A Conversation with Richmond Alake
I launched over forty courses at Deep Learning AI. Every Monday I’d watch the week’s upcoming course, take notes, and write the marketing. Forty courses. Forty instructors.
Richmond Alake was the only Nigerian, the only Black instructor I saw.
Every time I saw his name on the lineup, I thought: we’re in there.
The Two Heads Question
Richmond grew up in Lagos. At ten, he moved to London. But he carries something from the Lagos schooling system that he says he’s never talked about publicly before.
In Nigeria, you’re ranked. And you’re not celebrated unless you’re number one. Not number two. Not number five. Number one.
“My mom would say: ‘The person that got number one, do they have two heads?’”
That question: “do they have two heads?” is a Nigerian mother’s way of saying: you are capable of anything anyone else can do. Don’t stop. Don’t accept where you are. Keep going.
Richmond took that as a life philosophy. He doesn’t celebrate himself until he’s considered one of the best, or at least among them.
Then he chose AI – a field so vast, and with such concentrated genius, that he says he’ll probably never feel like the best.
“I’ve chosen a field where I’ll be very honest – I don’t think I can ever feel like I am the best. Just geniuses in this space. And that’s why I don’t think I’ll get bored of it.”
Learning in Public
The way Richmond built his career is the same way I’ve been building mine: by working through ideas in public before they’re polished.
After his master’s in computer vision and AI, after stints as a computer vision engineer and machine learning architect, he landed at MongoDB as a developer advocate. Not because he applied. Because he’d written over two hundred Medium articles. Because he’d been learning out loud for years.
“If you can’t explain something as simply as possible, then you don’t understand it,” he told me. “Writing helped me really understand what I was learning.”
Now those same writing skills are a secret weapon in his corporate strategy work at Oracle.
The Agent Memory Thesis
Richmond’s current obsession is agent memory. And his argument is simple: software is becoming agentic. The same way everything became an app, and before that everything became a website – now everything is going to be an agent. And an agent without memory isn’t really intelligent.
“Agent memory – it’s looking into what humans have, drawing inspiration from it, and using that as the seed to how we start to build reliable, believable, and capable agents.”
He works at Oracle’s database division. He says there’s no better place to be for this work — because forty-five hundred companies’ data sits on Oracle systems. The agent memory layer has to live somewhere.
Representation
Years ago, while still writing on Medium, Richmond saw a Black man listed as a Director or VP at Google. He didn’t know him. They’d never spoken. But that image – just the picture, just the title – made something shift.
“That representation just made it possible for me to realize that I myself can be that in a massive company.”
He’s made a pact to himself: he has no option but to represent.
I understood exactly what he meant.
Watch the Full Conversation
The full episode is embedded above (paid subscribers get early access to Practice Ground episodes).
Also on YouTube and Spotify/Apple Podcasts.
This is episode ten of Practice Ground.
Two Nigerians, both from Lagos, both in AI. Still going.
– Nifemi