The Morning Download: Tokenmaxxing Emerges as an AI Survival Skill

Good morning. Where do you stand on “tokenmaxxing?"

Don't let the Silicon Valley meets Gen-Z slang fool you. The practice of gamifying employee consumption of tokens—quantifiable chunks of information that AI models process—has emerged as a flashpoint over how best to embrace AI after an internal dashboard at Meta Platforms ranking employees by their individual token usage went viral.

The dashboard, a side project created by an individual employee earlier reported on by the Information, has since been taken down at the employee’s discretion, a Meta spokesperson said.

But the debate lives on. Is gamifying token consumption one path towards helping your company adapt to the AI age? Or is it a wasteful and expensive exercise, especially given today's computing capacity crunch?

The WSJ Leadership Institute's Isabelle Bousquette recently dove into the conversation.

Some business leaders see token counts as a vanity metric, almost performative. It's outcomes, not burning through tokens, that matter.

Others argue that the tokenmaxxing mindset is actually critical to their survival. Their companies don't get to billions of tokens without deeply learning how to orchestrate agents and integrate AI into real workflows. That's a valuable skill that develops fast under competitive pressure.

Critics of tokenmaxxing are missing the point, Sonya Huang, a partner at Sequoia Capital, tells Isabelle. Yes, it’s an imperfect metric, she admitted, but “the thing that matters for your company is: is my employee becoming insanely AI-pilled? And that requires getting them on this tokenmaxxing mindset.”

Are you in the tokenmaxxing mindset? Are your employees AI-pilled? Why or why not? Let us know. 

Anthropic's Mythos last month found a bug that had been hiding in an operating system used by firewalls, servers and network appliances for 27 years. And that was just one of thousands it uncovered in a single month. Welcome to what some are already calling "bug armageddon."

The WSJ reports how Mythos has set off a scramble among technology employees inside major companies, as many try to understand how the new model could upend cybersecurity and expose a range of new threats to their products. 

Mythos finds bugs, but it has also proved to be better at writing code that can exploit those vulnerabilities, and that's a potential nightmare scenario for security teams that are already overwhelmed. Fix times now stretch from 160 to 230 days on average, and bug reports are up 76% in a year. 

Some of the greatest risks to companies will likely come from dependencies on open-source tools.  Much of the internet runs on code maintained by volunteers who simply don't have the bandwidth to handle an AI-powered flood of bug reports. Companies also worry that previously ignored technology products might now become targets.

A surge in software stocks helped indexes open the week with gains Monday despite disturbing Iran-related news from the weekend, the Journal reports. Tech stocks led gains in the S&P 500, with the information-technology sector rising around 1% to its highest value since the war with Iran began at the end of February.

Some of Monday’s winners: Oracle, which rose over 12% and was the top-performing stock in the S&P 500; ServiceNow, which climbed more than 7%, and Adobe, which recorded its largest percent increase in a year. 

Meanwhile, shares of Dell Technologies and HP were up Monday on faint hopes that one PC maker could be in takeover talks with Nvidia, Barron's reports.

And CoreWeave stock spiked Monday. The AI cloud company last week unveiled an expanded $21 billion agreement to provide Meta Platforms with AI cloud capacity through 2032 and a multiyear deal with Anthropic for an undisclosed amount.
 

AI-related products accounted for nearly a quarter of U.S. imports in 2025. That’s according to Minneapolis Federal Reserve economist Michael Waugh, who used an LLM to classify more than 18,000 product codes in the trade data by their relevance to data-center construction and operations–chips and servers, but also cooling equipment . From 2023 to 2025, he calculates that AI imports grew 73%, while over the same period imports of non-AI-related products increased just 3%.

OpenAI revenue chief touts Amazon deal. OpenAI's new revenue chief Denise Dresser in a Sunday memo to employees praised the company’s new Amazon partnership, comparing it favorably against its long-standing agreement with Microsoft. “Our Microsoft partnership has been foundational to our success. But it has also limited our ability to meet enterprises where they are — for many that’s Bedrock,” she wrote, referring to AWS’s AI app development platform.

AI fuels Meta’s ad revenue surge. Meta credits its AI recommendation system for boosting watch time of its Instagram Reels in the U.S., enabling the platform to serve up more ads. Advertising research firm eMarketer projects that Meta will surpass Google in net ad revenue this year, reaching over $243.46 billion, edging past Google’s $239.54 billion. 

President Trump’s decision to post an AI-generated image depicting himself as a Christ-like figure sparked outrage on the religious right, triggering the most significant pushback from his Catholic and evangelical Christian supporters since he returned to the White House. (WSJ)

Saudi Arabia is pressing the U.S. to drop its blockade of the Strait of Hormuz and return to the negotiating table, fearing President Trump’s move to close it off could lead Iran to escalate and disrupt other important shipping routes, Arab officials said. (WSJ)

Global oil demand is set to contract this year as the near-closure of the Strait of Hormuz leaves nations scrambling to secure barrels from an increasingly tight pool of supply, the International Energy Agency said. (WSJ)

Home builders are pinching pennies to maintain profits in a stagnant housing market, choosing lower-end fixtures and synthetic over natural materials. (WSJ)

The WSJ Tech Council brings together CIOs, CTOs and CISOs advancing innovation and shaping the future. Join this trusted community where tech executives connect with peers to explore emerging trends and gain the perspective they need to stay ahead of disruption.

Request Information

The WSJ CIO Journal Team is Steven Rosenbush, Isabelle Bousquette and Belle Lin.

The editor, Tom Loftus, can be reached at thomas.loftus@wsj.com.