All this talk about MCP and all I can think is “ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ nailed it again.” lol, he talked about this a long time ago and it’s exciting seeing it happen. Props man. Check him out if you don’t already. Shameless plug.

I'm going to do a video on this, but here's my uber-simple explanation of MCPs. 1️⃣ Tons of applications and databases exist in the world 2️⃣ It takes a lot of manual effort to make them usable 3️⃣ MCP is a client/server protocol for making it easy 4️⃣ 👇🏼

I wish there was an explanation; but I’m not disagreeing

Not learning to code just because there are AI coding agents is like not learning how to think because there are talk shows. Writing = thinking. Creating = imagining. Coding = building. If you're in tech in 2025 and you can't do these things, your career is at risk. Adapt.

I love my Tesla vehicles, and I appreciate the work you and Department of Government Efficiency are doing to shine light on the greed and corruption in our government. I just ask that y’all take care to be as surgical as possible with these cuts, as there are many really good folks, from both sides of the

I’m addicted to research and teaching. I truly feel like it’s my calling.

My friends The SecOps Group just launched a fresh website design and are doing a giveaway you don’t wanna miss! 👇🏽 Win a FREE pentesting exam or get a massive 85% discount. Sid and team have doing some amazing stuff in the pentesting world. I am a huge fan of their C-AI/MLPen

Another tool release! Multidoc ...is a Go-based utility that processes input through multiple AI models concurrently (OpenAI, Claude, Gemini, and optionally Lambda Chat's Deepseek R1 671b) and provides a combined response, preserving the best parts of each individual output.

A little hype moment for us at Arcanum Information Security but.. We have been selected to speak at the 1st annual OpenAI Security Research Conference! We will be speaking on: “Comprehensive offensive‑testing methodology for AI systems” 🫡🥂🎉

Come by tomorrow to play Mario Kart against me! Beat me and win a free course! BSidesSF

(Freebie) AI hacking tips from the workshop today: Both on input to a model and output retuned to the user, enterprise implementations of genAI enabled applications use classifiers and “guards” to protect the integrity of conversations and apply security. Classifiers continue

Part of my ongoing research is applying prompt engineering for LLM's. I just dove into all of the current model system prompts and over 40 application system prompts that have been leaked. Some insights with a blog to come later! 1) Some things that stand out to me are that

We’ve submitted Fortnite to Apple for review so we can launch on the App Store in the U.S.

God kept tornadoes for his strongest soldiers – Americans.

🕺

Was just getting warmed up

Walmart to cut 1,500 jobs to lower expenses, WSJ reports #FinancialFreedom #Dividends #FYP #FIRE finance.yahoo.com/news/walmart-c…

Excited for the 1st beta run of this course tomorrow! I've synthesized all the model manufacturers tips and tricks, reverse engineered product prompts, raided the GPTStore, releasing all my proprietary techniques, and parsed SO MANY WHITEPAPERS. I'm pretty sure this will be the

Trump on Elon Musk: “I just wish him well.” Reserved. Poised. Classy. My President! 🫡