If you're not working with AI now, you won't catch up later

A founder in Zeeland put it like this the other day: the businesses not working with AI now, you'll simply run past them. Not competitors you'll catch up with later, but parties that drop out of view. That image stuck, because it matches exactly what I saw with the internet, with social media, with crypto. Two groups, one wave, and the distance between those groups grows every time once the first two years are behind us. With AI it's happening faster.

Most SMB founders I talk to are in the phase where they think they still have time. Next quarter. First close that one proposal. Open ChatGPT once the busy stretch is over. That's exactly the behaviour that produces the gap, not the cause of a temporary lag.

Why the distance can no longer be closed

Picture two businesses standing level today. One makes a habit of doing every work process with AI alongside it. The other waits until things calm down. After a month the difference is small. After a year it can no longer be closed.

That's because the lead compounds. A marketer running five solid campaigns a week with AI today will, a year from now, run them noticeably better-grounded for the same effort. The marketer without AI still runs five, at today's level. Every new AI capability stacks on the previous one, and whoever works with it daily picks up each improvement immediately. Whoever stands on the sidelines misses not one step, but the whole structure underneath it.

So the distance doesn't grow in a straight line. It accelerates.

Internet, social, crypto: the same pattern three times

With the internet, the split between companies that took it seriously and companies that waited took about six years. If you only got a real web strategy in 2002, you could still catch up, but the cost of closing the gap was already steep. With social media that window shrank to about three years. The companies that only got serious about LinkedIn and Instagram in 2013 paid agencies to build something others had grown organically.

Crypto played out differently because speculation was a big factor, but the pattern of two groups that can no longer bridge the gap was the same.

Same dynamic three times. First it looks optional. Then it turns out to be the standard. After that it's too expensive to catch up, not financially, but in terms of team, habits and infrastructure you never built.

With AI the window is smaller. Nobody knows exactly how small, but the early signals point to eighteen months to two years before the gap becomes structural. Not hypothetical: I'm already seeing it inside agencies that bill by the hour versus agencies that bill on outcome.

Two days a week is not a hobby

A sigh I hear often: "I just don't have time to put two days a week into this." Understandable, but the maths is off. The cost sits on the other side of the balance sheet. Doing nothing is the most expensive path right now, because the catch-up cost eighteen months from now is a multiple of the learning time today.

Blocking two days a week for AI doesn't mean two days of YouTube tutorials. It means doing your actual work with AI alongside you, making mistakes you won't have to make later, and spotting patterns that only become visible by building yourself. Anyone who keeps that up for a year has a way of working that someone starting later can no longer catch.

This isn't a plea for AI for AI's sake. It's a sober observation that the group that does invest will, two years out, run a different kind of business than the group that waits.

Who doesn't get it 'harder', but disappears?

The most uncomfortable part of this story: some roles disappear, they don't just get heavier. Developers without AI are already harder to place than a year ago. That isn't a raw labour-market forecast, it's what I see shifting in job descriptions.

The same dynamic is coming for marketers still setting up campaigns by hand, accountants who haven't automated their routine work, and agencies billing by the hour in a market where the outcome can be delivered in a fraction of those hours. The question isn't whether it will happen, the question is who's ready when the market shifts.

What you can do this month

The entry point is less dramatic than the section above suggests. Three things that make an immediate difference.

• Pick a process in your own week that you dislike. Writing proposals, answering email, building reports. Spend a week doing it with an AI model alongside you and track where it works and where it doesn't.
• Let your team look over your shoulder. Not as training, but as a habit. Skill with AI emerges in a team through observation and copying, not through a half-day workshop.
• Set a specific goal for three months out. For example: a recurring client task that takes four hours today, brought back to thirty minutes. A concrete goal beats vague enthusiasm.

If you want help with that first step, you can take an AI readiness scan to see where the first wins sit inside your own business. If you'd rather spar directly about how this applies in your sector, the contact page is open.

The gap between the two groups grows every month. That's not a marketing message, that's how waves work. The difference is that this is the first wave we know in advance is coming.