As companies rethink AI ROI, Replit's AI chief calls token leaderboards 'very dystopian'

Replit's AI chief Michele Catasta said "tokenmaxxing" is "irresponsible" and a poor measure of employee performance. Oisin McHugh/Web Summit via Sportsfile via Getty Images AI leaderboards ranking employees by token use became more popular. Replit's AI chief called the practice "very dystopian" and a poor performance metric. Tech firms like Amazon and Uber are questioning the ROI of AI spending. Employees competing to consume the most AI tokens may have been Silicon Valley's latest workplace flex. Now, some executives are pushing back. Speaking at Web Summit Rio this week, Michele Catasta, Replit's president and head of AI, criticized the recent trend of companies using leaderboards to rank employees based on how many AI tokens they use at work. He said some firms have built internal dashboards where "people are basically competing to be at the top of the ranking and showing everyone that they're burning more tokens than anyone else." "That is very dystopian," he said. The 'tokenmaxxing' backlash The comments come as a growing number of executives question whether maximizing AI usage is the right way to measure adoption. Last month, Amazon said it scrapped its internal AI-use leaderboard , with a spokesperson saying it "was never intended to promote the use of AI for usage's sake." Around the same time, Uber COO Andrew Macdonald said he was not seeing proportional productivity gains from increasing AI costs . Charles Holive, chief AI officer at BNP Paribas CIB, similarly dismissed "tokenmaxxing" as a " vanity metric ." Catasta said that token consumption is a poor measure of employee performance. Token usage is "not proportional to the amount of impact that those people are having within the company," he said. While Catasta said he was "very happy" about what he described as a recent "narrative flip" on "tokenmaxxing," he said that the practice can encourage wasteful behavior. "It's also irresponsible," he said, comparing excessive AI use to leaving the lights on at home and not caring about the electricity bill. "Using AI excessively comes with a certain level of impact: more energy is being used, there's less capacity for other use cases that other companies want to build on top of," he added. Read the original article on Business Insider