AI Browsers: What They Can Actually Do Right Now
By Venkata Anirudh Devireddy · Endoblog.dev
AI browsers went from a novelty to something several companies in the industry are shipping this year alone. So worth understanding what they can do for you.
The basic idea of AI browsers is that the browser will show you the website you want to view, but the AI browser will also understand that website, and perform actions on that website for you. It can take actions like clicking buttons, selecting items from forms, navigating to other websites, all based on the instruction you give it.
What actually works well today
Repetitive tasks with well-defined steps are where these tools shine. Stuffing information into a form with information you provide. Checking prices on a few different sites you specify. Pulling information from a page and organizing it. All of these tasks have a goal and a means of determining if the tool has successfully completed the task.
Where they still struggle
Any task that requires some level of judgment on an ambiguous page. If the website has a particularly confusing layout, contains a CAPTCHA, or requires some level of interaction with information outside of the page itself, these tools will trip up the AI browser toolset. Additionally, they are cautious around any form of payment or account changes which is correct, but limits the autonomy of the toolset in these instances.
Why this matters beyond convenience
Because an AI browser also extends what an AI system can do beyond just answering questions. It closes the gap between telling an AI what to do and having it do it. That's the same gap closing across most AI products right now, not in browsers specifically.
How to actually use one well
Give it tasks to do. "Find the cheapest flight from Seattle to Boston next Friday" is a task that will work better than just "help me plan a trip." The more specific you are with your tasks, the more reliably these tools will complete them for you.
