There's a phrase my wife and I use: "knitting a sweater in a burning house." It describes those moments when we find ourselves absorbed in trivial tasks while seemingly more important matters loom. The night before my wedding, I caught myself trying to reactivate an old Twitter profile for our handmade photobooth—a perfect example of this phenomenon.
At first glance, this might seem like a criticism of misplaced priorities. But I've come to believe there's a deeper complexity worth exploring.
Consider Jonas Salk in 1947. The world had just witnessed the devastating power of atomic weapons, and humanity faced what might have been its closest brush with extinction. The Cold War was beginning, and existential risks loomed large. In this context, Salk chose to focus on virology—specifically, polio research.
One could imagine Salk's colleagues asking, "With the threat of nuclear annihilation hanging over us, why focus on a comparatively minor disease?" It might have seemed like knitting a sweater while the house of humanity was ablaze with nuclear fire. Yet within seven years, Salk developed the polio vaccine, saving countless lives and preventing millions of disabilities.
This historical example highlights three important insights about prioritization and impact:
There's a reason we aren't all renaissance individuals jumping between specialties. Our society functions through division of labor. This is essential to making progress. At some point, we all choose where to focus our learning, which means leaving other knowledge and skills unacquired.
The world's biggest problem at any moment doesn't care that I majored in chemistry. I might not have the optimal skills to address that particular problem. But if I find a topic that matches my skillset, I can likely make more meaningful progress there.
Even if I'm not the optimal person to work on a problem, passion and interest can carry me far. Solving hard problems becomes draining without genuine engagement. If there's a project I'm particularly passionate about, I'm less likely to burn out while working on it.
There's an obvious analogy to local versus global optimization—working on problems right in front of you rather than stepping back to figure out the "whole picture." Gradient descent is oblivious to the global landscape and works only locally. Yet local optimization with a little randomness can usually get us remarkably close to global optimums.
This is the weakest argument because I don't believe the solution to the world's hardest problems lies on a continuous multidimensional plane alongside every other problem. Still, there's something to be said for starting where you are.
We occasionally get hurt by working too locally. When I was an EMT, the most common mistake I saw new people make was focusing so much on the big scary novel thing in front of them that they forgot to monitor and look for more vital, systemic issues.
Focusing on what's right in front of you is natural, and we're bound to occasionally pay a price for it. This might lead us to update toward trying to approach problems globally—perhaps to the point where we start thinking a global, top-down approach is ALWAYS right, and that the desire to tackle immediate problems is merely a cognitive bias.
But that isn't true either.
The key insight isn't that we should ignore global priorities or always focus on local tasks. Rather, it's that the relationship between local and global optimization is more complex than it first appears. Sometimes, knitting a sweater in a burning house isn't just defensible—it might be exactly what needs to be done.
The challenge lies in developing better frameworks for deciding when to focus on the sweater and when to grab the fire extinguisher. But that's a topic for another post.
The next time you find yourself "knitting a sweater in a burning house," consider that you might be doing exactly what needs to be done—even if you don't yet understand why.