A Roman Emperor's Private Journal Has Better ML Career Advice Than Most LinkedIn Posts
What Marcus Aurelius wrote to himself 2,000 years ago still hits harder than your last performance review.
Sunday. The one day you are allowed to be a little philosophical.
So let me tell you about a book I keep coming back to. It’s not an ML paper. It’s not a career framework. It’s the private journal of a Roman emperor who ruled one of the most powerful empires in history - and apparently needed to remind himself, constantly, to just do the work.
Marcus Aurelius never intended anyone to read the Meditations. That’s what makes them so good. No performance. No positioning. Just a man talking to himself about how to show up.
And somehow, almost every page sounds like a note I wish someone had left me at 3am before a model deployment.
“At dawn, when you have trouble getting out of bed...”
He actually wrote that. The most powerful man in the known world, writing notes to himself about motivation.
His answer: you were built to work. The ants work. The bees work. The plants grow. Do you think you were put here to stay warm under the blankets?
I think about this when I’m in the middle of a project that’s going sideways - a pipeline that won’t converge, a dataset that’s messier than anyone disclosed, a model that’s performing beautifully in the lab and falling apart in the clinic. The temptation to avoid it is real.
But avoidance isn’t rest. It’s just delayed discomfort with interest.
“Not to feel exasperated or defeated because your days aren’t packed with wise and moral actions.”
This one I needed badly.
There are days when everything I touch in ML and at work in general feels incremental. Another experiment. Another ablation. Another meeting where I explain why the AUC went up by 0.02 and why that is actually significant, yes, really.
Marcus’s answer was not to burn it all down and find meaning elsewhere. It was: get back up. Celebrate behaving like a human - however imperfectly - and fully embrace the pursuit.
The pursuit is the point. The 0.02 is part of it.
“Like a vine that produces grapes without looking for anything in return.”
This is the one I think about most when I am coaching clients.
There are people who help, then quietly catalogue the debt. There are people who help and feel quietly owed. And then there are the rare ones - the vine people - who produce, hand it off, and go back to growing.
In ML, in healthcare AI, in career development: the vine people are the ones you want to work with. They are the ones who share the dataset, the code, the framework, the connection - and then they are already onto the next thing.
Be the vine.
“What am I doing with my soul?”
He literally asked himself this. Regularly.
For us, the question is: what am I doing with my work?
Not “what’s my job title.” Not “what’s my team’s OKR.” What are you actually building, and does it matter to anyone who isn’t in your Slack channel?
I spent years building models that helped clinicians see things they couldn’t see before. That was my answer to the soul question. Not everyone has a healthcare angle - but everyone has an answer. And if you don’t have one yet, you’re allowed to sit with the discomfort of that for a little while.
Sunday is a good day for it.
The short version:
Marcus Aurelius ran an empire, fought wars, managed a court full of politics - and still found time to write himself honest notes about what actually mattered.
The notes boil down to this: show up, do the work, don’t keep score, and interrogate what you’re building often enough to stay honest about it.
That’s good ML career advice. That’s good life advice.
Now go do something useful :-)
Teodora
Teodora Szasz is a Staff ML Scientist and career coach at teodora.coach. Standout Systems posts every day about AI/ML, career development, and healthcare AI.