Some Logs You Ride Over. Some Change the Trail.

I was mountain biking this morning and came across a pretty good-sized log in the middle of the trail.

Logs or other obstacles are part of mountain biking. Some you ride over. Some can become a fun jump. Some you can easily ride around. And some are big enough, like today’s, that you need to find another way.

Market changes are a lot like that for founders.

Some don’t really change your business. Some actually create an opportunity. And every once in a while, something shows up that could fundamentally change the trail you’ve been riding.

AI agents may be one of those logs for Amazon.

Meta recently launched an AI shopping agent, and Amazon, while certainly not a startup, quickly blocked it from shopping on Amazon.

Amazon has said this is about authorization, transparency, privacy and security. I’m sure those are all considerations.

But there are some pretty obvious business questions here too.

Amazon made nearly $69 billion from advertising services last year. What happens to that business when an AI agent is doing the shopping and never sees the ads?

What happens to product discovery?

What happens to loyalty when an agent can compare the same product across Amazon, Walmart, Target and everywhere else in seconds?

I have no idea what the answers are. Amazon probably doesn’t either.

But I suspect Amazon has been thinking about these questions for quite a while.

It doesn’t take a genius to see that AI agents were eventually going to start shopping for us. Amazon couldn’t know when it would happen, who would get there first or exactly how it would work.

They could see the log coming.

And I think that’s the useful lesson for founders.

There are probably three or four things happening right now that could materially change your business over the next few years.

You don’t need to completely reinvent your business today based on something that may or may not happen.

But you can think through what your first move would be.

Amazon doesn’t have to figure out the future of ecommerce, advertising and AI agents today. For now, its first move was pretty straightforward.

Block the agent.

Maybe that’s temporary. Maybe Amazon eventually welcomes third-party agents. Maybe it builds its own. Maybe this ends up changing its business model in ways none of us can predict yet.

But Amazon didn’t have to figure all of that out before doing something. Founders don’t either.

Figure out which logs you can ride over. Which ones might make a pretty good jump. And which ones could block the trail.

For those, it’s probably worth having a first move ready.