Anthropic首席执行官达里奥表示，公司新近招聘的人才更多是冲着薪酬而来，而非认同公司的使命。[1]为了与OpenAI、Google、Meta等竞争对手争夺顶尖AI研究人员，Anthropic提供了颇具竞争力的薪酬方案：Performance Engineer岗位年薪范围为28万至85万美元（不含股权和奖金），L4工程师的总薪酬中位数约为130万美元，其中基本工资约40.5万美元，股权和奖金90万美元。[1]
然而，高薪的代价正在显现。行业内人才流动频繁，许多研究人员在获取经济回报后随即跳槽。翁荔曾离开Thinking Machines重返OpenAI，成为该公司一年内离开的第四位联合创始人。[1]类似的人员流动也发生在其他机构：Noam Shazeer从Google转向OpenAI，而John Jumper则从Google加入了Anthropic。[1]这反映出仅靠薪酬无法完全锁定顶尖AI人才的现实——除薪酬外，计算资源、技术自由度和行业地位等因素同样吸引着这些研究人员的注意。[1]
Anthropic CEO Dario Amodei has expressed concern that the company's recruitment efforts are drawing employees motivated primarily by compensation rather than commitment to the organization's mission [1]. Despite offering competitive packages to attract top AI talent, Amodei worries this approach may backfire by prioritizing financial incentives over cultural alignment.
To compete with rivals including OpenAI, Google, and Meta, Anthropic has substantially raised its compensation structure [1]. The company's Performance Engineer positions offer annual salaries ranging from $280,000 to $850,000 excluding equity and bonuses [1]. At the senior level, L4 engineers receive median total compensation of approximately $1.3 million annually, comprised of roughly $405,000 in base salary and $900,000 in equity and bonuses [1].
The talent competition in AI research has intensified dramatically, with multiple high-profile researchers changing employers in recent months. Wenli Weng departed Thinking Machines to rejoin OpenAI, making her the fourth co-founder to leave that organization within a year [1]. Additionally, Noam Shazeer transitioned from Google to OpenAI [1], while John Jumper moved from Google to Anthropic [1]. These moves underscore that top-tier AI researchers now evaluate opportunities based on multiple factors beyond salary alone, including access to computing resources, technical autonomy, and industry positioning [1].
