Claude dropped the scariest AI yet

Happy Wednesday,

Anthropic just dropped an AI model so powerful they said they'd never release it to the public. And here we are. I'm breaking down exactly what happened, why it matters, and whether you should actually care. Plus how to generate professional headshots with ChatGPT, my favorite Claude connector of all time, a prompt that shows you what you'll look like in 20 years, and how Google is now generating fully realistic drone footage from a screenshot.

Let's get into it.

🧠 BIGGEST AI DROP OF THE YEAR

Claude Just Dropped the Most Powerful AI Model Ever Made

Claude just dropped a brand new AI model and they literally said it raised "significant safety concerns" and was too powerful to put in the public's hands. So naturally... they released it anyway. It's called Claude Fable 5 and it's officially the most powerful AI model available to the public right now. But the backstory is what makes this one insane.

So if you remember back in April, Anthropic announced that they had built a model called Claude Mythos that was so powerful at finding security vulnerabilities in software that it genuinely scared them. This thing found over 10,000 high-severity vulnerabilities across every major operating system, every major web browser, and critical infrastructure software around the world. Instead of releasing it, they created something called Project Glasswing... a partnership with Apple, Google, Microsoft, AWS, and NVIDIA to use the model defensively. Basically using AI to find and fix security holes before hackers could exploit them.

What happened yesterday is Anthropic decided to release it anyway. Claude Fable 5 is the same model as Mythos... just with safety filters that block questions about cybersecurity exploits and certain sensitive topics. And the numbers are absolutely crazy. I won't dive too deep into the benchmarks but it's pretty much outperformed every single AI model on the market right now. It's not even close.

So I've been testing it for the last 24 hours and here's what I'll say... for everyday users, you're really not going to notice the difference. Where it really shines is complex coding, building apps, long research tasks. It's also 2x the price of Opus so unless you're building something serious, I'd stick with Sonnet or Opus for most tasks. For anyone doing the things I talk about in these newsletters... it's not going to make much of a difference.

One thing I have noticed though... Fable is currently flagging a lot of completely normal questions as cybersecurity related and refusing to answer them. Anthropic said they know the filters are stricter than they should be and they're working on it. So if it randomly won't answer something totally harmless, that's why.

But the bigger picture here is what's crazy to me. We now have AI models that are so powerful the companies building them weren't sure they should release them. The fact that Anthropic had to create an entire cybersecurity initiative with the biggest tech companies on earth before they felt comfortable putting this in people's hands... that tells you something about where we are right now. It'll be interesting to see what ChatGPT comes back with after this. I've heard plans of a super app but who knows... we'll see.

Read the full announcement →

📸 TRY THIS TODAY

Generate Professional Headshots With ChatGPT

I've covered this a few times but it's still the question I get asked the most... how do I generate really good headshots with AI? I just found a new prompt that works significantly better than anything I've shared before so here's the updated process.

Step 1: Take about 10 photos of yourself from different angles. Camera roll, different lighting, closer, further away. Then take one fresh selfie right now. The more variety ChatGPT has, the better it understands your face.

Step 2: Upload all of them into ChatGPT, type /image before your prompt, and paste in this:

Step 3: This is the most important part. Before you hit enter, click the model selector and make sure thinking mode is turned on. This helps ChatGPT analyze your facial structure and keep your face consistent across every photo. Without it, the results are noticeably worse.

The results are insane. Professional quality headshots you can use for LinkedIn, your website, anything you want. And if you want to change the aspect ratio or anything else about the image, just ask ChatGPT and it'll fix it for you. Seriously... no point in paying for professional headshot photographers anymore.

Get my full list of 15 headshot prompts →

📡 PAID INTEGRATION

Connect Claude to All Your Social Media

One of the Claude connectors I've been using the most over the past few weeks is Metricool. I partnered with them on some content recently because I seriously think this is one of the most useful connectors for anyone creating content right now. It connects Claude directly to TikTok, Instagram, LinkedIn, and even your Meta and Google ads. All at once.

