In partnership with

Hey AI Enthusiasts!

we've got two massive Google drops—Chrome becoming an autonomous browser agent AND Genie 3 generating playable 3D worlds from text. Plus viral AI assistant drama, a government cybersecurity fail and AI finding hidden treasures in 35 years of Hubble data.

Let’s dive in!

In today’s insights:

🌐 Chrome adds auto-browse, an AI that navigates the web for you
🦞 Clawdbot rebrands to Moltbot after Anthropic trademark pressure
🔐 US cyber chief shared sensitive documents with ChatGPT
🔭 AI uncovers 800+ hidden cosmic anomalies in Hubble data
🎮 Google Genie 3 lets you create playable 3D worlds from text prompts

Read time: 5 minutes.

🗞️ Recent Updates

The AI Field: Google just dropped "auto browse" for Chrome—a new feature that lets Gemini take control of your browser and complete tasks autonomously, like a human would.

Details:

  • Available now for Google AI Pro ($20/month) and AI Ultra ($250/month) subscribers in the US

  • Gemini can fill forms, log into websites, shop, make purchases, and research across multiple tabs

  • Uses Chrome's built-in password manager (with permission) for login tasks

  • Pauses and asks for confirmation before sensitive actions like purchases

  • New persistent Gemini sidebar for easier multitasking

Why This Matters: This is Google's answer to OpenAI's Atlas, Perplexity's Comet, and Anthropic's Claude for Chrome. The browser wars just became agent wars. Instead of you clicking through websites, your browser does it for you—researching flights, filling expense reports, scheduling appointments. We're watching the browser transform from a tool you use into an assistant that uses itself.

Want to get the most out of ChatGPT?

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

The AI Field: The open-source AI assistant that broke GitHub records last week has been forced to change its name from "Clawdbot" to "Moltbot" after Anthropic issued a trademark request.

Details:

  • Created by Peter Steinberger (founder of PSPDFKit, sold to Insight Partners)

  • Gained 17,830 GitHub stars in a single day—now over 100,000 total

  • Runs locally on your hardware (Mac Mini, Raspberry Pi, old laptop)

  • Connects to WhatsApp, Telegram, Slack and takes autonomous actions

  • Name was a play on Anthropic's "Claude"—Anthropic wasn't amused

The Chaos: During the rebrand, bots sniped the @clawdbot X handle and posted crypto wallets. A fake $CLAWD token hit $16M market cap before crashing 90%+. Security researchers found 181 leaked secrets in related repos, with 65 still valid.

Why This Matters: Moltbot represents a new wave of self-hosted AI agents that run on your own hardware, not in the cloud. It's powerful, open-source, and viral—but the security concerns are real. Running an AI with full shell access on your machine is what the creator calls "spicy."

The AI Field: Madhu Gottumukkala, acting director of CISA (the US Cybersecurity and Infrastructure Security Agency), uploaded at least 4 documents marked "for official use only" to a public instance of ChatGPT last summer.

Details:

  • Documents included agency contracting files designated for official use only

  • ChatGPT was broadly restricted across DHS systems at the time

  • Gottumukkala personally sought an exception to use the tool

  • CISA's automated monitoring tools detected the uploads and issued alerts

  • Internal review launched to determine if any security harm occurred

Why This Matters: The irony is almost too perfect—the head of America's cybersecurity agency potentially compromising security by using ChatGPT carelessly. This comes after reports that Gottumukkala didn't pass a polygraph test. Rep. Bennie Thompson called him "in over his head, if not unfit to lead." This is a reminder that even tech-savvy leaders can make basic AI security mistakes.

Image: ESA

The AI Field: ESA astronomers used an AI tool called "AnomalyMatch" to scan 35 years of Hubble Space Telescope archives—and found over 800 never-before-documented cosmic anomalies in just 2.5 days.

