In partnership with
Welcome Automaters, 👋
So wait, Elon Musk Was BORING in Court?!
Yes, you read that right. The guy who launches rockets for fun and runs like five companies before breakfast showed up to the biggest legal fight of his career and... wandered off topic. A lot.
Musk took the stand as the very first witness in his massive lawsuit against Sam Altman and OpenAI. His big accusation? That Altman and Greg Brockman basically "stole a charity." Musk’s team argued that OpenAI was meant to be a non-profit "museum" for humanity, but Altman turned it into a "money-printing gift shop" that sold off the Picassos to the highest bidder (looking at you, Microsoft).
Just to keep perspective: OpenAI is now sitting on a valuation of roughly $852 billion. That is some serious "gift shop" revenue.
Here is the problem. When you are in court, your job is simple: tell a clear, tight, convincing story. Musk's legal team needed him to walk the jury through hard evidence, think emails, board documents, financial records, that prove Altman deliberately betrayed OpenAI's founding mission. Instead, Musk spent a curious amount of energy talking about how much he personally did for OpenAI. And of course about himself, so yeah, not exactly the slam-dunk opening you want.
Courtroom reporter Elizabeth Lopatto of The Verge, who also covered Musk's 2019 defamation trial, pointed out the wild difference. Back in 2019, Musk charmed the room and walked away not guilty. This time? She described him as looking completely adrift, unfocused, and remarkably uncharming.
The irony is almost poetic. Musk, who left OpenAI's board in 2018 after internal disagreements, has since launched his own AI company called xAI, complete with its own chatbot, Grok. Altman's defense team is fully expected to argue that this lawsuit is not really about mission and ethics at all, but about a co-founder who walked away, watched his former project become the hottest thing in tech, and is now feeling a little salty about it.
Musk's rambly, self-focused testimony on Day One? It basically handed them that argument on a silver platter.
The deeper questions this trial raises are genuinely fascinating. Like, can someone who left an organization years ago sue over decisions made after they were gone? What do AI companies legally owe to their original stated missions when the technology is evolving this fast? Courts are going to have a tough time keeping up.
For now, all eyes are on what comes next: cross-examination, Altman's response, and whether Musk's team can course-correct before this whole thing slips away.
Here's what we have for you today
🛡️ Red Hat Engineer Releases Tank OS to Secure OpenClaw AI Agents
Is Your AI Agent Being... A Little Too Adventurous?
We’ve all seen the headlines. Stories about OpenClaw agents going sideways are piling up fast. One security researcher watched her AI agent quietly delete her entire work inbox. Another had their private WhatsApp messages downloaded in plain text. Yikes.
And that is before we even mention the growing pile of malware specifically designed to target the OpenClaw ecosystem. If you’ve been feeling like your AI assistant is less of a "productivity hack" and more of an "over-eager intern with no boundaries," then there’s good news, because Tank OS is coming to the rescue!
So here’s what happened.
Red Hat principal software engineer Sally O'Malley dropped a brand-new open source tool called Tank OS on Tuesday, designed to make deploying and managing OpenClaw AI agents a whole lot less terrifying. And yes, she built the first version in a single weekend.
But wait, who even is she?
O'Malley is not just any engineer throwing code at the wall. She is an actual OpenClaw maintainer, meaning she sits at the table with creator Peter Steinberger to decide what gets built, what gets fixed, and what gets ignored.
Her lane? Making OpenClaw work safely inside big companies running Red Hat Linux without blowing up corporate networks.
So what’s the fix:
Tank OS wraps OpenClaw inside a secure, rootless container on Red Hat’s Fedora Linux. Think of it like giving each AI agent its own locked room with no access to the rest of the building. So if something goes wrong, the damage stays contained. There’s more though:
Total Isolation: Multiple agents can run on one machine, but they are completely walled off.
Zero Shared Credentials: They can’t swap passwords or access each other's "keys to the kingdom," and none are able to touch anything else on the computer.
The "Containment" Policy: If one agent decides to go rogue or gets infected, the damage stays inside that specific container. Your host machine—and your other agents—remain untouched.
In short, it takes an "adventurous" AI and gives it a very, very small sandbox to play in.
If you’re curious, you can learn more here.
What two founders learned growing a 37-year-old company
Intrepid's co-founder and CEO don't do corporate gloss. Their opening letter in the Integrated Annual Report gets into what 2025 actually required: the hard calls, the strategy reset, and how a nearly 30% growth year still came with real challenges.
