What Amir Shneider Found in the New Siri Beta
Amir Shneider has been testing the beta of the new Siri, and his verdict is unambiguous: this is a real leap, not another incremental update. For the first time, he says, Siri feels less like a system waiting for an exact command and more like an assistant that understands what you are trying to achieve.
The difference shows in the very first exchange. Siri holds a more natural dialogue, remembers context and understands follow-up questions. Its general knowledge is broader too, and the gap in conversation quality compared with earlier versions is immediately noticeable.
What the New Siri Can Actually Do
The most significant capability is access to personal information. Siri locates details across messages, mail and photos, understands what is currently on screen, and takes action inside apps and across the device.
In everyday use this means you can ask it to find a detail from an earlier conversation, draft an email from scratch, or edit and share a group of photos. All of it without breaking the task into a sequence of small instructions, as has been the norm until now.
Why This Is the Difference Between a Voice Assistant and an AI Agent
Shneider sums up the shift in one sentence: Siri is moving from answering commands to understanding context and executing. A classic voice assistant waits to be told exactly what to do. An AI agent understands the goal, gathers the information it needs and gets it done.
That shift matters especially for organizations and professionals already working with AI agents on the desktop. Once the agent is also in your pocket, with access to messages, calendar and photos, a large share of the day's small tasks moves to it.
Why a Human Still Has to Stay in the Loop
Shneider does not hide the limitations. The version is still in beta, with glitches, misunderstandings and answers that need checking. The rule he repeats across all his posts about agents applies here as well: responsibility for the outcome stays with the human.
The practical takeaway is simple. It is worth starting to experiment now, learning where Siri excels and where it fails, and building work habits that include verification. Whoever arrives prepared for the stable release will save the time everyone else is still spending on the learning curve.