Gemini in Chrome: A Leap Toward Google’s…

<div id="zephr-anchor">
    <div class="duet--article--article-body-component">
        <p>I spent the morning exploring Gemini's integration with Chrome, which seamlessly brings the AI assistant directly into the browser. Instead of accessing the chatbot through its web app, you simply click the new Gemini button in the top-right corner of Chrome to start a chat. What sets this apart is that the browser's built-in assistant can "see" the content on your screen while you browse the web.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>From my perspective, Gemini's integration into Chrome appears to be the beginning of Google's efforts to enhance its AI capabilities. I found myself wishing it could do more. Currently, the Gemini early access version in Chrome is available only to AI Pro or AI Ultra subscribers using the Beta, Dev, or Canary versions of Chrome.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>I started by having Gemini summarize articles from <em>The Verge</em> and even locate gaming news on its homepage. It highlighted new Game Boy games for the Nintendo Switch Online, the upcoming <em>Elden Ring</em> film adaptation, and Valve's significant Steam Deck update.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>Gemini can only "see" visible content on your screen. For instance, if you want it to summarize <em>The Verge</em>'s comments section, you'll need to ensure that it's displayed first. It can follow you across tabs, but it can only extract information from one tab at a time.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>If you prefer speaking to typing, Gemini also has a "Live" feature. By selecting the button in the bottom-right corner of the dialog box, you can ask a question aloud, and Gemini will respond vocally.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>I found this especially useful while watching YouTube videos. For example, during a bathroom remodeling tutorial, I asked, “What tool is he using?” Gemini answered, “It appears he’s using a nail gun to fasten wood pieces together.” In another video, it identified a capacitor on a motherboard along with the tweezers and hot air tool used to remove it. While Gemini can summarize videos and provide details about parts you haven't watched, its accuracy may vary if a video lacks labeled chapters for reference.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>One of my favorite applications of this integration is having Gemini extract recipes from YouTube videos, saving me the hassle of manual note-taking or searching for links. It also came in handy when I asked it to identify waterproof bags on an Amazon search page.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>However, Gemini's performance wasn't always consistent. For instance, when I asked about MrBeast's location in a video about ancient Mayan cities, it initially said, “I don’t have access to real-time information,” only to later provide the location mentioned in the video description. In another instance, it couldn't find a link to buy a specific pair of pliers showcased in a video but did offer links to other products when requested.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>At times, I felt Gemini's responses were lengthy for a small Chrome pop-up window. While you can expand it, it still doesn't offer much space on my 13-inch MacBook Air. One of the main advantages of AI is its potential for quick, concise answers, which it doesn't always deliver unless prompted. Additionally, Gemini’s follow-up questions became somewhat repetitive.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>Despite these issues, I envision Google expanding Chrome’s Gemini features beyond basic Q&A. Google aims for its AI to become "agentic," capable of performing tasks for users, and Gemini in Chrome is likely to evolve in this direction. For instance, after asking Gemini to summarize a restaurant menu, I considered asking it to place a pickup order—something it currently can't do. In the future, I foresee it being able to bookmark travel-related pages or save YouTube recipe videos to my Watch Later playlist.</p>
    </div>
    <div class="duet--article--article-body-component">
        <p>Google seems to be progressing towards this goal with Project Mariner's "Agent Mode," which will enable Gemini to manage multiple tasks simultaneously and search the web for you. Perhaps one day, these capabilities will also