Evolve or Die

There are two kinds of employees right now.

The first group hears about a new AI tool and immediately starts experimenting. They try it on everything. They probably overuse it. Half the time it doesn't work, and they're fine with that because they're learning where it does.

The second group has a list of reasons not to bother. The model hallucinates. The output isn't good enough. The data is sensitive. The workflow doesn't fit. Every objection is technically valid. And every one of them is also an excuse.

Guess which group is pulling ahead.

I don't think this is really about AI. It's about a fear of the unknown dressed up as rigor. If you look for a reason not to try something, you'll always find one. That's been true forever.

What's different now is the speed.

Things are moving faster than at any point in my career, and the gap between people who experiment and people who don't is widening every quarter.

I just finished Boone Pickens' autobiography. One thing that stood out to me was how many times he reinvented himself. Different industries, different cycles, different energy markets, different strategies. Even into his 70s and 80s, he was still adapting.

Most people think reinvention is something you do when you're young. I think the best operators do it forever.

Durable Businesses, Adaptive Operators

Here's how I think about it at Filterbuy.

There are parts of the business that aren't going anywhere. Someone still has to manufacture the filter. Someone still has to show up at the house to service the HVAC system. Durable businesses solving real-world problems still matter. That's what I mean by boring money.

But the way you run those businesses — the back office, the customer service, the marketing, the hiring, the analysis layer sitting on top of the physical work — all of that is changing rapidly.

If you're not experimenting there, someone else is.

And in eighteen months, they'll probably have a lower cost structure than you.

Tight Feedback Loops Win

The thing I tell my team constantly is that the shorter your feedback loop, the faster you compound.

Try something.

See what happened.

Adjust.

Try again.

That's the whole game.

The people who pull ahead won't necessarily be the smartest people in the room. They'll be the people willing to test, adapt, and iterate while everyone else waits for certainty.

And certainty is usually fake anyway.

You don't have to bet the company on every new tool.

But you do have to be willing to look stupid for a week while you figure out if something works.

The people who win this cycle won't be the ones with the best AI opinions.

They'll be the operators who kept shipping while everyone else debated whether the tools were perfect.