Surf the Current: How to Win by Choosing Momentum and Embracing Disposable Code

What if the secret to winning is not building the biggest ship, but catching the fastest river?

Imagine two founders. One pours months into designing a luxurious yacht built to last for decades. The other grabs a simple raft, straps on a motor, and points into a powerful, unseen current. After a few weeks the raft is far downstream, winning customers and learning the terrain while the yacht is still getting its varnish. This is not a metaphor for laziness. It is a strategy for leverage.

At the same time, software has changed. Tools powered by large language models make it possible to generate vast amounts of code quickly. The new reality is that you can produce and discard code at a rate previously unimaginable. The ability to write 5000 lines, see they do not work, and move on is now normal. That capability is not a threat only; it is a new form of speed. The big question then becomes: where do you direct that speed? If speed is the motor, the current is the terrain.

This essay argues that the smartest teams and creators will combine two moves. First, choose markets and problems with existing momentum: find a fast current. Second, reframe software creation as rapid shipbuilding and ruthless selection: build many provisional vessels, sail them, and keep the ones that learn fastest. Together these choices create a compounding advantage that neither strategy achieves alone.

The tension: speed without direction is noise, direction without speed is slow

Two instincts collide in founders and engineers. One instinct values solidity: long term architectures, careful design, defensible moats. The other values velocity: quick experiments, large bets followed by fast abandonment, and maximal iteration. Both have merit. The problem is that velocity without a receptive environment is wasted effort, while a great environment without speed is an opportunity missed.

To make this concrete, consider two scenarios. A team invests in a new social network built with textbook engineering discipline, targeting a broad market that is not especially hungry for change. They will face slow adoption and fierce incumbents. Contrast this with another team that builds a small integration for a developer workflow that already experiences pain daily. The second team rides a current: users will notice, adopt, and tell colleagues, and the product can iterate rapidly into a living business.

Now mix in the generative code wave. The ability to produce working prototypes rapidly changes the calculus of exploration. You can sample many ideas in a week, but sampling is only valuable when the samples are poured into flowing water. If you spray effort across calm ponds, you will create ripples that fade. If you channel even modest effort into a strong current, the same effort will turn into movement, feedback, and real growth.

A practical framework: Current, Ship, Rudder

To turn these intuitions into action, use a three part framework to evaluate opportunities and structure work. Each part is a lever you can tune.

1. Current: How strong is the momentum in the market or problem space?

• Signals of a fast current include frequent pain, strong network effects, existing flows of attention or money, and low switching friction for early adopters.
• Examples: developer tools that save minutes every day, a niche enterprise workflow that spawns internal champions, a creator vertical with built in distribution.

2. Ship: How quickly can you build a vessel that proves value in that current?

• Here generative tools change everything. The goal is a minimal instrument that creates real feedback. You do not need elegant architecture. You need something that performs the primary job well enough to expose the user's judgement.
• Think in terms of disposable prototypes that are cheap to build and even cheaper to discard.

3. Rudder: Who or what steers the ship once it is in motion?

• Steering is selection, orchestration, and learning. It is where human judgment still matters most: choosing which prototypes to double down on, refining product direction, and scaling systems that work.
• Rudder work includes conversation with early users, metrics that measure the right leading indicators, and the ability to translate rough prototypes into durable product primitives when warranted.

These three elements create a flywheel. A strong current multiplies the effectiveness of a simple ship. Rapid shipbuilding generates many candidates for stewardship. Good rudder work preserves the right winners and turns them into lasting value.

Two concrete analogies that clarify the new craft

Analogy one: fly fishing vs ocean liners. A yacht is built to cross oceans. Fly fishing uses a light rod, nimble technique, and knowledge of where the trout run. In many modern markets, you do not need to cross oceans; you need to catch a tide. Fly fishing succeeds by going where the fish already are, using small cheap tools, then scaling the method when it proves effective.

Analogy two: a research lab that runs many cheap experiments. Old software engineering prized a single well designed experiment that you ran once. Modern practice favors many fast experiments, each designed to generate informative failure. The cost of an experiment has collapsed. That collapse reorders priorities: invest in how you interpret failure rather than in the single experiment that never saw sunlight.

These analogies highlight a counterintuitive result: creating fewer polished artifacts and more rapid throwaway artifacts often produces better outcomes when combined with good selection.

What changes for teams and individuals

If you accept the framework above, concrete roles and practices shift.

• The highest value work moves from pure craftsmanship into curation and steering. Writing a lot of code quickly is less valuable than choosing which of many pieces of code to turn into a product. The hard skill becomes judgement under uncertainty.

• Engineering judgment becomes a meta skill: how to design experiments that generate information, how to limit blast radius, how to instrument early prototypes with signals that matter. The ability to reframe a user reaction into a coherent next step is now more important than pristine architecture.

• The production cost of an idea falls, while the selection cost rises. You must build workflows that allow safe disposal of code, documentation, or models. This requires tooling, tests, and cultural norms that treat discard as normal rather than shameful.

• Speed creates an ethical and systems challenge: rapidly generated systems can behave unpredictably. The