Hey friends,

This one’s a big one. From final CLI battles with Juniper to liquid cooling tipping points, GPT-5 shaking up infrastructure, and even saying goodbye to AOL (😱 yes, really), there’s a lot to cover.

Let’s dive in 👇

🎥 1. A Tale of Two Tools (and One Final Juniper Video)

This is it! Our last (for now) video with the amazing folks at Juniper Networks. I’ve had a blast diving into intent-based networking, the awesome Marvis, and all the wild networking updates Juniper has delivered.

But for this final video, I look at a tale of two tools. Sometimes two tools are fine on their own, but together, they create more operational friction than freedom.

👉 Watch the video + tell me your thoughts

🔥 2. Why Liquid Cooling Is No Longer Optional

Between AI, HPC, and next-gen workloads, liquid cooling has gone from niche to necessary - fast.

That’s why I launched a new Data Center Frontier series + whitepaper.

• Part one: why liquid cooling is now the baseline.
• Part two: how we got here so quickly (spoiler: not just power-hungry GPUs).

👉 Read part one
👉 Read part two

3. The Speed of AI Adoption

The speed of the AI transition has many of us wondering what comes next, and what it all means for our lives and our work.

I had a great conversation with Rich Miller about AI adoption in the historical context of cloud computing - how early skepticism and trust issues eventually gave way to widespread usage as the technology matured.
👉 Watch the full discussion

🤖 4. Smarter Models, Smarter Infrastructure

Hold onto your GPT-generated cowboy hats - ChatGPT-5 is here. And it’s not just about sharper answers or longer context windows. It’s about efficiency, smarter batching, optimized transformer ops, and better performance-per-watt.

And here’s where it gets exciting: hierarchical reasoning models. Instead of flat outputs, these systems “think in layers,” tackling complexity step by step, just like humans.

⚡ 5. The Power Chat: AI’s Real Bottleneck

It’s not GPUs. It’s not cooling. The real AI bottleneck is power.

I sat down with Matt Vincent at Data Center Frontier for a no-fluff Power Chat on how AI’s insatiable energy needs are rewriting site selection, supply chains, and infrastructure strategies. We even coined a new word: braggawatts 🤓.

📟 6. Goodbye AOL, Hello Nostalgia

After 30 years, AOL is shutting down its dial-up service. For some of us, that’s more than just the end of a product, it’s the end of an era. Remember:

• AIM chats that went until 2AM
• “Free 100 Hours” CDs as coffee coasters
• That glorious screech-and-hiss handshake before the web unfolded

The internet felt smaller then. Stranger. Magical. And while the sound is gone, the memories live on.
👉 What’s your first AOL memory?

🌍 7. A New Chapter with iMasons

The future of infrastructure isn’t just about power, GPUs, or cooling, it’s about people.

I’m honored to now serve as Co-Chair of the Infrastructure Masons People Committee, alongside Dame Dawn Childs. Our mission: building intentional programs for education, inclusion, and workforce development.

From iMWomen and Emerging Talent groups to masterclasses and mentorship, this is about creating pathways for the next generation of digital builders.
👉 Learn more about the People Pillar

8. AC vs. DC Power: The Great Data Center Debate

AI workloads are rewriting the rules of data center power delivery. While AC has long been the standard, thanks to its ease of transmission and mature infrastructure - DC power is gaining traction as densities and performance demands soar.

Experts argue DC could be more efficient for tomorrow’s high‑density, GPU‑heavy environments.

👉 Read the full debate in Data Center Knowledge

Let’s keep building. And scaling. And questioning everything.
Until next time

Stay tuned, stay curious, and stay excited!

Best,
Bill Kleyman
CEO, Apolo
Program Chair, AFCOM Data Center World