Artificial intelligence eagerness has led to a competitive push among tech employees to use as much AI as possible. The controversial trend is called "tokenmaxxing," and companies are happily spending thousands to keep up with the output. Whether the practice is sustainable for the industry is up for debate.
What is it?
At the core of the AI workplace trend are tokens. The tokens represent small bits of text that AI models process during a prompt. The use of these tokens helps track AI usage and calculate costs. AI companies “typically charge a monthly subscription for a fixed allotment of tokens,” with additional usage billed separately or available in higher-tier plans, Built In said.
Tokenmaxxing is about “encouraging engineers to consume as many AI tokens as possible,” said Forbes. Companies argue that “token consumption is a key indicator for measuring employee and developer productivity.” There is a growing sentiment that “teams that aren’t burning enough tokens simply aren’t automating enough and get left behind.”
The Week
Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.
SUBSCRIBE & SAVE
Sign up for The Week's Free Newsletters
From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.
From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.
Latest Videos From
Employees rack up tokens by deploying multiple agentic AI models on separate projects simultaneously or by running longer prompts. The trend came to public attention after The Information uncovered that a Meta employee had created an internal leaderboard ranking employees by token usage. Employees were incentivized to use more tokens to outperform coworkers and earn rewards such as digital badges and exclusive titles like “Cache Wizard.” The highest-ranked individual user averaged 281 billion tokens, “which could cost in the hundreds or thousands of dollars,” said The Information. The leaderboard has since been taken down.
Leaderboards are just the icing on the AI-workplace cake. Token budgets are “becoming another form of employee compensation, alongside stock options and yearly bonuses,” said Built In. While some workers go through millions of tokens a week, employers are “happily footing the bill,” believing that “more AI use means more productivity and, of course, more money for the business in the long run.”
Is it worth it?
The popularity of tokenmaxxing “reflects a desire to incentivize AI usage” and presents the assumption that “tokens are the base unit for AI usage” meaning “greater consumption indicates higher value of AI,” Jim Rowan, the U.S. head of AI at Deloitte Consulting LLP, said to Forbes. While well-intentioned, there are “risks of turning tokens into a ‘vanity metric.’”
Still, some proponents of the competitive practice push back against such rhetoric. “We all should be tokenmaxxing,” Sonya Huang, a partner at Sequoia Capital, said to The Wall Street Journal. Artificial intelligence is an “insane new piece of technology that is fundamentally going to rewrite how we work.” What matters most for your company is: “Has my employee become insanely AI-pilled?” That requires “getting them on this tokenmaxxing mindset.”
The tokenmaxxing trend is a “crazy, rushed, temporary phase,” Michael Burry, the investor behind “The Big Short,” said in his Substack Short Thoughts. It is not “merely heavy AI use,” and it is “certainly not sustainable AI use.” It is “quota-driven, leaderboard-driven, management-mandated overconsumption.”
It’s true that the “cost of training AI models is falling, making AI tokens more affordable,” but people have started using “more tokens in their day-to-day tasks,” said The Week sister site Tom’s Hardware. Though AI is “indeed a useful tool,” some companies are “using it to replace people in a bid to cut labor costs.” If the number of tokens needed to accomplish tasks “outpaces the speed at which these tokens become cheaper, then that move might just backfire.”
