The Four Defaults Sabotaging Your AI/ML Career (And How to Override Them)
Why the smartest people in tech often make the worst career decisions—and a framework to change that
You’ve optimized neural networks, fine-tuned hyperparameters, and debugged code that would make most engineers cry. You can explain gradient descent to a five-year-old and architect ML systems at scale.
So why does your career still feel like it’s running on a random walk instead of gradient ascent?
Here’s the uncomfortable truth: The same brain that makes you brilliant at machine learning is actively sabotaging your career decisions.
The Hidden Bug in Your Decision-Making System
In ML, we spend enormous effort identifying and fixing bugs in our models. But there’s a bug in our own cognitive architecture that most of us never address.
Shane Parrish calls them defaults—automatic behavioral programs hardwired into our DNA that execute without conscious thought. They helped our ancestors survive the savanna, but in 2025, they’re often the reason smart people make terrible career moves.
There are four defaults that particularly devastate AI/ML careers:
1. The Emotion Default
You experience anxiety about job security, excitement about a hot startup, or frustration with your current role—and you feel compelled to act immediately.
How it manifests in your career:
Panic-applying to 100 jobs after hearing layoff rumors
Taking the first offer because the interview process is “exhausting”
Rage-quitting after a single bad performance review
Turning down a great opportunity because it feels “risky”
The ML parallel: It’s like training a model purely on the most recent batch of data, ignoring everything that came before. Your emotional spike becomes your entire training set.
“Emotions can multiply all of your progress by zero,” Parrish writes. “It doesn’t matter how much you’ve thought about or worked at something—it can all be undone in an instant.”
2. The Ego Default
This one is particularly dangerous for technical professionals. You’ve invested years mastering complex skills. Your identity is wrapped up in being the expert.
How it manifests in your career:
Refusing to pivot because you’ve “already invested so much” in your current path
Dismissing feedback because “they don’t understand the technical nuances”
Staying in a dying field to avoid admitting you chose wrong
Positioning yourself for roles you’re overqualified for because they’re “safer”
The ML parallel: This is overfitting to your past experience. You’ve memorized your training data so well that you can’t generalize to new opportunities.
“Our desire to feel right overpowers our desire to be right,” Parrish observes. The ego makes you more concerned with appearing successful than actually becoming successful.
