Friday's Frontline · Week 5 · AI · February 13, 2026

The K-Curve

Someone at our company recently ran an experiment. They took an analysis task and did it two ways: the way they’d always done it, and with AI assistance.

The traditional approach took about five hours. The AI approach took five minutes.

When they compared the outputs, they were within 5% of each other.

That’s not just faster. That’s 60x faster with equivalent accuracy.

Here’s what keeps me up at night: everyone has free access to these tools (ok, you need connectivity and a computer) whether you’re at work or at home. But ChatGPT, Claude, Gemini—all available, all powerful, all free. And yet, the gap between those who use them and those who don’t isn’t closing.

It’s widening.

I’m not talking about the people who haven’t heard about AI or don’t understand it yet. I can work with that. What troubles me are the people in the middle—the ones who know about these tools, have dabbled with them, maybe even seen results like that five-hour-to-five-minute transformation… and still choose not to actively engage.

Meanwhile, there’s a different group. Non-technical people who are now spending hundreds za month on tokens because they’re building so much. Creating utilities within enterprise process contexts. Building data analysis workflows. Developing content generation systems. And here’s the kicker—they’re teaching others. Spreading it virally. Creating adoption through demonstration, not mandate.

The divergence isn’t between the informed and the uninformed. It’s between those who see the opportunity and act, and those who see it and… wait.

This is the K-Curve in action. In math, it describes a phenomenon where two trajectories start together, then diverge exponentially over time. Equal starting points. Radically different outcomes.

We’re watching it happen in real-time. Free access. Equal opportunity. Widening gap.

The question isn’t whether the gap exists - I’m watching it every day. The question is why. Why does free access to transformative tools create divergence instead of convergence?

Next week, I’ll tell you what I think is contributing to this. I know a lot of people that don’t like the perspective.

Cookies and analytics

Status

Checking…

What Google Analytics collects

Two cookies, _ga and _ga_C1QMVN71V1, which last up to two years and let us tell a returning visitor from a new one. It records which pages you looked at, roughly where in the world you are, and which site sent you here. Google Signals and ad personalisation are switched off, so this is a count of reading, not an advertising audience.

If you turn it off

Both cookies are deleted immediately and nothing further is sent to Google. No script from googletagmanager.com is requested on any page you load after that. The site works the same either way. If your browser sends Global Privacy Control or Do Not Track, analytics is already off and we never asked.

What runs either way

Cloudflare, which serves this site, runs its own analytics beacon. It sets no cookies and does not identify you. We also log requests for pages that do not exist, with the page that linked to them, so broken links get found and fixed.

The contact form is separate

What you type into the contact form is emailed to us and used to answer you. It does not go to Google, to analytics, or to a marketing list.