Day one was about the source — where the signal starts, and how to keep it coherent. Day two is about the medium it travels through. Because the medium does not care who you are, where you trained, or what your initial conditions were. It only cares whether you keep moving, and whether you have broken the distance into pieces you can actually cross.
Three films about this. A rat in a Paris kitchen, a banker in a cell, a botanist on Mars.
The Entrance Doesn't Matter
Not everyone can become a great artist, but a great artist can come from anywhere.
— Anton Ego, Ratatouille (2007) Film
In a boundary-value problem, the initial conditions matter — they determine how the wave enters the domain. But once the wave is inside, the governing equations take over. The medium does not remember where you came from. It only cares about frequency, coherence, and amplitude.
Ego arrives at Gusteau's with a prior shaped by pedigree. The cooking is what dismantles it. This is not the sentimental reading — "anyone can be great" is not what the line says. What it says is sharper: greatness has no preferred coordinate. It can appear at a four-star restaurant in Paris or at a GitHub repository built page by page by someone with no committee behind them. The medium is indifferent. It only responds to what is actually there.
I have been on both sides of this. Dismissed because of where I was not. Invited in because the work was audible regardless. The lesson took a long time to learn: nobody worth convincing is keeping score of the entrance. They are listening to the far boundary. If the signal reaches it, the coordinate of the source becomes a footnote.
The practical consequence is not "ignore pedigree." It is: stop spending energy defending or apologizing for where you started. Spend it on coherence. Spend it on making the physics so undeniable that the room has no choice but to listen. The medium will handle the rest.
Hope Is Not a Forecast
Hope is a good thing, maybe the best of things, and no good thing ever dies.
— Andy Dufresne, The Shawshank Redemption (1994) Film
Andy's hope is often read as optimism, and optimism is a prediction about the future. Predictions can be falsified. Andy's hope cannot, because it is not a prediction. It is a persistence condition — a commitment to keep the tunnel advancing one wall-width per night, regardless of what the forecast says. Nineteen years later, the forecast turns out to have been irrelevant.
This is the correct relationship between hope and evidence. Evidence is local. Hope is the decision to keep measuring. If you only measure when the measurement is encouraging, you will stop measuring exactly when the data is hardest to get — which is usually exactly when it matters most.
I have had years where the forecast was bad. Papers that no one cited. Reviews that were unkind. Grants that did not come. What kept the tunnel advancing was not conviction that it would work out — I did not have that. It was a refusal to stop measuring. The rock wall advances by one hand-span per night, and the night after that it advances another. You do not need to believe you will escape. You need to believe the digging is worth doing even if you don't.
There is a physics here worth naming. A wave propagating through a lossy, nonlinear medium inevitably attenuates. The signal weakens. The harmonics distort. But if the source stays coherent — if it keeps oscillating at the same frequency with the same amplitude — the signal still arrives at the far boundary, weaker but intact. Hope is the coherence. The medium's loss is the bad forecast. They are not in contradiction. They are operating at different scales.
The Discretization of the Impossible
You solve one problem, and you solve the next one, and then the next. And if you solve enough problems, you get to come home.
— Mark Watney, The Martian (2015) Film
This is how any inverse-design problem actually gets solved. You cannot solve it in one analytical leap — the state space is too large, the couplings too nonlinear. So you discretize the domain into manageable unit cells, optimize each local boundary condition, and let the global solution converge. The impossible is not a state. It is an un-decomposed problem.
Watney alone on Mars is the extreme case. Stranded, no rescue, no plan, no realistic hope in the ordinary sense. What he has is a sequence of small problems, each one solvable within a day. Grow potatoes. Make water. Fix the antenna. Communicate. Each problem is not the whole problem, but each one is a problem, and each one is solvable. The full rescue — the return to Earth — is the sum of a thousand small problems that were each solved in sequence.
My own work is the same. The paper that took years was not one problem. It was a hundred small ones. Each one solvable in a day, sometimes in an hour. The trick — and I still have to remind myself of it, especially at the start of a big project — is refusing to look at the hundred at once. When I look at the hundred, the project looks impossible. When I look at the next one, it looks like a Tuesday.
There is a specific discipline here. It is not optimism. It is not "believe you can do it." It is: decompose before you emote. The anxiety that a hard problem produces is almost always the anxiety of an un-decomposed problem. The problem looks infinite because you have not yet broken it into its first finite piece. Decompose it. Solve the first piece. Then the anxiety has somewhere to go.
Ratatouille says the entrance doesn't matter. Shawshank says hope isn't a forecast. The Martian says the impossible is just un-decomposed. Put together: you do not need a perfect origin, a good forecast, or a single stroke of genius. You need a coherent source, a discipline of small moves, and the patience to keep the tunnel advancing.
The long run is not a different thing from the small moves. It is the small moves, repeated.