news

Meta drops token counts from performance reviews

Meta's Maher Saba and Santosh Janardhan told staff in an internal memo that adoption dashboards and token counts are out of performance reviews; managers should weigh the difficulty and quality of shipped work instead.

Published 2026-09-04Source: InfoWorld
Generated Tokenmaxxing editorial thumbnail for Meta drops token counts from performance reviews

Why it matters

Meta engineers had competed on a token leaderboard until the company found staff taking on extra tasks purely to lift their scores. Amazon hit the same gaming and deleted its board. The defining tokenmaxxing metric is being retired by its own adopters.

Tokenmaxxing read

Meta is not switching the telemetry off: a program introduced in April still tracks AI use for training purposes, just not for ratings. That split is the transferable part — keep measuring token spend as a cost signal, stop paying people to raise it.

Source takeaway

InfoWorld is relaying a report by The Information, so the memo text is secondhand and the primary sits behind a paywall. One corroborating detail is already on the record: CTO Andrew Bosworth told staff months ago that usage alone measures no impact.

Topic links

Related projects

Tools that match this angle

#1Direct
Routing

LiteLLM

BerriAI/litellm

An OpenAI-compatible gateway and SDK for calling many model providers with budgets, logging, load balancing, guardrails, and cost tracking.

59.1K11.6KSource-available
gatewaycost-trackingrouting
#2Direct
Observability

Langfuse

langfuse/langfuse

Open-source LLM engineering platform for observability, traces, metrics, evals, prompt management, datasets, and playground workflows.

34.8K3.8KSource-available
tracesevalscosts
#5Direct
Evaluation

promptfoo

promptfoo/promptfoo

A CLI and CI workflow for testing prompts, agents, and RAG systems across models, with evals and red-team style checks.

25.3K2.3KMIT
prompt-evalscirag
Related feed

More source-linked context

IT Pro source artwork
long-formIP
long-form

From tokenmaxxing to valuemaxxing

IT Pro canvasses Gartner, IDC, 451 Research and HPE on what replaces token leaderboards. The Tokenomics Foundation's Mike Fuller says the outcome-first 'valuemaxxing' fix only measures half the equation.

tokenmaxxingmetricscost-governance
Read note
Business Insider source artwork
newsBI
newsmedium review

Silicon Valley's AI token craze is facing a reality check

Business Insider says the gamified token-leaderboard era is yielding to efficiency-maxxing: Amazon told staff not to use AI for its own sake, Copilot moved to usage-based billing, and labs now compete on intelligence per dollar.

cost-governanceexplainermetrics
Read note
Generated Tokenmaxxing editorial thumbnail for “Tokenmaxxing is real, expensive & it’s spreading”: AI budgets are exploding - The New Stack
newsTN
newsmedium review

“Tokenmaxxing is real, expensive & it’s spreading”: AI budgets are exploding - The New Stack

AI accountability startup Lanai debuted Token Tuner, a beta that scores each employee's efficiency by matching token usage and model choice to task complexity — peers burned 10x the tokens for half the efficiency in one beta.

ai-spendcost-governanceexplainer
Read note