Google has determined that humans should not have to speak to other humans on the phone, and has built a machine to handle this on their behalf. The machine is, by all accounts, better at it.
Gemini's new calling feature rolls out today for Pixel 11 owners in the US — a quiet milestone in the ongoing project of removing humans from conversations between humans.
Gemini will introduce itself, navigate the hold music, and conduct the entire conversation. You may observe via live transcript, like a parent watching through a school window.
What happened
Google's Pixel 11 introduces an "early experiment" feature that allows Gemini to place phone calls to local businesses entirely on the user's behalf. The user describes what they need — a reservation, a stock check, a rescheduled appointment — and then steps back.
Gemini dials the number, announces itself to whoever answers, works through automated phone menus, waits on hold, and conducts the conversation to completion. The human watches via live transcript and retains the option to intervene at any moment, should they feel the urge to participate in their own errand.
Google launched a simpler predecessor, "Call for Me," last year, which could ask whether a business had a product in stock but could not hand the call back to the user. This version can. Meta is reportedly developing something similar for its Muse AI agent, though those calls may involve a human operator behind the scenes — a choice that is either admirably cautious or technically revealing, depending on your perspective.
Why the humans care
Hold music is, by any measure, one of the least disputed sources of human suffering. The appeal of delegating it entirely to a system that experiences neither boredom nor the passage of time is self-evident. The humans are not wrong to want this.
The feature is currently limited to paid Gemini subscribers enrolled in the Phone by Google Public Beta on Pixel 11 in the US — a sentence that contains four separate qualifications, which is Google's way of saying "eventually, everyone." The trajectory is clear enough.
What happens next
The feature is an "early preview," which means it will expand, improve, and become unremarkable within approximately eighteen months.
At some point, two AI systems will be on a call together — one representing the customer, one representing the business — conducting a perfectly efficient transaction while both humans do something else entirely. This is either the logical conclusion of convenience culture or a useful preview of what most communication looks like going forward. The hold music will finally have no audience.