What Are the Navier-Stokes Equations?

The Navier-Stokes equations describe how fluids move: water in a pipe, air over a wing, blood in an artery, smoke curling off a candle. They are essentially Newton's laws of motion written for a fluid, balancing the forces of pressure, friction, and inertia at every point in the flow.

In principle, if you know how a fluid is moving now, these equations tell you how it will move next. They sit behind weather prediction, aircraft design, and climate models. When Professor Juneja models climate, equations of exactly this family are in the machinery.

The trouble is a feature of the equations called non-linearity: the motion of the fluid feeds back on itself. Small changes can amplify, swirls spawn smaller swirls, and smooth flow can tip into turbulence. Turbulence, the chaotic churning you see in rapids or rising smoke, is often called the last great unsolved problem of classical physics.

Because of this feedback, we cannot in general write down a neat formula for the solution. We solve the equations approximately, on powerful computers, by chopping space and time into tiny steps. Those approximations are good enough to fly planes and forecast weather, but they are not the same as a full mathematical understanding.

In 2000, the Clay Mathematics Institute named seven Millennium Prize Problems, each carrying a one-million-dollar reward. Navier-Stokes is one of them. The precise question is technical, but the spirit is this: can we prove that smooth, sensible solutions to these equations always exist and stay well-behaved, or could a solution somehow blow up into nonsense in finite time?

Remarkably, for equations we use every single day, no one knows. We trust them in practice while being unable to prove, at the deepest level, that they always behave. That gap between 'works in practice' and 'proven in theory' is what makes it a prize problem.

Recently, researchers have brought AI and machine learning to bear on these equations, using them to search for unusual solutions, spot patterns, and guide where to look. This is a legitimate and active area of research, and it echoes the episode's theme of AI as a tool for exploration and search.

It is not a solution to the Millennium Problem. The equations remain unsolved in the formal sense. Keeping that distinction clear, between using a tool to investigate a problem and actually resolving it, is exactly the kind of careful reading the conversation encourages.