Chrome – Automate Actions on Your Website, Apple’s Siri AI Learns Only
Two of the biggest platforms in the world are putting an AI agent on the phone this month, and only one of them is coming to your website to do anything. That’s the difference that determines where your effort goes, and it’s what skips the launch. Apple introduced its new Siri AI at WWDC on June 8. Google’s Chrome Auto-Browse reached Android at the operating system level in late June. Both are real, both move on a large scale, and bet in opposite directions. Someone reads the web to answer you. Others visit websites to work for you.
2 agents, Same month, opposite locations
Within a few weeks, both prominent phone platforms have made the AI agent a default part of the operating system. The coverage is the result like the helper’s competition: his voice is very natural, his answers are intelligent. For anyone using the website, that’s the wrong scoreboard. The bottom line question is: Which of these agents actually come to my website and try to do something? Only one of them does, and you know which one tells you exactly what to prepare.
What Apple Siri AI Actually Does
Apple’s newsroom explains the new Siri AI in three steps. It can “make things happen across apps with even more app actions.” It can “answer questions related to the content on the user’s screen.” And “it can go out on the web to find the latest information using the world’s broadest knowledge and generate useful feedback.”
What Siri does is do within apps, read what’s on your screen, and then go to the web for information to write a response. What it doesn’t do, in Apple’s definition, is navigate to your website and complete a task on it. It is a web reader and an in-app actor. It runs on the device and on Apple’s Private Cloud Compute, a custom model from Google’s Gemini family, and is shipping this fall.
So Siri can completely reveal your content. The way it comes to you is the same discipline as being a clean source that any AI response engine can read: well-written content, served by a server that can pull you into a response. That is a real relationship, and it should be maintained. It’s not like an agent shows up to do the job.
What Google Auto-Browse Actually Does
Google’s Chrome Auto-Browse is another type of agent entirely. It fills out forms, books appointments, reserves parking, and conducts comparison shopping through a browser just like a human would. At the end of June, it comes to Android at the operating system level, coming inside Gemini in Chrome on Android rather than as a separate app or extension.
Two things make it different from all the agents before it. It is enabled by default, and has system-level permissions. It’s not waiting for someone to install an extension or open a different app. It is a browser, and it visits websites to work. If you want a full map of who else is in this category, I keep an active site for agent browsers, but Auto-Browse is the one that changes the stats, due to automation and scale.
One Of Them Appears On Your Website To Work
Apple Agent reads the web for responses and executes actions within applications. A Google agent visits your website and completes tasks on it. For the website owner, that’s the difference, and it’s bigger than any comparison factor.
The real change this month is bigger than two very intelligent assistants: an agent as a visitor, a machine that stays on your website to complete a task, from opt-in to automatic teleportation. Until now, every agent that works on websites was something that someone had to go and open. Auto browsing is turned on by default on hundreds of millions of Android phones. The visitor category that used to be the edge case for power users is becoming ambient, and the version of your website that we encounter is one that you haven’t crawled yet.
It’s another detail that you should grasp, because it says something about who holds the power. Both agents work in Gemini. Apple licenses a model based on Gemini, and Google Auto-Browse Gemini on Chrome. Google is a model under both front doors, which has not softened the question of access to sharpen it.
What to Do Before Discharge
You have two steps before Auto-Browse reaches Android at the operating system level in late June.
First, do your highest-value job comparison with the agent yourself. Open your checkout, your booking form, your lead capture, and try to complete each one the way a non-human would, without relying on visualization. Where it breaks down is where Auto Browsing will break down for your customers.
Second, always be a clean source of learning agents. Server-rendered content and clear structure is what allows Siri to go out into the web and come back with an answer instead of a competitor. The two agents are asking for two different things from your website, and you can use both, but only if you know which is which.
What you shouldn’t do is put this under the features of the assistant you’re going to watch. One of these agents will start completing tasks on your website automatically, at the rate of a phone. That’s not a feature. That’s a new guest with a job to complete.
The race that everyone is running is which assistant is smarter. The question for your website is small and urgent: Which agent shows up to do something, and does anything happen when it does? Currently, that agent belongs to Google; it automatically arrives at the end of June, and the version of your website you encounter is the one you haven’t checked.
Additional resources:
This post was originally published on No Hacks.
Featured image: SvetaZi/Shutterstock
