - The working conditions for annotators have deteriorated, with task times being reduced and expectations increasing, leading to anxiety and sleep issues among workers [9][10] - The pay for AI evaluators in the U.S. starts at $16 per hour, which, while higher than some regions, remains significantly lower than salaries for engineers in Silicon Valley [9]