# Google’s AI: Brilliant, Bloated, and Barreling Ahead

_If OpenAI is lurking in the shadows and Rabbit stumbled publicly, Google is going full-throttle—unleashing a torrent of AI models everywhere at once._

<p>If OpenAI is lurking in the shadows and Rabbit stumbled publicly, Google is going full-throttle—unleashing a torrent of AI models everywhere at once.</p>

<h2>Veo 3: Awesome and Terrifying</h2>

<p>Veo 3 is real. Launched in May 2025, it’s Google DeepMind’s new text-to-video model that <strong>generates 8-second 720p clips with synchronized audio</strong>—everything from dialogue to ambient sound.</p>

<p>There’s a “Fast” variant too: <strong>more than twice the speed</strong> for Gemini Pro and Flow users. It's available via Gemini mobile, Flow, Google Vids, Vertex AI, and Workspace integrations.</p>

<p>Reactions range from dazzled (“eerie, realistic scenes”) to alarmed ("fabricating realistic riots or election fraud")—a potential tool for misuse despite watermarks and content filters </p>

<h2>480 Trillion Tokens/month: Scale Gone Wild</h2>

<p>At Google I/O, Pichai revealed that Google’s AI pipeline now processes approximately <strong>480 trillion tokens per month</strong>, up from just 9.7 T a year ago—a roughly <strong>50× increase</strong>  .</p>

<p>That volume includes all modalities across Search (AI Mode), Workspace, Gemini APIs, Cloud, and mobile apps. The Gemini app alone serves over 400 million monthly users  .</p>

<p>What’s not clear: how tallies break down between <strong>input tokens</strong>, <strong>instruction prompts</strong>, <strong>output tokens</strong>, or modalities like text vs image vs audio vs video. And how much of that 480 T/month is video-related?</p>

<h2>So What?</h2>

<p>Google is deploying AI models across everything—Search, Docs, Android, Cloud, Gemini apps, Studios, the works . If you want AI in your workflow, Google’s got you. Ambitious, yes. But:</p>

<ul><li>No clear prioritization—this is AI sprawl.<br />Modalities like video/audio are <em>costly</em> tokens.</li><li>Hidden costs—compute, latency, and carbon footprint.<br />Risk of drowning in “AI slop”—low-value content dominates.<br /><br /></li></ul><p><strong>At <a href="/product/overview">Swept</a>, the rule holds:</strong></p>

<p><a href="/post/the-bold-vision-and-harsh-reality-of-the-humane-ai-pin">Humane</a> <em>shipped too soon</em>. <a href="/post/rabbit-r1-security-breach-highlights-the-need-for-robust-validation-mechanisms-in-ai-compani">Rabbit</a> <em>shipped too light</em>.</p>

<p>Google? It's shipping "everything, everywhere" at unprecedented scale.</p>

<p>We're watching the sanely useful signals inside the noise. Let's see what actually sticks. If you're trying to make sense of AI deployment at scale, <a href="/contact">we'd love to help</a>.</p>