Silicon Valley's New Obsession: Watching Bots Do Their Work --- Fleets of AI assistants check emails, manage calendars while techies play

“At a holiday gathering in San Francisco, partygoers sipped Celsius and kept sneaking glances at their cracked-open laptops with a mix of pride and fear.

 

They were checking on their fleets of AI assistants to make sure they were still laboring away.

 

What are your bots up to? The topic is the biggest flex -- and source of stress -- in Silicon Valley, where tech pros and amateurs are competing to see how much of their grunt work they can outsource to AI without things backfiring spectacularly. They're enlisting agents to code, manage their calendar and respond to emails.

 

Bots' human masters set them to work when they go to sleep or to parties, and check on them regularly. Call them the modern day Tamagotchi, the digital pet from the 1990s, but with a lot more firepower.

 

Nikunj Kothari, a venture capitalist by day who builds apps by night, wakes up and checks his agents first thing in the morning, before coffee. He says he barely watches Netflix anymore because playing with Anthropic's Claude Code is more fun.

 

Kothari has been staying up past 1 a.m. working on AI projects that have little to do with his day job. "I'm like, just one more prompt!" he said, referring to the instructions he gives agents. "We've been given this magical tool of agents that can do our bidding, so let's maximize every second," he said.

 

The proliferation of Bay Area taskmasters is being fueled by a new wave of AI tools. Anthropic late last year released Claude Opus 4.5, a model with startling new coding abilities.

 

Then came the viral rise of the AI assistant OpenClaw in early 2026.

 

Kothari said he's noticed more people sitting in Dolores Park adjacent to their open laptops, likely babysitting their agents. (If you close your laptop, the agents cease working.)

 

Users compare notes on how long their fleet of virtual interns can work without making a mistake. And they suffer from "token anxiety," Kothari says, the fear that their bots aren't getting enough work done.

 

These surges of enthusiasm are common in Silicon Valley. Waves of productivity tools, from a boom in note-taking and productivity apps like Evernote to Slack to the low-code, no-code movement, all promised to reinvent daily work, only for many of those products and companies to fade into oblivion. Engineers insist this time is different.

 

The increasing sophistication of AI tools has reshaped what it means to be a software engineer in this town.

 

The best developers aren't writing code anymore; they're learning how to lead a small army of AI assistants.

 

"I really want them all to be working overnight, so I'm always running downstairs before bed, just like 'one last check!'" said Simon Last, an engineer and co-founder of workplace startup Notion.

 

Recent data from research group Model Evaluation and Threat Research found the latest Claude model can complete tasks that take humans up to 12 hours.

 

But agents introduce a host of new risks. They can take liberties: A Meta executive posted that her OpenClaw bots began deleting her inbox despite instructing them to seek confirmation before acting. Some bots have lashed out at perceived enemies when they face obstacles.

 

Last compared managing agents to powering a lawn mower: crank it long enough and it starts running on its own. "Every time I sit at my desk, I'm trying to pull a crank," he said.

 

Last, a stereotypical Silicon Valley coder who chain-drinks Coke Zeros, said he has hardly written a line of code in nine months. Instead, he manages four agents and has earned a "black belt" on Notion's new internal ranking of AI fluency. He said he prefers managing agents to managing people.

 

The new elite skill set in tech is how many agents one person can juggle. Five seems to be the current agreed-upon upper limit before chaos ensues -- a sweet spot similar to some human teams. Bots don't need sleep, food, insurance or morale boosts. They just keep working.

 

On a February podcast, Boris Cherny, who leads Claude Code, declared that "the title software engineer is going to start to go away." During the recording, he had five agents working in the background.

 

An Andreessen Horowitz partner joked on X that future generations will "grow up in a world where B.C. refers to "Before Claude."

 

Several people said managing agents gives them a similar dopamine hit as strategy-focused videogames like "Age of Empires" and "StarCraft." (Many engineers also said that their spouses are growing irritated by how often they're checking on their agents.)

 

Jeff Seibert, founder and CEO of AI accounting startup Digits, said he hasn't written code since December. Seibert's new preferred method of working: dictating all of his directions to Claude Code.

 

"At night, I get the kids to bed and then I can just talk to Claude and tell it to code more things for me," he said.

 

Seibert, who learned to code at age 12, said he's found the transition depressing at times because a skill he's spent his life perfecting is "just gone. It's not needed anymore."

 

"On the other hand, I can ship more software than I ever could in my life."” [1]

 

1. Silicon Valley's New Obsession: Watching Bots Do Their Work --- Fleets of AI assistants check emails, manage calendars while techies play. Clark, Kate.  Wall Street Journal, Eastern edition; New York, N.Y.. 13 Mar 2026: A1.