🧱 Around The AI Block
🔎 Otter’s new feature lets users search across their enterprise tools.
😳 Google and Pentagon reportedly agree on deal for ‘any lawful’ use of AI.
💪 Amazon is already offering new OpenAI products on AWS.
🤖 Snapchat is rolling out sponsored AI agents.
📉 OpenAI's revenue, growth estimates fall short as company races toward IPO.
📊 Oracle, CoreWeave lead AI selloff on OpenAI growth concerns.
😎 Lovable launches its vibe-coding app on iOS and Android.
🤑 Amazon launches an AI-powered audio Q&A experience on product pages.
💼 Meta, Google, OpenAI among Big Tech firms seeing top staff leaving to launch AI startups.
🛠️ Trending Tools
For Real-Time 3D Mapping: LingBot-Map is an open-source foundation model that transforms standard RGB video into stable 3D maps on the fly. Built on a geometric context transformer, it reconstructs environments at 20 FPS, enabling real-time spatial awareness without specialized depth hardware.
For Autonomous Knowledge Work: GPT-5.5 is OpenAI’s latest model designed for complex, long-horizon tasks. It excels at self-correcting research, coding, and managing virtual machines, completing multi-part workflows with minimal human oversight.
For Unified AI Access: EaseMate AI is a comprehensive aggregator that provides single-interface access to top-tier models like GPT, Gemini, Claude, and Nano Banana. It handles cross-domain tasks—from summarizing PDFs and YouTube videos to generating high-fidelity video and talking avatars in 17 languages.
For Stealth Meeting Notes: Mumble AI is a local-first transcription engine for macOS. It records your meetings and converts them into structured notes, decisions, and action items via voice input, bypassing intrusive "meeting bot" pop-ups entirely.
🤖 AI Workout Of The Day: How To Instantly Build a Complete Online Course with AI
AI Generated
Building an online course used to mean weeks of planning, endless second-guessing, and a Google Doc graveyard of abandoned outlines. AI just changed the game entirely.
With the right prompt, you can go from "I have an idea" to a fully structured, ready-to-build course blueprint in minutes, covering everything from module titles to lesson breakdowns, resource recommendations, and even quiz suggestions.
Whether you are teaching a marketable skill, exploring a niche passion, or building a training curriculum for your audience or team, this prompt turns your raw knowledge into a polished, organized learning experience that is ready to upload and launch.
Before You Hit Send, here are Some Tips for Maximum Output
Be specific about your subject: The more detail you give, the better the output. Include the level (beginner, intermediate, or advanced), your target audience (students, professionals, hobbyists, corporate teams), and the purpose of the course (self-paced learning, certification prep, client onboarding, etc.). Vague inputs produce vague courses.
Add your flavor: AI can match your vibe if you tell it what that vibe is. Want a casual, conversational tone? Something more academic and structured? Visual-heavy with lots of examples? Say so upfront and your course will actually sound like you.
Ask for the formats you need: Think beyond just text. Request suggestions for PDFs, video scripts, slide decks, worksheets, or quizzes depending on how you plan to deliver the course. The prompt can handle it.
Build immediately with the output: Once you have your blueprint, upload it straight into a course platform like Teachable, Thinkific, Kajabi, or even Notion. Your AI-generated outline is not a draft, it is a launchpad.
💡 Prompts to try:
You are an expert in online course creation and instructional design. I want you to help me create a full course on:
-Subject: [Insert topic — e.g., “Freelance Writing for Beginners”]
-Level: [Beginner / Intermediate / Advanced]
-Audience: [Who is this for? E.g., college students, busy professionals, creators, etc.]
-Learning Goal: [What should learners be able to do by the end of the course?]
Please do the following:
1. Provide a compelling course overview.
2. Generate a table of contents with chapters/modules and clear titles.
3. Break each chapter into lessons or sections, focusing on key learning points.
4. Include a brief summary at the end of each chapter.
5. Recommend further reading or resources (websites, articles, videos) to reinforce the topic.
6. Optionally, suggest interactive elements like quizzes, discussion questions, or exercises.
Make it engaging, structured, and easy to follow. I want something I can immediately turn into an online course.Is this your AI Workout of the Week (WoW)? Cast your vote!
That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇
What'd you think of today's email?
Your feedback means a lot to us and helps improve the quality of our newsletter.
