Elon Musk just said his AI agent costs too much to serve. His own company has been showing how the same agent saves money.
A Grok Bot customer had run through their allowance with four days left before the reset. When they asked for it early, Musk’s explanation was direct: “The cost of serving Grok Bot was unsustainably high relative to revenue.”

In a September procurement write-up, xAI reported that Grok Bot found 43 paid seats for one software product with no activity in 90 days, and that review and downgrades produced $14,220 in savings.
Both things can be true. An agent can create substantial value for its user while costing its provider too much to supply at the current price. The cost of doing the work, the price charged for doing it, and the value of getting it done are three different numbers.
We’re about to have to get much better at thinking about all three.
The question underneath the Grok bill is what happens when an individual can afford to direct work that previously required assembling a team. I think this is one of the biggest changes agents bring to ordinary working life. Which work to fund, how much effort it deserves, and whether the result justified the expense are becoming decisions that more of us make ourselves.
You can investigate more possibilities, build things you’d otherwise leave in a notebook, or ask for help with a problem before it becomes urgent enough to justify everybody else’s time. Or you can spend a surprising amount of money generating work that nobody needs, then spend your evenings reviewing it.
In the video that goes with this piece, I promised to help you work out what your time and your agents’ work are worth. We’ll do that, including an interview you can give your AI to work through your own situation.
Here’s what’s inside:
Reopen the work you stopped considering. Find the ideas you shelved because the work made them impractical, and see how an agent changes the cost of investigating them.
Put a working range on your hour. Value your time by its next use, then test ordinary decisions against a range like $20 to $50.
Run the numbers on the inherited weekly report. See how the setup pays back by the second week, and why a $20 option can beat a $5 one once your attention counts.
Budget an idea you’ve never tried. Choose the maximum money and human attention you’re willing to spend finding out whether people will use it.
Have your AI interview you. It asks one question at a time and ends with a decision note and a brief for the agent doing the work.
We need to start with the work you’ve already decided you can’t afford to attempt.
Listen to this episode with a 7-day free trial
Subscribe to Nate’s Substack to listen to this post and get 7 days of free access to the full post archives.











