By ranking engineers on AI token consumption, Meta is experiencing Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure." Employees reportedly build bots to needlessly burn tokens for status, demonstrating how gamifying a proxy metric can backfire and disconnect from actual business impact.

While celebrated for high output, 'cracked engineers' can be a double-edged sword. Their focus on speed can create a 'trail of bugs' and technical debt that burdens the team. This superstar culture also risks overlooking essential 'glue work' and may reward individuals who take credit for team efforts, creating an antisocial environment.

A trend called "tokenmaxxing" is emerging in Silicon Valley, where companies like Meta use leaderboards to track employee AI token usage. This reflects a corporate bet that higher token consumption correlates with increased productivity, turning AI usage into a new, albeit gameable, performance metric for engineers.

Gamifying AI token consumption via internal leaderboards, as seen at Meta, creates perverse incentives. Employees may burn tokens to climb the ranks rather than to solve real business problems. This "tokenmaxxing" promotes conspicuous consumption of compute, a vanity metric that masks true productivity and ROI.

Traditional big tech ladders often promote based on scope and cross-team influence, encouraging politics. A better system focuses on skill gradients like "truth-seeking." It rewards being right about foundational decisions, not just being loud or well-positioned, which fosters a healthier engineering culture.

Intense pressure to hit goals corrupts data-driven cultures. Teams may block improvements to A/B testing tools if accurate results threaten a 'win'. This pathology extends to shipping features solely to meet a deadline, with a plan to delete the code immediately after the performance review cycle ends.

Both Meta and Google lacked a formal process for an employee to voluntarily take a lower-level role. The speaker's request was a challenge for recruiters and HR because systems are designed for upward mobility. It required special exceptions and created suspicion, as it's an unconventional career move.

To escape dysfunctional promotion incentives, engineers can join teams with a reputation for a higher technical bar, like Meta's PyTorch. These teams attract talent passionate about the craft, not just advancement. While promotions may be slower, the team's strong reputation can create better long-term career outcomes.

A Meta engineer was denied a promotion despite a "Greatly Exceeds" rating due to a behavioral gap in cross-functional collaboration. This shows that lagging promotions hinge on consistently demonstrating the behaviors of the next level, not just delivering high impact at the current level.