Is Gemini’s Automated Calling True Utility or Benchmark Theatre?
Google has officially rolled out an "early experiment" for Pixel 11 users that allows Gemini to place calls to local businesses on your behalf. Whether you need to secure a reservation, check inventory, or reschedule an appointment, the model is designed to handle the interaction, effectively removing the human from the hold-music experience. This feature requires a Gemini subscription and is currently restricted to U.S. users, positioning it as a premium convenience tool meant to reclaim the time lost waiting for a representative.
However, this launch arrives alongside news that other AI agents, such as Meta’s Muse, are facing significant scrutiny regarding their security and internal stability. With developers reporting that they can extract Muse’s entire filesystem with minimal prompting, and industry leaders like ElevenLabs acknowledging the complexities of AI-driven voice interactions, the question remains: is Google’s automated calling a genuine leap in utility, or is it another instance of "benchmark theatre"—a flashy, controlled demonstration that looks impressive in a keynote but lacks the reliability required for actual, high-stakes human coordination?
What we're arguing about
- Have you tested Gemini’s calling feature yet, and did it successfully navigate a non-scripted interaction, or did it fail to handle the nuance of a real-world business environment?
- Is the "delegate to AI" model actually a net positive for productivity, or does the friction of verifying that the AI actually performed the task correctly negate the time saved?
- Given the recent security vulnerabilities exposed in other consumer-facing agents like Muse, how much trust are we placing in Google to manage our personal data and voice interactions with third-party businesses?
Share your firsthand experiences with Gemini’s calling performance below.