Everyone at B2B Ignite agreed about AI. That's what worried me.

Last week I sat in a packed room at B2B Ignite in London and listened to some of the sharpest people in our industry talk about AI. Session after session, a consensus formed. Agents. Automation. Orchestration. Efficiency. More content, produced faster, by fewer people.

Nobody was wrong, exactly. The efficiency gains are real, and any marketer ignoring them is leaving money on the table. But somewhere around the third talk, a thought got stuck in my head and wouldn't leave: a whole room of smart people had completely agreed about a technology, and the thing they agreed on was the smallest thing it can do.

For a while I wondered if I was the one missing something. Then I realised what I was actually watching.

We all see the mirror, not the machine

Ask a room full of doctors what AI is good for and they'll tell you diagnosis. Ask accountants and they'll tell you reconciliation. Ask lawyers and they'll say document review. Every profession, handed the most general-purpose technology in history, describes... its own to-do list.

Marketers are no different. We looked at AI and saw a faster way to do the marketing we already do. Not because we lack imagination (marketers have imagination to spare) but because that's how people size up anything new. Through the lens of our own job. The limit isn't the technology's imagination. It's ours.

But with AI and marketing, I think something deeper is going on. Because there's one thing this technology can do that almost nobody in that room wanted to talk about.

The part of the job we won't let it near

Here's the uncomfortable bit. We're handing AI everything around the craft: the production, the workflows, the scheduling, the scale. And we're keeping one thing walled off: the thinking. The empathy. The understanding of what a buyer actually cares about, fears, and needs to hear.

And honestly? I get why. That understanding is the core of who we are professionally. It's what we tell ourselves, correctly, makes a great marketer great. Take away the thing that defines someone's working life and you shouldn't be surprised when they push back. It's not stubbornness. It's self-preservation. If a machine can genuinely help with insight and empathy, what's left that's ours?

So we've settled on a story that keeps everyone comfortable: AI is a multiplier of what we do, never a player in how we think. Efficiency is the safe job to give it. It threatens nobody's sense of self.

The trouble is, comfort and competitive advantage are rarely found in the same place.

What Pythagoras would have done with a calculator

My ten-year-old has a calculator. It does not make him Pythagoras. The genius, seeing the pattern and asking the right question, is still entirely, safely human.

But flip it around. Imagine Pythagoras with a calculator. He wouldn't have used it to do his old proofs a bit faster. He'd have used it to explore territory he could never reach before: questions he'd never had time to ask, patterns he'd never been able to test.

Same tool. Completely different ambition. The difference isn't the calculator. It's what the mind holding it asks of it.

That's where B2B marketing is right now. Most of the industry is using AI like the ten-year-old: same marketing, fewer hours. The marketers who pull ahead will be the ones who use it like Pythagoras, doing marketing that wasn't possible before.

And here's the kicker: AI isn't even a calculator. A calculator only computes. This is the first tool in history that can help with the understanding itself. How a CFO weighs the same proposal differently from a CIO. Why a buying group stalls. What a specific reader in a specific industry in a specific market actually needs to believe before they'll move.

Understanding was always the job

So what does that unreachable territory look like for us? Forrester found that 65% of B2B content goes unused. Not because we don't make enough, but because most of it is written for an average reader who doesn't exist. Meanwhile Gartner's research shows most buying committees disagree internally, and a huge share of deals simply fizzle out because no one resolved it. The content that wins isn't the loudest or the most plentiful. It's the piece that makes each person on that committee feel individually understood.

No marketer, however brilliant, can hold a deep understanding of every reader in every account in every market in their head. That was never a talent problem. It was a capacity problem. And capacity problems are exactly what this technology solves, if we let it work on understanding, not just output.

This is the shift I wish that room had spent a day on. Not "how do we produce more content with AI" but "how do we see the world through each buyer's eyes, at a scale no human ever could." Not more. Deeper. It's the belief we've built Ada on: that AI's real power in marketing isn't doing the same things faster, it's understanding people better.

The marketers who cross that line won't find their craft diminished. They'll find it amplified in exactly the direction their CEOs keep asking for: relevance that earns trust, insight that moves buying groups, marketing that visibly touches revenue. Their judgement, taste and strategy become more valuable, because for the first time they're applied with real knowledge of every reader, not a hopeful average.

Who you are as a marketer was never really the process. It was the understanding. The only question is whether you'll use the most powerful understanding tool ever built, or keep handing it the photocopying.

Pythagoras wouldn't have hesitated.