I Thought I Was Selling AI Training. Then I Listened to What People Were Actually Asking For.

When I started The AI Training Group, the pitch was straightforward. I'd go into a business, run workshops, teach the team to use AI tools, and leave them better equipped than I found them.

It made sense at the time. AI was new, people were confused, and education felt like the right response to confusion.

Except something kept happening in those early conversations with leaders. I'd sit down with a CEO or a COO, ready to talk about training programs and team workshops, and they'd say something like:

"Look, I just don't have the headspace for this. I know AI is important. I know we're probably behind. But I can't spend six months getting my team up to speed on tools that might change next year anyway. Can you just come in, figure out where it fits, put it in, and show us how to use it?"

The first few times I heard it, I thought they were being a bit passive. The fifth time, I realised they were being completely rational.

The people who want to learn AI are already learning it.

This is the thing worth understanding. There is a type of person who genuinely wants to get into AI tools: curious, self-driven, willing to experiment and break things and figure it out. Those people don't need training programs. They're already watching tutorials at 11pm, building their own workflows, and quietly becoming the most productive person in the office.

The rest of the organisation is not those people. And that's not a criticism. A good operations manager doesn't need to understand how a language model works any more than they need to understand how their accounting software's database is structured. They need it to work, they need to know how to use it for their specific job, and they need someone else to have made the right decisions about where it belongs.

Trying to turn the second group into the first group with a half-day workshop is, at best, an expensive way to produce a small amount of temporary enthusiasm.

What businesses actually need is someone to do the thinking for them.

Not because they're incapable. Because they're busy running a business.

The leaders I talk to are smart people operating under genuine constraints. They have staff to manage, clients to service, and about forty things on the list that were urgent last week and are somehow still not done. When AI comes up, it gets filed under "important but not urgent" and it stays there.

What changes that is someone removing the thinking overhead entirely. Instead of "you need to learn how to evaluate AI tools, map your workflows, identify opportunities, build business cases, manage change, and train your staff," the conversation becomes: "we found three places where this can save you real time, here's what we built, here's what your team needs to know to use it."

That second conversation takes thirty minutes. The first one takes six months and usually doesn't happen.

What this meant for how I work

I stopped designing training programs and started designing solutions.

The shift is bigger than it sounds. A training program asks "what do your people need to know?" A solution asks "what is currently costing you the most time and money, and what's the simplest thing that fixes it?"

Those are very different starting points and they lead to very different outcomes.

The businesses I work with now don't have AI-literate teams in the traditional sense. They have specific tools built for their specific processes, their people know exactly how to use those tools, and the results show up in the numbers rather than in post-workshop surveys.

One client's team went from spending ninety minutes per job on manual data entry to spending three minutes. They didn't learn AI. They got a platform where the AI does the reading, populates the fields, and flags anything it's not sure about. The team's job changed from "read and type" to "review and confirm." That took about twenty minutes to explain.

None of which means training has no value.

It does. For the self-driven person who wants to understand the tools more deeply, good AI training is genuinely useful. For a leadership team that wants to think strategically about where AI fits in their industry over the next five years, that conversation is worth having.

But those are specific needs for specific situations. They're not a universal answer to "how do we get value from AI?"

The universal answer, as far as I can tell, is simpler and less exciting: find the manual process that's costing you the most, build something that fixes it, and help the people affected by it understand the ten things they need to know. Then find the next one.

That's it. No transformation program required.