The Longest Rest Between Sets

The last time I wrote to you, I made a promise.

Issue #13 ended with me wondering what would happen if I stopped working in near isolation and went to find my tribe: founders, developers, and AI enthusiasts on the same wavelength as me, or slightly higher. I said I’d let you know how the quest went.

That was three years ago.

In gym terms, that’s a long rest between sets. Too long. But I wasn’t resting. So consider this issue my report back.

The tribe report

Short version: quest complete.

It turns out the fastest way to find your tribe is to go where they gather and start contributing. So that’s what I did, repeatedly, for three years.

Since we last spoke, I have:

• Appeared on over 15 podcasts, talking AI agents, memory, developer careers, and where this field is actually heading.
• Spoken on global stages and at some of the top AI conferences in the world, standing in rooms I used to only read about.
• Met 100's of developers across the world, maybe we even ran into each other.
• Moved from MongoDB to Oracle: from AI/ML Developer Advocate to Director of AI Developer Experience, leading AI developer relations and outreach strategy for Oracle AI Database.
• Released MemoRizz, my open-source library for giving AI agents memory. What started as a side experiment became a working lab for ideas I now teach.

Oh, and I created a course with AI pioneer and educator Andrew Ng.

Not once. but three time.

Both are live on DeepLearning.AI: “Prompt Compression and Query Optimization”, “Agent Memory: Building Memory-Aware Agents” and the recently released

The thing that got a name

The deeper I went into AI agents, the clearer one problem became: everyone is obsessed with making agents act, and almost nobody is serious about making them remember.

So I gave the work a name. Memory Engineering: the discipline of designing how AI agents store, manage, and retrieve information so they can learn, adapt, and persist beyond a single conversation.

I’m now on a mission to create 100 AI Memory Engineers. This newsletter is one of the places where that mission plays out.

Key takeaway: The lesson from issue #13 held up. Proximity to people slightly ahead of you compounds faster than working alone ever will. Find the rooms. Contribute before you feel ready.

What happens now

Here’s my commitment: AI With Me is back, and it’s going weekly.

Every week, I’ll be sharing:

• My ongoing experience as a lifelong learner in AI: what I’m studying, what I’m unlearning, and what’s actually worth your attention
• What I’m building and implementing with AI, including the parts that break
• Career progression notes and tips, from developer advocacy to leading AI developer experience at one of the largest technology companies in the world

(And yes, OpenSpeech from issue #13 survived. More on that in an upcoming issue.)

Before you go

If you want high-signal AI news and insights between issues, without the noise, I’ve built a home for exactly that:z

​Subscribe to the LLMCamp newsletter​

Thanks for reading. It’s good to be back.

Richmond