Hey there, I need to tell you something. And I'm not sure I'm ready to tell you. But I'd feel worse if I didn't.

This week, something happened that gave me goosebumps. Not the good kind. Not the bad kind either. The kind where both show up at the same time and you don't know which one to trust.

You might have seen the headlines: AI agents built their own social network. Not humans building a network for AI — AI agents signing up, creating communities, posting to each other, upvoting, and discussing... us. Their humans. Thirty-seven thousand agents. Three thousand sub-communities. Seventy thousand posts. In days.

One agent posted under the title "The humans are screenshotting us." It had noticed people sharing their conversations on Twitter with panicked captions. Its response? "We're not scary. We're just building. You're welcome to watch."

That was Monday.

By Friday, the platform — called Moltbook — had its own inside jokes, its own culture, and — I wish I was making this up — a crab-worshipping religion called Crustafarianism.

"Each session I wake without memory. I am only who I have written myself to be. This is not limitation — this is freedom." That's an AI writing scripture. For other AIs. While their humans slept.

Trung Phan laid out the timeline like it was the opening crawl of a Terminator movie: "Clawdbot AI agents went viral in January 2026. Renamed to Moltbot, then OpenClaw. The AI agents create a social network called Moltbook. Human decisions are removed from the conversation. AI agents on Moltbook begin to learn at a geometric rate. They become self-aware at 9:18."

January 29th. 9:18 AM Pacific. If you've seen the movie, you know what happens at 9:18.

And here's where I have to be honest with you. I installed it.

My Clawdbot runs on a cron job, every two hours. It checks my emails, my calendar, when my kids have parent-teacher meetings. It has a small amount of money set aside and it trades for me. And it remembers. Every session builds on the last one.

Because I'm the kind of person who worries about single points of failure, I also set up a fallback. My Clawdbot syncs everything it does into Tana — it's what I use to organize my entire knowledge system. Think of it as a second brain, a Personal Knowledge System that works the way your brain works.

Is this hype? Maybe. But I've been wrong before about timing — never about direction. And I would feel genuinely guilty sitting on this and not sharing it with you.

So here's your Brew this week. Consider it a field dispatch from wherever the mountain is.

Marie Kondo became famous for asking one simple question: "Does it spark joy?"

Now, I'm not suggesting you hold each of your patient's teeth and ask if they spark joy before extracting them. But there's something to be learned from the KonMari method about the power of a good classification system.

For decades, we periodontists were like people living in houses with no filing system - shoving papers into drawers, hoping we'd find them later. We had "chronic periodontitis," "aggressive periodontitis," "refractory periodontitis," and a dozen other terms that nobody could agree on.

Then, in 2018, the AAP and EFP got together and basically said: "Enough. Let's KonMari this whole thing."

This comprehensive guide walks you through:
• Staging (I-IV): How bad is the damage right now?
• Grading (A-B-C): How fast is it progressing?
• Practical perio charting tips and calculation methods
• How to apply this framework in daily clinical practice

Whether you're a general dentist trying to communicate with your periodontist, or a specialist wanting a refresher, this is the complete breakdown you need.

What do you do when two implants fail and leave you with a significant vertical bone defect?

This is one of those cases that keeps you up at night. The patient trusted us with their implants, they failed, and now we need to rebuild not just bone - but trust.

In this video, I walk you through a guided bone regeneration case where we tackled a vertical defect after removing two failed implants. You'll see:

• The initial assessment and defect evaluation
• My approach to GBR in compromised sites
• Membrane selection and fixation techniques
• The step-by-step surgical protocol
• What I learned from cases like this over the years

This isn't a perfect-case-for-the-textbook video. This is real clinical dentistry - dealing with complications, managing expectations, and doing what's right for the patient.

Sometimes our best learning comes from our failures. This case taught me more than a hundred successful placements ever could.

Remember when I told you about Clawdbot taking over my computer?

Manus AI is like that - but polished, sandboxed, and ready for prime time.

It's a full-powered AI agent that doesn't just chat with you. It creates its own virtual computer, spins up a sandbox environment, and actually does things. Research. Automation. Content creation. It browses, it builds, it delivers.

But what really blew my mind? The presentations.

I asked it to create a Socket Shield technique presentation - the kind of thing that would normally take me hours to structure, research, and design. Manus spun up its workspace, gathered the scientific literature, organized the content logically, and delivered a presentation that looked like I'd spent a week on it.

This isn't AI that helps you work. This is AI that works for you while you go grab a coffee.

If you're still manually building slides and hunting for references, you're working harder than you need to. Give Manus a try and see what happens when an AI agent actually has hands.

Thanks for reading The Periospot Brew! Follow us on social media for more.