John Wong

At a Menlo Park Starbucks down the road from all the big venture capital firms, John Wong hosts a meetup for the AI agent Curious.

Huang is a tech industry veteran who runs his own startup investment network.

And he spent the last few months trying to automate as much busy work as he can.

Wong used to pay for a human assistant in the Philippines.

Now, he just pays for tokens.

That's how AI usage is measured.

A crowd of about a dozen showed up on this Wednesday night last month.

Entrepreneurs, big tech workers, and hobbyist tinkerers.

Kevin State runs sales strategy for an AI startup.

Right now, he's mostly using agents to organize his emails and write research reports.

But he's working on building a custom networking assistant to track who he meets, map connections, and prompt him to follow up.

A sense that you're falling behind if you're not 10x-ing productivity even while you sleep has taken over the industry, says Nikunj Kothari, a venture capital investor in San Francisco.

He calls it token anxiety.

People keeping tabs on their agents during parties, at bars, even while outside at the park.

Qatari himself is running agents for email, market research, data analysis.

His evening Netflix time has been replaced with Claude Code, dreaming up new tasks to automate just for fun.

And it's not just scrappy founder types.

Some of the biggest companies in tech are reportedly token maxing, pushing employees to burn through as many AI credits as possible in an effort to churn out new features and products at an ever faster pace.

There is genuine excitement, but also fear, according to Eric Weber, who spent years leading data and AI teams, most recently at Grammarly.

AI is automating the very skills that tech workers spent their entire careers developing.