A new method for evaluating employee performance is emerging in major tech companies. Employees are now being ranked not based on their achieved results, but on how intensively they use AI models. Specifically, the Claudeonomics system created by engineers at Meta tracks the volume of tokens consumed by employees and grants statuses like "Token Legend" to the most active users. This is reported by Ixbt.com reports .
This practice, known as "tokenmaxxing," is being discussed not only at Meta but also at influential organizations like OpenAI, Anthropic, Shopify, and Sequoia Capital. NVIDIA CEO Jensen Huang predicted that in the future, employee token consumption would be tied to their salaries, and management would resemble the control of computing resources.
However, experts and philosophers are sharply criticizing this approach. Philosopher C. T. Nguyen points out that no measurement system is neutral and that it shapes employee behavior. As a result, the actual value of work may become less important than the sheer volume of interaction with AI.
Experts compare this situation to the errors in the banking system before the 2008 financial crisis. Back then, KPI metrics encouraged employees to issue loans without assessing risks. Today, there is a risk that measuring token counts will incentivize the consumption of computing power rather than the quality of work.
Currently, alternative approaches are being observed in the industry, such as attempts to measure energy consumption or final outcomes instead of tokens. Time will tell whether "tokenmaxxing" will remain a temporary trend in the corporate world or become a new standard for managing digital labor.
Read “Zamin” on Telegram!