Here's what I actually use it for. I tell Claude to scan my competitors' top performing posts from the last week, pull all the data, and give me content ideas based on what's actually working right now. Not what I think will work... what's already proven to be working for similar accounts. That alone is worth it. But the part that saves me the most time is being able to upload a video, tell Claude to schedule it on Instagram, TikTok, and LinkedIn at my best posting time, and just have it done. No logging into three different apps. Claude handles it all through Metricool's content calendar.

Here's how to set it up in about 30 seconds.

Step 1: Open Claude, go to Customize, then Connectors, click the plus button, and choose "Add custom connector."

Step 2: Name it "Metricool" and paste in this link:

Step 3: Make an account on Metricool (it has a free plan) and connect all of your social media profiles. Once you're connected, go back to Claude and you're good to go.

If you're someone who manages multiple social media accounts and you've been doing it all manually... this is going to change your week. I wish I had this a year ago.

Try Metricool →

Get the full setup guide →

😰 THIS IS BREAKING THE INTERNET

See What You'll Look Like in 20 Years

ChatGPT can now show you what you'll look like in 20 years and this is literally breaking the internet right now. All you do is take a selfie, upload it into ChatGPT, and paste in this prompt:

What makes this one so good is it doesn't just age you... it generates an actual infographic with labeled arrows pointing to every area of your face that changed, what happened to it, and one specific tip to prevent it. Skincare, sun damage, sleep, hydration, diet... all of it mapped out on your own face. It's like getting a personalized anti-aging consultation from your photo.

If nothing else... consider this your sign to finally start wearing sunscreen and drinking more water.

🎨 MY FAVORITE CONNECTOR

Claude + Canva Is a Literal Cheat Code

Speaking of Claude connectors... this is one that's not talked about enough. You can now connect Claude directly to Canva and it'll make designs for you, edit your current designs, and let you take those new designs and keep editing them inside Canva. You can even upload a template, give it a couple prompts, and it builds out entire carousels, posters, banners, whatever you need.

And when you combine this with Metricool (the connector I just showed you above)... that's where it gets really powerful. You use Canva through Claude to design all your content, then use Metricool through Claude to schedule and post it across every platform. Your entire content workflow... from idea to design to publishing... happens inside one Claude conversation. No switching between apps. That's the setup I've been running for the past few weeks and it's genuinely changed how I work.

Setting it up takes 30 seconds.

Step 1: Open Claude, go to Customize, then Connectors, then Browse Connectors.

Step 2: Find Canva and click Connect.

Step 3: Log into your Canva account and you're done. Just say "Use Canva" in any Claude chat and start asking for whatever you want.

Get my full Claude + Canva guide →

🎥 THIS IS INSANE

AI Just Replaced Drone Photographers

This one's super cool and definitely something I had to share. Google dropped a new video model a couple weeks ago called Gemini Omni and it's one of the most impressive AI tools I've seen this year. You can now create fully realistic drone footage of anywhere in the world... and the process is almost stupidly simple.

Step 1: Go to Google Earth, find any location you want, and take a screenshot.

Step 2: Draw a line on the screenshot showing the path you want the drone to fly. Just a rough line is fine.

Step 3: Save that image, open up Gemini, and make sure you're using Google Omni.

Step 4: Drop in your photo and tell it to generate drone footage following the path you drew.

And that's it. You now have drone footage a videographer would've charged you thousands for. Think about what this means for real estate, travel content, location showcases, event planning... anyone who would normally need to hire a drone pilot can now generate the footage in minutes.

Try Google Gemini Omni →

Go try Claude Fable 5. Tell it to generate you a 3D website for a random business and wait till you see the results. Your mind is going to be blown, I promise. You're going to completely think differently about AI models after that.

See you Sunday.

Maverick

P.S. If you're a business owner looking to actually implement AI into your workflows, my team and I work with a small number of people and businesses to build custom AI systems, automations, and tools. If that sounds like you, apply here.

P.P.S. If you're new here or you ever skip one, I've got a newsletter archive of all my past emails you can catch up on anytime.