Details:

  • Scanned nearly 100 million image cutouts from the Hubble Legacy Archive

  • Found 1,300+ odd objects—800+ never documented in scientific literature

  • Discoveries include galaxy mergers, gravitational lenses, jellyfish galaxies, and star-forming clumps

  • Dozens of objects were so bizarre they don't fit any existing classification

  • Paper published in Astronomy & Astrophysics

Why This Matters: Hubble has been capturing images since 1990, but humans simply can't process that much data. AI just proved it can find things we've been missing for decades. And this is just the beginning—NASA's upcoming Roman Space Telescope and other new observatories will generate even more data. Tools like AnomalyMatch will become essential for discovery.

Learn AI in 5 minutes a day

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

  1. Sign up for The Rundown AI newsletter

  2. They send you 5-minute email updates on the latest AI news and how to use it

  3. You learn how to become 2x more productive by leveraging AI

The AI Field: Google DeepMind just launched Project Genie publicly, an AI that generates fully interactive 3D environments you can explore in real-time, just from a text description.

Details:

  • Type a prompt → get a playable 3D world you can walk, fly, or drive through

  • Runs at 24fps, 720p resolution, generates the world as you move

  • Choose first-person or third-person camera

  • Simulates physics and object interactions in real-time

  • Available now for Google AI Ultra subscribers ($250/month) in the US

  • Sessions limited to 60 seconds currently

Why This Matters: This isn't just video generation, it's interactive world generation. You describe a forest, a city, a Mars base, and then you actually explore it. The AI creates what's ahead of you as you move. Google calls this a "world model" and says it's a key stepping stone toward AGI. Game development, architecture visualization, training simulations—the use cases are massive. This is the closest we've gotten to "holodeck on demand."

👉 Read more here

🎬 Deevid.ai – Create videos from text, images, or existing clips with AI-powered templates and effects, perfect for fast content creation and social posts.

🤖 CustomGPT - Build custom AI chatbots trained on your business content with no coding required.

🎙️ ElevenLabs - Generate ultra-realistic AI voices, clone your own voice, and dub videos in 70+ languages.

Base44 - Turn ideas into fully functional web apps in minutes using natural language prompts.

🎨 Adcreative - Generate high-converting ad creatives, banners, and videos using AI for any platform.

📧 Plusvibe - AI-powered cold email automation with inbox warm-up, lead enrichment, and deliverability optimization.

✍️ writesonic - AI writing and SEO platform for articles, ad copy, and search-optimized content.

🚀 Emergent - Build full-stack web and mobile apps in minutes using natural language prompts, backed by Y Combinator.

Gamma - Create stunning presentations, websites, and documents with AI-powered design—no design skills needed.

🗞️ More AI Hits

🤖 Top engineers at OpenAI and Anthropic say AI now writes 100% of their code - Anthropic's Claude Code lead Boris Cherny says he hasn't written any code in over two months. OpenAI researcher "Roon" confirmed the same, AI handles everything now. Dario Amodei predicts we're 6-12 months from AI doing all software engineering.

🏥 Anthropic launches Claude for Healthcare - Following OpenAI's ChatGPT Health reveal, Anthropic released Claude for Healthcare — tools that summarize medical history, explain test results in plain language, detect patterns in health data, and prep questions for doctor appointments.

🤖 Robotics startup Skild AI raises $1.4B, triples valuation to $14B in 7 months - Skild AI is building an "omni-bodied" brain that can operate any robot for any task. SoftBank led the round with Nvidia, Bezos Expeditions, and others joining. Their last round was just 7 months ago at $4.5B.

🇨🇳 Chinese AI companies race to release new models one year after DeepSeek - Alibaba, Moonshot, and other Chinese firms are accelerating AI model rollouts as they compete for the next edge in AI applications — marking one year since DeepSeek shook up the landscape.

WRAP-UP

That's all for todays roundup!

  • Google Chrome becomes an autonomous agent with "auto browse"

  • Google Genie 3 generates playable 3D worlds from text prompts

  • Viral open-source AI assistant drama: Clawdbot → Moltbot after Anthropic trademark request

  • CISA chief uploaded sensitive docs to ChatGPT—internal review underway

  • AI found 800+ cosmic anomalies hidden in 35 years of Hubble archives

Login or Subscribe to participate

Until next time!
Olle | Founder of The AI Field

Keep Reading

No posts found