Build like you're sailing. Never bet against AI.

Imagine your company is a sailboat. There are other boats in the water. Every once in a while, a strong wind comes through. The wind is AI got better. Your job is to be the boat that gets pushed the farthest each time the wind blows.

There’s a phrase I keep hearing in San Francisco: never bet against AI. Never bet that this technology stops improving, or that it gets more expensive, or that it caps out at today’s capability. As a founder, if you bet against AI, you’re betting against the wind. Build with the wind, or get capsized.

        wind ──────────→  AI capability compounding


                          ╱│
                         ╱ │
                        ╱  │   ← sail
                       ╱   │     size = team's belief × AI usage
                      ╱    │     angle = which problem you pick
                     ╱─────┤
                     │     │
                     └─────┘   ← hull = your company
                  ═══════════
              ~~~ ~~~ ~~~ ~~~ ~~~
           ~~~~~~~~~~~~~~~~~~~~~~~~~~

   small sail or wrong angle → the gust capsizes you
   big sail + right angle     → the same gust shoots you forward

Two things matter for the boat.

The sail has to be as big as possible. Big sail means: your team uses AI like crazy, your team believes deeply in the trajectory, your team is humble enough to follow the model where it leads. It means having a mix of ages and prefrontal cortex states (see the other essay) so the team isn’t crippled by collective caution. Your team should include some “Dumb CEOs” — people who let AI decide. Big sail.

The sail has to point the right direction. Direction means: don’t pour your effort into problems that the next model release will solve for free. Don’t build a thin layer that an OpenAI release wipes out. Pick the problems where AI capability needs to be connected to a specific niche, a specific region, a specific customer relationship — that’s where you point. The capability comes from upstream; the connection to ground truth is yours.

If you do both — big sail, right direction — you’ll have months that look bad, even years that look bad. Then a wind comes through (a model release, a price drop, a capability jump) and your boat lurches forward in a way that no amount of effort would have produced. The boats that did the prep work catch the wind. The ones still adjusting their sails when the wind hits are at risk.

The frequency of these gusts is increasing. You can’t be the founder who waits for the wind to start before adjusting. You have to be in position constantly, tweaking direction every day, because the gust is always closer than it looks.

Big sail. Right direction. Never bet against AI.

▮